Abstract
Background: Colorectal cancer (CRC) is among the four cancers with the highest incidence and mortality in Colombia. CRC screening is currently recommended in national guidelines starting at age 50. The American Cancer Society has suggested beginning screening at age 45.
Objectives: To estimate the cost-utility ratio of initiating CRC screening at age 45 compared to age 50 in Colombia.
Methods: A Markov model simulating the natural history of CRC was used to compare the initiation of screening at ages 45 and 50. The cost-utility analysis incorporated costs and utilized Colombia’s per capita gross domestic product (GDP) threshold (COP 21,500,000).
Results: Starting colonoscopy screening at age 45 was 0.33% less costly, yielded a gain of 3.49 quality-adjusted life years (QALYs), and reduced mortality by 37% compared to initiating at age 50 (costs: COP 92,364,407 vs. COP 92,669,231; QALYs: 19.40 vs. 15.91; and mortality proportion: 3.4% vs. 5.4%, respectively).
Conclusion: Based on the Markov model analysis and the assumptions made, initiating CRC screening at age 45 was cost-effective. Costs were lower, more QALYs were gained, and the incremental cost-effectiveness ratio was below Colombia’s GDP threshold. A budget impact analysis is recommended to support the implementation of policies in Colombia.
Keywords: Colon cancer, screening, cost-effectiveness, quality-adjusted life years.
Resumen
Antecedentes: El cáncer colorrectal (CCR) se encuentra entre los cuatro cánceres de mayor incidencia y mortalidad en Colombia. La tamización de CCR está establecida en las guías nacionales a partir de los 50 años. La Sociedad Estadounidense de Cáncer ha recomendado iniciarla a los 45 años.
Objetivos: Estimar la relación entre el costo y la utilidad de iniciar a los 45 años la tamización de CCR en Colombia.
Métodos: Se realizó un modelo de Markov que simula la historia natural del CCR para comparar el inicio de la tamización a los 45 y 50 años mediante un estudio de costo y utilidad que incluye costos y usa el umbral del producto interno bruto (PIB) per cápita de Colombia (COP 21.500.000).
Resultados: Iniciar la tamización de la colonoscopia a los 45 años fue 0,33% menos costosa, produjo una ganancia de 3,49 años de vida ajustados por calidad (AVAC) y una proporción de mortalidad 37% menor que si se inicia a los 50 años (costos: COP 92.364.407 frente a COP 92.669.231, AVAC: 19,40 frente a 15,91, y proporción de mortalidad: 3,4% frente a 5,4%, respectivamente).
Conclusión: Con el análisis del modelo de Markov y los supuestos establecidos, la tamización de CCR a los 45 años fue rentable. Los costos fueron menores y se ganaron más AVAC, además de que la rentabilidad incremental fue inferior al PIB establecido como umbral para Colombia. Se sugiere realizar un análisis del impacto presupuestal para implementar políticas en Colombia.
Palabras clave: Cáncer de colon, tamización (cribado), costo-efectividad, años de vida ajustados por calidad de vida.
Original articles
Cost-Utility Analysis of Colorectal Cancer Screening Starting at Age 45 Compared to Age 50 in Colombia
Análisis de costo y utilidad de la tamización del cáncer colorrectal a partir de los 45 años en comparación con la tamización a partir de los 50 años en Colombia
Gerardo Andrés Puentes-Leal gandrespl@yahoo.com.ar
Esperanza Peña-Torres
Received: 18 June 2024
Accepted: 17 August 2024
Colorectal cancer (CRC) is the third leading cause of cancer-related deaths in Colombia, following breast and prostate cancer, contributing significantly to morbidity, mortality, and high public health costs worldwide1-3. Approximately 80% of CRC cases are preceded by adenomatous polyps, considered precancerous lesions. When removed, these polyps significantly reduce CRC incidence1,4,5. Adenomas are histologically classified into tubular (87%), tubulovillous (8%), and villous (5%) types, with the latter having a higher potential for malignancy over a prolonged latency period of 10 to 15 years. Age is a known risk factor for their development5-8.
The mortality rate from CRC in Colombia is 7.4%, and this figure has increased in recent years1. In 2021, 7,579 new cases were reported, with a five-year prevalence of 18,654 cases (equivalent to 36.6 cases per 100,000 inhabitants)2. Early-stage interventions can significantly alter cancer incidence9-12. According to Colombian legislation, CRC is recognized as a public health concern and a national priority. Early and integrated prevention of risk factors, along with effective, timely, and continuous detection, improves quality of life, reduces economic impact, and enhances social outcomes for patients (Article 5, Law 1384 of 2010)13-16. CRC screening is a recommended intervention outlined in local guidelines and endorsed by scientific cancer societies worldwide. Colonoscopy is the gold standard for CRC detection, offering the highest sensitivity and specificity as both a diagnostic and therapeutic tool for identifying adenomas. Colombian guidelines recommend negative-screening colonoscopies be repeated every 10 years, starting at age 50 and continuing until age 7517. For cases with positive adenoma findings, follow-up intervals depend on polyp size and histological characteristics. However, colonoscopy has certain disadvantages: it is invasive, requires bowel preparation, carries a risk of complications, and may be costly or have limited availability for large-scale implementation18.
In Latin America, CRC screening is recommended starting at age 5018-20. The 2013 Colombian Clinical Practice Guideline (CPG) for colon and rectal cancer from the Ministry of Health and Social Protection, along with the 2015 CPG for colorectal cancer detection by the Colombian Association of Gastroenterology and the Colombian Association of Coloproctology, establish that CRC screening should begin at age 50 (a favorable recommendation with moderate-quality evidence) as a best practice. The choice of screening test should be individualized based on medical judgment. Colonoscopy is recommended as the first option, but fecal occult blood tests (FOBT), including the fecal immunochemical test (FIT) and the guaiac-based FOBT (g-FOBT), are acceptable alternatives17. When comparing colonoscopy to g-FOBT/FIT, colonoscopy demonstrated higher CRC detection rates (relative risk [RR]: 5.91, confidence interval [CI]: 70); however, compliance rates were higher with g-FOBT/FIT (RR: 0.57, 95% CI: 0.42-0.78)17,18.
In 2018, the American College of Gastroenterology (ACG), the American Society for Gastrointestinal Endoscopy (ASGE), and the American Cancer Society published CRC screening guidelines recommending screening start at age 45. This was based on the rising incidence of CRC over the past two decades, which increased from 1.3% to 2.3% among individuals under 55 years of age21. Screening at age 45 could prevent 277,000 CRC cases and 203,000 deaths by 203022. In 2019, a study conducted by Stanford University, the University of California, and the University of Pittsburgh reported that CRC screening at age 45 in a cohort of 1,000 individuals prevented four CRC cases, two CRC-related deaths, and gained 14 quality-adjusted life years (QALYs). However, earlier screening required 758 additional colonoscopies, thereby increasing costs23. A Colombian study documented adenoma findings in one out of every five screening colonoscopies in individuals aged 45 to 50 years (18.5% prevalence), supporting the justification for CRC screening starting at age 4524.
Early CRC screening has economic implications for society12. In Europe, the economic burden of CRC is estimated at EUR 13 billion25, while in the United States, the cost of diagnosing and treating CRC is approximately USD 100,00026. In Colombia, the economic burden of CRC and the cost of screening starting at age 45 remain unknown. U.S. recommendations are not directly applicable to Colombia due to differences in CRC diagnosis and treatment costs. Therefore, to evaluate and decide on the implementation of CRC screening in Colombia, cost-utility and cost-effectiveness studies using local monetary values are warranted27.
The aim of this study was to assess whether initiating CRC screening at age 45 in Colombia is cost-effective compared to the threshold established by the World Health Organization (WHO) based on the gross domestic product (GDP). Additionally, the study evaluated its impact on mortality, QALYs, and costs.
A microeconomic evaluation was conducted to analyze the cost-effectiveness27 of initiating CRC screening at age 45 compared to age 50 from the perspective of a third-party payer. Costs and mortality were compared, expected outcomes for each age group were estimated using simulated mathematical modeling, and the incremental cost-utility ratio (ICUR) was calculated28-30. Screening was considered cost-effective when the ICUR was below the per capita GDP threshold, as defined by the Colombian Institute for Health Technology Assessment (IETS in Spanish) and the World Health Organization (WHO)30. A Markov model (Figure 1) was designed to compare three screening strategies: annual FIT-based FOBT, annual guaiac-based FOBT, and colonoscopy every 10 years. A positive FOBT result was followed by a colonoscopy.
A systematic literature review was performed using various MeSH terms in the Cochrane Library, Medline, SciELO, and Google Scholar databases. There were no language restrictions, and gray literature was included.
A 31-year time horizon was defined, starting screening at age 45 and continuing to age 76, the life expectancy in Colombia according to the National Administrative Department of Statistics (DANE in Spanish). The evaluation was conducted from the perspective of Colombia’s General System of Social Security in Health (SGSSS in Spanish), incorporating the value of all direct resources associated with the use of the technology and the perceived health benefits directly experienced by patients. A 5% annual discount rate, recommended by the IETS, was applied to costs and health effectiveness outcomes. The primary outcomes were the proportion of CRC-related mortality in both groups and quality-adjusted life years (QALYs).
To estimate costs, the Medication Price Information System (SISMED in Spanish) database was consulted to determine weighted averages and reported unit quantities. Micro-costing was conducted, and monetary valuation was calculated using the official fee manual of Colombia’s mandatory traffic accident insurance (SOAT in Spanish). Direct costs for each stage were considered, including procedural costs and treatment costs for complications (Tables 1-5).
Based on published literature, frequencies for findings from screening colonoscopy were constructed, including probabilities for the health states proposed in the model for the disease at ages 45 and 50. Frequencies of findings in screening colonoscopy for patients aged 45 and 50 were obtained from a Chilean study (population similar to Colombia’s) on a multicenter CRC screening program31, a Colombian study on adenoma prevalence in individuals aged 45-5024, and Korea’s national cancer screening program32 (Table 6).
Screening begins at ages 45 and 50 and ends at age 76 or upon death.
There is a probability that, after undergoing the procedure, the patient remains in the same stage they were in initially.
Colonoscopies meet quality standards, preventing the occurrence of interval cancers within the model.
Health stages are successive and non-reversible unless intervention with polypectomy or surgery occurs.
The two absorbing states are death from CRC (F) and death from other causes (G).
Colonoscopy follow-ups are conducted in the first, third, fifth, seventh, and tenth years based on the type of adenoma documented, following international guidelines on digestive endoscopy and cancer33.
The proposed model defines the transition probabilities for each stage, as presented in the transition probability matrix, which reflects the natural history of CRC.
The literature indicates that a patient undergoing a screening colonoscopy with normal results has a 97% probability of remaining in this normal stage for one year and a 1% probability of progressing to an adenoma smaller than 10 mm31,32,34-38. A person with an adenoma smaller than 10 mm has a 97% probability of remaining in the same stage for one year and a 1% probability of progressing to an adenoma larger than 10 mm35,37,38. A patient with an adenoma larger than 10 mm has a 5% probability of progressing to advanced regional cancer within one year and a 94% probability of remaining in the same stage34,36-38. A patient with advanced regional cancer identified during a screening colonoscopy has a 77% probability of remaining in this stage for one year, a 13% probability of progressing to metastatic cancer, and a 9% probability of progressing to death from CRC within one year39. A patient diagnosed with metastatic CRC during screening has a 62% probability of surviving for one year and a 37% probability of progressing to death within that time39 (Table 7).
Using the parameters described above (Table 8), a transition matrix was constructed for a cohort of 1,000 patients. This included 31 cycles for patients starting screening at age 45 and 26 cycles for those starting at age 50 (Tables 9 and 10).
Quality-adjusted life years (QALYs) were used as the utility outcome. To calculate the quality of life for each stage, the EuroQol (EQ-5D-5L) scale was applied to patients undergoing CRC screening. A value of 1 was assigned to normal quality of life (stages A and B), 0.74 to patients undergoing interventions such as polypectomies, surgeries, hospitalizations, chemotherapy, or radiotherapy (stages C, D, and E), and 0 to patients who died (stages F and G)40.
The cost-utility ratio (C/U) and incremental cost-utility ratio (ICUR) were calculated for the two screening strategies. Life years gained were adjusted according to the effectiveness of the two screening approaches: 0.82 for screening starting at age 45 and 0.75 for screening from ages 50 to 76 (Tables 9 and 10).
The CRC screening strategy for Colombian adults starting at age 45 was 0.329% less expensive compared to screening starting at age 50 (COP 92,364,407 versus COP 92,669,231). Screening at age 45 yielded 19.40 QALYs, compared to 15.91 QALYs for screening at age 50, resulting in a gain of 3.49 QALYs with earlier screening.
The incremental cost-utility ratio (ICUR) was COP 87,243 per QALY for screening at age 45, making it more effective by gaining more QALYs and less costly. This positions it as a dominant strategy over screening at age 50 (Table 11).
Mortality was 3.4% for screening starting at age 45, compared to 5.4% for screening at age 50. This represents a 37% reduction in mortality with earlier screening at age 45.
To assess the influence of parameter uncertainty on the final outcome of the mathematical model, a deterministic sensitivity analysis was performed using probabilities ranging from 0 to 1. Monte Carlo simulations (1,000 iterations) were conducted in Visual Basic Excel, varying parameter values randomly according to statistical probability distributions. Seed variables were input based on the frequency table of screening findings for CRC in patients aged 45 and 50, and a probability matrix was constructed with the number of individuals transitioning through each stage, guided by the previously described transition probabilities. This iterative process continued until cycle 31 for patients starting screening at age 45 and cycle 26 for those starting at age 50, representing the point when the patient reaches 76 years of age or dies.
Once data for each screening group were collected, scenarios were created using QALY impact as the variable to observe the distribution of incremental utility data and incremental costs relative to the per capita GDP, given the absence of explicit thresholds in Colombia. Cost adjustments were calculated at 12%, and quality adjustments were based on the EQ-5D-5L scale, assigning a value of 0.617 for the stages of adenoma larger than 10 mm, advanced cancer, and metastatic cancer, following the EuroQol (EQ-5D-5L) results41.
The sensitivity analysis revealed that colonoscopy screening for adults aged 45 generated 20.35 QALYs, compared to 19.85 QALYs for those aged 50. Additionally, with a 12% cost adjustment (applying the extreme discount rate), screening starting at age 45 was found to be less expensive than starting at age 50 (COP 48,284,262 versus COP 49,935,820) (Table 12).
This study addresses a societal need by examining a public health issue, providing results that benefit the entire population by evaluating and analyzing screening strategies and their cost-effectiveness. While the study adheres to the research guidelines of Colombia’s IETS and follows economic evaluation methodology processes, it is a review study that relies on secondary sources. Therefore, its results must be interpreted cautiously and within the methodological framework to avoid conclusions or judgments that exceed the reproducibility of the findings.
To compare costs and benefits between two age groups (45 years and older versus 50 years and older), comprehensive data for all stages in both groups are necessary. A key limitation is the scarcity of studies reporting screening findings in patients over 45, as screening colonoscopy for individuals under 50 is not yet a recommended practice in Colombia.
This study arises from the need to evaluate the cost-effectiveness and cost-utility of initiating CRC screening at age 45, as recommended by the American Cancer Society, due to the increasing incidence of CRC in individuals under 50 in Colombia. On one hand, this decision has several implications for the healthcare system, such as increasing the demand for screening colonoscopies among a larger number of asymptomatic individuals, which will lead to higher resource consumption. On the other hand, the goal of screening is to diagnose precancerous lesions early, allowing for timely intervention and altering the natural history of a high-cost disease like CRC, ultimately contributing to resource savings.
The earlier screening begins, the larger the number of individuals to be screened, based on the population pyramid, and the higher the likelihood of not finding disease in these patients. To reduce this uncertainty, national studies are needed to estimate the prevalence of adenomas and CRC in individuals under 50.
Another issue to consider is that CRC screening in Colombia is opportunity-based. Consequently, there is a population of individuals over 50 who lack access to screening. It could be argued that a greater benefit might be achieved by including more patients from the 50+ age group rather than expanding the screening age range42-45. Healthcare organizations and government entities should prioritize cancer screening as part of their agenda, as its cost-effectiveness is well-documented. This prioritization could positively impact future healthcare cost distribution by mitigating the economic burden of advanced-stage CRC.
In a screening scenario based on the assumptions established in this study, initiating CRC screening at age 45 was shown to dominate the strategy of screening individuals over 50 in Colombia. This suggests that beginning CRC detection studies in this population yields greater benefits and lower overall costs, making it a cost-effective strategy.
The cost difference between the two strategies was 0.329%. Screening at age 45 resulted in higher utility measured in quality-adjusted life years (QALYs) and reduced mortality by 2% compared to starting screening at age 50.
A budget impact analysis (BIA) is required to support the decision to implement a national public health policy for universal CRC screening at age 45.
*Correspondence: Gerardo Andrés Puentes-Leal. gandrespl@yahoo.com.ar
