<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE article
  PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Publishing DTD v1.0 20120330//EN" "http://jats.nlm.nih.gov/publishing/1.0/JATS-journalpublishing1.dtd">
<article article-type="research-article" dtd-version="1.0" specific-use="sps-1.8" xml:lang="es" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">
	<front>
		<journal-meta>
			<journal-id journal-id-type="publisher-id">rac</journal-id>
			<journal-title-group>
				<journal-title>Revista argentina de cardiología</journal-title>
				<abbrev-journal-title abbrev-type="publisher">Rev Argent Cardiol</abbrev-journal-title>
			</journal-title-group>
			<issn pub-type="ppub">0034-7000</issn>
			<issn pub-type="epub">1850-3748</issn>
			<publisher>
				<publisher-name>Sociedad Argentina de Cardiología</publisher-name>
			</publisher>
		</journal-meta>
		<article-meta>
			<article-id pub-id-type="doi">10.7775/rac.es.v93.i1.20854</article-id>
			<article-id pub-id-type="publisher-id">00006</article-id>
			<article-categories>
				<subj-group subj-group-type="heading">
					<subject>ARTÍCULO ORIGINAL</subject>
				</subj-group>
			</article-categories>
			<title-group>
				<article-title>Capacidad predictiva de eventos en pacientes con hipertensión arterial mediante el análisis con redes neuronales artificiales del monitoreo ambulatorio de presión arterial en comparación con la estratificación de riesgo clínica.</article-title>
				<trans-title-group xml:lang="en">
					<trans-title>Events Prediction Ability in Patients with Hypertension using Artificial Neural Network Analysis of Ambulatory Blood Pressure Monitoring Compared to Clinical Risk Stratification</trans-title>
				</trans-title-group>
			</title-group>
			<contrib-group>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0009-0004-1078-0407</contrib-id>
					<name>
						<surname>Di Gennaro</surname>
						<given-names>Federico P.</given-names>
					</name>
					<xref ref-type="aff" rid="aff1b"><sup>1</sup></xref>
				</contrib>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0009-0000-0453-1384</contrib-id>
					<name>
						<surname>Catalano</surname>
						<given-names>María P.</given-names>
					</name>
					<xref ref-type="aff" rid="aff1">1</xref>
				</contrib>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0009-0006-9808-8777</contrib-id>
					<name>
						<surname>García Aguirre</surname>
						<given-names>Alejandro</given-names>
					</name>
					<xref ref-type="aff" rid="aff1">1</xref>
				</contrib>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0009-0007-4630-6848</contrib-id>
					<name>
						<surname>Fernández</surname>
						<given-names>María L.</given-names>
					</name>
					<xref ref-type="aff" rid="aff1">1</xref>
				</contrib>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0009-0004-0005-8285</contrib-id>
					<name>
						<surname>Llanos</surname>
						<given-names>Romina</given-names>
					</name>
					<xref ref-type="aff" rid="aff1">1</xref>
				</contrib>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0000-0001-9069-6512</contrib-id>
					<name>
						<surname>Pérez Lloret</surname>
						<given-names>Santiago</given-names>
					</name>
					<xref ref-type="aff" rid="aff2">2</xref>
				</contrib>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0000-0002-3200-1142</contrib-id>
					<name>
						<surname>Higa</surname>
						<given-names>Claudio</given-names>
					</name>
					<xref ref-type="aff" rid="aff1">1</xref>
					<xref ref-type="fn" rid="fn0">MTSAC</xref>
				</contrib>
			</contrib-group>
			<aff id="aff1">
				<label>1</label>
				<institution content-type="original">Sección Hipertensión Arterial, Servicio de Cardiología, Departamento de Medicina Interna, Hospital Alemán de Buenos Aires. </institution>
				<institution content-type="orgdiv2">Servicio de Cardiología</institution>
				<institution content-type="orgdiv1">Departamento de Medicina Interna</institution>
				<institution content-type="normalized">Hospital Alemán de Buenos Aires</institution>
				<country country="AR">Argentina</country>
			</aff>
			<aff id="aff2">
				<label>2</label>
				<institution content-type="original">Observatorio de Salud, Universidad Católica Argentina, Consejo de Investigación Científicas y Técnicas (CONICET).</institution>
				<institution content-type="orgdiv2">Observatorio de Salud</institution>
				<institution content-type="orgdiv1">Universidad Católica Argentina</institution>
				<institution content-type="normalized">Consejo de Investigación Científicas y Técnicas</institution>
				<country country="AR">Argentina</country>
			</aff>
			<aff id="aff1b">
				<label>1</label>
				<institution content-type="original">Sección Hipertensión Arterial, Servicio de Cardiología, Departamento de Medicina Interna, Hospital Alemán de Buenos Aires. </institution>
				<institution content-type="orgdiv2">Servicio de Cardiología</institution>
				<institution content-type="orgdiv1">Departamento de Medicina Interna</institution>
				<institution content-type="normalized">Hospital Alemán de Buenos Aires</institution>
				<country country="AR">Argentina</country>
				<email>fpdigennaro@hospitalaleman.com</email>
			</aff>
			<author-notes>
				<corresp id="c1">
					<label>Dirección para correspondencia</label><bold>:</bold> Federico Di Gennaro, Hospital Alemán, Av. Pueyrredón 1640 (1112), Buenos Aires, Argentina. <bold>E-mail:</bold><email>fpdigennaro@hospitalaleman.com</email>
				</corresp>
				<fn fn-type="other" id="fn0">
					<label>MTSAC</label>
					<p>Miembro Titular de la Sociedad Argentina de Cardiología</p>
				</fn>
				<fn fn-type="conflict" id="fn1">
					<label>Declaración de conflicto de intereses</label>
					<p> Los autores declaran que no tienen conflicto de intereses. (Véase formulario de conflictos de interés del autor en la Web).</p>
				</fn>
			</author-notes>
			<!--<pub-date date-type="pub" publication-format="electronic">
				<day>26</day>
				<month>02</month>
				<year>2025</year>
			</pub-date>
			<pub-date date-type="collection" publication-format="electronic">
				<season>Jan-Feb</season>
				<year>2025</year>
			</pub-date>-->
			<pub-date pub-type="epub-ppub">
				<season>Jan-Feb</season>
				<year>2025</year>
			</pub-date>
			<volume>93</volume>
			<issue>1</issue>
			<fpage>33</fpage>
			<lpage>42</lpage>
			<history>
				<date date-type="received">
					<day>10</day>
					<month>12</month>
					<year>2024</year>
				</date>
				<date date-type="accepted">
					<day>12</day>
					<month>01</month>
					<year>2025</year>
				</date>
			</history>
			<permissions>
				<license license-type="open-access" xlink:href="https://creativecommons.org/licenses/by-nc/4.0/" xml:lang="es">
					<license-p>Este es un artículo publicado en acceso abierto bajo una licencia Creative Commons</license-p>
				</license>
			</permissions>
			<abstract>
				<title>RESUMEN</title>
				<sec>
					<title>Introducción:</title>
					<p> No hay evidencia disponible sobre la comparación del valor predictivo de eventos graves en el seguimiento de pacientes hipertensos mediante el análisis con redes neuronales artificiales (RNA) de las mediciones del monitoreo ambulatorio de presión arterial (MAPA) en comparación con la estratificación de riesgo clínica (EC). </p>
				</sec>
				<sec>
					<title>Material y métodos:</title>
					<p> Se analizaron estudios de MAPA que incluyeron 27 mediciones cada uno: presión arterial media sistólica, diastólica, presión del pulso y frecuencia cardiaca de 24 hs, diurnas y nocturnas; carga hipertensiva; desvíos estándar de presiones y frecuencia cardíaca; ritmo circadiano. La variable dependiente fue el punto final combinado de muerte, accidente cerebrovascular, infarto agudo de miocardio, insuficiencia cardíaca e insuficiencia renal. Para la EC de cada paciente se utilizó como modelo el Consenso Argentino de Hipertensión Arterial. Se evaluaron la capacidad discriminativa del punto final con RNA-MAPA y con EC por análisis de regresión logística a través del análisis del área bajo la curva ROC (ABCR). Se compararon ambas áreas bajo la curva ROC mediante test de De Long. Para los análisis estadísticos y el modelaje de las RNA se usó el programa SPSS 23.0 Statistics. </p>
				</sec>
				<sec>
					<title>Resultados: </title>
					<p>Se analizó la información de 491 estudios de MAPA; edad media: 69 ± 14 años, 53 % mujeres, 11,6% diabéticos, 51% dislipidémicos, media de índice de masa corporal 26 ± 4 kg/m<sup>2</sup>, 14,3% fumadores. La mediana del seguimiento fue 6,6 años (rango intercuartílico 4,5-8). El modelo de RNA con mejor capacidad predictiva fue el Perceptrón Multicapa con una capa oculta; arquitectura neuronal (27/7/2). La presión arterial sistólica (PAS) nocturna presentó una importancia normalizada independiente del 100% para la determinación del modelo. El ABCR para la discriminación del punto final fue, con el análisis con RNA del MAPA, 0,81 (IC 95% 0,77-0,90); con la estratificación de riesgo clínico fue de 0,67 (IC 95% 0,56-0,77); test de De Long p &lt; 0,001. </p>
				</sec>
				<sec>
					<title>Conclusión:</title>
					<p> Observamos una mayor capacidad discriminativa de eventos mediante el análisis con RNA de las variables del MAPA vs la estratificación de riesgo clínico, lo cual constituye una hipótesis de investigación a validar prospectivamente. </p>
				</sec>
			</abstract>
			<trans-abstract xml:lang="en">
				<title>ABSTRACT</title>
				<sec>
					<title>Background:</title>
					<p> There is no available evidence comparing the predictive value of an artificial neural network (ANN)-based analysis method that integrates ambulatory blood pressure monitoring (ABPM) variables versus clinical risk stratification (CRS) for serious events in hypertensive patients at follow-up. </p>
				</sec>
				<sec>
					<title>Methods:</title>
					<p> We analyzed ABPM studies that included 27 measurements each one. The variables were daytime, nighttime and 24-hour mean, systolic and diastolic blood pressure, pulse pressure and heart rate; hypertensive load; standard deviations of pressures and heart rate; circadian rhythm. The dependent variable was the combined endpoint of death, stroke, acute myo cardial infarction, heart failure and kidney disease. For clinical risk stratification, the Argentine Consensus on Hypertension was used as a model. We evaluated the discriminative ability to predict the endpoint using ANN-ABPM and CRS by logistic regression through the analysis of the area under the receiver operating characteristic curve (AUC-ROC). Both AUC-ROC were compared by De Long test. SPSS 23.0 Statistics was used for statistical analyses and ANN modelling. </p>
				</sec>
				<sec>
					<title>Results: </title>
					<p>Data from 491 ABPM studies were analyzed. Mean age was 69 ± 14 years; 53% of population was female; 11.6% had diabetes; 51% had dyslipidemia; mean body mass index was 26 ± 4 kg/m<sup>2</sup>; 14.3% were smokers. Median follow-up was 6.6 years (interquartile range 4.5-8). The best predictive ANN model was the Multilayer Perceptron one with a hidden layer; neuronal architecture (27/7/2). Nocturnal systolic blood pressure (SBP) had 100% independent normalized importance for modelling. The AUC-ROC for the combined endpoint was 0.81 (95% CI 0.77-0.90) using neural network analysis with ABPM variables, and 0.67 (95% CI 0.56-0.77) using CRS; De Long's test p &lt; 0.001. </p>
				</sec>
				<sec>
					<title>Conclusion:</title>
					<p> We observed a higher discriminative ability to predict events at follow-up using ANN analysis with ABPM vari ables compared to conventional CRS. This observation raises a research hypothesis to be validated prospectively to optimize risk stratification and treatment in hypertensive patients. </p>
				</sec>
			</trans-abstract>
			<kwd-group xml:lang="en">
				<title>Key words:</title>
				<kwd>Risk assessments</kwd>
				<kwd>Artificial neural networks</kwd>
				<kwd>Hypertension</kwd>
			</kwd-group>
			<kwd-group xml:lang="es">
				<title>Palabras clave:</title>
				<kwd>Estratificación del riesgo</kwd>
				<kwd>Redes neuronales artificiales</kwd>
				<kwd>Hipertensión</kwd>
			</kwd-group>
			<counts>
				<fig-count count="3"/>
				<table-count count="3"/>
				<equation-count count="0"/>
				<ref-count count="41"/>
				<page-count count="10"/>
			</counts>
		</article-meta>
	</front>
	<body>
		<sec sec-type="intro">
			<title>INTRODUCCIÓN</title>
			<p>Las enfermedades cardiovasculares continúan siendo actualmente la principal causa de morbilidad y mortalidad a nivel mundial. La predicción de eventos cardiovasculares (ECV) es de vital importancia para la identificación temprana de individuos en riesgo y a través de ella la implementación de intervenciones preventivas más apropiadas. (<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B2">2</xref>)</p>
			<p>Por ese motivo se recomienda estimar el riesgo cardiovascular (RCV) global en todos los pacientes hipertensos para decidir conductas terapéuticas y de control de los factores de riesgo cardiovascular. La información obtenida a partir de la anamnesis, el examen físico, la medición de la presión arterial en el consultorio (PAC) y los resultados de estudios complementarios recomendados, determinan la presencia de factores de riesgo asociados, compromiso o daño de órgano blanco y antecedentes de eventos cardiovasculares. Con esta información es posible estratificar el RCV global del paciente hipertenso y determinar su riesgo como bajo, moderado, alto y muy alto. El conocimiento de la estratificación del RCV global del paciente individual representa una importante información pronóstica, facilita el enfoque global de la prevención y un tratamiento farmacológico adecuado. (<xref ref-type="bibr" rid="B3">3</xref>,<xref ref-type="bibr" rid="B4">4</xref>) </p>
			<p>Una serie de fórmulas o <italic>scores</italic> de riesgo son propuestos a fin de calcular el RCV. Los calculadores que de ellas surgen constituyen un grupo heterogéneo con diferentes limitaciones (variables cualitativas, estudios complementarios que no se utilizan en la práctica clínica cotidiana) y muchos de ellos sin validación para la población de nuestro país. (<xref ref-type="bibr" rid="B3">3</xref>,<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B5">5</xref>)</p>
			<p>El Consenso Argentino de Hipertensión Arterial propone un enfoque similar al utilizado por la Sociedad Europea de Hipertensión y adaptado para nuestro medio. (<xref ref-type="bibr" rid="B5">5</xref>) </p>
			<p>Si bien en la actualidad la medición de la presión arterial en el consultorio es el método diagnóstico recomendado, no se encuentra exenta de significativa variabilidad y sesgos por imprecisión en la técnica de medición. Por tal motivo, diferentes guías nacionales e internacionales recomiendan obtener mediciones fuera del consultorio, mediante el monitoreo ambulatorio de presión arterial (MAPA) para confirmar el diagnóstico de hipertensión arterial (HTA) y aportar información pronóstica más precisa. (<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B7">7</xref>)</p>
			<p>En los últimos años, el MAPA se ha convertido en un estudio complementario de gran utilidad para el diagnóstico y evaluación pronóstica de ECV en pacientes hipertensos en comparación con las mediciones aisladas realizadas en el consultorio. Además, el MAPA puede proporcionar datos adicionales, como la variabilidad de la presión arterial, los patrones de descenso nocturno y los valores de presión arterial promedio en diferentes períodos del día. (<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B9">9</xref>,<xref ref-type="bibr" rid="B10">10</xref>)</p>
			<p>Aunque los modelos de predicción del riesgo cardiovascular han mejorado en precisión a lo largo de los años, todavía existe cierta incertidumbre en las estimaciones. En la actualidad no se consideran las variables hemodinámicas que aporta el MAPA para la estratificación del riesgo cardiovascular en pacientes hipertensos. (<xref ref-type="bibr" rid="B11">11</xref>,<xref ref-type="bibr" rid="B12">12</xref>)</p>
			<p>En este sentido, es importante destacar la necesidad de herramientas más precisas en su capacidad predictiva que incorporen las diferentes variables de presión arterial estudiadas en el MAPA.</p>
			<p>Una de las herramientas metodológicas para el análisis predictivo en diferentes áreas de la medicina, más difundidas y en pleno desarrollo en la actualidad son las redes neuronales artificiales (RNA). El análisis con RNA como modelo de inteligencia artificial (IA) ha demostrado superioridad en la precisión pronóstica al compararlas con herramientas estadísticas (en particular cuando existen asociaciones no lineales), que utilizamos habitualmente, como el análisis multivariado y la regresión logística. (<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref>,<xref ref-type="bibr" rid="B15">15</xref>)</p>
			<p>Las RNA pueden detectar características relevantes en los datos y ajustar sus pesos sinápticos y conexiones para mejorar el rendimiento predictivo, que depende entre otras cosas de la cantidad de variables ingresadas y al entrenamiento que reciban, permitiéndoles hacer predicciones más precisas. (<xref ref-type="bibr" rid="B16">16</xref>,<xref ref-type="bibr" rid="B17">17</xref>,<xref ref-type="bibr" rid="B18">18</xref>)</p>
			<p>La aplicación de diferentes modelos de máquinas de aprendizaje, ha tenido como objetivo la detección precoz y el tamizaje para identificar a aquellos que desarrollan hipertensión arterial. (<xref ref-type="bibr" rid="B19">19</xref>,<xref ref-type="bibr" rid="B20">20</xref>,<xref ref-type="bibr" rid="B21">21</xref>,<xref ref-type="bibr" rid="B22">22</xref>)</p>
			<p>El análisis mediante RNA integrando las variables estudiadas en el MAPA podrían mejorar la capacidad predictiva y proporcionar información para diseñar una estratificación del RCV más precisa y completa en comparación con los modelos existentes.</p>
			<p>El objetivo de este estudio fue evaluar la capacidad discriminativa de eventos graves en el seguimiento de pacientes hipertensos con el análisis de RNA integrando las variables del MAPA, en comparación con la estratificación de riesgo clínico convencional.</p>
		</sec>
		<sec sec-type="materials|methods">
			<title>MATERIAL Y MÉTODOS</title>
			<p>Se analizó una base de datos con las mediciones registradas en estudios de MAPA con los siguientes criterios de inclusión: pacientes adultos (mayores de 18 años), con diagnóstico de hipertensión arterial esencial con tratamiento farmacológico estudiados con MAPA para evaluar la eficacia terapéutica. Con seguimiento completo a través de la historia clínica informatizada y consultas clínicas de un Hospital de comunidad.</p>
			<p>Se incluyeron estudios consecutivos de MAPA realizados entre septiembre de 2013 y abril de 2020 con un seguimiento clínico completo hasta noviembre del 2022. La transferencia de los datos de los informes de los estudios de MAPA (se tuvieron en cuenta los promedios de cada una de las variables analizadas) a una planilla de cálculo y su procesamiento se realizó mediante la aplicación de los programas informáticos Visual Basic y SQL.</p>
			<p>Para realizar la estratificación del riesgo cardiovascular se utilizó como modelo las variables propuestas en el Consenso Argentino de Hipertensión Arterial (Sociedad Argentina de Cardiología, Sociedad Argentina de Hipertensión Arterial, Federación Argentina de Cardiología). (<xref ref-type="bibr" rid="B5">5</xref>)</p>
			<p>Las variables consideradas son las siguientes: a) Factores de riesgo: edad, género, antecedentes de dislipemia, diabetes, tabaquismo, obesidad; b) Compromiso de órgano blanco: diagnóstico de hipertrofia ventricular izquierda (HVI) por ecocardiograma, insuficiencia renal crónica (estadios 1 y 2); c) Condiciones clínicas asociadas o antecedente de eventos cardiovasculares: infarto agudo de miocardio (IAM), insuficiencia cardiaca (IC), accidente cerebrovascular y/o ataque isquémico transitorio (ACV/AIT), enfermedad coronaria, revascularización miocárdica, insuficiencia renal crónica (estadios 3,4 y 5). </p>
			<p>Se definió como Bajo riesgo: pacientes con un factor de riesgo asociado; Moderado riesgo: pacientes con dos factores de riesgo asociados; Alto riesgo: pacientes con tres o más factores de riesgo asociados y/o diabetes y/o daño de órgano blanco; Muy alto riesgo: pacientes con antecedentes de eventos cardiovasculares o condiciones clínicas asociadas. (<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B5">5</xref>)</p>
			<p>Se definió un punto final combinado de eventos serios (ES) en el seguimiento compuesto de la ocurrencia de muerte y/o IAM no fatal y/o ACV y/o AIT y/o IC y/o insuficiencia renal crónica, constatados en la historia clínica informatizada por médicos especialistas en Medicina Interna y Cardiología según las guías nacionales e internacionales vigentes. (<xref ref-type="bibr" rid="B3">3</xref>,<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B5">5</xref>)</p>
			<p>Se desarrollaron modelos de algoritmos de una red neuronal que incluyó a las variables del MAPA como cofactores independientes para su ingreso a la RNA y el de ES como evento dependiente (capa de salida). Un algoritmo NN es un tipo especial de regresión no lineal que presenta múltiples valores mínimos locales. Por lo tanto, cada vez que se ejecute el algoritmo de entrenamiento, convergerá en un modelo diferente. Para elegir el mejor modelo, el proceso de entrenamiento se repitió 50 veces. Sólo se seleccionaron para la comparación los modelos con el mejor poder de discriminación por regresión logística o RNA.</p>
			<p>Para los análisis estadísticos y el modelaje de la RNA se usó el programa SPSS 26.0 Statistics. Se compararon diferentes modelos, arquitectura y funciones de activación para seleccionar la de mejor rendimiento en la discriminación del punto final. </p>
			<p>Las variables categóricas se expresaron como porcentajes con su IC 95%, mientras que las continuas de acuerdo con su distribución (paramétrica o no paramétrica) como medias y su respectiva desviación estándar o mediana y su rango intercuartílico (RIC) 25-75. </p>
			<p>Se evaluó la capacidad discriminativa de la variable dependiente con análisis del área bajo la curva ROC (ABCR) con el análisis del MAPA. Para la comparación de las Áreas Bajo la Curva ROC se utilizó el test de De Long con el programa MEDCALC versión 23.0.9</p>
			<p>A fin de identificar las variables con mayor peso y utilidad en el desarrollo de la red neuronal, se realizó un análisis de sensibilidad determinando las importancias normalizadas en el modelo de RNA.</p>
			<sec>
				<title>Consideraciones éticas</title>
				<p>El estudio fue revisado y aprobado por el Comité de Ética institucional e independiente a esta investigación. Ante la naturaleza observacional del presente análisis no se requirió de consentimiento informado. De acuerdo con la Declaración de Helsinki de la Asociación Médica Mundial,<xref ref-type="bibr" rid="B23">23</xref> se tomaron todas las precauciones para proteger la privacidad y la confidencialidad de toda la información utilizada. </p>
			</sec>
		</sec>
		<sec sec-type="results">
			<title>RESULTADOS</title>
			<p>Se analizó en total la información de 491 estudios de MAPA que incluyeron 27 variables numéricas provenientes de cada estudio que se describen a continuación: presión arterial sistólica media de 24 hs (PASm24), presión arterial diastólica media de 24 hs (PADm 24), presión arterial media 24 hs (PAM 24), presión de pulso media de 24 hs (PPm 24), frecuencia cardíaca media de 24 hs (FCm 24), presión arterial sistólica media diurna (PASm Día), presión arterial diastólica media diurna (PADm Día), presión arterial media diurna (PAMm Día), presión de pulso media diurna (PPm Día), frecuencia cardíaca media diurna (FCm Día), presión arterial sistólica media nocturna (PASm Noche), presión arterial diastólica media nocturna (PADm Noche), presión arterial media nocturna (PAMm Noche), presión de pulso media nocturna (PPm Noche), frecuencia cardíaca media nocturna (FCm Noche), variabilidad de la presión arterial sistólica en 24 hs (PASsd 24), variabilidad de la presión arterial diastólica en 24 hs (PADsd 24), variabilidad de la presión de pulso en 24 hs (PPsd 24), variabilidad de la presión arterial media en 24 hs (PAMsd 24), variabilidad de la frecuencia cardíaca en 24 hs (FCsd 24), carga hipertensiva sistólica diurna (Carga HtaSist Día), carga hipertensiva diastólica diurna (Carga HtaDiast Día), carga hipertensiva sistólica nocturna (Carga HtaSist Noche), carga hipertensiva diastólica nocturna (Carga HtaDiast Noche), presión arterial diurna ≥ 135/85 mmHg (HTA diurna), presión arterial nocturna ≥ 120/70 mmHg (HTA nocturna), ritmo circadiano con caída nocturna de PAS y/o PAD &lt; 10% (patrón <italic>non dipper</italic>).</p>
			<p>En la <xref ref-type="table" rid="t1">Tabla 1</xref> se detallan los promedios de cada una de las variables de los estudios de MAPA utilizados para el modelaje de las RNA para el punto final combinado. </p>
			<p>
				<table-wrap id="t1">
					<label>Tabla 1</label>
					<caption>
						<title>Descripción de los valores promedios de las variables analizadas en los estudios de MAPA.</title>
					</caption>
					<table frame="hsides" rules="groups">
						<colgroup>
							<col/>
							<col/>
						</colgroup>
						<thead>
							<tr style="border: 0; background-color:#ab0534;color:#ffffff;">
								<th align="center"> Variables del MAPA</th>
								<th align="center">Valores promedios</th>
							</tr>
						</thead>
						<tbody>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PASm 24</td>
								<td align="center">126,16±11,65 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PADm 24</td>
								<td align="center">79,22 ± 9,30 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PAMm 24</td>
								<td align="center">94,87 ± 9,4 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">FCm 24</td>
								<td align="center">75,21 ± 8,5 lpm</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PPm 24</td>
								<td align="center">46,93 ± 7,62 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PASsd 24</td>
								<td align="center">17,92 ± 5,6 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PADsd 24</td>
								<td align="center">14,52 ± 4,2 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PPsd 24</td>
								<td align="center">16,9 ± 6,1 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PAMsd 24</td>
								<td align="center">13,51 ± 3,8 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">FCsd 24</td>
								<td align="center">11,59 ± 3,1 lpm</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PASm Día</td>
								<td align="center">130,13 ± 12,2 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PADm Día</td>
								<td align="center">82,35 ± 9,9 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PAMm Día</td>
								<td align="center">98,17 ± 10,14 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">FCm Día</td>
								<td align="center">78,85 ± 9,1 lpm</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PPm Día</td>
								<td align="center">47,77 ± 8,08 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PASm Noche</td>
								<td align="center">117,34± 13,47 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PADm Noche</td>
								<td align="center">72,23 ± 9,90 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PPmNoche</td>
								<td align="center">48,10 ± 8,99 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">PAMm Noche</td>
								<td align="center">87,27 ± 10,50 mmHg</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">FCm Noche</td>
								<td align="center">67,47 ± 9,2 lpm</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">Carga Hta Sist Día</td>
								<td align="center">34,86%</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">Carga Hta Diast Día</td>
								<td align="center">42,25%</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">Carga Hta Sist Noche</td>
								<td align="center">39,47%</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">Carga Hta Diast Noche</td>
								<td align="center">49,68%</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">HTA Noche</td>
								<td align="center">58,4% (IC 95% 55-72)</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">HTA Día</td>
								<td align="center">43,9% (IC 95% 40-56)</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="justify">Non dipper</td>
								<td align="center">42,6% (IC 95% 39-55)</td>
							</tr>
						</tbody>
					</table>
					<table-wrap-foot>
						<fn id="TFN1">
							<p>Carga HtaDiast Día: carga hipertensiva diastólica diurna; Carga HtaDiast Noche: carga hipertensiva diastólica nocturna; Carga HtaSist Día: carga hipertensiva sistólica diurna; Carga HtaSist Noche: carga hipertensiva sistólica nocturna; FCm 24: frecuencia cardíaca media de 24 hs; FCm Día: frecuencia cardíaca media diurna; FCm Noche: frecuencia cardíaca media nocturna; FCsd 24: variabilidad de la frecuencia cardíaca en 24 hs; HTA diurna: presión arterial diurna ≥ 135/85 mmHg; HTA nocturna: presión arterial nocturna ≥120/70 mmHg; MAPA: monitoreo ambulatorio de la presión arterial; Non dipper: ritmo circadiano con caída nocturna de PAS y/o PAD &lt; 10%; PADm 24: presión arterial Diastólica media de 24 hs; PADm Día: Presión Arterial Diastólica media diurna; PADm Noche: Presión Arterial diastólica media nocturna; PADsd 24: variabilidad de la presión arterial diastólica en 24 hs; PAM 24: presión arterial media 24 hs; PAMm Día: presión arterial media diurna; PAMm Noche: presión arterial media nocturna; PAMsd 24: variabilidad de la presión arterial media en 24 hs; PASm 24: presión arterial sistólica media de 24 hs; PASm Día: presión arterial sistólica media diurna; PASm Noche: presión arterial sistólica media nocturna; PASsd 24: variabilidad de la presión arterial sistólica en 24 hs; PPm 24: presión de pulso media de 24 hs; PPm Día: presión de pulso media diurna; PPm Noche: presión de pulso media nocturna; PPsd 24: variabilidad de la presión de pulso en 24 hs.</p>
						</fn>
					</table-wrap-foot>
				</table-wrap>
			</p>
			<p>La edad media de la población fue de 64 ± 14 años, 47% eran mujeres, 12 % tenían diabetes, 11% eran tabaquistas activos, 52% dislipidémicos y el índice de masa corporal promedio fue de 26 ± 4 kg/m<sup>2</sup>.</p>
			<p>La mediana del seguimiento de los pacientes fue de 6,6 años (RIC 4.5-8). La incidencia del punto final en el seguimiento fue 2,6%. En la <xref ref-type="table" rid="t2">Tabla 2</xref> se detallan los mejores modelos de RNA con sus funciones de activación neuronales de la capa oculta y de salida y su Área Bajo la Curva ROC. </p>
			<p>
				<table-wrap id="t2">
					<label>Tabla 2</label>
					<caption>
						<title>Modelos de redes neuronales de mejor rendimiento (perceptrón multicapa con una capa oculta, dos capas ocultas y con modelo de base radial) con sus funciones de activación neuronal de la capa oculta y la capa de salida, la arquitectura neuronal y su área bajo la curva ROC.</title>
					</caption>
					<table frame="hsides" rules="groups">
						<colgroup>
							<col/>
							<col/>
							<col/>
							<col/>
							<col/>
						</colgroup>
						<thead>
							<tr style="border: 0; background-color:#ab0534;color:#ffffff;">
								<th align="center">Modelo</th>
								<th align="center">Función de activación capa oculta</th>
								<th align="center">Función de activación capa de salida</th>
								<th align="center">Arquitectura neuronal</th>
								<th align="center">Área bajo la curva ROC</th>
							</tr>
						</thead>
						<tbody>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">Perceptrón multicapa con una capa oculta</td>
								<td align="center">Hiperbólica tangente</td>
								<td align="center">Softmax</td>
								<td align="center">27/7/2</td>
								<td align="center">0,81 (IC 95% 0,77-0,90)</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">Perceptrón multicapa con dos capas ocultas</td>
								<td align="center">Hiperbólica tangente</td>
								<td align="center">Softmax</td>
								<td align="center">27/20/15/2</td>
								<td align="center">0,75 (IC 95% 0,68-0,80)</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">Base radial</td>
								<td align="center">Softmax</td>
								<td align="center">Identity</td>
								<td align="center">27/6/2</td>
								<td align="center">0,68 (IC 95% 0,61-0,70)</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">Test de De Long</td>
								<td align="center">Perceptrón multicapa 1 capa oculta vs 2 capas ocultas</td>
								<td align="center">Perceptrón multicapa 1 capa oculta vs Base radial</td>
								<td align="center" colspan="2">Base radial vs Perceptrón multicapa 2 capas ocultas </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">Valor de p</td>
								<td align="center">0,040</td>
								<td align="center">0,002</td>
								<td align="center" colspan="2">0,001</td>
							</tr>
						</tbody>
					</table>
				</table-wrap>
			</p>
			<p>Los modelos de mejor rendimiento fueron el de Perceptrón Multicapa de una capa oculta (función de activación de las neuronas de la capa oculta de tipo hiperbólica tangente) y los de la capa de salida de tipo <italic>softmax</italic> con una arquitectura neuronal (27/7/2) que describen los nodos de cada una de las capas (<xref ref-type="fig" rid="f1">Figura 1</xref>). En la <xref ref-type="table" rid="t3">Tabla 3</xref> se describen los pesos sinápticos con los que se construye y entrena la red neuronal para la predicción del punto final. Se utilizó como función de activación de la capa de entrada la de tangente hiperbólica y la <italic>softmax</italic> para la capa oculta. Se dividió la muestra con una segmentación de 70% del grupo entrenamiento y 30% de validación de los algoritmos. Se tuvieron en cuenta los pesos o ponderaciones sinápticas estimadas para el desarrollo y testeo de un modelo de perceptrón multicapa para el punto final combinado basados en la entrada de las 27 variables de los estudios de MAPA analizados. </p>
			<p>
				<fig id="f1">
					<label>Fig. 1</label>
					<caption>
						<title>Arquitectura de la red neuronal del tipo perceptrón multicapa con una capa oculta con 27 neuronas en la capa de entrada, 7 neuronas en la capa oculta y 2 neuronas en la capa de salida. </title>
					</caption>
					<graphic xlink:href="1850-3748-rac-93-01-33-gf1.jpg"/>
				</fig>
			</p>
			<p>
				<table-wrap id="t3">
					<label>Tabla 3</label>
					<caption>
						<title>Descripción de los pesos sinápticos con los que se construye y entrena la red neuronal para la predicción del punto final.</title>
					</caption>
					<table frame="hsides" rules="groups">
						<colgroup>
							<col span="2"/>
							<col span="10"/>
						</colgroup>
						<thead>
							<tr style="border: 0; background-color:#ab0534;color:#ffffff;">
								<th align="left" colspan="2">
 </th>
								<th align="center" colspan="10">Predichos </th>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<th align="left" colspan="2">
 </th>
								<th align="center" colspan="7">Capa oculta 1 </th>
								<th align="center" colspan="3">Capa de salida </th>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<th align="center" colspan="2">Predictor+A1:L42 </th>
								<th align="center">O (1:1)</th>
								<th align="center">O (1:2)</th>
								<th align="center">O (1:3)</th>
								<th align="center">O (1:4)</th>
								<th align="center">O (1:5)</th>
								<th align="center">O (1:6)</th>
								<th align="center">O (1:7)</th>
								<th align="center">PF</th>
								<th align="center">[PF=0]</th>
								<th align="center">[PF=1]</th>
							</tr>
						</thead>
						<tbody>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left" colspan="2">Capa de (Sesgo) </td>
								<td align="center">,526</td>
								<td align="center">,579</td>
								<td align="center">-,457</td>
								<td align="center">-,125</td>
								<td align="center">-,394</td>
								<td align="center">-,600</td>
								<td align="center">,459</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left" rowspan="30">entrada </td>
								<td align="left">nondipper 0 </td>
								<td align="center">,560</td>
								<td align="center">,134</td>
								<td align="center">-,105</td>
								<td align="center">-,501</td>
								<td align="center">-,245</td>
								<td align="center">,258</td>
								<td align="center">,314</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">nondipper 1 </td>
								<td align="center">,293</td>
								<td align="center">-,190</td>
								<td align="center">,180</td>
								<td align="center">-,293</td>
								<td align="center">-,418</td>
								<td align="center">-,298</td>
								<td align="center">-,112</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">HTAnoche 0 </td>
								<td align="center">-,018</td>
								<td align="center">-,055</td>
								<td align="center">,410</td>
								<td align="center">-,073</td>
								<td align="center">,170</td>
								<td align="center">-,531</td>
								<td align="center">,356</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">HTAnoche 1 </td>
								<td align="center">,046</td>
								<td align="center">,100</td>
								<td align="center">-,270</td>
								<td align="center">,251</td>
								<td align="center">-,128</td>
								<td align="center">-,358</td>
								<td align="center">,424</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">HTAdia01 0 </td>
								<td align="center">,209</td>
								<td align="center">-,108</td>
								<td align="center">-,165</td>
								<td align="center">-,299</td>
								<td align="center">,332</td>
								<td align="center">-,427</td>
								<td align="center">,505</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">HTAdia01 1 </td>
								<td align="center">,459</td>
								<td align="center">-,275</td>
								<td align="center">,110</td>
								<td align="center">,368</td>
								<td align="center">,169</td>
								<td align="center">-,669</td>
								<td align="center">,200</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PASm24 </td>
								<td align="center">,119</td>
								<td align="center">-,452</td>
								<td align="center">-,159</td>
								<td align="center">,315</td>
								<td align="center">,005</td>
								<td align="center">-,141</td>
								<td align="center">,388</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PADm24 </td>
								<td align="center">,334</td>
								<td align="center">-,166</td>
								<td align="center">,080</td>
								<td align="center">-,167</td>
								<td align="center">,241</td>
								<td align="center">,417</td>
								<td align="center">-,365</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">FCm24 </td>
								<td align="center">,228</td>
								<td align="center">-,080</td>
								<td align="center">,435</td>
								<td align="center">-,323</td>
								<td align="center">,136</td>
								<td align="center">-,457</td>
								<td align="center">,551</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PAMm24 </td>
								<td align="center">,486</td>
								<td align="center">,475</td>
								<td align="center">,216</td>
								<td align="center">,019</td>
								<td align="center">,036</td>
								<td align="center">-,313</td>
								<td align="center">,321</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PPm24 </td>
								<td align="center">-,361</td>
								<td align="center">-,470</td>
								<td align="center">-,308</td>
								<td align="center">-,269</td>
								<td align="center">-,080</td>
								<td align="center">,020</td>
								<td align="center">-,005</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PASsd24 </td>
								<td align="center">-,179</td>
								<td align="center">-,346</td>
								<td align="center">-,027</td>
								<td align="center">,319</td>
								<td align="center">-,312</td>
								<td align="center">,002</td>
								<td align="center">,427</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PADsd24 </td>
								<td align="center">-,275</td>
								<td align="center">,037</td>
								<td align="center">-,240</td>
								<td align="center">,050</td>
								<td align="center">-,378</td>
								<td align="center">,697</td>
								<td align="center">-,199</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">FCsd24 </td>
								<td align="center">-,184</td>
								<td align="center">,178</td>
								<td align="center">-,381</td>
								<td align="center">,133</td>
								<td align="center">,228</td>
								<td align="center">-,230</td>
								<td align="center">-,109</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PAMsd24 </td>
								<td align="center">-,208</td>
								<td align="center">,170</td>
								<td align="center">-,290</td>
								<td align="center">,268</td>
								<td align="center">-,269</td>
								<td align="center">,153</td>
								<td align="center">-,023</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PPsd24 </td>
								<td align="center">,088</td>
								<td align="center">,233</td>
								<td align="center">-,445</td>
								<td align="center">,141</td>
								<td align="center">-,167</td>
								<td align="center">,184</td>
								<td align="center">,347</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PASmDIA </td>
								<td align="center">-,089</td>
								<td align="center">-,398</td>
								<td align="center">,480</td>
								<td align="center">,291</td>
								<td align="center">,325</td>
								<td align="center">,331</td>
								<td align="center">-,216</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PADmDIA </td>
								<td align="center">-,330</td>
								<td align="center">,307</td>
								<td align="center">-,107</td>
								<td align="center">,391</td>
								<td align="center">,243</td>
								<td align="center">-,350</td>
								<td align="center">,110</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">FCmDIA </td>
								<td align="center">-,427</td>
								<td align="center">,294</td>
								<td align="center">-,401</td>
								<td align="center">-,089</td>
								<td align="center">-,330</td>
								<td align="center">-,263</td>
								<td align="center">,154</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PAMmDIA </td>
								<td align="center">-,470</td>
								<td align="center">,010</td>
								<td align="center">,345</td>
								<td align="center">,318</td>
								<td align="center">-,147</td>
								<td align="center">-,041</td>
								<td align="center">-,604</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PPmDIA </td>
								<td align="center">,053</td>
								<td align="center">,222</td>
								<td align="center">,373</td>
								<td align="center">,179</td>
								<td align="center">,133</td>
								<td align="center">-,169</td>
								<td align="center">-,001</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">htaloadsbpDIA </td>
								<td align="center">,476</td>
								<td align="center">,488</td>
								<td align="center">-,056</td>
								<td align="center">-,249</td>
								<td align="center">,460</td>
								<td align="center">-,051</td>
								<td align="center">,265</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">htaloaddbpDIA </td>
								<td align="center">,265</td>
								<td align="center">,474</td>
								<td align="center">,331</td>
								<td align="center">-,092</td>
								<td align="center">,126</td>
								<td align="center">,365</td>
								<td align="center">,372</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PASmNOCHE </td>
								<td align="center">-,266</td>
								<td align="center">-,077</td>
								<td align="center">-,196</td>
								<td align="center">-,281</td>
								<td align="center">-,514</td>
								<td align="center">,146</td>
								<td align="center">-,317</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PADmNOCHE </td>
								<td align="center">,134</td>
								<td align="center">,278</td>
								<td align="center">-,099</td>
								<td align="center">,426</td>
								<td align="center">-,488</td>
								<td align="center">,554</td>
								<td align="center">,269</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">FCmNOCHE </td>
								<td align="center">-,429</td>
								<td align="center">-,124</td>
								<td align="center">-,280</td>
								<td align="center">-,184</td>
								<td align="center">,066</td>
								<td align="center">,307</td>
								<td align="center">-,257</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PAMmNOCHE </td>
								<td align="center">-,260</td>
								<td align="center">,253</td>
								<td align="center">,435</td>
								<td align="center">,511</td>
								<td align="center">-,116</td>
								<td align="center">-,308</td>
								<td align="center">-,046</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">PPmNOCHE </td>
								<td align="center">,313</td>
								<td align="center">-,343</td>
								<td align="center">-,294</td>
								<td align="center">,278</td>
								<td align="center">-,272</td>
								<td align="center">-,245</td>
								<td align="center">-,196</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">htaloadsbpNOCHE </td>
								<td align="center">-,376</td>
								<td align="center">-,294</td>
								<td align="center">,233</td>
								<td align="center">-,010</td>
								<td align="center">-,285</td>
								<td align="center">-,195</td>
								<td align="center">-,305</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">htaloaddbpNOCHE </td>
								<td align="center">,189</td>
								<td align="center">-,134</td>
								<td align="center">-,062</td>
								<td align="center">-,336</td>
								<td align="center">,122</td>
								<td align="center">-,123</td>
								<td align="center">,276</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left" rowspan="8">Capa oculta 1 </td>
								<td align="left">(Bias) </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="center">-1,302</td>
								<td align="center">1,206</td>
								<td align="center">-,093</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">O (1:1)</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="center">-,348</td>
								<td align="center">,844</td>
								<td align="center">,712</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">O (1:2)</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="center">-,230</td>
								<td align="center">,008</td>
								<td align="center">,005</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">O (1:3)</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="center">,121</td>
								<td align="center">-,058</td>
								<td align="center">-,129</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">O (1:4)</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="center">,158</td>
								<td align="center">-,023</td>
								<td align="center">-,129</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">O (1:5)</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="center">-,031</td>
								<td align="center">,045</td>
								<td align="center">-,504</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">O (1:6)</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="center">,589</td>
								<td align="center">-,891</td>
								<td align="center">-,178</td>
							</tr>
							<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
								<td align="left">O (1:7)</td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="left"> </td>
								<td align="center">-,553</td>
								<td align="center">,691</td>
								<td align="center">-,168</td>
							</tr>
						</tbody>
					</table>
					<table-wrap-foot>
						<fn id="TFN2">
							<p>FCm 24: frecuencia cardíaca media de 24 hs; FCm Día: frecuencia cardíaca media diurna; FCm Noche: frecuencia cardíaca media nocturna; FCsd 24: variabilidad de la frecuencia cardíaca en 24 hs; HTA día: presión arterial diurna ≥ 135/85 mmHg; HTA noche: presión arterial nocturna &gt;=120/70 mmHg; htaloaddbp DIA: carga hipertensiva diastólica diurna; htaloaddbp NOCHE: carga hipertensiva diastólica nocturna; htaloadsbp DIA: carga hipertensiva sistólica diurna; htaloadsbp NOCHE: carga hipertensiva sistólica nocturna; Non dipper: ritmo circadiano con caída nocturna de PAS y/o PAD &lt; 10%; PADm 24: presión arterial diastólica media de 24 hs; PADm Día: presión arterial diastólica media diurna; PADm Noche: presión arterial diastólica media nocturna; PADsd 24: variabilidad de la presión arterial diastólica en 24 hs; PAM 24: presión arterial media 24 hs; PAMm Día: presión arterial media diurna; PAMm Noche: presión arterial media nocturna; PAMsd 24: variabilidad de la presión arterial media en 24 hs; PASm 24: presión arterial sistólica media de 24 hs; PASm Día: presión arterial sistólica media diurna; PASm Noche: presión arterial sistólica media nocturna; PASsd 24: variabilidad de la presión arterial sistólica en 24 hs; PPm 24: presión de pulso media de 24 hs; PPm Día: presión de pulso media diurna; PPm Noche: presión de pulso media nocturna; PPsd 24: variabilidad de la presión de pulso en 24 hs. </p>
						</fn>
					</table-wrap-foot>
				</table-wrap>
			</p>
			<p>El área bajo la curva ROC del análisis de las variables del MAPA mediante redes neuronales fue 0,81 (IC 95% 0,77-0,90) versus con la estratificación de riesgo clínico 0,67 (IC 95% 0,56-0,77) para el punto final combinado; test de De Long p &lt; 0,001 (<xref ref-type="fig" rid="f2">Figura 2</xref>).</p>
			<p>
				<fig id="f2">
					<label>Fig. 2</label>
					<caption>
						<title>Comparación de las Áreas Bajo la Curva ROC. Variables del MAPA analizados con redes neuronales artificiales: 0,81 (IC 95% 0,77-0,90) versus estratificación de riesgo clínico 0,67 (IC 95% 0,56-0,77). Test de De Long p &lt; 0,001.</title>
					</caption>
					<graphic xlink:href="1850-3748-rac-93-01-33-gf2.jpg"/>
				</fig>
			</p>
			<p>Las variables con mayor peso y utilidad en el desarrollo de la red neuronal o importancias normalizadas en este modelo fueron las siguientes: presión arterial sistólica nocturna (PASm Noche) con un valor del 100%, seguida por la presión arterial sistólica de 24 hs (PASm 24), el IMC, la presión arterial media de 24 hs (PAMm 24), la presión arterial media nocturna (PAMm Noche) y la presión de pulso media de 24 hs (PPm 24) (<xref ref-type="fig" rid="f3">Figura 3</xref>)</p>
			<p>
				<fig id="f3">
					<label>Fig. 3</label>
					<caption>
						<title>Gráfico de barras representando las variables con mayor importancia normalizada para el desarrollo de la red neuronal artificial en este modelo. </title>
					</caption>
					<graphic xlink:href="1850-3748-rac-93-01-33-gf3.jpg"/>
				</fig>
			</p>
		</sec>
		<sec sec-type="discussion">
			<title>DISCUSIÓN</title>
			<p>A nuestro entender, el presente análisis es el primero en reportar la capacidad predictiva de eventos a largo plazo en paciencias hipertensos a través del análisis de las variables de los estudios de MAPA mediante redes neuronales artificiales y su superioridad en comparación con la estratificación de riesgo clínica.</p>
			<p>La importancia de estratificar los pacientes hipertensos de acuerdo con el riesgo estimado de presentar un evento cardiovascular permite ajustar el tratamiento en función de dicho riesgo y no solo de las cifras de la PAC. Entre las escalas de riesgo más difundidas están la ecuación de Framingham, el sistema SCORE (<italic>Systematic Coronary Risk Evaluation</italic>), el QRISK, calculadoras de diferentes sociedades médicas de los Estados Unidos, de la Organización Mundial de la Salud; esta última adaptada para diferentes zonas geográficas. En nuestro país podemos aplicar la estratificación recomendada por el Consenso Argentino de Hipertensión Arterial. (<xref ref-type="bibr" rid="B3">3</xref>,<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B5">5</xref>)</p>
			<p>Sin embargo, los scores tienen varias limitaciones que pueden afectar su precisión y aplicabilidad, entre ellas: las diferencias entre las poblaciones según las regiones geográficas, las características ambientales y socioeconómicas, la subestimación del riesgo en pacientes jóvenes; en muchos casos la diferencia cualitativa de factores de riesgo; la falta de identificación de pacientes con diagnóstico de hipertensión enmascarada. La mayoría de los scores no consideran el tratamiento antihipertensivo en la reducción del riesgo o la predicción de riesgo a corto plazo. (<xref ref-type="bibr" rid="B15">15</xref>,<xref ref-type="bibr" rid="B16">16</xref>)</p>
			<p>A diferencia de la metodología convencional por análisis multivariado, el algoritmo entrenado de la RNA se presenta como herramienta superadora respecto a la capacidad pronóstica de eventos, ya que ingresa todas las variables disponibles en los estudios de MAPA y evita de esta manera un sesgo en la selección de las variables. </p>
			<p>Las redes neuronales de tipo Perceptrón Multicapa se encuentran entre las arquitecturas de red más poderosas y populares. Están formadas por una capa de entrada, un número arbitrario de capas ocultas, y una capa de salida. Cada una de las neuronas ocultas o de salida recibe una entrada de las neuronas de la capa previa (conexiones hacia atrás), pero no existen conexiones laterales entre las neuronas dentro de cada capa. La capa de entrada contiene tantas neuronas como categorías correspondan a las variables independientes (categóricas y continuas). La capa de salida corresponde a la variable respuesta, que en este caso es una variable categórica (punto final combinado).</p>
			<p>Las redes neuronales de Base Radial son aquellas cuyas funciones de activación en los nodos ocultos son radialmente simétricas. Se dice que una función es radialmente simétrica si su salida depende de la distancia entre un vector que almacena los datos de entrada y un vector de pesos sinápticos, que recibe el nombre de centro o centroide. (<xref ref-type="bibr" rid="B17">17</xref>,<xref ref-type="bibr" rid="B18">18</xref>,<xref ref-type="bibr" rid="B19">19</xref>,<xref ref-type="bibr" rid="B20">20</xref>)</p>
			<p>Estudios previos aplicando las máquinas de aprendizaje analizaron su utilidad como método de tamizaje y detección precoz de hipertensión en diferentes poblaciones en el mundo, mientras que otros grupos de investigación las aplicaron para la predicción de eventos en pacientes hipertensos. (<xref ref-type="bibr" rid="B21">21</xref>,<xref ref-type="bibr" rid="B22">22</xref>,<xref ref-type="bibr" rid="B24">24</xref>,<xref ref-type="bibr" rid="B25">25</xref>,<xref ref-type="bibr" rid="B26">26</xref>)</p>
			<p>Otros autores han reportado la utilidad de las máquinas de aprendizaje para optimizar la toma de decisiones con el objetivo de mejorar el tratamiento en pacientes hipertensos con base en datos clínicos, logrando una alta precisión al predecir la probabilidad individual de alcanzar los objetivos de la presión arterial en el consultorio con diferentes tratamientos. Las áreas bajo la curva ROC estuvieron muy cerca de 0,90, indicando una elevada precisión de predicción y los coeficientes Kappa resultaron cercanos a 0,8, mostrando elevados niveles de concordancia entre los resultados de objetivos observados y previstos. (<xref ref-type="bibr" rid="B27">27</xref>,<xref ref-type="bibr" rid="B28">28</xref>)</p>
			<p>El área bajo la curva ROC evidencia que las RNA detectan relaciones no lineales entre variables independientes y dependientes más allá del alcance de la regresión logística. Estos resultados respaldan la utilidad de las RNA como una metodología para el análisis en la predicción de eventos graves en pacientes hipertensos. (<xref ref-type="bibr" rid="B24">24</xref>,<xref ref-type="bibr" rid="B27">27</xref>)</p>
			<p>Los modelos de predicción basados en RNA demuestran ser robustos y confiables, podrían implementarse en la práctica clínica como herramientas de apoyo a la toma de decisiones. Esta metodología identificaría de manera temprana a los pacientes hipertensos en mayor riesgo de desarrollar eventos graves y permitiría intervenciones preventivas más efectivas. (<xref ref-type="bibr" rid="B22">22</xref>, <xref ref-type="bibr" rid="B24">24</xref>,<xref ref-type="bibr" rid="B25">25</xref>,<xref ref-type="bibr" rid="B26">26</xref>)</p>
			<p>Destacamos que la presión arterial sistólica media nocturna (PASm Noche) presentó una importancia normalizada independiente para la determinación del modelo. La presión arterial nocturna ha demostrado ser una variable de riesgo independiente en la predicción de eventos cardiovasculares en pacientes hipertensos, asociada con una mayor incidencia de accidente cerebrovascular, infarto de miocardio, insuficiencia cardíaca e insuficiencia renal. <xref ref-type="bibr" rid="B29">29</xref>
			</p>
			<p>Durante las horas de la noche, la presión arterial sigue un patrón circadiano característico, con una disminución fisiológica. En pacientes hipertensos, este descenso puede ser insuficiente o incluso invertido, lo que se conoce como <italic>non-dipper</italic> o <italic>reverse-dipper</italic>. Estos patrones anormales de presión arterial nocturna se han asociado con un mayor riesgo y eventos cardiovasculares. (<xref ref-type="bibr" rid="B29">29</xref>,<xref ref-type="bibr" rid="B30">30</xref>,<xref ref-type="bibr" rid="B31">31</xref>)</p>
			<p>Diferentes publicaciones describieron que el aumento de la PASm Noche se asocia con un mayor riesgo de eventos cardiovasculares en paciente hipertensos, incluso después de ajustar por otros factores de riesgo conocidos. (<xref ref-type="bibr" rid="B30">30</xref>,<xref ref-type="bibr" rid="B31">31</xref>,<xref ref-type="bibr" rid="B32">32</xref>,<xref ref-type="bibr" rid="B33">33</xref>,<xref ref-type="bibr" rid="B34">34</xref>,<xref ref-type="bibr" rid="B35">35</xref>,<xref ref-type="bibr" rid="B36">36</xref>,<xref ref-type="bibr" rid="B37">37</xref>,<xref ref-type="bibr" rid="B38">38</xref>) </p>
			<p>Teniendo en cuenta observaciones previas y los resultados obtenidos en el presente estudio mediante el análisis con RNA, podría considerarse a la PASm Noche como un marcador útil en la estratificación del riesgo cardiovascular en pacientes hipertensos. Su evaluación ayudaría a identificar a pacientes con mayor riesgo de complicaciones y permitiría intervenciones clínicas para reducirlos. </p>
			<p>La implementación efectiva de los modelos actuales de redes neuronales puede realizarse mediante sistemas de software o hardware. De este modo, los pesos sinápticos de las arquitecturas propuestas, pueden ser entrenados y calculados a través de un algoritmo en Python. A su vez, este algoritmo debe ser capaz de leer automáticamente conjuntos de datos de registros electrónicos de estudios.</p>
			<p>Este estudio muestra que el análisis de variables obtenidas en el MAPA mediante el uso de RNA presenta un valor pronóstico significativo para predecir eventos graves en pacientes hipertensos; sugiere la importancia de considerar las variables del MAPA para la estratificación de riesgo en pacientes hipertensos y muestra el uso de las RNA como una herramienta efectiva para el análisis predictivo de eventos cardiovasculares con una adecuada precisión.</p>
			<sec>
				<title>Limitaciones</title>
				<p>Si bien se utilizaron diversas variables clínicas y las obtenidas en el MAPA, la inclusión de otras mediciones como por ejemplo la microalbuminuria o la enfermedad vascular periférica, podrían proporcionar información adicional para mejorar la capacidad de predicción. La estratificación de riesgo se realizó con todos los datos disponibles en la historia clínica informatizada. </p>
				<p>Consideramos que el análisis retrospectivo constituye una limitante por los sesgos concomitantes que inciden en la calidad de la evidencia. Por su carácter de estudio unicéntrico, limitando la transferibilidad de nuestros resultados al mundo real.</p>
				<p>Un mayor tamaño de la muestra con un análisis prospectivo, validaciones externas en diferentes cohortes de pacientes hipertensos y la participación de diversos centros asistenciales optimizarían la robustez de esta hipótesis de estudio. </p>
				<p>Los algoritmos de las redes neuronales han sido criticados en múltiples oportunidades por ser considerados como una “caja negra” con capacidad limitada para identificar las posibles relaciones causales. En el presente estudio identificamos los factores más influyentes en la modelación del algoritmo, a través de los valores de importancia estandarizada.</p>
				<p>La calidad del funcionamiento de las máquinas de aprendizaje y sus algoritmos se asocia a una proporción de eventos mucho mayores que la del presente estudio al igual que con volúmenes de casos de la población total superiores a lo que hemos reportado. Cabe aclarar que la incidencia de eventos registrada en nuestro trabajo es consistente con lo reportado en la literatura en la población de hipertensos ambulatorios. <xref ref-type="bibr" rid="B39"><sup>39</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B40"><sup>40</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B41"><sup>41</sup></xref>
				</p>
			</sec>
		</sec>
		<sec sec-type="conclusions">
			<title>CONCLUSIONES</title>
			<p>La originalidad de este estudio se basa en ser el primero publicar la capacidad predictiva de las RNA integrando las variables analizadas en los MAPA para predecir eventos a largo plazo en comparación con la estratificación de riesgo cardiovascular recomendada en la actualidad. </p>
			<p>Un dato de interés adicional es que observamos que la PAS media nocturna fue variable con mayor ponderación en el funcionamiento de la RNA.</p>
			<p>Consideramos este análisis como generador de una hipótesis de investigación a evaluarse con futuros estudios multicéntricos, con una potencia adecuada y con representatividad del mundo real para la transferencia de sus resultados.</p>
			<p>Estos algoritmos pueden integrarse al resultado de cada estudio de MAPA, permitiendo calcular automáticamente una probabilidad de riesgo de eventos mayores en el seguimiento de pacientes hipertensos, y así poder colaborar en la toma de decisiones del médico tratante.</p>
		</sec>
	</body>
	<back>
		<ack>
			<title>Agradecimientos</title>
			<p>Los autores agradecen al Ingeniero Roberto Bunge (Director de Ingeniería en Inteligencia Artificial de la Universidad San Andrés) y a la Ingeniera Trinidad Monreal (Investigación y Desarrollo del Laboratorio de Inteligencia Artificial y Robótica de la Universidad San Andrés) por sus aportes y conocimientos en inteligencia artificial y redes neuronales.</p>
		</ack>
		<ref-list>
			<title>BIBLIOGRAFÍA</title>
			<ref id="B1">
				<label>1</label>
				<mixed-citation>GBD 2017 Causes of Death Collaborators. Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2018;392:1736-88<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/S0140-6736(18)32203-7">https://doi.org/10.1016/S0140-6736(18)32203-7</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<collab>GBD 2017 Causes of Death Collaborators</collab>
					</person-group>
					<article-title>Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980-2017: a systematic analysis for the Global Burden of Disease Study 2017</article-title>
					<source>Lancet</source>
					<year>2018</year>
					<volume>392</volume>
					<fpage>1736</fpage>
					<lpage>1788</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/S0140-6736(18)32203-7">https://doi.org/10.1016/S0140-6736(18)32203-7</ext-link>
				</element-citation>
			</ref>
			<ref id="B2">
				<label>2</label>
				<mixed-citation>Mills KT, Bundy JD, Kelly TN, Reed JE, Kearney PM, Reynolds K. Global Disparities of Hypertension Prevalence and Control: A Systematic Analysis of Population-Based Studies From 90 Countries. Circulation 2016;134:441-50. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/CIRCULATIONAHA.115.018912">https://doi.org/10.1161/CIRCULATIONAHA.115.018912</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Mills</surname>
							<given-names>KT</given-names>
						</name>
						<name>
							<surname>Bundy</surname>
							<given-names>JD</given-names>
						</name>
						<name>
							<surname>Kelly</surname>
							<given-names>TN</given-names>
						</name>
						<name>
							<surname>Reed</surname>
							<given-names>JE</given-names>
						</name>
						<name>
							<surname>Kearney</surname>
							<given-names>PM</given-names>
						</name>
						<name>
							<surname>Reynolds</surname>
							<given-names>K</given-names>
						</name>
					</person-group>
					<article-title>Global Disparities of Hypertension Prevalence and Control: A Systematic Analysis of Population-Based Studies From 90 Countries</article-title>
					<source>Circulation</source>
					<year>2016</year>
					<volume>134</volume>
					<fpage>441</fpage>
					<lpage>450</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/CIRCULATIONAHA.115.018912">https://doi.org/10.1161/CIRCULATIONAHA.115.018912</ext-link>
				</element-citation>
			</ref>
			<ref id="B3">
				<label>3</label>
				<mixed-citation>Williams B, Mancia G, Spiering W, Agabiti Rosei E, Azizi M, Burnier MAuthors/Task Force Members:. 2018 ESC/ESH Guidelines for the management of arterial hypertension: The Task Force for the management of arterial hypertension of the European Society of Cardiology and the European Society of Hypertension: The Task Force for the management of arterial hypertension of the European Society of Cardiology and the European Society of Hypertension. J Hypertens 2018;36:1953-2041. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000001940">https://doi.org/10.1097/HJH.0000000000001940</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Williams</surname>
							<given-names>B</given-names>
						</name>
						<name>
							<surname>Mancia</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>Spiering</surname>
							<given-names>W</given-names>
						</name>
						<name>
							<surname>Agabiti Rosei</surname>
							<given-names>E</given-names>
						</name>
						<name>
							<surname>Azizi</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>Burnier</surname>
							<given-names>M</given-names>
						</name>
					</person-group>
					<comment>Authors/Task Force Members:</comment>
					<article-title>2018 ESC/ESH Guidelines for the management of arterial hypertension: The Task Force for the management of arterial hypertension of the European Society of Cardiology and the European Society of Hypertension: The Task Force for the management of arterial hypertension of the European Society of Cardiology and the European Society of Hypertension</article-title>
					<source>J Hypertens</source>
					<year>2018</year>
					<volume>36</volume>
					<fpage>1953</fpage>
					<lpage>2041</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000001940">https://doi.org/10.1097/HJH.0000000000001940</ext-link>
				</element-citation>
			</ref>
			<ref id="B4">
				<label>4</label>
				<mixed-citation>Whelton PK, Carey RM, Aronow WS, Casey DEJr, Collins KJ, Dennison Himmelfarb C. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation 2017;138:e484-e594. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/CIR.0000000000000596">https://doi.org/10.1161/CIR.0000000000000596</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Whelton</surname>
							<given-names>PK</given-names>
						</name>
						<name>
							<surname>Carey</surname>
							<given-names>RM</given-names>
						</name>
						<name>
							<surname>Aronow</surname>
							<given-names>WS</given-names>
						</name>
						<name>
							<surname>Casey</surname>
							<given-names>DE</given-names>
							<suffix>Jr</suffix>
						</name>
						<name>
							<surname>Collins</surname>
							<given-names>KJ</given-names>
						</name>
						<name>
							<surname>Dennison Himmelfarb</surname>
							<given-names>C</given-names>
						</name>
					</person-group>
					<article-title>2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines</article-title>
					<source>Circulation</source>
					<year>2017</year>
					<volume>138</volume>
					<fpage>e484</fpage>
					<lpage>e594</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/CIR.0000000000000596">https://doi.org/10.1161/CIR.0000000000000596</ext-link>
				</element-citation>
			</ref>
			<ref id="B5">
				<label>5</label>
				<mixed-citation>Marín M, Bendersky M, Páez O, Obregón S, Salazar M, Rodríguez Py cols. Consenso argentino de hipertensión arterial. Rev Argent Cardiol 2018;86s21-53</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Marín</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>Bendersky</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>Páez</surname>
							<given-names>O</given-names>
						</name>
						<name>
							<surname>Obregón</surname>
							<given-names>S</given-names>
						</name>
						<name>
							<surname>Salazar</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>Rodríguez</surname>
							<given-names>P</given-names>
						</name>
					</person-group>
					<article-title>Consenso argentino de hipertensión arterial</article-title>
					<source>Rev Argent Cardiol</source>
					<year>2018</year>
					<volume>86</volume>
					<supplement>s2</supplement>
					<fpage>1</fpage>
					<lpage>53</lpage>
				</element-citation>
			</ref>
			<ref id="B6">
				<label>6</label>
				<mixed-citation>O'Brien E, Asmar R, Beilin L, Imai Y, Mallion JM, Mancia GEuropean Society of Hypertension Working Group on Blood Pressure Monitoring. European Society of Hypertension recommendations for conventional, ambulatory and home blood pressure measurement. J Hypertens 2003;21:821-48. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/00004872-200305000-00001">https://doi.org/10.1097/00004872-200305000-00001</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>O'Brien</surname>
							<given-names>E</given-names>
						</name>
						<name>
							<surname>Asmar</surname>
							<given-names>R</given-names>
						</name>
						<name>
							<surname>Beilin</surname>
							<given-names>L</given-names>
						</name>
						<name>
							<surname>Imai</surname>
							<given-names>Y</given-names>
						</name>
						<name>
							<surname>Mallion</surname>
							<given-names>JM</given-names>
						</name>
						<name>
							<surname>Mancia</surname>
							<given-names>G</given-names>
						</name>
					</person-group>
					<comment>European Society of Hypertension Working Group on Blood Pressure Monitoring</comment>
					<article-title>European Society of Hypertension recommendations for conventional, ambulatory and home blood pressure measurement</article-title>
					<source>J Hypertens</source>
					<year>2003</year>
					<volume>21</volume>
					<fpage>821</fpage>
					<lpage>848</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/00004872-200305000-00001">https://doi.org/10.1097/00004872-200305000-00001</ext-link>
				</element-citation>
			</ref>
			<ref id="B7">
				<label>7</label>
				<mixed-citation>Parati G, Stergiou GS, Asmar R, Bilo G, de Leeuw P, Imai YESH Working Group on Blood Pressure Monitoring. European Society of Hypertension practice guidelines for home blood pressure monitoring. J Hum Hypertens 2010;24:779-85. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1038/jhh.2010.54">https://doi.org/10.1038/jhh.2010.54</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Parati</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>Stergiou</surname>
							<given-names>GS</given-names>
						</name>
						<name>
							<surname>Asmar</surname>
							<given-names>R</given-names>
						</name>
						<name>
							<surname>Bilo</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>de Leeuw</surname>
							<given-names>P</given-names>
						</name>
						<name>
							<surname>Imai</surname>
							<given-names>Y</given-names>
						</name>
					</person-group>
					<comment>ESH Working Group on Blood Pressure Monitoring</comment>
					<article-title>European Society of Hypertension practice guidelines for home blood pressure monitoring</article-title>
					<source>J Hum Hypertens</source>
					<year>2010</year>
					<volume>24</volume>
					<fpage>779</fpage>
					<lpage>785</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1038/jhh.2010.54">https://doi.org/10.1038/jhh.2010.54</ext-link>
				</element-citation>
			</ref>
			<ref id="B8">
				<label>8</label>
				<mixed-citation>Sega R, Facchetti R, Bombelli M, Cesana G, Corrao G, Grassi G. Prognostic value of ambulatory and home blood pressures compared with office blood pressure in the general population: follow-up results from the Pressioni Arteriose Monitorate e Loro Associazioni (PAMELA) study. Circulation 2005;111:1777-83. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/01.CIR.0000160923.04524.5B">https://doi.org/10.1161/01.CIR.0000160923.04524.5B</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Sega</surname>
							<given-names>R</given-names>
						</name>
						<name>
							<surname>Facchetti</surname>
							<given-names>R</given-names>
						</name>
						<name>
							<surname>Bombelli</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>Cesana</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>Corrao</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>Grassi</surname>
							<given-names>G</given-names>
						</name>
					</person-group>
					<article-title>Prognostic value of ambulatory and home blood pressures compared with office blood pressure in the general population: follow-up results from the Pressioni Arteriose Monitorate e Loro Associazioni (PAMELA) study</article-title>
					<source>Circulation</source>
					<year>2005</year>
					<volume>111</volume>
					<fpage>1777</fpage>
					<lpage>1783</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/01.CIR.0000160923.04524.5B">https://doi.org/10.1161/01.CIR.0000160923.04524.5B</ext-link>
				</element-citation>
			</ref>
			<ref id="B9">
				<label>9</label>
				<mixed-citation>Kikuya M, Ohkubo T, Asayama K, Metoki H, Obara T, Saito S. Ambulatory blood pressure and 10-year risk of cardiovascular and noncardiovascular mortality: the Ohasama study. Hypertension 2005;45:240-5. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/01.HYP.0000152079.04553.2c">https://doi.org/10.1161/01.HYP.0000152079.04553.2c</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Kikuya</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>Ohkubo</surname>
							<given-names>T</given-names>
						</name>
						<name>
							<surname>Asayama</surname>
							<given-names>K</given-names>
						</name>
						<name>
							<surname>Metoki</surname>
							<given-names>H</given-names>
						</name>
						<name>
							<surname>Obara</surname>
							<given-names>T</given-names>
						</name>
						<name>
							<surname>Saito</surname>
							<given-names>S</given-names>
						</name>
					</person-group>
					<article-title>Ambulatory blood pressure and 10-year risk of cardiovascular and noncardiovascular mortality: the Ohasama study</article-title>
					<source>Hypertension</source>
					<year>2005</year>
					<volume>45</volume>
					<fpage>240</fpage>
					<lpage>245</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/01.HYP.0000152079.04553.2c">https://doi.org/10.1161/01.HYP.0000152079.04553.2c</ext-link>
				</element-citation>
			</ref>
			<ref id="B10">
				<label>10</label>
				<mixed-citation>Banegas JR, Ruilope LM, de la Sierra A, Vinyoles E, Gorostidi M, de la Cruz JJ. Relationship between Clinic and Ambulatory Blood-Pressure Measurements and Mortality. N Engl J Med 2018;378:1509-20. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1056/NEJMoa1712231">https://doi.org/10.1056/NEJMoa1712231</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Banegas</surname>
							<given-names>JR</given-names>
						</name>
						<name>
							<surname>Ruilope</surname>
							<given-names>LM</given-names>
						</name>
						<name>
							<surname>de la Sierra</surname>
							<given-names>A</given-names>
						</name>
						<name>
							<surname>Vinyoles</surname>
							<given-names>E</given-names>
						</name>
						<name>
							<surname>Gorostidi</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>de la Cruz</surname>
							<given-names>JJ</given-names>
						</name>
					</person-group>
					<article-title>Relationship between Clinic and Ambulatory Blood-Pressure Measurements and Mortality</article-title>
					<source>N Engl J Med</source>
					<year>2018</year>
					<volume>378</volume>
					<fpage>1509</fpage>
					<lpage>1520</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1056/NEJMoa1712231">https://doi.org/10.1056/NEJMoa1712231</ext-link>
				</element-citation>
			</ref>
			<ref id="B11">
				<label>11</label>
				<mixed-citation>Parikh NI, Pencina MJ, Wang TJ, Benjamin EJ, Lanier KJ, Levy D. A risk score for predicting near-term incidence of hypertension: the Framingham Heart Study. Ann Intern Med 2008;148:102-10. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.7326/0003-4819-148-2-200801150-00005">https://doi.org/10.7326/0003-4819-148-2-200801150-00005</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Parikh</surname>
							<given-names>NI</given-names>
						</name>
						<name>
							<surname>Pencina</surname>
							<given-names>MJ</given-names>
						</name>
						<name>
							<surname>Wang</surname>
							<given-names>TJ</given-names>
						</name>
						<name>
							<surname>Benjamin</surname>
							<given-names>EJ</given-names>
						</name>
						<name>
							<surname>Lanier</surname>
							<given-names>KJ</given-names>
						</name>
						<name>
							<surname>Levy</surname>
							<given-names>D</given-names>
						</name>
					</person-group>
					<article-title>A risk score for predicting near-term incidence of hypertension: the Framingham Heart Study</article-title>
					<source>Ann Intern Med</source>
					<year>2008</year>
					<volume>148</volume>
					<fpage>102</fpage>
					<lpage>110</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.7326/0003-4819-148-2-200801150-00005">https://doi.org/10.7326/0003-4819-148-2-200801150-00005</ext-link>
				</element-citation>
			</ref>
			<ref id="B12">
				<label>12</label>
				<mixed-citation>Conroy RM, Pyörälä K, Fitzgerald AP, Sans S, Menotti A, De Backer GSCORE project group. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J 2003;24:987-1003. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/S0195-668X(03)00114-3">https://doi.org/10.1016/S0195-668X(03)00114-3</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Conroy</surname>
							<given-names>RM</given-names>
						</name>
						<name>
							<surname>Pyörälä</surname>
							<given-names>K</given-names>
						</name>
						<name>
							<surname>Fitzgerald</surname>
							<given-names>AP</given-names>
						</name>
						<name>
							<surname>Sans</surname>
							<given-names>S</given-names>
						</name>
						<name>
							<surname>Menotti</surname>
							<given-names>A</given-names>
						</name>
						<name>
							<surname>De Backer</surname>
							<given-names>G</given-names>
						</name>
					</person-group>
					<comment>SCORE project group</comment>
					<article-title>Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project</article-title>
					<source>Eur Heart J</source>
					<year>2003</year>
					<volume>24</volume>
					<fpage>987</fpage>
					<lpage>1003</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/S0195-668X(03)00114-3">https://doi.org/10.1016/S0195-668X(03)00114-3</ext-link>
				</element-citation>
			</ref>
			<ref id="B13">
				<label>13</label>
				<mixed-citation>Itchhaporia D, Snow PB, Almassy RJ, Oetgen WJ. Artificial neural networks: current status in cardiovascular medicine. J Am Coll Cardiol 1996;28:515-21. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/0735-1097(96)00174-X">https://doi.org/10.1016/0735-1097(96)00174-X</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Itchhaporia</surname>
							<given-names>D</given-names>
						</name>
						<name>
							<surname>Snow</surname>
							<given-names>PB</given-names>
						</name>
						<name>
							<surname>Almassy</surname>
							<given-names>RJ</given-names>
						</name>
						<name>
							<surname>Oetgen</surname>
							<given-names>WJ</given-names>
						</name>
					</person-group>
					<article-title>Artificial neural networks: current status in cardiovascular medicine</article-title>
					<source>J Am Coll Cardiol</source>
					<year>1996</year>
					<volume>28</volume>
					<fpage>515</fpage>
					<lpage>521</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/0735-1097(96)00174-X">https://doi.org/10.1016/0735-1097(96)00174-X</ext-link>
				</element-citation>
			</ref>
			<ref id="B14">
				<label>14</label>
				<mixed-citation>Ning G, Su J, Li Y, Wang X, Li C, Yan W. Artificial neural network based model for cardiovascular risk stratification in hypertension. Med Biol Eng Comput 2006;44:202-8. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1007/s11517-006-0028-2">https://doi.org/10.1007/s11517-006-0028-2</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Ning</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>Su</surname>
							<given-names>J</given-names>
						</name>
						<name>
							<surname>Li</surname>
							<given-names>Y</given-names>
						</name>
						<name>
							<surname>Wang</surname>
							<given-names>X</given-names>
						</name>
						<name>
							<surname>Li</surname>
							<given-names>C</given-names>
						</name>
						<name>
							<surname>Yan</surname>
							<given-names>W</given-names>
						</name>
					</person-group>
					<article-title>Artificial neural network based model for cardiovascular risk stratification in hypertension</article-title>
					<source>Med Biol Eng Comput</source>
					<year>2006</year>
					<volume>44</volume>
					<fpage>202</fpage>
					<lpage>208</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1007/s11517-006-0028-2">https://doi.org/10.1007/s11517-006-0028-2</ext-link>
				</element-citation>
			</ref>
			<ref id="B15">
				<label>15</label>
				<mixed-citation>Cui J. Overview of risk prediction models in cardiovascular disease research. Ann Epidemiol 2009;19:711-7. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.annepidem.2009.05.005">https://doi.org/10.1016/j.annepidem.2009.05.005</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Cui</surname>
							<given-names>J</given-names>
						</name>
					</person-group>
					<article-title>Overview of risk prediction models in cardiovascular disease research</article-title>
					<source>Ann Epidemiol</source>
					<year>2009</year>
					<volume>19</volume>
					<fpage>711</fpage>
					<lpage>717</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.annepidem.2009.05.005">https://doi.org/10.1016/j.annepidem.2009.05.005</ext-link>
				</element-citation>
			</ref>
			<ref id="B16">
				<label>16</label>
				<mixed-citation>Elosua R. Cardiovascular risk functions: usefulness and limitations. Rev Esp Cardiol (Engl Ed) 2014;67:77-9. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.rec.2013.09.012">https://doi.org/10.1016/j.rec.2013.09.012</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Elosua</surname>
							<given-names>R</given-names>
						</name>
					</person-group>
					<article-title>Cardiovascular risk functions: usefulness and limitations</article-title>
					<source>Rev Esp Cardiol (Engl Ed)</source>
					<year>2014</year>
					<volume>67</volume>
					<fpage>77</fpage>
					<lpage>79</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.rec.2013.09.012">https://doi.org/10.1016/j.rec.2013.09.012</ext-link>
				</element-citation>
			</ref>
			<ref id="B17">
				<label>17</label>
				<mixed-citation>LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015;521:436-44. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1038/nature14539">https://doi.org/10.1038/nature14539</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>LeCun</surname>
							<given-names>Y</given-names>
						</name>
						<name>
							<surname>Bengio</surname>
							<given-names>Y</given-names>
						</name>
						<name>
							<surname>Hinton</surname>
							<given-names>G</given-names>
						</name>
					</person-group>
					<article-title>Deep learning</article-title>
					<source>Nature</source>
					<year>2015</year>
					<volume>521</volume>
					<fpage>436</fpage>
					<lpage>444</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1038/nature14539">https://doi.org/10.1038/nature14539</ext-link>
				</element-citation>
			</ref>
			<ref id="B18">
				<label>18</label>
				<mixed-citation>Sargent DJ. Comparison of artificial neural networks with other statistical approaches: results from medical data sets. Cancer 2001;91:1636-42. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1002/1097-0142(20010415)91:8+&lt;1636::AID-CNCR1176&gt;3.0.CO;2-D">https://doi.org/10.1002/1097-0142(20010415)91:8+&lt;1636::AID-CNCR1176&gt;3.0.CO;2-D</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Sargent</surname>
							<given-names>DJ</given-names>
						</name>
					</person-group>
					<article-title>Comparison of artificial neural networks with other statistical approaches: results from medical data sets</article-title>
					<source>Cancer</source>
					<year>2001</year>
					<volume>91</volume>
					<fpage>1636</fpage>
					<lpage>1642</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1002/1097-0142(20010415)91:8+&lt;1636::AID-CNCR1176&gt;3.0.CO;2-D">https://doi.org/10.1002/1097-0142(20010415)91:8+&lt;1636::AID-CNCR1176&gt;3.0.CO;2-D</ext-link>
				</element-citation>
			</ref>
			<ref id="B19">
				<label>19</label>
				<mixed-citation>Trujillano J, March J, Sorribas A. Aproximación metodológica al uso de redes neuronales artificiales para la predicción de resultados en medicina. Med Clin (Barc) 2004;122:59-67. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1157/13057536">https://doi.org/10.1157/13057536</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Trujillano</surname>
							<given-names>J</given-names>
						</name>
						<name>
							<surname>March</surname>
							<given-names>J</given-names>
						</name>
						<name>
							<surname>Sorribas</surname>
							<given-names>A</given-names>
						</name>
					</person-group>
					<article-title>Aproximación metodológica al uso de redes neuronales artificiales para la predicción de resultados en medicina</article-title>
					<source>Med Clin (Barc)</source>
					<year>2004</year>
					<volume>122</volume>
					<fpage>59</fpage>
					<lpage>67</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1157/13057536">https://doi.org/10.1157/13057536</ext-link>
				</element-citation>
			</ref>
			<ref id="B20">
				<label>20</label>
				<mixed-citation>Ye X, Zeng QT, Facelli JC, Brixner DI, Conway M, Bray BE. Predicting Optimal Hypertension Treatment Pathways Using Recurrent Neural Networks. Int J Med Inform 2020;139:104122. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.ijmedinf.2020.104122">https://doi.org/10.1016/j.ijmedinf.2020.104122</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Ye</surname>
							<given-names>X</given-names>
						</name>
						<name>
							<surname>Zeng</surname>
							<given-names>QT</given-names>
						</name>
						<name>
							<surname>Facelli</surname>
							<given-names>JC</given-names>
						</name>
						<name>
							<surname>Brixner</surname>
							<given-names>DI</given-names>
						</name>
						<name>
							<surname>Conway</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>Bray</surname>
							<given-names>BE</given-names>
						</name>
					</person-group>
					<article-title>Predicting Optimal Hypertension Treatment Pathways Using Recurrent Neural Networks</article-title>
					<source>Int J Med Inform</source>
					<year>2020</year>
					<volume>139</volume>
					<fpage>104122</fpage>
					<lpage>104122</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.ijmedinf.2020.104122">https://doi.org/10.1016/j.ijmedinf.2020.104122</ext-link>
				</element-citation>
			</ref>
			<ref id="B21">
				<label>21</label>
				<mixed-citation>Yan Y, Chen R, Yang Z, Ma Y, Huang J, Luo L. Application of back propagation neural network model optimized by particle swarm algorithm in predicting the risk of hypertension. J Clin Hypertens (Greenwich) 2022;24:1606-17. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1111/jch.14597">https://doi.org/10.1111/jch.14597</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Yan</surname>
							<given-names>Y</given-names>
						</name>
						<name>
							<surname>Chen</surname>
							<given-names>R</given-names>
						</name>
						<name>
							<surname>Yang</surname>
							<given-names>Z</given-names>
						</name>
						<name>
							<surname>Ma</surname>
							<given-names>Y</given-names>
						</name>
						<name>
							<surname>Huang</surname>
							<given-names>J</given-names>
						</name>
						<name>
							<surname>Luo</surname>
							<given-names>L</given-names>
						</name>
					</person-group>
					<article-title>Application of back propagation neural network model optimized by particle swarm algorithm in predicting the risk of hypertension</article-title>
					<source>J Clin Hypertens (Greenwich)</source>
					<year>2022</year>
					<volume>24</volume>
					<fpage>1606</fpage>
					<lpage>1617</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1111/jch.14597">https://doi.org/10.1111/jch.14597</ext-link>
				</element-citation>
			</ref>
			<ref id="B22">
				<label>22</label>
				<mixed-citation>Huang S, Xu Y, Yue L, Wei S, Liu L, Gan X. Evaluating the risk of hypertension using an artificial neural network method in rural residents over the age of 35 years in a Chinese area. Hypertens Res 2010;33:722-6. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1038/hr.2010.73">https://doi.org/10.1038/hr.2010.73</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Huang</surname>
							<given-names>S</given-names>
						</name>
						<name>
							<surname>Xu</surname>
							<given-names>Y</given-names>
						</name>
						<name>
							<surname>Yue</surname>
							<given-names>L</given-names>
						</name>
						<name>
							<surname>Wei</surname>
							<given-names>S</given-names>
						</name>
						<name>
							<surname>Liu</surname>
							<given-names>L</given-names>
						</name>
						<name>
							<surname>Gan</surname>
							<given-names>X</given-names>
						</name>
					</person-group>
					<article-title>Evaluating the risk of hypertension using an artificial neural network method in rural residents over the age of 35 years in a Chinese area</article-title>
					<source>Hypertens Res</source>
					<year>2010</year>
					<volume>33</volume>
					<fpage>722</fpage>
					<lpage>726</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1038/hr.2010.73">https://doi.org/10.1038/hr.2010.73</ext-link>
				</element-citation>
			</ref>
			<ref id="B23">
				<label>23</label>
				<mixed-citation>World Medical Association. World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects. JAMA 2013;310:2191-4. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1001/jama.2013.281053">https://doi.org/10.1001/jama.2013.281053</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<collab>World Medical Association</collab>
					</person-group>
					<article-title>World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects</article-title>
					<source>JAMA</source>
					<year>2013</year>
					<volume>310</volume>
					<fpage>2191</fpage>
					<lpage>2194</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1001/jama.2013.281053">https://doi.org/10.1001/jama.2013.281053</ext-link>
				</element-citation>
			</ref>
			<ref id="B24">
				<label>24</label>
				<mixed-citation>Islam MM, Rahman MJ, Chandra Roy D, Tawabunnahar M, Jahan R, Ahmed NAMF. Machine learning algorithm for characterizing risks of hypertension, at an early stage in Bangladesh. Diabetes Metab Syndr 2021;15:877-84. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.dsx.2021.03.035">https://doi.org/10.1016/j.dsx.2021.03.035</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Islam</surname>
							<given-names>MM</given-names>
						</name>
						<name>
							<surname>Rahman</surname>
							<given-names>MJ</given-names>
						</name>
						<name>
							<surname>Chandra Roy</surname>
							<given-names>D</given-names>
						</name>
						<name>
							<surname>Tawabunnahar</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>Jahan</surname>
							<given-names>R</given-names>
						</name>
						<name>
							<surname>Ahmed</surname>
							<given-names>NAMF</given-names>
						</name>
					</person-group>
					<article-title>Machine learning algorithm for characterizing risks of hypertension, at an early stage in Bangladesh</article-title>
					<source>Diabetes Metab Syndr</source>
					<year>2021</year>
					<volume>15</volume>
					<fpage>877</fpage>
					<lpage>884</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.dsx.2021.03.035">https://doi.org/10.1016/j.dsx.2021.03.035</ext-link>
				</element-citation>
			</ref>
			<ref id="B25">
				<label>25</label>
				<mixed-citation>Leoncini G, Sacchi G, Viazzi F, Ravera M, Parodi D, Ratto E. Microalbuminuria identifies overall cardiovascular risk in essential hypertension: an artificial neural network-based approach. J Hypertens 2002;20:1315-21. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/00004872-200207000-00018">https://doi.org/10.1097/00004872-200207000-00018</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Leoncini</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>Sacchi</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>Viazzi</surname>
							<given-names>F</given-names>
						</name>
						<name>
							<surname>Ravera</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>Parodi</surname>
							<given-names>D</given-names>
						</name>
						<name>
							<surname>Ratto</surname>
							<given-names>E</given-names>
						</name>
					</person-group>
					<article-title>Microalbuminuria identifies overall cardiovascular risk in essential hypertension: an artificial neural network-based approach</article-title>
					<source>J Hypertens</source>
					<year>2002</year>
					<volume>20</volume>
					<fpage>1315</fpage>
					<lpage>1321</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/00004872-200207000-00018">https://doi.org/10.1097/00004872-200207000-00018</ext-link>
				</element-citation>
			</ref>
			<ref id="B26">
				<label>26</label>
				<mixed-citation>Viazzi F, Leoncini G, Sacchi G, Parodi D, Ratto E, Falqui V. Predicting cardiovascular risk using creatinine clearance and an artificial neural network in primary hypertension. J Hypertens 2006;24:1281-6. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/01.hjh.0000234107.08368.e5">https://doi.org/10.1097/01.hjh.0000234107.08368.e5</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Viazzi</surname>
							<given-names>F</given-names>
						</name>
						<name>
							<surname>Leoncini</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>Sacchi</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>Parodi</surname>
							<given-names>D</given-names>
						</name>
						<name>
							<surname>Ratto</surname>
							<given-names>E</given-names>
						</name>
						<name>
							<surname>Falqui</surname>
							<given-names>V</given-names>
						</name>
					</person-group>
					<article-title>Predicting cardiovascular risk using creatinine clearance and an artificial neural network in primary hypertension</article-title>
					<source>J Hypertens</source>
					<year>2006</year>
					<volume>24</volume>
					<fpage>1281</fpage>
					<lpage>1286</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/01.hjh.0000234107.08368.e5">https://doi.org/10.1097/01.hjh.0000234107.08368.e5</ext-link>
				</element-citation>
			</ref>
			<ref id="B27">
				<label>27</label>
				<mixed-citation>Mandrekar JN. Receiver operating characteristic curve in diagnostic test assessment. J Thorac Oncol 2010;5:1315-6. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/JTO.0b013e3181ec173d">https://doi.org/10.1097/JTO.0b013e3181ec173d</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Mandrekar</surname>
							<given-names>JN</given-names>
						</name>
					</person-group>
					<article-title>Receiver operating characteristic curve in diagnostic test assessment</article-title>
					<source>J Thorac Oncol</source>
					<year>2010</year>
					<volume>5</volume>
					<fpage>1315</fpage>
					<lpage>1316</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/JTO.0b013e3181ec173d">https://doi.org/10.1097/JTO.0b013e3181ec173d</ext-link>
				</element-citation>
			</ref>
			<ref id="B28">
				<label>28</label>
				<mixed-citation>McHugh ML. Interrater reliability: the kappa statistic. Biochem Med (Zagreb) 2012;22:276-82. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.11613/BM.2012.031">https://doi.org/10.11613/BM.2012.031</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>McHugh</surname>
							<given-names>ML</given-names>
						</name>
					</person-group>
					<article-title>Interrater reliability: the kappa statistic</article-title>
					<source>Biochem Med (Zagreb)</source>
					<year>2012</year>
					<volume>22</volume>
					<fpage>276</fpage>
					<lpage>282</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.11613/BM.2012.031">https://doi.org/10.11613/BM.2012.031</ext-link>
				</element-citation>
			</ref>
			<ref id="B29">
				<label>29</label>
				<mixed-citation>Kario K. Home Blood Pressure Monitoring: Current Status and New Developments. Am J Hypertens 2021;34:783-794. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1093/ajh/hpab017">https://doi.org/10.1093/ajh/hpab017</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Kario</surname>
							<given-names>K</given-names>
						</name>
					</person-group>
					<article-title>Home Blood Pressure Monitoring: Current Status and New Developments</article-title>
					<source>Am J Hypertens</source>
					<year>2021</year>
					<volume>34</volume>
					<fpage>783</fpage>
					<lpage>794</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1093/ajh/hpab017">https://doi.org/10.1093/ajh/hpab017</ext-link>
				</element-citation>
			</ref>
			<ref id="B30">
				<label>30</label>
				<mixed-citation>Tadic M, Cuspidi C, Celic V, Petrovic O, Pencic B, Mancia G. The prognostic importance of right ventricular remodeling and the circadian blood pressure pattern on the long-term cardiovascular outcome. J Hypertens 2020;38:1525-30. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000002432">https://doi.org/10.1097/HJH.0000000000002432</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Tadic</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>Cuspidi</surname>
							<given-names>C</given-names>
						</name>
						<name>
							<surname>Celic</surname>
							<given-names>V</given-names>
						</name>
						<name>
							<surname>Petrovic</surname>
							<given-names>O</given-names>
						</name>
						<name>
							<surname>Pencic</surname>
							<given-names>B</given-names>
						</name>
						<name>
							<surname>Mancia</surname>
							<given-names>G</given-names>
						</name>
					</person-group>
					<article-title>The prognostic importance of right ventricular remodeling and the circadian blood pressure pattern on the long-term cardiovascular outcome</article-title>
					<source>J Hypertens</source>
					<year>2020</year>
					<volume>38</volume>
					<fpage>1525</fpage>
					<lpage>1530</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000002432">https://doi.org/10.1097/HJH.0000000000002432</ext-link>
				</element-citation>
			</ref>
			<ref id="B31">
				<label>31</label>
				<mixed-citation>Kario K, Williams B. Nocturnal Hypertension and Heart Failure: Mechanisms, Evidence, and New Treatments. Hypertension 2021;78:564-77. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.121.17440">https://doi.org/10.1161/HYPERTENSIONAHA.121.17440</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Kario</surname>
							<given-names>K</given-names>
						</name>
						<name>
							<surname>Williams</surname>
							<given-names>B</given-names>
						</name>
					</person-group>
					<article-title>Nocturnal Hypertension and Heart Failure: Mechanisms, Evidence, and New Treatments</article-title>
					<source>Hypertension</source>
					<year>2021</year>
					<volume>78</volume>
					<fpage>564</fpage>
					<lpage>577</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.121.17440">https://doi.org/10.1161/HYPERTENSIONAHA.121.17440</ext-link>
				</element-citation>
			</ref>
			<ref id="B32">
				<label>32</label>
				<mixed-citation>Cardoso CRL, Salles GF. Associations of the nocturnal blood pressure fall and morning surge with cardiovascular events and mortality in individuals with resistant hypertension. J Hypertens 2021;39:1177-8710.1097/HJH.0000000000002775</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Cardoso</surname>
							<given-names>CRL</given-names>
						</name>
						<name>
							<surname>Salles</surname>
							<given-names>GF</given-names>
						</name>
					</person-group>
					<article-title>Associations of the nocturnal blood pressure fall and morning surge with cardiovascular events and mortality in individuals with resistant hypertension</article-title>
					<source>J Hypertens</source>
					<year>2021</year>
					<volume>39</volume>
					<fpage>1177</fpage>
					<lpage>1187</lpage>
					<pub-id pub-id-type="doi">10.1097/HJH.0000000000002775</pub-id>
				</element-citation>
			</ref>
			<ref id="B33">
				<label>33</label>
				<mixed-citation>Fagard RH, Celis H, Thijs L, Staessen JA, Clement DL, De Buyzere ML. Daytime and nighttime blood pressure as predictors of death and cause-specific cardiovascular events in hypertension. Hypertension 2008;5:55-61. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000002775">https://doi.org/10.1097/HJH.0000000000002775</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Fagard</surname>
							<given-names>RH</given-names>
						</name>
						<name>
							<surname>Celis</surname>
							<given-names>H</given-names>
						</name>
						<name>
							<surname>Thijs</surname>
							<given-names>L</given-names>
						</name>
						<name>
							<surname>Staessen</surname>
							<given-names>JA</given-names>
						</name>
						<name>
							<surname>Clement</surname>
							<given-names>DL</given-names>
						</name>
						<name>
							<surname>De Buyzere</surname>
							<given-names>ML</given-names>
						</name>
					</person-group>
					<article-title>Daytime and nighttime blood pressure as predictors of death and cause-specific cardiovascular events in hypertension</article-title>
					<source>Hypertension</source>
					<year>2008</year>
					<volume>5</volume>
					<fpage>55</fpage>
					<lpage>61</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000002775">https://doi.org/10.1097/HJH.0000000000002775</ext-link>
				</element-citation>
			</ref>
			<ref id="B34">
				<label>34</label>
				<mixed-citation>Draman MS, Dolan E, van der Poel L, Tun TK, McDermott JH, Sreenan S. The importance of night-time systolic blood pressure in diabetic patients: Dublin Outcome Study. J Hypertens 2015;33:1373-7. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.107.100727">https://doi.org/10.1161/HYPERTENSIONAHA.107.100727</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Draman</surname>
							<given-names>MS</given-names>
						</name>
						<name>
							<surname>Dolan</surname>
							<given-names>E</given-names>
						</name>
						<name>
							<surname>van der Poel</surname>
							<given-names>L</given-names>
						</name>
						<name>
							<surname>Tun</surname>
							<given-names>TK</given-names>
						</name>
						<name>
							<surname>McDermott</surname>
							<given-names>JH</given-names>
						</name>
						<name>
							<surname>Sreenan</surname>
							<given-names>S</given-names>
						</name>
					</person-group>
					<article-title>The importance of night-time systolic blood pressure in diabetic patients: Dublin Outcome Study</article-title>
					<source>J Hypertens</source>
					<year>2015</year>
					<volume>33</volume>
					<fpage>1373</fpage>
					<lpage>1377</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.107.100727">https://doi.org/10.1161/HYPERTENSIONAHA.107.100727</ext-link>
				</element-citation>
			</ref>
			<ref id="B35">
				<label>35</label>
				<mixed-citation>Kario K, Kanegae H, Tomitani N, Okawara Y, Fujiwara T, Yano Y. Nighttime Blood Pressure Measured by Home Blood Pressure Monitoring as an Independent Predictor of Cardiovascular Events in General Practice. Hypertension 2019;73:1240-8. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.118.12740">https://doi.org/10.1161/HYPERTENSIONAHA.118.12740</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Kario</surname>
							<given-names>K</given-names>
						</name>
						<name>
							<surname>Kanegae</surname>
							<given-names>H</given-names>
						</name>
						<name>
							<surname>Tomitani</surname>
							<given-names>N</given-names>
						</name>
						<name>
							<surname>Okawara</surname>
							<given-names>Y</given-names>
						</name>
						<name>
							<surname>Fujiwara</surname>
							<given-names>T</given-names>
						</name>
						<name>
							<surname>Yano</surname>
							<given-names>Y</given-names>
						</name>
					</person-group>
					<article-title>Nighttime Blood Pressure Measured by Home Blood Pressure Monitoring as an Independent Predictor of Cardiovascular Events in General Practice</article-title>
					<source>Hypertension</source>
					<year>2019</year>
					<volume>73</volume>
					<fpage>1240</fpage>
					<lpage>1248</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.118.12740">https://doi.org/10.1161/HYPERTENSIONAHA.118.12740</ext-link>
				</element-citation>
			</ref>
			<ref id="B36">
				<label>36</label>
				<mixed-citation>Hansen TW, Li Y, Boggia J, Thijs L, Richart T, Staessen JA. Predictive role of the nighttime blood pressure. Hypertension 2011;57:3-10. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.109.133900">https://doi.org/10.1161/HYPERTENSIONAHA.109.133900</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Hansen</surname>
							<given-names>TW</given-names>
						</name>
						<name>
							<surname>Li</surname>
							<given-names>Y</given-names>
						</name>
						<name>
							<surname>Boggia</surname>
							<given-names>J</given-names>
						</name>
						<name>
							<surname>Thijs</surname>
							<given-names>L</given-names>
						</name>
						<name>
							<surname>Richart</surname>
							<given-names>T</given-names>
						</name>
						<name>
							<surname>Staessen</surname>
							<given-names>JA</given-names>
						</name>
					</person-group>
					<article-title>Predictive role of the nighttime blood pressure</article-title>
					<source>Hypertension</source>
					<year>2011</year>
					<volume>57</volume>
					<fpage>3</fpage>
					<lpage>10</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.109.133900">https://doi.org/10.1161/HYPERTENSIONAHA.109.133900</ext-link>
				</element-citation>
			</ref>
			<ref id="B37">
				<label>37</label>
				<mixed-citation>Salles GF, Reboldi G, Fagard RH, Cardoso CR, Pierdomenico SD, Verdecchia PABC-H Investigators. Prognostic Effect of the Nocturnal Blood Pressure Fall in Hypertensive Patients: The Ambulatory Blood Pressure Collaboration in Patients With Hypertension (ABC-H) Meta-Analysis. Hypertension 2016;67:693-700. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.115.06981">https://doi.org/10.1161/HYPERTENSIONAHA.115.06981</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Salles</surname>
							<given-names>GF</given-names>
						</name>
						<name>
							<surname>Reboldi</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>Fagard</surname>
							<given-names>RH</given-names>
						</name>
						<name>
							<surname>Cardoso</surname>
							<given-names>CR</given-names>
						</name>
						<name>
							<surname>Pierdomenico</surname>
							<given-names>SD</given-names>
						</name>
						<name>
							<surname>Verdecchia</surname>
							<given-names>P</given-names>
						</name>
					</person-group>
					<person-group person-group-type="author"/>
					<article-title>Prognostic Effect of the Nocturnal Blood Pressure Fall in Hypertensive Patients: The Ambulatory Blood Pressure Collaboration in Patients With Hypertension (ABC-H) Meta-Analysis</article-title>
					<source>Hypertension</source>
					<year>2016</year>
					<volume>67</volume>
					<fpage>693</fpage>
					<lpage>700</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.115.06981">https://doi.org/10.1161/HYPERTENSIONAHA.115.06981</ext-link>
				</element-citation>
			</ref>
			<ref id="B38">
				<label>38</label>
				<mixed-citation>Zhu J, Hao X, Tang H, Xu J, Wang A, Zhang X. Time point of nocturnal trough systolic blood pressure as an independent predictor of cardiovascular events. J Clin Hypertens (Greenwich) 2022;24:283-91. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1111/jch.14404">https://doi.org/10.1111/jch.14404</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Zhu</surname>
							<given-names>J</given-names>
						</name>
						<name>
							<surname>Hao</surname>
							<given-names>X</given-names>
						</name>
						<name>
							<surname>Tang</surname>
							<given-names>H</given-names>
						</name>
						<name>
							<surname>Xu</surname>
							<given-names>J</given-names>
						</name>
						<name>
							<surname>Wang</surname>
							<given-names>A</given-names>
						</name>
						<name>
							<surname>Zhang</surname>
							<given-names>X</given-names>
						</name>
					</person-group>
					<article-title>Time point of nocturnal trough systolic blood pressure as an independent predictor of cardiovascular events</article-title>
					<source>J Clin Hypertens (Greenwich)</source>
					<year>2022</year>
					<volume>24</volume>
					<fpage>283</fpage>
					<lpage>291</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1111/jch.14404">https://doi.org/10.1111/jch.14404</ext-link>
				</element-citation>
			</ref>
			<ref id="B39">
				<label>39</label>
				<mixed-citation>Huguet N, Green BB, Voss RW, Larson AE, Angier H, Miguel M. Factors Associated With Blood Pressure Control Among Patients in Community Health Centers. Am J Prev Med 2023;64:631-41. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.amepre.2022.11.002">https://doi.org/10.1016/j.amepre.2022.11.002</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Huguet</surname>
							<given-names>N</given-names>
						</name>
						<name>
							<surname>Green</surname>
							<given-names>BB</given-names>
						</name>
						<name>
							<surname>Voss</surname>
							<given-names>RW</given-names>
						</name>
						<name>
							<surname>Larson</surname>
							<given-names>AE</given-names>
						</name>
						<name>
							<surname>Angier</surname>
							<given-names>H</given-names>
						</name>
						<name>
							<surname>Miguel</surname>
							<given-names>M</given-names>
						</name>
					</person-group>
					<article-title>Factors Associated With Blood Pressure Control Among Patients in Community Health Centers</article-title>
					<source>Am J Prev Med</source>
					<year>2023</year>
					<volume>64</volume>
					<fpage>631</fpage>
					<lpage>641</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.amepre.2022.11.002">https://doi.org/10.1016/j.amepre.2022.11.002</ext-link>
				</element-citation>
			</ref>
			<ref id="B40">
				<label>40</label>
				<mixed-citation>Jaeger BC, Bress AP, Bundy JD, Cheung AK, Cushman WC, Drawz PE. Longer-Term All-Cause and Cardiovascular Mortality With Intensive Blood Pressure Control: A Secondary Analysis of a Randomized Clinical Trial. JAMA Cardiol 2022;7:1138-46. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1001/jamacardio.2022.3345">https://doi.org/10.1001/jamacardio.2022.3345</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Jaeger</surname>
							<given-names>BC</given-names>
						</name>
						<name>
							<surname>Bress</surname>
							<given-names>AP</given-names>
						</name>
						<name>
							<surname>Bundy</surname>
							<given-names>JD</given-names>
						</name>
						<name>
							<surname>Cheung</surname>
							<given-names>AK</given-names>
						</name>
						<name>
							<surname>Cushman</surname>
							<given-names>WC</given-names>
						</name>
						<name>
							<surname>Drawz</surname>
							<given-names>PE</given-names>
						</name>
					</person-group>
					<article-title>Longer-Term All-Cause and Cardiovascular Mortality With Intensive Blood Pressure Control: A Secondary Analysis of a Randomized Clinical Trial</article-title>
					<source>JAMA Cardiol</source>
					<year>2022</year>
					<volume>7</volume>
					<fpage>1138</fpage>
					<lpage>1146</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1001/jamacardio.2022.3345">https://doi.org/10.1001/jamacardio.2022.3345</ext-link>
				</element-citation>
			</ref>
			<ref id="B41">
				<label>41</label>
				<mixed-citation>Mancia G, Kreutz R, Brunström M, Burnier M, Grassi G, Januszewicz A. 2023 ESH Guidelines for the management of arterial hypertension The Task Force for the management of arterial hypertension of the European Society of Hypertension: Endorsed by the International Society of Hypertension (ISH) and the European Renal Association (ERA). J Hypertens 2023;41:1874-2071. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000003480">https://doi.org/10.1097/HJH.0000000000003480</ext-link>
				</mixed-citation>
				<element-citation publication-type="journal">
					<person-group person-group-type="author">
						<name>
							<surname>Mancia</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>Kreutz</surname>
							<given-names>R</given-names>
						</name>
						<name>
							<surname>Brunström</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>Burnier</surname>
							<given-names>M</given-names>
						</name>
						<name>
							<surname>Grassi</surname>
							<given-names>G</given-names>
						</name>
						<name>
							<surname>Januszewicz</surname>
							<given-names>A</given-names>
						</name>
					</person-group>
					<article-title>2023 ESH Guidelines for the management of arterial hypertension The Task Force for the management of arterial hypertension of the European Society of Hypertension: Endorsed by the International Society of Hypertension (ISH) and the European Renal Association (ERA)</article-title>
					<source>J Hypertens</source>
					<year>2023</year>
					<volume>41</volume>
					<fpage>1874</fpage>
					<lpage>2071</lpage>
					<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000003480">https://doi.org/10.1097/HJH.0000000000003480</ext-link>
				</element-citation>
			</ref>
		</ref-list>
		<fn-group>
			<fn fn-type="other" id="fn2">
				<label>2</label>
				<p>Financiamiento Los autores declaran no haber recibido financiamiento alguno para la realización de este trabajo.</p>
			</fn>
			<fn fn-type="other" id="fn3">
				<label>3</label>
				<p>Política de uso de la inteligencia artificial No se utilizaron recursos de inteligencia artificial para la escritura del texto, ni para la generación de tablas ni gráficos.</p>
			</fn>
		</fn-group>
	</back>
	<!--<sub-article article-type="translation" id="s1" xml:lang="en">
		<front-stub>
			<article-categories>
				<subj-group subj-group-type="heading">
					<subject>ORIGINAL ARTICLE</subject>
				</subj-group>
			</article-categories>
			<title-group>
				<article-title>Events Prediction Ability in Patients with Hypertension using Artificial Neural Network Analysis of Ambulatory Blood Pressure Monitoring Compared to Clinical Risk Stratification</article-title>
			</title-group>
			<contrib-group>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0009-0004-1078-0407</contrib-id>
					<name>
						<surname>Di Gennaro</surname>
						<given-names>Federico P.</given-names>
					</name>
					<xref ref-type="aff" rid="aff3"><sup>1</sup></xref>
				</contrib>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0009-0000-0453-1384</contrib-id>
					<name>
						<surname>Catalano</surname>
						<given-names>María P.</given-names>
					</name>
					<xref ref-type="aff" rid="aff3">1</xref>
				</contrib>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0009-0006-9808-8777</contrib-id>
					<name>
						<surname>García Aguirre</surname>
						<given-names>Alejandro</given-names>
					</name>
					<xref ref-type="aff" rid="aff3">1</xref>
				</contrib>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0009-0007-4630-6848</contrib-id>
					<name>
						<surname>Fernández</surname>
						<given-names>María L.</given-names>
					</name>
					<xref ref-type="aff" rid="aff3">1</xref>
				</contrib>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0009-0004-0005-8285</contrib-id>
					<name>
						<surname>Llanos</surname>
						<given-names>Romina</given-names>
					</name>
					<xref ref-type="aff" rid="aff3">1</xref>
				</contrib>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0000-0001-9069-6512</contrib-id>
					<name>
						<surname>Pérez Lloret</surname>
						<given-names>Santiago</given-names>
					</name>
					<xref ref-type="aff" rid="aff4">2</xref>
				</contrib>
				<contrib contrib-type="author">
					<contrib-id contrib-id-type="orcid">0000-0002-3200-1142</contrib-id>
					<name>
						<surname>Higa</surname>
						<given-names>Claudio</given-names>
					</name>
					<xref ref-type="aff" rid="aff3"><sup>1</sup></xref>
					<xref ref-type="fn" rid="fn10"><sup>MTSAC</sup></xref>
				</contrib>
			</contrib-group>
			<aff id="aff3">
				<label>1</label>
				<institution content-type="original"> Arterial Hypertension Area, Cardiology Service, Department of Internal Medicine, Hospital Alemán de Buenos Aires. </institution>
				<institution content-type="orgname">Hospital Alemán de Buenos Aires</institution>
			</aff>
			<aff id="aff4">
				<label>2</label>
				<institution content-type="original"> Health Observatory , Universidad Católica Argentina, Consejo de Investigación Científicas y Técnicas (CONICET).</institution>
				<institution content-type="orgdiv2">Health Observatory</institution>
				<institution content-type="orgdiv1">Universidad Católica Argentina</institution>
				<institution content-type="orgname">Consejo de Investigación Científicas y Técnicas</institution>
			</aff>
			<author-notes>
				<corresp id="c2">
					<label>Correspondence:</label> Federico Di Gennaro, Hospital Alemán, Av. Pueyrredón 1640 (1112), Buenos Aires, Argentina. <bold>E-mail:</bold><email>fpdigennaro@hospitalaleman.com</email>
				</corresp>
				<fn fn-type="other" id="fn10">
					<label>MTSAC</label>
					<p>Miembro Titular de la Sociedad Argentina de Cardiología</p>
				</fn>
				<fn fn-type="conflict" id="fn4">
					<label>Conflicts of Interest</label>
					<p> None declared. (See the author's conflict of interests forms on the Web).</p>
				</fn>
			</author-notes>
			<abstract>
				<title>ABSTRACT </title>
				<sec>
					<title>Background:</title>
					<p> There is no available evidence comparing the predictive value of an artificial neural network (ANN)-based analysis method that integrates ambulatory blood pressure monitoring (ABPM) variables versus clinical risk stratification (CRS) for serious events in hypertensive patients at follow-up. </p>
				</sec>
				<sec>
					<title>Methods:</title>
					<p> We analyzed ABPM studies that included 27 measurements each one. The variables were daytime, nighttime and 24-hour mean, systolic and diastolic blood pressure, pulse pressure and heart rate; hypertensive load; standard deviations of pressures and heart rate; circadian rhythm. The dependent variable was the combined endpoint of death, stroke, acute myo cardial infarction, heart failure and kidney disease. For clinical risk stratification, the Argentine Consensus on Hypertension was used as a model. We evaluated the discriminative ability to predict the endpoint using ANN-ABPM and CRS by logistic regression through the analysis of the area under the receiver operating characteristic curve (AUC-ROC). Both AUC-ROC were compared by De Long test. SPSS 23.0 Statistics was used for statistical analyses and ANN modelling. </p>
				</sec>
				<sec>
					<title>Results: </title>
					<p>Data from 491 ABPM studies were analyzed. Mean age was 69 ± 14 years; 53% of population was female; 11.6% had diabetes; 51% had dyslipidemia; mean body mass index was 26 ± 4 kg/m<sup>2</sup>; 14.3% were smokers. Median follow-up was 6.6 years (interquartile range 4.5-8). The best predictive ANN model was the Multilayer Perceptron one with a hidden layer; neuronal architecture (27/7/2). Nocturnal systolic blood pressure (SBP) had 100% independent normalized importance for modelling. The AUC-ROC for the combined endpoint was 0.81 (95% CI 0.77-0.90) using neural network analysis with ABPM variables, and 0.67 (95% CI 0.56-0.77) using CRS; De Long's test p &lt; 0.001. </p>
				</sec>
				<sec>
					<title>Conclusion:</title>
					<p> We observed a higher discriminative ability to predict events at follow-up using ANN analysis with ABPM vari ables compared to conventional CRS. This observation raises a research hypothesis to be validated prospectively to optimize risk stratification and treatment in hypertensive patients. </p>
				</sec>
			</abstract>
			<kwd-group xml:lang="en">
				<title>Key words: </title>
				<kwd>Risk assessments </kwd>
				<kwd>Artificial neural networks </kwd>
				<kwd>Hypertension</kwd>
			</kwd-group>
		</front-stub>
		<body>
			<sec sec-type="intro">
				<title>INTRODUCTION</title>
				<p>Cardiovascular diseases remain the leading cause of morbidity and mortality worldwide. Prediction of cardiovascular (CV) events is essential for early identification of individuals at risk and implementation of more appropriate preventive interventions. (<xref ref-type="bibr" rid="B42">1</xref>,<xref ref-type="bibr" rid="B43">2</xref>)</p>
				<p>Therefore, it is recommended that all hypertensive patients be assessed to determine their overall cardiovascular risk (CVR) in order to define therapeutic and cardiovascular risk factor control measures. The information from medical history, physical examination, office blood pressure (BP) measurements and the results of recommended complementary studies will determine the presence of associated risk factors, target organ damage, and history of cardiovascular events. With this data it is possible to stratify the overall CVR of hypertensive patients and classify their risk as low, moderate, high and very high. Knowledge of an individual patient’s overall CVR stratification provides important predictive information, a global approach to prevention and the appropriate drug therapy. (<xref ref-type="bibr" rid="B44">3</xref>,<xref ref-type="bibr" rid="B45">4</xref>)</p>
				<p>A series of formulas or risk scores have been proposed to calculate CVR. The calculators arising from them are a heterogeneous group with various limitations (qualitative variables, complementary studies that are not used in daily clinical practice). Many of them have not been validated in our population. (<xref ref-type="bibr" rid="B44">3</xref>,<xref ref-type="bibr" rid="B45">4</xref>,<xref ref-type="bibr" rid="B46">5</xref>)</p>
				<p>The Argentine Consensus on Arterial Hypertension proposes an approach similar to that used by the European Society of Hypertension and adapted to our setting. (<xref ref-type="bibr" rid="B46">5</xref>)</p>
				<p>Although office blood pressure measurement is currently the recommended diagnostic method, it is not free of significant variability and bias due to the inaccuracy of the measurement technique. Thus, several national and international guidelines recommend the use of out-of-office measurements by means of ambulatory blood pressure monitoring (ABPM) to confirm the diagnosis of hypertension (HT) and provide more accurate predictive information. (<xref ref-type="bibr" rid="B46">5</xref>,<xref ref-type="bibr" rid="B47">6</xref>,<xref ref-type="bibr" rid="B48">7</xref>)</p>
				<p>In recent years, ABPM has become a useful adjunctive study for the diagnosis and prognostic assessment of CV events in hypertensive patients compared to isolated office measurements. In addition, ABPM can provide additional data, such as blood pressure variability, blood pressure dipping patterns, and mean blood pressure values at different periods of the day. (<xref ref-type="bibr" rid="B49">8</xref>,<xref ref-type="bibr" rid="B50">9</xref>,<xref ref-type="bibr" rid="B51">10</xref>)</p>
				<p>Although the accuracy of cardiovascular risk prediction models has improved over the years, some uncertainty still remains in the estimates. At present, the hemodynamic variables provided by the ABPM are not considered for cardiovascular risk stratification in hypertensive patients. (<xref ref-type="bibr" rid="B52">11</xref>,<xref ref-type="bibr" rid="B53">12</xref>)</p>
				<p>In this regard, it is important to highlight the need for more accurate predictive tools that incorporate the different blood pressure variables derived from ABPM.</p>
				<p>One of the most widely used methodological tools for predictive analysis in different areas of medicine, currently in full development, is artificial neural network (ANN). Analysis with ANN as an artificial intelligence (AI) model has shown to be superior in terms of prognostic accuracy (especially in the presence of non-linear associations) to the statistical tools we usually use, such as multivariate analysis and logistic regression. (<xref ref-type="bibr" rid="B54">13</xref>,<xref ref-type="bibr" rid="B55">14</xref>,<xref ref-type="bibr" rid="B56">15</xref>)</p>
				<p>ANNs can recognize relevant features in the data and adjust their synaptic weights and connections to improve their predictive performance, which depends on the number of input variables and their training. This allows ANN to make more accurate predictions. (<xref ref-type="bibr" rid="B57">16</xref>,<xref ref-type="bibr" rid="B58">17</xref>,<xref ref-type="bibr" rid="B59">18</xref>)</p>
				<p>The application of different machine learning models has been aimed at early detection and screening to identify those who will develop hypertension. (<xref ref-type="bibr" rid="B60">19</xref>,<xref ref-type="bibr" rid="B61">20</xref>,<xref ref-type="bibr" rid="B62">21</xref>,<xref ref-type="bibr" rid="B63">22</xref>)</p>
				<p>The ANN-based analysis model which integrates ABPM variables could improve the predictive ability and provide information to design a more accurate and complete stratification of CVR compared to the existing models.</p>
				<p>This study aimed to evaluate the ability to predict serious events in hypertensive patients at follow-up using an ANN-based analysis model which integrates ABPM variables compared to conventional clinical risk stratification.</p>
			</sec>
			<sec sec-type="methods">
				<title>METHODS</title>
				<p>A database with measurements from ABPM studies was analyzed according to the following inclusion criteria: adult patients (over 18 years of age), with a diagnosis of essential arterial hypertension under drug therapy, who underwent ABPM studies to evaluate treatment efficacy. They were required to have complete follow-up through electronic medical records and clinical visits to a community hospital.</p>
				<p>Consecutive ABPM studies performed between September 2013 and April 2020 with complete clinical follow-up until November 2022 were included. Data from the ABPM study reports (the averages of each of the variables were considered) were transferred to a spreadsheet and processed using Visual Basic and SQL softwares.</p>
				<p>Cardiovascular risk stratification was performed using the variables proposed by the Argentine Consensus on Arterial Hypertension (Argentine Society of Cardiology, Argentine Society of Arterial Hypertension, Argentine Federation of Cardiology) as a model. (<xref ref-type="bibr" rid="B46">5</xref>)</p>
				<p>The variables considered were the following: a) Risk factors: age, gender, history of dyslipidemia, diabetes, smoking, obesity; b) Target organ damage: diagnosis of left ventricular hypertrophy (LVH) confirmed by echocardiogram, chronic renal failure (stages 1 and 2); c) Associated clinical conditions or history of cardiovascular events: acute myocardial infarction (AMI), heart failure (HF), stroke and/or transient ischemic attack (stroke/TIA), coronary artery disease, myocardial revascularization, chronic kidney disease (stages 3, 4 and 5). </p>
				<p>Low risk was defined as patients with one associated risk factor; moderate risk: patients with two associated risk factors; high risk: patients with three or more associated risk factors and/or diabetes and/or target organ damage; very high risk: patients with history of cardiovascular events or associated clinical conditions. (<xref ref-type="bibr" rid="B45">4</xref>,<xref ref-type="bibr" rid="B46">5</xref>)</p>
				<p>A combined endpoint of serious events (SE) was defined as the occurrence of death and/or non-fatal AMI and/or stroke and/or TIA and/or HF and/or chronic renal failure validated in the electronic medical record by specialists in Internal Medicine and Cardiology according to current national and international guidelines. (<xref ref-type="bibr" rid="B44">3</xref>,<xref ref-type="bibr" rid="B45">4</xref>,<xref ref-type="bibr" rid="B46">5</xref>)</p>
				<p>Neural network algorithm models were developed including the ABPM variables as independent cofactors for input to the ANN, and SAE as the dependent event (output layer). An NN algorithm is a special type of non-linear regression that has multiple local minimum values. Therefore, every time the training algorithm runs, it will converge in a different model. To choose the best model, the training process was repeated 50 times. Only models showing the best discriminative power by logistic regression or ANN were selected for comparison.</p>
				<p>SPSS 26.0 Statistics was used for statistical analysis and ANN modelling. Different models, architecture and activation functions were compared to select the one with the best performance in terms of discrimination to predict the endpoint. </p>
				<p>Categorical variables were expressed as percentages with 95% CI and continuous variables were expressed as means and their respective standard deviation or median and interquartile range (IQR) 25-75, according to their distribution (parametric or non-parametric). </p>
				<p>The discriminative ability of ABPM vs. CRS to predict SE was assessed with the area under the receiver operating characteristic curve (AUC-ROC) analysis. For the comparison of the AUC-ROC, the De Long test was used with the MEDCALC program, version 23.0.9.</p>
				<p>In order to identify the variables with the greatest weight and usefulness in the development of the ANN, a sensitivity analysis was performed to determine their normalized importance in the model.</p>
				<sec>
					<title>Ethical considerations</title>
					<p>The study was reviewed and approved by the institutional and independent Ethics Committee. Due to the observational nature of this analysis, informed consent was not required. According to the Declaration of Helsinki of the World Medical Association, (<xref ref-type="bibr" rid="B64">23</xref>) every precaution was taken to protect the privacy and confidentiality of all personal information. </p>
				</sec>
			</sec>
			<sec sec-type="results">
				<title>RESULTS</title>
				<p>We analyzed data from 491 ABPM studies that included 27 numerical variables from each study: means of 24-hour systolic blood pressure (24-h mSBP), 24-hour diastolic blood pressure (24-h mDBP), 24-hour mean blood pressure (24-h mMBP), 24-hour pulse pressure (24-h mPP), 24-hour heart rate (24-h mHR), daytime systolic blood pressure (day mSBP), daytime diastolic blood pressure (day mDBP), daytime mean blood pressure (day mMBP), daytime pulse pressure (day mPP), daytime heart rate (day mHR), nighttime systolic blood pressure (night mSBP), nighttime diastolic blood pressure (night mDBP), nighttime mean blood pressure (night mMBP), nighttime pulse pressure (night mPP), nighttime heart rate (night mHR); variability of 24-hour systolic blood pressure (24-h SBPsd), 24-hour diastolic blood pressure (24-h DBPsd), 24-hour pulse pressure (24-h PPsd), 24-hour mean blood pressure (24-h MBPsd), 24-hour heart rate (24-h HRsd); daytime SBP load (day SBPL), daytime DBP load (day DBPL), nighttime SBP load (night SBPL), nighttime DBP load (night DBPL); daytime blood pressure ≥135/85 mmHg (day HT), nighttime blood pressure ≥120/70 mmHg (night HT), circadian rhythm with nocturnal SBP and/or DBP fall &lt;10% (<italic>non-dipping</italic> pattern).</p>
				<p>
					<xref ref-type="table" rid="t4">Table 1</xref> details the mean values of each variable from the ABPM studies used for ANN modelling for the combined endpoint. </p>
				<p>
					<table-wrap id="t4">
						<label>Table 1</label>
						<caption>
							<title>Description of the mean values of the ABPM variables.</title>
						</caption>
						<table frame="hsides" rules="groups">
							<colgroup>
								<col/>
								<col/>
							</colgroup>
							<thead>
								<tr style="border: 0; background-color:#ab0534;color:#ffffff;">
									<th align="left"><bold>ABPM variables</bold></th>
									<th align="center">Mean values</th>
								</tr>
							</thead>
							<tbody>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">24-h mSBP</td>
									<td align="center">126.16 ± 11.65 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">24-h mDBP</td>
									<td align="center">79.22 ± 9.30 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">24-h mMBP</td>
									<td align="center">94.87 ± 9.4 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">24-h mHR</td>
									<td align="center">75.21 ± 8.5 bpm</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">24-h mPP</td>
									<td align="center">46.93 ± 7.62 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">24-h SBPsd</td>
									<td align="center">17.92 ± 5.6 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">24-h DBPsd</td>
									<td align="center">14.52 ± 4.2 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">24-h PPsd</td>
									<td align="center">16.9 ± 6.1 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">24-h MBPsd</td>
									<td align="center">13.51 ± 3.8 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">24-h HRsd</td>
									<td align="center">11.59 ± 3.1 bpm</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">day mSBP</td>
									<td align="center">130.13 ± 12.2 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">day mDBP</td>
									<td align="center">82.35 ± 9.9 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">day mMBP</td>
									<td align="center">98.17 ± 10.14 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">day mHR</td>
									<td align="center">78.85 ± 9.1 bpm</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">day mPP</td>
									<td align="center">47.77 ± 8.08 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">night mSBP</td>
									<td align="center">117.34 ± 13.47 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">night mDBP</td>
									<td align="center">72.23 ± 9.90 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">night mPP</td>
									<td align="center">48.10 ± 8.99 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">night mMBP</td>
									<td align="center">87.27 ± 10.50 mmHg</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">night mHR</td>
									<td align="center">67.47 ± 9.2 bpm</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">day SBPL</td>
									<td align="center">34.86%</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">day DBPL</td>
									<td align="center">42.25%</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">night SBPL</td>
									<td align="center">39.47%</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">night DBPL</td>
									<td align="center">49.68%</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">night HT</td>
									<td align="center">58.4% (95% IC 55-72)</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">day HT</td>
									<td align="center">43.9% (95% IC 40-56)</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">Non-dipping</td>
									<td align="center">42.6% (95% IC 39-55)</td>
								</tr>
							</tbody>
						</table>
						<table-wrap-foot>
							<fn id="TFN3">
								<p>24-h DBPsd: 24-hour diastolic blood pressure variability; 24-h HRsd: 24hour heart rate variability; 24-h mMBP: mean of 24-hour mean blood pressure; 24-h MBPsd: 24-hour mean blood pressure variability; 24-h mDBP: 24-hour mean diastolic blood pressure; 24-h mHR: 24-hour mean heart rate; 24-h mPP: 24-hour mean pulse pressure; 24-h mSBP: 24-hour mean systolic blood pressure; 24-h PPsd: 24-hour pulse pressure variability; 24-h SBPsd: 24-hour systolic blood pressure variability; ABPM: ambulatory blood pressure monitoring; day DBPL: daytime DBP load; day HT: daytime blood pressure ≥135/85 mmHg; day mDBP: daytime mean diastolic blood pressure; day mHR: daytime mean heart rate; day mMBP: mean of daytime mean blood pressure: day mPP: daytime mean pulse pressure; day mSBP: daytime mean systolic blood pressure; day SBPL: daytime SBP load; night DBPL: nighttime DBP load; night HT: nighttime blood pressure ≥120/70 mmHg; night mDBP: nighttime mean diastolic blood pressure; night mHR: nighttime mean heart rate; night mMBP: mean of nighttime mean blood pressure: night mPP: nighttime mean pulse pressure; night mSBP: nighttime mean systolic blood pressure; night SBPL: nighttime SBP load; non-dipping: circadian rhythm with nocturnal SBP and/or DBP fall &lt;10%. </p>
							</fn>
						</table-wrap-foot>
					</table-wrap>
				</p>
				<p>The mean age of the population was 64 ± 14 years, 47% were women, 12% had diabetes, 11% were active smokers, 52% had dyslipidemia and the mean body mass index was 26 ± 4 kg/m<sup>2</sup>.</p>
				<p>The median patient follow-up was 6.6 years (IQR 4.5-8). The endpoint incidence at follow-up was 2.6%. <xref ref-type="table" rid="t5">Table 2</xref> details the best ANN models with their neuronal activation functions of the hidden and output layers and their AUC-ROC. </p>
				<p>
					<table-wrap id="t5">
						<label>Table 2</label>
						<caption>
							<title>Neural network models with best performance (multilayer perceptron with one hidden layer, two hidden layers and with radial basis model) with their neural activation functions of the hidden layer and the output layer, the neural architecture, and their area under the ROC curve.</title>
						</caption>
						<table frame="hsides" rules="groups">
							<colgroup>
								<col/>
								<col/>
								<col/>
								<col/>
								<col/>
							</colgroup>
							<thead>
								<tr style="border: 0; background-color:#ab0534;color:#ffffff;">
									<th align="left">Model</th>
									<th align="center">Hidden layer activation function</th>
									<th align="center">Output layer activation function</th>
									<th align="center">Neural architecture</th>
									<th align="center">Area under the ROC curve</th>
								</tr>
							</thead>
							<tbody>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">Multilayer perceptron with a hidden layer</td>
									<td align="center">Hyperbolic tangent</td>
									<td align="center">Softmax</td>
									<td align="center">27/7/2002</td>
									<td align="center">0.81 (95% CI 0.77-0.90)</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">Multilayer perceptron with two hidden layers</td>
									<td align="center">Hyperbolic tangent</td>
									<td align="center">Softmax</td>
									<td align="center">27/20/15/2</td>
									<td align="center">0.75 (95% CI 0.68-0.80)</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">Radial basis</td>
									<td align="center">Softmax</td>
									<td align="center">Identity</td>
									<td align="center">27/6/2002</td>
									<td align="center">0.68 (95% CI 0.61-0.70)</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">De Long’s test</td>
									<td align="center">Multilayer perceptron 1 hidden layer vs; 2 hidden layers</td>
									<td align="center">Multilayer perceptron 1 hidden layer vs; radial basis</td>
									<td align="center" colspan="2">Radial basis vs; multilayer perceptron 2 hidden layers </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">p-value</td>
									<td align="center">0.040</td>
									<td align="center">0.002</td>
									<td align="center" colspan="2">0.001</td>
								</tr>
							</tbody>
						</table>
						<table-wrap-foot>
							<fn id="TFN4">
								<p>95% CI: 95% confidence interval</p>
							</fn>
						</table-wrap-foot>
					</table-wrap>
				</p>
				<p>The best performing models were the Multilayer Perceptron of a hidden layer (activation function of the hidden layer neurons of hyperbolic tangent type) and those of the output layer of <italic>softmax</italic> type with a neural architecture (27/7/2) describing the nodes of each of the layers (<xref ref-type="fig" rid="f4">Figure 1</xref>). <xref ref-type="table" rid="t6">Table 3</xref> describes the synaptic weights with which the neural network is constructed and trained to predict the endpoint. The hyperbolic tangent function was used as the activation function for the input layer and the <italic>softmax</italic> for the hidden layer. The sample was split with a segmentation of 70% training group and 30% validation of the algorithms. The estimated synaptic weights were considered for the development and testing of a multilayer perceptron model for the combined endpoint based on the input of the 27 variables from the ABPM studies.</p>
				<p>
					<fig id="f4">
						<label>Figure 1</label>
						<caption>
							<title>Architecture of the multilayer perceptron type neural network with a hidden layer with 27 neurons in the input layer, 7 neurons in the hidden layer and 2 neurons in the output layer.</title>
						</caption>
						<graphic xlink:href="1850-3748-rac-93-01-33-gf4.jpg"/>
					</fig>
				</p>
				<p>
					<table-wrap id="t6">
						<label>Table 3</label>
						<caption>
							<title>N Description of the synaptic weights with which the neural network is constructed and trained to predict the endpoint.</title>
						</caption>
						<table frame="hsides" rules="groups">
							<colgroup>
								<col/>
								<col/>
								<col span="10"/>
							</colgroup>
							<thead>
								<tr style="border: 0; background-color:#ab0534;color:#ffffff;">
									<th align="left"> </th>
									<th align="left"> </th>
									<th align="center" colspan="10">Predicted </th>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<th align="center"> </th>
									<th align="left"> </th>
									<th align="center" colspan="7">Hidden layer 1 </th>
									<th align="center" colspan="3">Output layer </th>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<th align="left" colspan="2">Predictor+A1:L42 </th>
									<th align="center">O (1:1)</th>
									<th align="center">O (1:2)</th>
									<th align="center">O (1:3) </th>
									<th align="center">O (1:4)</th>
									<th align="center">O (1:5)</th>
									<th align="center">O (1:6)</th>
									<th align="center">O (1:7)</th>
									<th align="center">PF</th>
									<th align="center">[PF = 0]</th>
									<th align="center">[PF = 1]</th>
								</tr>
							</thead>
							<tbody>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">Input layer</td>
									<td align="left">(Bias)</td>
									<td align="center">0.526</td>
									<td align="center">0.579</td>
									<td align="center">-0.457</td>
									<td align="center">-0.125</td>
									<td align="center">-0.394</td>
									<td align="center">-0.600</td>
									<td align="center">0.459</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">non-dipping 0</td>
									<td align="center">0.560</td>
									<td align="center">0.134</td>
									<td align="center">-0.105</td>
									<td align="center">-0.501</td>
									<td align="center">-0.245</td>
									<td align="center">0.258</td>
									<td align="center">0.314</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">non-dipping 1</td>
									<td align="center">0.293</td>
									<td align="center">-0.190</td>
									<td align="center">0.180</td>
									<td align="center">-0.293</td>
									<td align="center">-0.418</td>
									<td align="center">-0.298</td>
									<td align="center">-0.112</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">night HT 0</td>
									<td align="center">-0.018</td>
									<td align="center">-0.055</td>
									<td align="center">0.410</td>
									<td align="center">-0.073</td>
									<td align="center">0.170</td>
									<td align="center">-0.531</td>
									<td align="center">0.356</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">night HT 1</td>
									<td align="center">0.046</td>
									<td align="center">0.100</td>
									<td align="center">-0.270</td>
									<td align="center">0.251</td>
									<td align="center">-0.128</td>
									<td align="center">-0.358</td>
									<td align="center">0.424</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">day HT01 0</td>
									<td align="center">0.209</td>
									<td align="center">-0.108</td>
									<td align="center">-0.165</td>
									<td align="center">-0.299</td>
									<td align="center">0.332</td>
									<td align="center">-0.427</td>
									<td align="center">0.505</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">day HT01 1</td>
									<td align="center">0.459</td>
									<td align="center">-0.275</td>
									<td align="center">0.110</td>
									<td align="center">0.368</td>
									<td align="center">0.169</td>
									<td align="center">-0.669</td>
									<td align="center">0.200</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">24-h mSBP</td>
									<td align="center">0.119</td>
									<td align="center">-0.452</td>
									<td align="center">-0.159</td>
									<td align="center">0.315</td>
									<td align="center">0.005</td>
									<td align="center">-0.141</td>
									<td align="center">0.388</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">24-h mDBP</td>
									<td align="center">0.334</td>
									<td align="center">-0.166</td>
									<td align="center">0.080</td>
									<td align="center">-0.167</td>
									<td align="center">0.241</td>
									<td align="center">0.417</td>
									<td align="center">-0.365</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">24-h mHR</td>
									<td align="center">0.228</td>
									<td align="center">-0.080</td>
									<td align="center">0.435</td>
									<td align="center">-0.323</td>
									<td align="center">0.136</td>
									<td align="center">-0.457</td>
									<td align="center">0.551</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">24-h mMBP</td>
									<td align="center">0.486</td>
									<td align="center">0.475</td>
									<td align="center">0.216</td>
									<td align="center">0.019</td>
									<td align="center">0.036</td>
									<td align="center">-0.313</td>
									<td align="center">0.321</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">24-h mPP</td>
									<td align="center">-0.361</td>
									<td align="center">-0.470</td>
									<td align="center">-0.308</td>
									<td align="center">-0.269</td>
									<td align="center">-0.080</td>
									<td align="center">0.020</td>
									<td align="center">-0.005</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">24-h SBPsd</td>
									<td align="center">-0.179</td>
									<td align="center">-0.346</td>
									<td align="center">-0.027</td>
									<td align="center">0.319</td>
									<td align="center">-0.312</td>
									<td align="center">0.002</td>
									<td align="center">0.427</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">24-h DBPsd</td>
									<td align="center">-0.275</td>
									<td align="center">0.037</td>
									<td align="center">-0.240</td>
									<td align="center">0.050</td>
									<td align="center">-0.378</td>
									<td align="center">0.697</td>
									<td align="center">-0.199</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">24-h HRsd</td>
									<td align="center">-0.184</td>
									<td align="center">0.178</td>
									<td align="center">-0.381</td>
									<td align="center">0.133</td>
									<td align="center">0.228</td>
									<td align="center">-0.230</td>
									<td align="center">-0.109</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">24-h MBPsd</td>
									<td align="center">-0.208</td>
									<td align="center">0.170</td>
									<td align="center">-0.290</td>
									<td align="center">0.268</td>
									<td align="center">-0.269</td>
									<td align="center">0.153</td>
									<td align="center">-0.023</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">24-h PPsd</td>
									<td align="center">0.088</td>
									<td align="center">0.233</td>
									<td align="center">-0.445</td>
									<td align="center">0.141</td>
									<td align="center">-0.167</td>
									<td align="center">0.184</td>
									<td align="center">0.347</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">day mSBP</td>
									<td align="center">-0.089</td>
									<td align="center">-0.398</td>
									<td align="center">0.480</td>
									<td align="center">0.291</td>
									<td align="center">0.325</td>
									<td align="center">0.331</td>
									<td align="center">-0.216</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">day mDBP</td>
									<td align="center">-0.330</td>
									<td align="center">0.307</td>
									<td align="center">-0.107</td>
									<td align="center">0.391</td>
									<td align="center">0.243</td>
									<td align="center">-0.350</td>
									<td align="center">0.110</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">day mHR</td>
									<td align="center">-0.427</td>
									<td align="center">0.294</td>
									<td align="center">-0.401</td>
									<td align="center">-0.089</td>
									<td align="center">-0.330</td>
									<td align="center">-0.263</td>
									<td align="center">0.154</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">day mMBP</td>
									<td align="center">-0.470</td>
									<td align="center">0.010</td>
									<td align="center">0.345</td>
									<td align="center">0.318</td>
									<td align="center">-0.147</td>
									<td align="center">-0.041</td>
									<td align="center">-0.604</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">day mPP</td>
									<td align="center">0.053</td>
									<td align="center">0.222</td>
									<td align="center">0.373</td>
									<td align="center">0.179</td>
									<td align="center">0.133</td>
									<td align="center">-0.169</td>
									<td align="center">-0.001</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">day SBP load</td>
									<td align="center">0.476</td>
									<td align="center">0.488</td>
									<td align="center">-0.056</td>
									<td align="center">-0.249</td>
									<td align="center">0.460</td>
									<td align="center">-0.051</td>
									<td align="center">0.265</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">day DBP load</td>
									<td align="center">0.265</td>
									<td align="center">0.474</td>
									<td align="center">0.331</td>
									<td align="center">-0.092</td>
									<td align="center">0.126</td>
									<td align="center">0.365</td>
									<td align="center">0.372</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">night mSBP</td>
									<td align="center">-0.266</td>
									<td align="center">-0.077</td>
									<td align="center">-0.196</td>
									<td align="center">-0.281</td>
									<td align="center">-0.514</td>
									<td align="center">0.146</td>
									<td align="center">-0.317</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">night mDBP</td>
									<td align="center">0.134</td>
									<td align="center">0.278</td>
									<td align="center">-0.099</td>
									<td align="center">0.426</td>
									<td align="center">-0.488</td>
									<td align="center">0.554</td>
									<td align="center">0.269</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">night mHR</td>
									<td align="center">-0.429</td>
									<td align="center">-0.124</td>
									<td align="center">-0.280</td>
									<td align="center">-0.184</td>
									<td align="center">0.066</td>
									<td align="center">0.307</td>
									<td align="center">-0.257</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">night mMBP</td>
									<td align="center">-0.260</td>
									<td align="center">0.253</td>
									<td align="center">0.435</td>
									<td align="center">0.511</td>
									<td align="center">-0.116</td>
									<td align="center">-0.308</td>
									<td align="center">-0.046</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">night mPP</td>
									<td align="center">0.313</td>
									<td align="center">-0.343</td>
									<td align="center">-0.294</td>
									<td align="center">0.278</td>
									<td align="center">-0.272</td>
									<td align="center">-0.245</td>
									<td align="center">-0.196</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">night SBP load</td>
									<td align="center">-0.376</td>
									<td align="center">-0.294</td>
									<td align="center">0.233</td>
									<td align="center">-0.010</td>
									<td align="center">-0.285</td>
									<td align="center">-0.195</td>
									<td align="center">-0.305</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">night DBP load</td>
									<td align="center">0.189</td>
									<td align="center">-0.134</td>
									<td align="center">-0.062</td>
									<td align="center">-0.336</td>
									<td align="center">0.122</td>
									<td align="center">-0.123</td>
									<td align="center">0.276</td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="left">Hidden layer 1</td>
									<td align="left">(Bias)</td>
									<td align="left"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center">-1.302</td>
									<td align="center">1.206</td>
									<td align="center">-0.093</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">O (1:1)</td>
									<td align="left"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center">-0.348</td>
									<td align="center">0.844</td>
									<td align="center">0.712</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">O (1:2)</td>
									<td align="left"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center">-0.230</td>
									<td align="center">0.008</td>
									<td align="center">0.005</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">O (1:3)</td>
									<td align="left"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center">0.121</td>
									<td align="center">-0.058</td>
									<td align="center">-0.129</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">O (1:4)</td>
									<td align="left"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center">0.158</td>
									<td align="center">-0.023</td>
									<td align="center">-0.129</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">O (1:5)</td>
									<td align="left"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center">-0.031</td>
									<td align="center">0.045</td>
									<td align="center">-0.504</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">O (1:6)</td>
									<td align="left"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center">0.589</td>
									<td align="center">-0.891</td>
									<td align="center">-0.178</td>
								</tr>
								<tr style="border-bottom: 2px solid white; background-color: #e3aea9;">
									<td align="center"> </td>
									<td align="left">O (1:7)</td>
									<td align="left"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center"> </td>
									<td align="center">-0.553</td>
									<td align="center">0.691</td>
									<td align="center">-0.168</td>
								</tr>
							</tbody>
						</table>
						<table-wrap-foot>
							<fn id="TFN5">
								<p>24-h DBPsd: 24-hour diastolic blood pressure variability; 24-h HRsd: 24-hour heart rate variability; 24-h mMBP: mean of 24-hour mean blood pressure; 24-h MBPsd: 24-hour mean blood pressure variability; 24-h mDBP: 24-hour mean diastolic blood pressure; 24-h mHR: 24-hour mean heart rate; 24-h mPP: 24-hour mean pulse pressure; 24-h mSBP: 24-hour mean systolic blood pressure; 24-h PPsd: 24-hour pulse pressure variability; 24-h SBPsd: 24-hour systolic blood pressure variability; day DBPL: daytime DBP load; day HT: daytime blood pressure ≥135/85 mmHg; day mMBP: mean of daytime mean blood pressure; day mDBP: daytime mean diastolic blood pressure; day mHR: daytime mean heart rate; day mPP: daytime mean pulse pressure; day mSBP: daytime mean systolic blood pressure; day SBPL: daytime SBP load; night DBPL: nighttime DBP load; night HT: nighttime blood pressure ≥120/70 mmHg; night mMBP: mean of nighttime mean blood pressure; night mDBP: nighttime mean diastolic blood pressure; night mHR: nighttime mean heart rate; night mPP: nighttime mean pulse pressure; night mSBP: nighttime mean systolic blood pressure; night SBPL: nighttime SBP load; non-dipping: circadian rhythm with nocturnal SBP and/or DBP fall &lt;10%.</p>
							</fn>
						</table-wrap-foot>
					</table-wrap>
				</p>
				<p>The AUC-ROC of the analysis of ABPM variables using neural networks was 0.81 (95% CI 0.77-0.90) compared to clinical risk stratification, 0.67 (95% CI 0.56-0.77), for the combined endpoint; De Long's test p &lt; 0.001 (<xref ref-type="fig" rid="f5">Figure 2</xref>).</p>
				<p>
					<fig id="f5">
						<label>Figure 2;</label>
						<caption>
							<title>Comparison of the areas under the ROC curve; ABPM variables analyzed with artificial neural networks: 0.81 (95% CI 0.77-0.90) versus clinical risk stratification: 0.67 (95% CI 0.56-0.77); De Long's test p &lt; 0;001</title>
						</caption>
						<graphic xlink:href="1850-3748-rac-93-01-33-gf5.jpg"/>
					</fig>
				</p>
				<p>The variables with the greatest weight and usefulness in the development of the neural network or normalized importances in this model were: nighttime mean systolic blood pressure (night mSBP) with a value of 100%, followed by 24-hour mean systolic blood pressure (24-h mSBP), BMI, mean of 24-hour mean blood pressure (24-h mMBP), nighttime mean blood pressure (night mMBP) and 24-hour mean pulse pressure (24-h mPP) (<xref ref-type="fig" rid="f6">Figure 3</xref>).</p>
				<p>
					<fig id="f6">
						<label>Figure 3</label>
						<caption>
							<title>Bar graph representing the variables with the highest normalized importance for the development of the artificial neural network in this model </title>
						</caption>
						<graphic xlink:href="1850-3748-rac-93-01-33-gf6.jpg"/>
					</fig>
				</p>
			</sec>
			<sec sec-type="discussion">
				<title>DISCUSSION</title>
				<p>To our knowledge, this analysis is the first to report the predictive ability of the analysis of variables from ABPM studies using ANN and their superiority compared to clinical risk stratification for long-term events in hypertensive patients.</p>
				<p>The importance of stratifying hypertensive patients according to their estimated risk of a cardiovascular events allows treatment to be adapted to this risk rather than to the office BP levels alone. Among the most widely used risk scales are the Framingham equation, the SCORE (<italic>Systematic Coronary Risk Evaluation</italic>) system, the QRISK, calculators from different United States medical societies and from the World Health Organization, the latter adapted to different geographical areas. In our country, the stratification recommended by the Argentine Consensus on Arterial Hypertension is applied. (<xref ref-type="bibr" rid="B44">3</xref>,<xref ref-type="bibr" rid="B45">4</xref>,<xref ref-type="bibr" rid="B46">5</xref>)</p>
				<p>However, the scores have several limitations that may affect their accuracy and applicability, such as differences between populations according to geographic regions, environmental and socioeconomic characteristics, underestimation of the risk in young patients, qualitative difference in risk factors in many cases, and failure to identify patients with a diagnosis of masked hypertension. Most scores do not consider antihypertensive treatment to reduce o predict risk in the short term. (<xref ref-type="bibr" rid="B56">15</xref>,<xref ref-type="bibr" rid="B57">16</xref>)</p>
				<p>In contrast to the conventional multivariate analysis methods, the trained ANN algorithm is presented as a powerful tool to predict events, as it includes all the variables available in the ABPM studies which avoids bias in the selection of variables. </p>
				<p>Multi-layer Perceptron neural networks are among the most powerful and popular network architectures. They consist of an input layer, an arbitrary number of hidden layers, and an output layer. Each of the hidden or output neurons receives an input from neurons in the previous layer (backward connections), but there are no lateral connections between neurons within each layer. The input layer contains as many neurons as categories corresponding to the independent variables (categorical and continuous). The output layer corresponds to the response variable, which in this case is a categorical variable (combined endpoint).</p>
				<p>Radial basis neural networks are those whose activation functions at the hidden nodes are radially symmetric. A function is said to be radially symmetric if its output depends on the distance between a vector that stores the input data and a vector of synaptic weights, which is called the center or centroid. (<xref ref-type="bibr" rid="B58">17</xref>,<xref ref-type="bibr" rid="B59">18</xref>,<xref ref-type="bibr" rid="B60">19</xref>,<xref ref-type="bibr" rid="B61">20</xref>)</p>
				<p>Previous studies using machine learning have studied its usefulness as a method to screen and early detect hypertension in different populations around the world, while other research groups have used it to predict events in hypertensive patients. (<xref ref-type="bibr" rid="B62">21</xref>,<xref ref-type="bibr" rid="B63">22</xref>,<xref ref-type="bibr" rid="B65">24</xref>,<xref ref-type="bibr" rid="B66">25</xref>,<xref ref-type="bibr" rid="B67">26</xref>)</p>
				<p>Other authors have reported the usefulness of learning machines to optimize decision making in order to improve the treatment of hypertensive patients based on clinical data, and to achieve high accuracy when predicting the individual probability of achieving office blood pressure targets with different treatments. The AUC-ROC were very close to 0.90, indicating high prediction accuracy, and the Kappa coefficients were close to 0.8, showing high levels of agreement between observed and predicted target outcomes. (<xref ref-type="bibr" rid="B68">27</xref>,<xref ref-type="bibr" rid="B69">28</xref>)</p>
				<p>The AUC-ROC shows that the ANNs detect nonlinear relationships between independent and dependent variables beyond the scope of logistic regression. These results support the usefulness of the ANN as a method of analysis in the prediction of serious events in hypertensive patients. (<xref ref-type="bibr" rid="B65">24</xref>,<xref ref-type="bibr" rid="B68">27</xref>)</p>
				<p>ANN-based prediction models prove to be robust and reliable and could be implemented in clinical practice as decision support tools. This method would early identify those hypertensive patients at higher risk of developing serious events and would allow more effective preventive interventions. (<xref ref-type="bibr" rid="B63">22</xref>, <xref ref-type="bibr" rid="B65">24</xref>,<xref ref-type="bibr" rid="B66">25</xref>,<xref ref-type="bibr" rid="B67">26</xref>)</p>
				<p>We highlight that nighttime mean systolic blood pressure (night mSBP) presented an independent normalized significance for the determination of the model. Nocturnal blood pressure has been shown to be an independent risk variable in the prediction of cardiovascular events in hypertensive patients, associated with a higher incidence of stroke, myocardial infarction, heart failure and renal failure. (<xref ref-type="bibr" rid="B70">29</xref>) </p>
				<p>During nighttime hours, blood pressure follows a characteristic circadian pattern, with a physiological decrease. In hypertensive patients, this decrease may be insufficient or even reversed, known as non-dipping or reverse dipping pattern. These abnormal nocturnal blood pressure patterns have been associated with increased cardiovascular risk and events. (<xref ref-type="bibr" rid="B70">29</xref>,<xref ref-type="bibr" rid="B71">30</xref>,<xref ref-type="bibr" rid="B72">31</xref>)</p>
				<p>Different publications have described that increased nighttime mSBP is associated with higher risk of cardiovascular events in hypertensive patients, even after adjusting for other known risk factors. (<xref ref-type="bibr" rid="B71">30</xref>,<xref ref-type="bibr" rid="B72">31</xref>,<xref ref-type="bibr" rid="B73">32</xref>,<xref ref-type="bibr" rid="B74">33</xref>,<xref ref-type="bibr" rid="B75">34</xref>,<xref ref-type="bibr" rid="B76">35</xref>,<xref ref-type="bibr" rid="B77">36</xref>,<xref ref-type="bibr" rid="B78">37</xref>,<xref ref-type="bibr" rid="B79">38</xref>)</p>
				<p>Considering previous observations and the results of this study using the ANN analysis, nighttime mSBP could be considered a useful marker for cardiovascular risk stratification in hypertensive patients. Its evaluation would help to identify patients at higher risk of complications and would allow their reduction by clinical interventions.</p>
				<p>The effective implementation of the current neural network models can be done using software or hardware systems. Thus, the synaptic weights of the proposed architectures can be trained and calculated through a Python algorithm. In turn, this algorithm should be able to automatically read data sets from electronic study records.</p>
				<p>The present study shows that the analysis of ABPM variables using ANNs has a significant predictive value of serious events in hypertensive patients. This suggests the importance of considering ABPM variables for risk stratification in hypertensive patients and shows that the ANNs are an effective tool for the predictive analysis of cardiovascular events with adequate accuracy.</p>
				<sec>
					<title>Limitations</title>
					<p>Although several clinical and ABPM variables were used, the inclusion of other measurements, such as microalbuminuria or peripheral vascular disease, may provide additional information to improve the predictive capacity. Risk stratification was performed using all data available in the electronic medical record. </p>
					<p>In our opinion, the retrospective nature of the analysis is a limitation due to the associated biases that affect the quality of the evidence. The single-center nature of this study limits the transfer of our results to the real world.</p>
					<p>A larger sample size with a prospective analysis, external validations in different cohorts of hypertensive patients and the participation of various healthcare centers would optimize the robustness of this study hypothesis. </p>
					<p>Neural network algorithms have been criticized on several occasions for being &quot;black boxes&quot; with limited ability to identify possible causal relationships. In the present study we identify the most influential factors in the algorithm modelling using standardized importance values.</p>
					<p>The performance quality of learning machines and their algorithms is associated with a larger proportion of events and a higher volume of cases in the total population compared to this study. It should be noted that the incidence of events registered in our work is consistent with that reported in the literature for the outpatient hypertensive population. (<xref ref-type="bibr" rid="B80">39</xref>,<xref ref-type="bibr" rid="B81">40</xref>,<xref ref-type="bibr" rid="B82">41</xref>)</p>
				</sec>
			</sec>
			<sec sec-type="conclusions">
				<title>CONCLUSION</title>
				<p>The novelty of this study is that it was the first to publish the use of ANNs integrating the ABPM variables to predict long-term events compared to currently recommended cardiovascular risk stratification. </p>
				<p>Of additional interest, we observed that nighttime mean SBP was the variable with the greatest weight in the performance of the ANN.</p>
				<p>In our view, this analysis is a generator of a research hypothesis to be evaluated in future multicenter studies using adequate power and real-world representativeness to transfer its results.</p>
				<p>These algorithms can be integrated into the results of each ABPM study to automatically calculate the probability of risk of serious events in hypertensive patients at follow-up, thus supporting physician’s decision making.</p>
			</sec>
		</body>
		<back>
			<ack>
				<title>Acknowledgments</title>
				<p>The authors would like to thank Engineer Roberto Bunge (Director of Artificial Intelligence Engineering, Universidad San Andrés) and Engineer Trinidad Monreal (Research and Development, Artificial Intelligence and Robotics Laboratory, Universidad San Andrés) for their contributions and expertise in artificial intelligence and neural networks.</p>
			</ack>
			<ref-list>
				<title>BIBLIOGRAFÍA</title>
				<ref id="B42">
					<label>1</label>
					<mixed-citation>1 GBD 2017 Causes of Death Collaborators. Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2018;392:1736-88. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/S0140-6736(18)32203-7">https://doi.org/10.1016/S0140-6736(18)32203-7</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<collab>GBD 2017 Causes of Death Collaborators</collab>
						</person-group>
						<article-title>Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980-2017: a systematic analysis for the Global Burden of Disease Study 2017</article-title>
						<source>Lancet</source>
						<year>2018</year>
						<volume>392</volume>
						<fpage>1736</fpage>
						<lpage>1788</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/S0140-6736(18)32203-7">https://doi.org/10.1016/S0140-6736(18)32203-7</ext-link>
					</element-citation>
				</ref>
				<ref id="B43">
					<label>2</label>
					<mixed-citation>2 Mills KT, Bundy JD, Kelly TN, Reed JE, Kearney PM, Reynolds K. Global Disparities of Hypertension Prevalence and Control: A Systematic Analysis of Population-Based Studies From 90 Countries. Circulation 2016;134:441-50. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/CIRCULATIONAHA.115.018912">https://doi.org/10.1161/CIRCULATIONAHA.115.018912</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Mills</surname>
								<given-names>KT</given-names>
							</name>
							<name>
								<surname>Bundy</surname>
								<given-names>JD</given-names>
							</name>
							<name>
								<surname>Kelly</surname>
								<given-names>TN</given-names>
							</name>
							<name>
								<surname>Reed</surname>
								<given-names>JE</given-names>
							</name>
							<name>
								<surname>Kearney</surname>
								<given-names>PM</given-names>
							</name>
							<name>
								<surname>Reynolds</surname>
								<given-names>K</given-names>
							</name>
						</person-group>
						<article-title>Global Disparities of Hypertension Prevalence and Control: A Systematic Analysis of Population-Based Studies From 90 Countries</article-title>
						<source>Circulation</source>
						<year>2016</year>
						<volume>134</volume>
						<fpage>441</fpage>
						<lpage>450</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/CIRCULATIONAHA.115.018912">https://doi.org/10.1161/CIRCULATIONAHA.115.018912</ext-link>
					</element-citation>
				</ref>
				<ref id="B44">
					<label>3</label>
					<mixed-citation>3 Williams B, Mancia G, Spiering W, Agabiti Rosei E, Azizi M, Burnier MAuthors/Task Force Members:. 2018 ESC/ESH Guidelines for the management of arterial hypertension: The Task Force for the management of arterial hypertension of the European Society of Cardiology and the European Society of Hypertension: The Task Force for the management of arterial hypertension of the European Society of Cardiology and the European Society of Hypertension. J Hypertens 2018;36:1953-2041. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000001940">https://doi.org/10.1097/HJH.0000000000001940</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Williams</surname>
								<given-names>B</given-names>
							</name>
							<name>
								<surname>Mancia</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>Spiering</surname>
								<given-names>W</given-names>
							</name>
							<name>
								<surname>Agabiti Rosei</surname>
								<given-names>E</given-names>
							</name>
							<name>
								<surname>Azizi</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>Burnier</surname>
								<given-names>M</given-names>
							</name>
						</person-group>
						<comment>Authors/Task Force Members:</comment>
						<article-title>2018 ESC/ESH Guidelines for the management of arterial hypertension: The Task Force for the management of arterial hypertension of the European Society of Cardiology and the European Society of Hypertension: The Task Force for the management of arterial hypertension of the European Society of Cardiology and the European Society of Hypertension</article-title>
						<source>J Hypertens</source>
						<year>2018</year>
						<volume>36</volume>
						<fpage>1953</fpage>
						<lpage>2041</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000001940">https://doi.org/10.1097/HJH.0000000000001940</ext-link>
					</element-citation>
				</ref>
				<ref id="B45">
					<label>4</label>
					<mixed-citation>4 Whelton PK, Carey RM, Aronow WS, Casey DEJr, Collins KJ, Dennison Himmelfarb C. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation 2017;138:e484-e594. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/CIR.0000000000000596">https://doi.org/10.1161/CIR.0000000000000596</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Whelton</surname>
								<given-names>PK</given-names>
							</name>
							<name>
								<surname>Carey</surname>
								<given-names>RM</given-names>
							</name>
							<name>
								<surname>Aronow</surname>
								<given-names>WS</given-names>
							</name>
							<name>
								<surname>Casey</surname>
								<given-names>DE</given-names>
								<suffix>Jr</suffix>
							</name>
							<name>
								<surname>Collins</surname>
								<given-names>KJ</given-names>
							</name>
							<name>
								<surname>Dennison Himmelfarb</surname>
								<given-names>C</given-names>
							</name>
						</person-group>
						<article-title>2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines</article-title>
						<source>Circulation</source>
						<year>2017</year>
						<volume>138</volume>
						<fpage>e484</fpage>
						<lpage>e594</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/CIR.0000000000000596">https://doi.org/10.1161/CIR.0000000000000596</ext-link>
					</element-citation>
				</ref>
				<ref id="B46">
					<label>5</label>
					<mixed-citation>5 Marín M, Bendersky M, Páez O, Obregón S, Salazar M, Rodríguez Py cols. Argentine consensus on arterial hypertension. Argentine Society of Cardiology. Rev Argent Cardiol 2018;86s21-53</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Marín</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>Bendersky</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>Páez</surname>
								<given-names>O</given-names>
							</name>
							<name>
								<surname>Obregón</surname>
								<given-names>S</given-names>
							</name>
							<name>
								<surname>Salazar</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>Rodríguez</surname>
								<given-names>P</given-names>
							</name>
						</person-group>
						<article-title>Argentine consensus on arterial hypertension. Argentine Society of Cardiology</article-title>
						<source>Rev Argent Cardiol</source>
						<year>2018</year>
						<volume>86</volume>
						<supplement>s2</supplement>
						<fpage>1</fpage>
						<lpage>53</lpage>
					</element-citation>
				</ref>
				<ref id="B47">
					<label>6</label>
					<mixed-citation>6 O'Brien E, Asmar R, Beilin L, Imai Y, Mallion JM, Mancia GEuropean Society of Hypertension Working Group on Blood Pressure Monitoring. European Society of Hypertension recommendations for conventional, ambulatory and home blood pressure measurement. J Hypertens 2003;21:821-48. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/00004872-200305000-00001">https://doi.org/10.1097/00004872-200305000-00001</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>O'Brien</surname>
								<given-names>E</given-names>
							</name>
							<name>
								<surname>Asmar</surname>
								<given-names>R</given-names>
							</name>
							<name>
								<surname>Beilin</surname>
								<given-names>L</given-names>
							</name>
							<name>
								<surname>Imai</surname>
								<given-names>Y</given-names>
							</name>
							<name>
								<surname>Mallion</surname>
								<given-names>JM</given-names>
							</name>
							<name>
								<surname>Mancia</surname>
								<given-names>G</given-names>
							</name>
						</person-group>
						<comment>European Society of Hypertension Working Group on Blood Pressure Monitoring</comment>
						<article-title>European Society of Hypertension recommendations for conventional, ambulatory and home blood pressure measurement</article-title>
						<source>J Hypertens</source>
						<year>2003</year>
						<volume>21</volume>
						<fpage>821</fpage>
						<lpage>848</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/00004872-200305000-00001">https://doi.org/10.1097/00004872-200305000-00001</ext-link>
					</element-citation>
				</ref>
				<ref id="B48">
					<label>7</label>
					<mixed-citation>7 Parati G, Stergiou GS, Asmar R, Bilo G, de Leeuw P, Imai YESH Working Group on Blood Pressure Monitoring. European Society of Hypertension practice guidelines for home blood pressure monitoring. J Hum Hypertens 2010;24:779-85. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1038/jhh.2010.54">https://doi.org/10.1038/jhh.2010.54</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Parati</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>Stergiou</surname>
								<given-names>GS</given-names>
							</name>
							<name>
								<surname>Asmar</surname>
								<given-names>R</given-names>
							</name>
							<name>
								<surname>Bilo</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>de Leeuw</surname>
								<given-names>P</given-names>
							</name>
							<name>
								<surname>Imai</surname>
								<given-names>Y</given-names>
							</name>
						</person-group>
						<comment>ESH Working Group on Blood Pressure Monitoring</comment>
						<article-title>European Society of Hypertension practice guidelines for home blood pressure monitoring</article-title>
						<source>J Hum Hypertens</source>
						<year>2010</year>
						<volume>24</volume>
						<fpage>779</fpage>
						<lpage>785</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1038/jhh.2010.54">https://doi.org/10.1038/jhh.2010.54</ext-link>
					</element-citation>
				</ref>
				<ref id="B49">
					<label>8</label>
					<mixed-citation>8 Sega R, Facchetti R, Bombelli M, Cesana G, Corrao G, Grassi G. Prognostic value of ambulatory and home blood pressures compared with office blood pressure in the general population: follow-up results from the Pressioni Arteriose Monitorate e Loro Associazioni (PAMELA) study. Circulation 2005;111:1777-83. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/01.CIR.0000160923.04524.5B">https://doi.org/10.1161/01.CIR.0000160923.04524.5B</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Sega</surname>
								<given-names>R</given-names>
							</name>
							<name>
								<surname>Facchetti</surname>
								<given-names>R</given-names>
							</name>
							<name>
								<surname>Bombelli</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>Cesana</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>Corrao</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>Grassi</surname>
								<given-names>G</given-names>
							</name>
						</person-group>
						<article-title>Prognostic value of ambulatory and home blood pressures compared with office blood pressure in the general population: follow-up results from the Pressioni Arteriose Monitorate e Loro Associazioni (PAMELA) study</article-title>
						<source>Circulation</source>
						<year>2005</year>
						<volume>111</volume>
						<fpage>1777</fpage>
						<lpage>1783</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/01.CIR.0000160923.04524.5B">https://doi.org/10.1161/01.CIR.0000160923.04524.5B</ext-link>
					</element-citation>
				</ref>
				<ref id="B50">
					<label>9</label>
					<mixed-citation>9 Kikuya M, Ohkubo T, Asayama K, Metoki H, Obara T, Saito S. Ambulatory blood pressure and 10-year risk of cardiovascular and noncardiovascular mortality: the Ohasama study. Hypertension 2005;45:240-5. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/01.HYP.0000152079.04553.2c">https://doi.org/10.1161/01.HYP.0000152079.04553.2c</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Kikuya</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>Ohkubo</surname>
								<given-names>T</given-names>
							</name>
							<name>
								<surname>Asayama</surname>
								<given-names>K</given-names>
							</name>
							<name>
								<surname>Metoki</surname>
								<given-names>H</given-names>
							</name>
							<name>
								<surname>Obara</surname>
								<given-names>T</given-names>
							</name>
							<name>
								<surname>Saito</surname>
								<given-names>S</given-names>
							</name>
						</person-group>
						<article-title>Ambulatory blood pressure and 10-year risk of cardiovascular and noncardiovascular mortality: the Ohasama study</article-title>
						<source>Hypertension</source>
						<year>2005</year>
						<volume>45</volume>
						<fpage>240</fpage>
						<lpage>245</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/01.HYP.0000152079.04553.2c">https://doi.org/10.1161/01.HYP.0000152079.04553.2c</ext-link>
					</element-citation>
				</ref>
				<ref id="B51">
					<label>10</label>
					<mixed-citation>10 Banegas JR, Ruilope LM, de la Sierra A, Vinyoles E, Gorostidi M, de la Cruz JJ. Relationship between Clinic and Ambulatory Blood-Pressure Measurements and Mortality. N Engl J Med 2018;378:1509-20. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1056/NEJMoa1712231">https://doi.org/10.1056/NEJMoa1712231</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Banegas</surname>
								<given-names>JR</given-names>
							</name>
							<name>
								<surname>Ruilope</surname>
								<given-names>LM</given-names>
							</name>
							<name>
								<surname>de la Sierra</surname>
								<given-names>A</given-names>
							</name>
							<name>
								<surname>Vinyoles</surname>
								<given-names>E</given-names>
							</name>
							<name>
								<surname>Gorostidi</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>de la Cruz</surname>
								<given-names>JJ</given-names>
							</name>
						</person-group>
						<article-title>Relationship between Clinic and Ambulatory Blood-Pressure Measurements and Mortality</article-title>
						<source>N Engl J Med</source>
						<year>2018</year>
						<volume>378</volume>
						<fpage>1509</fpage>
						<lpage>1520</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1056/NEJMoa1712231">https://doi.org/10.1056/NEJMoa1712231</ext-link>
					</element-citation>
				</ref>
				<ref id="B52">
					<label>11</label>
					<mixed-citation>11 Parikh NI, Pencina MJ, Wang TJ, Benjamin EJ, Lanier KJ, Levy D. A risk score for predicting near-term incidence of hypertension: the Framingham Heart Study. Ann Intern Med 2008;148:102-10. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.7326/0003-4819-148-2-200801150-00005">https://doi.org/10.7326/0003-4819-148-2-200801150-00005</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Parikh</surname>
								<given-names>NI</given-names>
							</name>
							<name>
								<surname>Pencina</surname>
								<given-names>MJ</given-names>
							</name>
							<name>
								<surname>Wang</surname>
								<given-names>TJ</given-names>
							</name>
							<name>
								<surname>Benjamin</surname>
								<given-names>EJ</given-names>
							</name>
							<name>
								<surname>Lanier</surname>
								<given-names>KJ</given-names>
							</name>
							<name>
								<surname>Levy</surname>
								<given-names>D</given-names>
							</name>
						</person-group>
						<article-title>A risk score for predicting near-term incidence of hypertension: the Framingham Heart Study</article-title>
						<source>Ann Intern Med</source>
						<year>2008</year>
						<volume>148</volume>
						<fpage>102</fpage>
						<lpage>110</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.7326/0003-4819-148-2-200801150-00005">https://doi.org/10.7326/0003-4819-148-2-200801150-00005</ext-link>
					</element-citation>
				</ref>
				<ref id="B53">
					<label>12</label>
					<mixed-citation>12 Conroy RM, Pyörälä K, Fitzgerald AP, Sans S, Menotti A, De Backer GSCORE project group. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J 2003;24:987-1003. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/S0195-668X(03)00114-3">https://doi.org/10.1016/S0195-668X(03)00114-3</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Conroy</surname>
								<given-names>RM</given-names>
							</name>
							<name>
								<surname>Pyörälä</surname>
								<given-names>K</given-names>
							</name>
							<name>
								<surname>Fitzgerald</surname>
								<given-names>AP</given-names>
							</name>
							<name>
								<surname>Sans</surname>
								<given-names>S</given-names>
							</name>
							<name>
								<surname>Menotti</surname>
								<given-names>A</given-names>
							</name>
							<name>
								<surname>De Backer</surname>
								<given-names>G</given-names>
							</name>
						</person-group>
						<comment>SCORE project group</comment>
						<article-title>Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project</article-title>
						<source>Eur Heart J</source>
						<year>2003</year>
						<volume>24</volume>
						<fpage>987</fpage>
						<lpage>1003</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/S0195-668X(03)00114-3">https://doi.org/10.1016/S0195-668X(03)00114-3</ext-link>
					</element-citation>
				</ref>
				<ref id="B54">
					<label>13</label>
					<mixed-citation>13 Itchhaporia D, Snow PB, Almassy RJ, Oetgen WJ. Artificial neural networks: current status in cardiovascular medicine. J Am Coll Cardiol 1996;28:515-21. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/0735-1097(96)00174-X">https://doi.org/10.1016/0735-1097(96)00174-X</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Itchhaporia</surname>
								<given-names>D</given-names>
							</name>
							<name>
								<surname>Snow</surname>
								<given-names>PB</given-names>
							</name>
							<name>
								<surname>Almassy</surname>
								<given-names>RJ</given-names>
							</name>
							<name>
								<surname>Oetgen</surname>
								<given-names>WJ</given-names>
							</name>
						</person-group>
						<article-title>Artificial neural networks: current status in cardiovascular medicine</article-title>
						<source>J Am Coll Cardiol</source>
						<year>1996</year>
						<volume>28</volume>
						<fpage>515</fpage>
						<lpage>521</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/0735-1097(96)00174-X">https://doi.org/10.1016/0735-1097(96)00174-X</ext-link>
					</element-citation>
				</ref>
				<ref id="B55">
					<label>14</label>
					<mixed-citation>14 Ning G, Su J, Li Y, Wang X, Li C, Yan W. Artificial neural network based model for cardiovascular risk stratification in hypertension. Med Biol Eng Comput 2006;44:202-8. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1007/s11517-006-0028-2">https://doi.org/10.1007/s11517-006-0028-2</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Ning</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>Su</surname>
								<given-names>J</given-names>
							</name>
							<name>
								<surname>Li</surname>
								<given-names>Y</given-names>
							</name>
							<name>
								<surname>Wang</surname>
								<given-names>X</given-names>
							</name>
							<name>
								<surname>Li</surname>
								<given-names>C</given-names>
							</name>
							<name>
								<surname>Yan</surname>
								<given-names>W</given-names>
							</name>
						</person-group>
						<article-title>Artificial neural network based model for cardiovascular risk stratification in hypertension</article-title>
						<source>Med Biol Eng Comput</source>
						<year>2006</year>
						<volume>44</volume>
						<fpage>202</fpage>
						<lpage>208</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1007/s11517-006-0028-2">https://doi.org/10.1007/s11517-006-0028-2</ext-link>
					</element-citation>
				</ref>
				<ref id="B56">
					<label>15</label>
					<mixed-citation>15 Cui J. Overview of risk prediction models in cardiovascular disease research. Ann Epidemiol 2009;19:711-7. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.annepidem.2009.05.005">https://doi.org/10.1016/j.annepidem.2009.05.005</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Cui</surname>
								<given-names>J</given-names>
							</name>
						</person-group>
						<article-title>Overview of risk prediction models in cardiovascular disease research</article-title>
						<source>Ann Epidemiol</source>
						<year>2009</year>
						<volume>19</volume>
						<fpage>711</fpage>
						<lpage>717</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.annepidem.2009.05.005">https://doi.org/10.1016/j.annepidem.2009.05.005</ext-link>
					</element-citation>
				</ref>
				<ref id="B57">
					<label>16</label>
					<mixed-citation>16 Elosua R. Cardiovascular risk functions: usefulness and limitations. Rev Esp Cardiol (Engl Ed) 2014;67:77-9. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.rec.2013.09.012">https://doi.org/10.1016/j.rec.2013.09.012</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Elosua</surname>
								<given-names>R</given-names>
							</name>
						</person-group>
						<article-title>Cardiovascular risk functions: usefulness and limitations</article-title>
						<source>Rev Esp Cardiol (Engl Ed)</source>
						<year>2014</year>
						<volume>67</volume>
						<fpage>77</fpage>
						<lpage>79</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.rec.2013.09.012">https://doi.org/10.1016/j.rec.2013.09.012</ext-link>
					</element-citation>
				</ref>
				<ref id="B58">
					<label>17</label>
					<mixed-citation>17 LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015;521:436-44. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1038/nature14539">https://doi.org/10.1038/nature14539</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>LeCun</surname>
								<given-names>Y</given-names>
							</name>
							<name>
								<surname>Bengio</surname>
								<given-names>Y</given-names>
							</name>
							<name>
								<surname>Hinton</surname>
								<given-names>G</given-names>
							</name>
						</person-group>
						<article-title>Deep learning</article-title>
						<source>Nature</source>
						<year>2015</year>
						<volume>521</volume>
						<fpage>436</fpage>
						<lpage>444</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1038/nature14539">https://doi.org/10.1038/nature14539</ext-link>
					</element-citation>
				</ref>
				<ref id="B59">
					<label>18</label>
					<mixed-citation>18 Sargent DJ. Comparison of artificial neural networks with other statistical approaches: results from medical data sets. Cancer 2001;91:1636-42. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1002/1097-0142(20010415)91:8+&lt;1636::AID-CNCR1176&gt;3.0.CO;2-D">https://doi.org/10.1002/1097-0142(20010415)91:8+&lt;1636::AID-CNCR1176&gt;3.0.CO;2-D</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Sargent</surname>
								<given-names>DJ</given-names>
							</name>
						</person-group>
						<article-title>Comparison of artificial neural networks with other statistical approaches: results from medical data sets</article-title>
						<source>Cancer</source>
						<year>2001</year>
						<volume>91</volume>
						<fpage>1636</fpage>
						<lpage>1642</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1002/1097-0142(20010415)91:8+&lt;1636::AID-CNCR1176&gt;3.0.CO;2-D">https://doi.org/10.1002/1097-0142(20010415)91:8+&lt;1636::AID-CNCR1176&gt;3.0.CO;2-D</ext-link>
					</element-citation>
				</ref>
				<ref id="B60">
					<label>19</label>
					<mixed-citation>19 Trujillano J, March J, Sorribas A. Aproximación metodológica al uso de redes neuronales artificiales para la predicción de resultados en medicina. Med Clin (Barc) 2004;122:59-67. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1157/13057536">https://doi.org/10.1157/13057536</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Trujillano</surname>
								<given-names>J</given-names>
							</name>
							<name>
								<surname>March</surname>
								<given-names>J</given-names>
							</name>
							<name>
								<surname>Sorribas</surname>
								<given-names>A</given-names>
							</name>
						</person-group>
						<article-title>Aproximación metodológica al uso de redes neuronales artificiales para la predicción de resultados en medicina</article-title>
						<source>Med Clin (Barc)</source>
						<year>2004</year>
						<volume>122</volume>
						<fpage>59</fpage>
						<lpage>67</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1157/13057536">https://doi.org/10.1157/13057536</ext-link>
					</element-citation>
				</ref>
				<ref id="B61">
					<label>20</label>
					<mixed-citation>20 Ye X, Zeng QT, Facelli JC, Brixner DI, Conway M, Bray BE. Predicting Optimal Hypertension Treatment Pathways Using Recurrent Neural Networks. Int J Med Inform 2020;139:104122. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.ijmedinf.2020.104122">https://doi.org/10.1016/j.ijmedinf.2020.104122</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Ye</surname>
								<given-names>X</given-names>
							</name>
							<name>
								<surname>Zeng</surname>
								<given-names>QT</given-names>
							</name>
							<name>
								<surname>Facelli</surname>
								<given-names>JC</given-names>
							</name>
							<name>
								<surname>Brixner</surname>
								<given-names>DI</given-names>
							</name>
							<name>
								<surname>Conway</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>Bray</surname>
								<given-names>BE</given-names>
							</name>
						</person-group>
						<article-title>Predicting Optimal Hypertension Treatment Pathways Using Recurrent Neural Networks</article-title>
						<source>Int J Med Inform</source>
						<year>2020</year>
						<volume>139</volume>
						<fpage>104122</fpage>
						<lpage>104122</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.ijmedinf.2020.104122">https://doi.org/10.1016/j.ijmedinf.2020.104122</ext-link>
					</element-citation>
				</ref>
				<ref id="B62">
					<label>21</label>
					<mixed-citation>21 Yan Y, Chen R, Yang Z, Ma Y, Huang J, Luo L. Application of back propagation neural network model optimized by particle swarm algorithm in predicting the risk of hypertension. J Clin Hypertens (Greenwich) 2022;24:1606-17. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1111/jch.14597">https://doi.org/10.1111/jch.14597</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Yan</surname>
								<given-names>Y</given-names>
							</name>
							<name>
								<surname>Chen</surname>
								<given-names>R</given-names>
							</name>
							<name>
								<surname>Yang</surname>
								<given-names>Z</given-names>
							</name>
							<name>
								<surname>Ma</surname>
								<given-names>Y</given-names>
							</name>
							<name>
								<surname>Huang</surname>
								<given-names>J</given-names>
							</name>
							<name>
								<surname>Luo</surname>
								<given-names>L</given-names>
							</name>
						</person-group>
						<article-title>Application of back propagation neural network model optimized by particle swarm algorithm in predicting the risk of hypertension</article-title>
						<source>J Clin Hypertens (Greenwich)</source>
						<year>2022</year>
						<volume>24</volume>
						<fpage>1606</fpage>
						<lpage>1617</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1111/jch.14597">https://doi.org/10.1111/jch.14597</ext-link>
					</element-citation>
				</ref>
				<ref id="B63">
					<label>22</label>
					<mixed-citation>22 Huang S, Xu Y, Yue L, Wei S, Liu L, Gan X. Evaluating the risk of hypertension using an artificial neural network method in rural residents over the age of 35 years in a Chinese area. Hypertens Res 2010;33:722-6. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1038/hr.2010.73">https://doi.org/10.1038/hr.2010.73</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Huang</surname>
								<given-names>S</given-names>
							</name>
							<name>
								<surname>Xu</surname>
								<given-names>Y</given-names>
							</name>
							<name>
								<surname>Yue</surname>
								<given-names>L</given-names>
							</name>
							<name>
								<surname>Wei</surname>
								<given-names>S</given-names>
							</name>
							<name>
								<surname>Liu</surname>
								<given-names>L</given-names>
							</name>
							<name>
								<surname>Gan</surname>
								<given-names>X</given-names>
							</name>
						</person-group>
						<article-title>Evaluating the risk of hypertension using an artificial neural network method in rural residents over the age of 35 years in a Chinese area</article-title>
						<source>Hypertens Res</source>
						<year>2010</year>
						<volume>33</volume>
						<fpage>722</fpage>
						<lpage>726</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1038/hr.2010.73">https://doi.org/10.1038/hr.2010.73</ext-link>
					</element-citation>
				</ref>
				<ref id="B64">
					<label>23</label>
					<mixed-citation>23 World Medical Association. World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects. JAMA 2013;310:2191-4. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1001/jama.2013.281053">https://doi.org/10.1001/jama.2013.281053</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<collab>World Medical Association</collab>
						</person-group>
						<article-title>World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects</article-title>
						<source>JAMA</source>
						<year>2013</year>
						<volume>310</volume>
						<fpage>2191</fpage>
						<lpage>2194</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1001/jama.2013.281053">https://doi.org/10.1001/jama.2013.281053</ext-link>
					</element-citation>
				</ref>
				<ref id="B65">
					<label>24</label>
					<mixed-citation>24 Islam MM, Rahman MJ, Chandra Roy D, Tawabunnahar M, Jahan R, Ahmed NAMF. Machine learning algorithm for characterizing risks of hypertension, at an early stage in Bangladesh. Diabetes Metab Syndr 2021;15:877-84. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.dsx.2021.03.035">https://doi.org/10.1016/j.dsx.2021.03.035</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Islam</surname>
								<given-names>MM</given-names>
							</name>
							<name>
								<surname>Rahman</surname>
								<given-names>MJ</given-names>
							</name>
							<name>
								<surname>Chandra Roy</surname>
								<given-names>D</given-names>
							</name>
							<name>
								<surname>Tawabunnahar</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>Jahan</surname>
								<given-names>R</given-names>
							</name>
							<name>
								<surname>Ahmed</surname>
								<given-names>NAMF</given-names>
							</name>
						</person-group>
						<article-title>Machine learning algorithm for characterizing risks of hypertension, at an early stage in Bangladesh</article-title>
						<source>Diabetes Metab Syndr</source>
						<year>2021</year>
						<volume>15</volume>
						<fpage>877</fpage>
						<lpage>884</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.dsx.2021.03.035">https://doi.org/10.1016/j.dsx.2021.03.035</ext-link>
					</element-citation>
				</ref>
				<ref id="B66">
					<label>25</label>
					<mixed-citation>25 Leoncini G, Sacchi G, Viazzi F, Ravera M, Parodi D, Ratto E. Microalbuminuria identifies overall cardiovascular risk in essential hypertension: an artificial neural network-based approach. J Hypertens 2002;20:1315-21. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/00004872-200207000-00018">https://doi.org/10.1097/00004872-200207000-00018</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Leoncini</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>Sacchi</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>Viazzi</surname>
								<given-names>F</given-names>
							</name>
							<name>
								<surname>Ravera</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>Parodi</surname>
								<given-names>D</given-names>
							</name>
							<name>
								<surname>Ratto</surname>
								<given-names>E</given-names>
							</name>
						</person-group>
						<article-title>Microalbuminuria identifies overall cardiovascular risk in essential hypertension: an artificial neural network-based approach</article-title>
						<source>J Hypertens</source>
						<year>2002</year>
						<volume>20</volume>
						<fpage>1315</fpage>
						<lpage>1321</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/00004872-200207000-00018">https://doi.org/10.1097/00004872-200207000-00018</ext-link>
					</element-citation>
				</ref>
				<ref id="B67">
					<label>26</label>
					<mixed-citation>26 Viazzi F, Leoncini G, Sacchi G, Parodi D, Ratto E, Falqui V. Predicting cardiovascular risk using creatinine clearance and an artificial neural network in primary hypertension. J Hypertens 2006;24:1281-6. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/01.hjh.0000234107.08368.e5">https://doi.org/10.1097/01.hjh.0000234107.08368.e5</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Viazzi</surname>
								<given-names>F</given-names>
							</name>
							<name>
								<surname>Leoncini</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>Sacchi</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>Parodi</surname>
								<given-names>D</given-names>
							</name>
							<name>
								<surname>Ratto</surname>
								<given-names>E</given-names>
							</name>
							<name>
								<surname>Falqui</surname>
								<given-names>V</given-names>
							</name>
						</person-group>
						<article-title>Predicting cardiovascular risk using creatinine clearance and an artificial neural network in primary hypertension</article-title>
						<source>J Hypertens</source>
						<year>2006</year>
						<volume>24</volume>
						<fpage>1281</fpage>
						<lpage>1286</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/01.hjh.0000234107.08368.e5">https://doi.org/10.1097/01.hjh.0000234107.08368.e5</ext-link>
					</element-citation>
				</ref>
				<ref id="B68">
					<label>27</label>
					<mixed-citation>27 Mandrekar JN. Receiver operating characteristic curve in diagnostic test assessment. J Thorac Oncol 2010;5:1315-6. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/JTO.0b013e3181ec173d">https://doi.org/10.1097/JTO.0b013e3181ec173d</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Mandrekar</surname>
								<given-names>JN</given-names>
							</name>
						</person-group>
						<article-title>Receiver operating characteristic curve in diagnostic test assessment</article-title>
						<source>J Thorac Oncol</source>
						<year>2010</year>
						<volume>5</volume>
						<fpage>1315</fpage>
						<lpage>1316</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/JTO.0b013e3181ec173d">https://doi.org/10.1097/JTO.0b013e3181ec173d</ext-link>
					</element-citation>
				</ref>
				<ref id="B69">
					<label>28</label>
					<mixed-citation>28 McHugh ML. Interrater reliability: the kappa statistic. Biochem Med (Zagreb) 2012;22:276-82. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.11613/BM.2012.031">https://doi.org/10.11613/BM.2012.031</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>McHugh</surname>
								<given-names>ML</given-names>
							</name>
						</person-group>
						<article-title>Interrater reliability: the kappa statistic</article-title>
						<source>Biochem Med (Zagreb)</source>
						<year>2012</year>
						<volume>22</volume>
						<fpage>276</fpage>
						<lpage>282</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.11613/BM.2012.031">https://doi.org/10.11613/BM.2012.031</ext-link>
					</element-citation>
				</ref>
				<ref id="B70">
					<label>29</label>
					<mixed-citation>29 Kario K. Home Blood Pressure Monitoring: Current Status and New Developments. Am J Hypertens 2021;34:783-794. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1093/ajh/hpab017">https://doi.org/10.1093/ajh/hpab017</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Kario</surname>
								<given-names>K</given-names>
							</name>
						</person-group>
						<article-title>Home Blood Pressure Monitoring: Current Status and New Developments</article-title>
						<source>Am J Hypertens</source>
						<year>2021</year>
						<volume>34</volume>
						<fpage>783</fpage>
						<lpage>794</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1093/ajh/hpab017">https://doi.org/10.1093/ajh/hpab017</ext-link>
					</element-citation>
				</ref>
				<ref id="B71">
					<label>30</label>
					<mixed-citation>30 Tadic M, Cuspidi C, Celic V, Petrovic O, Pencic B, Mancia G. The prognostic importance of right ventricular remodeling and the circadian blood pressure pattern on the long-term cardiovascular outcome. J Hypertens 2020;38:1525-30. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000002432">https://doi.org/10.1097/HJH.0000000000002432</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Tadic</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>Cuspidi</surname>
								<given-names>C</given-names>
							</name>
							<name>
								<surname>Celic</surname>
								<given-names>V</given-names>
							</name>
							<name>
								<surname>Petrovic</surname>
								<given-names>O</given-names>
							</name>
							<name>
								<surname>Pencic</surname>
								<given-names>B</given-names>
							</name>
							<name>
								<surname>Mancia</surname>
								<given-names>G</given-names>
							</name>
						</person-group>
						<article-title>The prognostic importance of right ventricular remodeling and the circadian blood pressure pattern on the long-term cardiovascular outcome</article-title>
						<source>J Hypertens</source>
						<year>2020</year>
						<volume>38</volume>
						<fpage>1525</fpage>
						<lpage>1530</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000002432">https://doi.org/10.1097/HJH.0000000000002432</ext-link>
					</element-citation>
				</ref>
				<ref id="B72">
					<label>31</label>
					<mixed-citation>31 Kario K, Williams B. Nocturnal Hypertension and Heart Failure: Mechanisms, Evidence, and New Treatments. Hypertension 2021;78:564-77. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.121.17440">https://doi.org/10.1161/HYPERTENSIONAHA.121.17440</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Kario</surname>
								<given-names>K</given-names>
							</name>
							<name>
								<surname>Williams</surname>
								<given-names>B</given-names>
							</name>
						</person-group>
						<article-title>Nocturnal Hypertension and Heart Failure: Mechanisms, Evidence, and New Treatments</article-title>
						<source>Hypertension</source>
						<year>2021</year>
						<volume>78</volume>
						<fpage>564</fpage>
						<lpage>577</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.121.17440">https://doi.org/10.1161/HYPERTENSIONAHA.121.17440</ext-link>
					</element-citation>
				</ref>
				<ref id="B73">
					<label>32</label>
					<mixed-citation>32 Cardoso CRL, Salles GF. Associations of the nocturnal blood pressure fall and morning surge with cardiovascular events and mortality in individuals with resistant hypertension. J Hypertens 2021;39:1177-8710.1097/HJH.0000000000002775</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Cardoso</surname>
								<given-names>CRL</given-names>
							</name>
							<name>
								<surname>Salles</surname>
								<given-names>GF</given-names>
							</name>
						</person-group>
						<article-title>Associations of the nocturnal blood pressure fall and morning surge with cardiovascular events and mortality in individuals with resistant hypertension</article-title>
						<source>J Hypertens</source>
						<year>2021</year>
						<volume>39</volume>
						<fpage>1177</fpage>
						<lpage>1187</lpage>
						<pub-id pub-id-type="doi">10.1097/HJH.0000000000002775</pub-id>
					</element-citation>
				</ref>
				<ref id="B74">
					<label>33</label>
					<mixed-citation>33 Fagard RH, Celis H, Thijs L, Staessen JA, Clement DL, De Buyzere ML. Daytime and nighttime blood pressure as predictors of death and cause-specific cardiovascular events in hypertension. Hypertension 2008;5:55-61. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000002775">https://doi.org/10.1097/HJH.0000000000002775</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Fagard</surname>
								<given-names>RH</given-names>
							</name>
							<name>
								<surname>Celis</surname>
								<given-names>H</given-names>
							</name>
							<name>
								<surname>Thijs</surname>
								<given-names>L</given-names>
							</name>
							<name>
								<surname>Staessen</surname>
								<given-names>JA</given-names>
							</name>
							<name>
								<surname>Clement</surname>
								<given-names>DL</given-names>
							</name>
							<name>
								<surname>De Buyzere</surname>
								<given-names>ML</given-names>
							</name>
						</person-group>
						<article-title>Daytime and nighttime blood pressure as predictors of death and cause-specific cardiovascular events in hypertension</article-title>
						<source>Hypertension</source>
						<year>2008</year>
						<volume>5</volume>
						<fpage>55</fpage>
						<lpage>61</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000002775">https://doi.org/10.1097/HJH.0000000000002775</ext-link>
					</element-citation>
				</ref>
				<ref id="B75">
					<label>34</label>
					<mixed-citation>34 Draman MS, Dolan E, van der Poel L, Tun TK, McDermott JH, Sreenan S. The importance of night-time systolic blood pressure in diabetic patients: Dublin Outcome Study. J Hypertens 2015;33:1373-7. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.107.100727">https://doi.org/10.1161/HYPERTENSIONAHA.107.100727</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Draman</surname>
								<given-names>MS</given-names>
							</name>
							<name>
								<surname>Dolan</surname>
								<given-names>E</given-names>
							</name>
							<name>
								<surname>van der Poel</surname>
								<given-names>L</given-names>
							</name>
							<name>
								<surname>Tun</surname>
								<given-names>TK</given-names>
							</name>
							<name>
								<surname>McDermott</surname>
								<given-names>JH</given-names>
							</name>
							<name>
								<surname>Sreenan</surname>
								<given-names>S</given-names>
							</name>
						</person-group>
						<article-title>The importance of night-time systolic blood pressure in diabetic patients: Dublin Outcome Study</article-title>
						<source>J Hypertens</source>
						<year>2015</year>
						<volume>33</volume>
						<fpage>1373</fpage>
						<lpage>1377</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.107.100727">https://doi.org/10.1161/HYPERTENSIONAHA.107.100727</ext-link>
					</element-citation>
				</ref>
				<ref id="B76">
					<label>35</label>
					<mixed-citation>35 Kario K, Kanegae H, Tomitani N, Okawara Y, Fujiwara T, Yano Y. Nighttime Blood Pressure Measured by Home Blood Pressure Monitoring as an Independent Predictor of Cardiovascular Events in General Practice. Hypertension 2019;73:1240-8. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.118.12740">https://doi.org/10.1161/HYPERTENSIONAHA.118.12740</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Kario</surname>
								<given-names>K</given-names>
							</name>
							<name>
								<surname>Kanegae</surname>
								<given-names>H</given-names>
							</name>
							<name>
								<surname>Tomitani</surname>
								<given-names>N</given-names>
							</name>
							<name>
								<surname>Okawara</surname>
								<given-names>Y</given-names>
							</name>
							<name>
								<surname>Fujiwara</surname>
								<given-names>T</given-names>
							</name>
							<name>
								<surname>Yano</surname>
								<given-names>Y</given-names>
							</name>
						</person-group>
						<article-title>Nighttime Blood Pressure Measured by Home Blood Pressure Monitoring as an Independent Predictor of Cardiovascular Events in General Practice</article-title>
						<source>Hypertension</source>
						<year>2019</year>
						<volume>73</volume>
						<fpage>1240</fpage>
						<lpage>1248</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.118.12740">https://doi.org/10.1161/HYPERTENSIONAHA.118.12740</ext-link>
					</element-citation>
				</ref>
				<ref id="B77">
					<label>36</label>
					<mixed-citation>36 Hansen TW, Li Y, Boggia J, Thijs L, Richart T, Staessen JA. Predictive role of the nighttime blood pressure. Hypertension 2011;57:3-10. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.109.133900">https://doi.org/10.1161/HYPERTENSIONAHA.109.133900</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Hansen</surname>
								<given-names>TW</given-names>
							</name>
							<name>
								<surname>Li</surname>
								<given-names>Y</given-names>
							</name>
							<name>
								<surname>Boggia</surname>
								<given-names>J</given-names>
							</name>
							<name>
								<surname>Thijs</surname>
								<given-names>L</given-names>
							</name>
							<name>
								<surname>Richart</surname>
								<given-names>T</given-names>
							</name>
							<name>
								<surname>Staessen</surname>
								<given-names>JA</given-names>
							</name>
						</person-group>
						<article-title>Predictive role of the nighttime blood pressure</article-title>
						<source>Hypertension</source>
						<year>2011</year>
						<volume>57</volume>
						<fpage>3</fpage>
						<lpage>10</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.109.133900">https://doi.org/10.1161/HYPERTENSIONAHA.109.133900</ext-link>
					</element-citation>
				</ref>
				<ref id="B78">
					<label>37</label>
					<mixed-citation>37 Salles GF, Reboldi G, Fagard RH, Cardoso CR, Pierdomenico SD, Verdecchia PABC-H Investigators. Prognostic Effect of the Nocturnal Blood Pressure Fall in Hypertensive Patients: The Ambulatory Blood Pressure Collaboration in Patients With Hypertension (ABC-H) Meta-Analysis. Hypertension 2016;67:693-700. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.115.06981">https://doi.org/10.1161/HYPERTENSIONAHA.115.06981</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Salles</surname>
								<given-names>GF</given-names>
							</name>
							<name>
								<surname>Reboldi</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>Fagard</surname>
								<given-names>RH</given-names>
							</name>
							<name>
								<surname>Cardoso</surname>
								<given-names>CR</given-names>
							</name>
							<name>
								<surname>Pierdomenico</surname>
								<given-names>SD</given-names>
							</name>
							<name>
								<surname>Verdecchia</surname>
								<given-names>P</given-names>
							</name>
						</person-group>
						<person-group person-group-type="author"/>
						<article-title>Prognostic Effect of the Nocturnal Blood Pressure Fall in Hypertensive Patients: The Ambulatory Blood Pressure Collaboration in Patients With Hypertension (ABC-H) Meta-Analysis</article-title>
						<source>Hypertension</source>
						<year>2016</year>
						<volume>67</volume>
						<fpage>693</fpage>
						<lpage>700</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1161/HYPERTENSIONAHA.115.06981">https://doi.org/10.1161/HYPERTENSIONAHA.115.06981</ext-link>
					</element-citation>
				</ref>
				<ref id="B79">
					<label>38</label>
					<mixed-citation>38 Zhu J, Hao X, Tang H, Xu J, Wang A, Zhang X. Time point of nocturnal trough systolic blood pressure as an independent predictor of cardiovascular events. J Clin Hypertens (Greenwich) 2022;24:283-91. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1111/jch.14404">https://doi.org/10.1111/jch.14404</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Zhu</surname>
								<given-names>J</given-names>
							</name>
							<name>
								<surname>Hao</surname>
								<given-names>X</given-names>
							</name>
							<name>
								<surname>Tang</surname>
								<given-names>H</given-names>
							</name>
							<name>
								<surname>Xu</surname>
								<given-names>J</given-names>
							</name>
							<name>
								<surname>Wang</surname>
								<given-names>A</given-names>
							</name>
							<name>
								<surname>Zhang</surname>
								<given-names>X</given-names>
							</name>
						</person-group>
						<article-title>Time point of nocturnal trough systolic blood pressure as an independent predictor of cardiovascular events</article-title>
						<source>J Clin Hypertens (Greenwich)</source>
						<year>2022</year>
						<volume>24</volume>
						<fpage>283</fpage>
						<lpage>291</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1111/jch.14404">https://doi.org/10.1111/jch.14404</ext-link>
					</element-citation>
				</ref>
				<ref id="B80">
					<label>39</label>
					<mixed-citation>39 Huguet N, Green BB, Voss RW, Larson AE, Angier H, Miguel M. Factors Associated With Blood Pressure Control Among Patients in Community Health Centers. Am J Prev Med 2023;64:631-41. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.amepre.2022.11.002">https://doi.org/10.1016/j.amepre.2022.11.002</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Huguet</surname>
								<given-names>N</given-names>
							</name>
							<name>
								<surname>Green</surname>
								<given-names>BB</given-names>
							</name>
							<name>
								<surname>Voss</surname>
								<given-names>RW</given-names>
							</name>
							<name>
								<surname>Larson</surname>
								<given-names>AE</given-names>
							</name>
							<name>
								<surname>Angier</surname>
								<given-names>H</given-names>
							</name>
							<name>
								<surname>Miguel</surname>
								<given-names>M</given-names>
							</name>
						</person-group>
						<article-title>Factors Associated With Blood Pressure Control Among Patients in Community Health Centers</article-title>
						<source>Am J Prev Med</source>
						<year>2023</year>
						<volume>64</volume>
						<fpage>631</fpage>
						<lpage>641</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1016/j.amepre.2022.11.002">https://doi.org/10.1016/j.amepre.2022.11.002</ext-link>
					</element-citation>
				</ref>
				<ref id="B81">
					<label>40</label>
					<mixed-citation>40 Jaeger BC, Bress AP, Bundy JD, Cheung AK, Cushman WC, Drawz PE. Longer-Term All-Cause and Cardiovascular Mortality With Intensive Blood Pressure Control: A Secondary Analysis of a Randomized Clinical Trial. JAMA Cardiol 2022;7:1138-46. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1001/jamacardio.2022.3345">https://doi.org/10.1001/jamacardio.2022.3345</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Jaeger</surname>
								<given-names>BC</given-names>
							</name>
							<name>
								<surname>Bress</surname>
								<given-names>AP</given-names>
							</name>
							<name>
								<surname>Bundy</surname>
								<given-names>JD</given-names>
							</name>
							<name>
								<surname>Cheung</surname>
								<given-names>AK</given-names>
							</name>
							<name>
								<surname>Cushman</surname>
								<given-names>WC</given-names>
							</name>
							<name>
								<surname>Drawz</surname>
								<given-names>PE</given-names>
							</name>
						</person-group>
						<article-title>Longer-Term All-Cause and Cardiovascular Mortality With Intensive Blood Pressure Control: A Secondary Analysis of a Randomized Clinical Trial</article-title>
						<source>JAMA Cardiol</source>
						<year>2022</year>
						<volume>7</volume>
						<fpage>1138</fpage>
						<lpage>1146</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1001/jamacardio.2022.3345">https://doi.org/10.1001/jamacardio.2022.3345</ext-link>
					</element-citation>
				</ref>
				<ref id="B82">
					<label>41</label>
					<mixed-citation>41 Mancia G, Kreutz R, Brunström M, Burnier M, Grassi G, Januszewicz A. 2023 ESH Guidelines for the management of arterial hypertension The Task Force for the management of arterial hypertension of the European Society of Hypertension: Endorsed by the International Society of Hypertension (ISH) and the European Renal Association (ERA). J Hypertens 2023;41:1874-2071. <ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000003480">https://doi.org/10.1097/HJH.0000000000003480</ext-link>
					</mixed-citation>
					<element-citation publication-type="journal">
						<person-group person-group-type="author">
							<name>
								<surname>Mancia</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>Kreutz</surname>
								<given-names>R</given-names>
							</name>
							<name>
								<surname>Brunström</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>Burnier</surname>
								<given-names>M</given-names>
							</name>
							<name>
								<surname>Grassi</surname>
								<given-names>G</given-names>
							</name>
							<name>
								<surname>Januszewicz</surname>
								<given-names>A</given-names>
							</name>
						</person-group>
						<article-title>2023 ESH Guidelines for the management of arterial hypertension The Task Force for the management of arterial hypertension of the European Society of Hypertension: Endorsed by the International Society of Hypertension (ISH) and the European Renal Association (ERA)</article-title>
						<source>J Hypertens</source>
						<year>2023</year>
						<volume>41</volume>
						<fpage>1874</fpage>
						<lpage>2071</lpage>
						<ext-link ext-link-type="uri" xlink:href="https://doi.org/10.1097/HJH.0000000000003480">https://doi.org/10.1097/HJH.0000000000003480</ext-link>
					</element-citation>
				</ref>
			</ref-list>
			<fn-group>
				<fn fn-type="other" id="fn5">
					<label>Financing</label>
					<p> The authors declare they have not received any funding for this work.</p>
				</fn>
				<fn fn-type="other" id="fn6">
					<label>Policy on the use of artificial intelligence</label>
					<p> Artificial intelligence resources have not been used to write the text or to create the tables or graphs.</p>
				</fn>
			</fn-group>
		</back>
	</sub-article>-->
</article>