Pharmacists targeted were community pharmacists involved in the provision of diabetes care to patients, in full-time employment but non-diabetes credentialed pharmacists. The literature on sample size determinants for a qualitative study suggested a sample size between five and 25. 17, 18 Taking this factor into consideration and pharmacists’ potential time constraints, 13 potential participants were approached through personal contacts of the researchers and a snowball recruitment process. 19, 20 They were invited by email and were provided with an information sheet about the research and had the opportunity to ask questions before providing consent.

Participants were required to complete a training session. The overall goal of the training was to enhance participants’ understanding of the pharmacist’s role in providing diabetes care and incorporated demonstrating how the Simpler™ tool facilitated the provision of structured diabetes care. Emphasis was placed on how the tool assisted in identifying the reasons for therapeutic failure and resolve the issues by providing evidence-based suggestions through application of a systematic approach. The training program was developed by the primary author (SA) and the overall syllabus details are presented in Table 1.

Pre- and post-evaluation questions (in the format of questionnaires) and training modules were peer-reviewed by three pharmacist academics with specialist diabetes knowledge. The face, content and usefulness were subsequently validated, and pilot tested by two Australian and two Malaysian pharmacists experienced in the management of diabetes patients who provided further feedback. Adjustments to the modules were subsequently made. Some pharmacists had a face-to-face workshop while others received online training. Since the first author was in Australia at the time of the study, face-to-face training sessions for the Australian participants were offered in the first instance, followed by e-learning sessions for the Malaysian participants. The same presentation slides were used for both the face-to-face and the online training sessions. In addition, the voice-over of the presentation slides followed a standardised script. The recorded training modules were uploaded to a cloud file storage service which allowed large file viewing. Sharing and access to the file was provided to participants via email. Pharmacists had the opportunity to ask questions during face-to-face workshops and those doing the online training through various channels including social media. Upon completion, participants attending the face-to-face training completed the post-training questionnaire while participants who followed the online training informed the researcher (SA) and were subsequently sent the post-training questionnaire by email. Both groups had access to notes and the Simpler™ tool when completing the post-training questionnaire.

The pre-training questionnaire consisted of two sections: Section A included five closed-ended questions directed at participants’ training background and current practices, and Section B consisted of two open-ended questions on a patient’s case scenario aimed to test participants’ knowledge of diabetes guidelines and their skills in suggesting medication management interventions. The post-training questionnaire contained the same questions in Section B of the pre-training questionnaire. The pre-and post-training questionnaire is shown in

online Appendix (pdf)

. The questionnaire was face and content validated by the same pharmacists who pilot tested the training modules. Participants’ answers were marked by two markers using a written marking scheme validated by an independent pharmacist. Each answer had point/s awarded and the scores were marked out of 27.

Upon completion of the training, the same participants were given one month to apply the tool in their practice settings. They were provided with a template to record the number of times the tool was used on patients and the types of interventions conducted by utilising the tool. A unique identification number was allocated indicating where the participant originated: participants were assigned the letter P and numbered 1 to 6. The letter A was assigned to participants from Australia (example P1A) and those from Malaysia the letter M (example P1M). This allowed to differentiate participants’ perception of the tool from both countries as the two healthcare systems differed.

Semi-structured interviews were conducted by SA (July to August 2015). Face-to-face interviews were conducted with two pharmacists at their workplaces and two at a university. Telephone interviews were conducted with the remaining eight pharmacists. The interview process followed Kvale’s seven stages for conducting interviews and the requirements of consolidated criteria for reporting qualitative research guidelines (COREQ). 21, 22 The interview guide consisted of three sections, sections A, B and C, presented in Table 2. The interviewer followed the interview guide while allowing opportunity for probing questions and clarifications. The interview guide was pilot tested with two independent pharmacists.

Participants could raise points during the interview that were not included in the interview guide if these were relevant to the overall aim of the study. The interviews ended when all questions were exhausted and no new information was obtained (interviews reached a saturation point). 23 Interviews were audio recorded and transcribed verbatim by SA. Audio recordings were saved with a unique identification code to protect participants’ anonymity. A project supervisor (HLH) conducted quality checks of transcripts against audio recordings.

Differences between the pre- and post-training questionnaire responses were analysed using the Wilcoxon Signed Rank Test for non-parametric testing as the sample did not meet the requirements for normal distribution. SPSS statistical package version 22 was used for the quantitative analysis. 24

Descriptive analyses were used for closed-ended interview questions (Sections A and B of the interview guide) whilst thematic analysis was used for the open-ended questions (Section C) to gain insight into pharmacists’ opinions, views and perceptions of the Simpler™ tool. In addition, the open-ended questions were used as a guide to identify emerging patterns. An inductive process was followed throughout the analysis and recurring topics from the interview data were investigated using the qualitative framework method as suggested by Boyatzis. 25 Participants’ raw data were highlighted in order to determine sentences or keywords which were then assigned a label called ‘codes’. The codes were then sorted into topics. Different views under the same topic were grouped as a subtopic. Transcripts were then scrutinised again for new or emerging topics. The coding process was performed by SA and project team members verified the analytical process before finalising the analysis. NVivo qualitative analysis software version 10 was used to categorise and organise the qualitative data. 26

Of the 13 pharmacists approached, 12 consented to undertake the study. There was equal representation of participants from Malaysia (n=6) and Australia (n=6). Most participants (75%, 9/12) had less than three years’ experience of conducting diabetes management. Table 3 presents participants’ demographic data and practice experiences.

Interestingly, the majority of participants (66.7%; 8/12) had never or only sometimes referred to the Australian or Malaysian diabetes practice guidelines when providing diabetes MMS. 2, 3, 27 Regarding the question “What motivated you to participate in this research?” most participants ranked interest in the subject (83.3%;10/12) and improve patients’ outcomes (91.7%; 11/12) as the main incentive.

1) Pre- and post-training questionnaire

There was a significant improvement in post-training questionnaire scores (P=0.002) by both markers. Before attending the training session, the participants’ median test score was 6.5/27, interquartile range (IQR) 1.4 (1st marker) and 5.3/27, IQR 2.0 (2nd marker). After attending the training session, the scores doubled to 14.3/27, IQR 4.5 (1st marker) and 11.3/27, IQR 3.1 (2nd marker), showing significant improvements (p=0.002). Pharmacists initially struggled to frame better questions to make meaningful interventions. However, post-training results showed a marked improvement in addressing the seven diabetes factors to facilitate the intervention process.

All 12 pharmacists participated in the semi-structured interviews. The average duration of the interviews was 32 minutes with the face-to-face interviews ranging between 19 to 32 minutes (mean 26 minutes) and the telephone interviews between 16 to 54 minutes (mean 36 minutes). Most participants (91.7%; 11/12) used the Simpler™ tool to facilitate their intervention process. Those included: to add a statin to achieve cholesterol targets; initiate metformin in patients with uncontrolled diabetes; dose adjustments and improving medication adherence. One participant did not use the tool as this participant only focused on lifestyle factors during patient consultations. The participant therefore expected more detailed counselling points on lifestyle management. Participants reported making interventions using one or more tool indicators. The types and number of interventions made are provided in Table 4 with supporting quotations.

Interview analysis revealed patterns that were grouped into three main topics. Those were:

• Perception of training program (interview guide question 5 of Section C),

• Perceived effectiveness of the Simpler™ tool (from various questions), and

• Barriers to the Simpler™ tool utilisation (interview guide questions 1,3,4,6 of Section C).

1) Perception of the training program

Most participants (83.3%; 10/12) commented that the training module content was adequate and relevant. In addition, they found the length of training appropriate. The majority believed the Simpler™ training session increased their knowledge and confidence in evidence-based diabetes management and for some it served as a refresher. Participants provided positive comments on the training sessions overall. The supporting participants’ quotations on the training session are presented in Table 5.

Improvements to the Simpler™ training modules included to: 1) add an intervention recording template to document interventions in patients’ medical records (PMR), 2) develop a flow chart to illustrate the information gathering process before the Simpler™ tool application, 3) include more slides on identifying medication related problems, 4) add information on glucagon use, and 5) add materials on lifestyle management.

2) Perceived effectiveness of the Simpler™ tool in practice

All 12 participants found the Simpler™ tool to be beneficial when conducting medication reviews with patients. Participants used words such as ‘organised’, ‘sequential’, ‘straight to the point’, ‘my accounting made relevant’, ‘compact’, ‘complete’ and ‘easiest tool’ to describe the benefits. Participants from both Malaysia and Australia expressed their reliance on the Simpler™ tool when conducting MMS as they considered it to be a point of reference. All participants expressed the tool as an ‘aide memoir’ in recollection of the factors associated with diabetes management.

Eight specific issues were identified on the perceived effectiveness of the tool, summarised with corresponding quotations in Table 5. The Simpler™ tool allowed participants to conduct more comprehensive reviews during consultations. Of specific interest was that one participant found that the tool made diabetes medication reviews more purposeful as improving patients’ glycaemia levels became the focus. However, three of the participants found the tool time consuming to use. However, they indicated that the benefits of being able to conduct detailed and organised patient assessments outweighed the time factor. A common view amongst participants was that the tool facilitated the writing of interventions in PMR (Malaysia) and in software programs (Australia). In addition, one participant felt the tool promoted her to have a more specialised role in diabetes management and thus found the tool specifically targeted for diabetes management.

3) Perceived barriers

Two specific issues were identified on the perceived barriers to the effective use of the tool, as summarised in Table 5. Two Australian participants found the limited access to patient’s medical data a barrier and was therefore unable to make a meaningful intervention while one believed accredited pharmacists providing home medicine reviews service were better suited to make interventions.

Participants provided the following suggestions to further refine the Simpler™ tool:

• Use visual prompts

• Larger font for headings

• Use either Malaysian or Australian targets

• Use terms like Asian or Caucasian for body mass index targets

Based on these suggestions, the tool was further refined as presented in Table 6.