ORIGINAL ARTICLE
Community pharmacist-led educational intervention to recommend ophthalmology visits for patients with diabetes: a cluster randomized controlled trial
2026 年 8 巻 2 号 p. 37-45
詳細
2026 年 8 巻 2 号 p. 37-45
BACKGROUND
It remains unclear whether pharmacist-led educational programs in community pharmacies could increase appropriate ophthalmic visits for patients with diabetes. We assessed efficacy of pharmacist-led education on ophthalmology visits for diabetes.
METHODS
We conducted a cluster randomized controlled trial at 32 community pharmacies in Japan, targeting individuals with diabetes with no ophthalmic visits over a year. Pharmacists in the intervention group received online training on diabetic retinopathy and educated patients, while the control group received a pamphlet. The primary outcome was ophthalmic visits during the follow-up period. Generalized estimating equations were performed with two adjusted models: age and sex (model 1), and additionally diabetic retinopathy factors (model 2). Key secondary outcomes were changes in behavior for ophthalmic visits and glycated hemoglobin (HbA1c) levels.
RESULTS
Overall, 268 patients were included (133 intervention and 135 control). Participants’ mean age was 60.1 years, and HbA1c level was 7.5%. Ophthalmic visits occurred in 18.8% (25/133) of the intervention and 20.7% (28/135) of the control group, yielding no significant difference (model 1, risk difference [RD] −0.03 [−0.14 to 0.08], risk ratio [RR] 0.88 (0.54 to 1.45); model 2, RD −0.07 [−0.21 to 0.08], RR 0.73 [0.41 to 1.30]). There was no significant difference between the two groups in the mean changes from baseline to 6 months in behavior for ophthalmic visits (0.07 [−0.23 to 0.37]) and HbA1c levels (−0.28 [−0.76 to 0.20]).
CONCLUSIONS
Pharmacist-led education on diabetic retinopathy did not increase ophthalmology visits or improve diabetes-related outcomes. Effective strategies to encourage ophthalmology visits are required.
The International Diabetes Federation estimated that there were 536 million individuals worldwide living with diabetes in 2021, representing 10.5% of adults aged 20–80 years1). Among them, 22.3% (approximately 103.1 million people) were affected by diabetic retinopathy, with expectations that this number will rise to 160.5 million by 20452). Although diabetic retinopathy can lead to blindness3),4), it is a preventable cause of visual impairment when dilated fundus examinations are performed by an ophthalmologist for early detection and treatment5). However, it is estimated that in 2020, approximately 9 million adults aged ≥50 years lost their vision due to diabetic retinopathy6). Addressing the issue of individuals with diabetes who do not receive ophthalmic care remains a pressing challenge7),8).
Community pharmacies play a pivotal role in patient care, where pharmacists are well positioned to assess the overall health of patients, prescribed medications, and offer crucial disease education9–11). While previous research has shown that educational interventions by community pharmacists can improve glycemic control in individuals with diabetes12),13), studies focused on diabetic retinopathy education remain limited14). Moreover, some studies have demonstrated the efficacy of educational interventions in increasing ophthalmology visits in patients with diabetes. These interventions include repeated educational sessions monthly for four months15) or telephone interventions repeated up to seven times over six months16). However, these interventions face challenges in terms of their feasibility within clinical settings. It remains unclear whether pharmacist-led practical educational programs in community pharmacies can increase the number of appropriate ophthalmic visits among individuals with diabetes.
We conducted a cluster randomized controlled trial (RCT) to evaluate the efficacy of educational interventions provided by community pharmacists. Using an educational guide on diabetic retinopathy, we aimed to encourage ophthalmology visits and increase the number of individuals with diabetes who could access ophthalmic care at appropriate times.
This study is a cluster RCT evaluating the efficacy of community pharmacist-led education on diabetic retinopathy prevention, targeting individuals with diabetes who had no eye care visits for over a year. The study was conducted from February 2023 to November 2024 at the Total Medical Services Corporation, which has 49 pharmacies in Western Japan. The pharmacy group covers various hospital types in Japan, including university hospitals, regional core hospitals, and clinics, and handles prescriptions from all medical departments. Among these, we included 32 pharmacies (cluster units) where at least 20 patients were prescribed diabetes medications every year.
This study adhered to the Consolidated Standards of Reporting Trials guidelines. It was approved by the Ethics Committee of the Hokkaido Institute for Pharmacy Benefit Co., Ltd. (approved number: 2022050), and registered with the UMIN Clinical Trials Registry (registration number: UMIN000049274) in January 2023. All patients and pharmacists who participated in this study provided written informed consent.
PARTICIPANT ENROLLMENTIndividuals with diabetes who met the following inclusion criteria were recruited: (1) aged between 20 and 75 years; (2) diagnosed with diabetes for at least 6 months; (3) prescribed one or more diabetes medications (either oral or injectable); and (4) had not undergone an eye examination for diabetic retinopathy in the past year or longer. As the guidelines recommend annual eye examinations for diabetic retinopathy17–19), we defined inappropriate visits with an ophthalmologist as patients who had not undergone an eye examination for diabetic retinopathy in the past year or longer. Patients were excluded if they were unable to provide informed consent, did not attend the pharmacy visit themselves, or were blind bilaterally.
RANDOMIZATION AND MASKINGThe pharmacies were centrally allocated and randomly assigned to the intervention and control groups in a 1:1 ratio using computer-generated random numbers, which was stratified based on pharmacy size (whether the average number of prescriptions handled exceeded 100 per day)20) and whether there was an ophthalmology department in the hospital where the pharmacy primarily accepted prescriptions. Due to the nature of the educational intervention, it was not possible to fully mask the pharmacists delivering the intervention or the patients receiving it, thus only the data analysts were masked.
INTERVENTION AND CONTROLThe intervention involved providing online pre-intervention training on diabetic retinopathy to pharmacists, who then educated their patients at the pharmacy on the importance of ophthalmic examinations for diabetic retinopathy using a guidebook (Supplement 1). The online training was designed to equip pharmacists with fundamental knowledge of diabetes and diabetic retinopathy, as well as the importance of adhering to the ophthalmic follow-up intervals recommended in the guidelines. The training also focused on developing the communication skills necessary for educating patients and delivering standardized information using the guidebook. The online education was conducted in a 30-minute workshop format by pharmacists experienced in diabetes care. The initial 15 minutes focused on teaching the following fundamental knowledge: (1) pathophysiology of diabetic retinopathy; (2) epidemiology of diabetic retinopathy; and (3) follow-up criteria for ophthalmic examinations based on the guidelines. In the remaining 15 minutes, participating pharmacists practiced explaining the procedures using the guidebook in role-playing format. All pharmacists in the intervention group completed this online training.
The pharmacists in the intervention group executed the educational intervention using the guidebook and spent at least 3 minutes educating patients on diabetic retinopathy, following the trained method. The educational content covered aforementioned three fundamental knowledges related to diabetic retinopathy, following the guidelines19), with the educational time set to three minutes based on clinical feasibility and prior research31). Patients were also provided with a pamphlet (Supplement 2), which summarizes the key information about diabetic retinopathy, after the educational intervention. In the control group, patients only received the pamphlet.
OUTCOMESThe primary outcome was the proportion of patients who visited the ophthalmology for diabetic retinopathy within 6 months15),16). The secondary outcomes assessed at 6 months included the followings: changes in the behavior regarding ophthalmic visits, diet and exercise, medication adherence, glycemic control based on glycated hemoglobin (HbA1c) levels, and body weight.
Outcome assessments were conducted through self-administered surveys at two time points: at the start of the study before receiving educational intervention and/or pamphlets (baseline data) and 6 months later (end of the follow-up period). The 6-month survey was administered at the first visit after 6 months had passed since the initial survey. If participants did not visit within 1 month of the 6-month mark, the researcher contacted them by phone and requested that they complete and return the survey by mail. Participants were considered lost to follow-up if the survey results were not obtained after two monthly contact attempts. To prevent interviewer bias, participants completed the survey in a setting without pharmacists, ensuring that individual responses could not be identified.
DATA COLLECTIONWe collected the following patient information through self-administered surveys at the aforementioned two time points: age, sex, weight, HbA1c level, number of family members living together, occupation, highest level of education, household income, duration since diagnosis of diabetes, history of dilated fundus examinations, history of diabetic retinopathy diagnosis, family history of diabetic retinopathy, details of diabetes treatment (medication, diet, and exercise therapy), comorbidities unrelated to diabetes, access to ophthalmic care, medication adherence as measured using the Visual Analogue Scale, stages of behavior change related to ophthalmic visits, and the type of healthcare facility visited for diabetes treatment (hospital with or without ophthalmology department, or private clinic). The information collected from pharmacies included the average number of prescriptions per day, the years of experience of the pharmacists, and whether they held diabetes-related certifications.
STATISTICAL ANALYSISThe sample size was calculated based on the hypothesis that the proactive educational intervention by community pharmacists using a study guide would increase the ophthalmology visits of individuals with diabetes compared to the control group. According to a previous educational study on individuals with diabetes, the intervention and control groups achieved 50% and 30% visits, respectively15). Referring to a previous cluster RCT conducted in community pharmacies, we set the intra-cluster correlation coefficient to 0.0521). With an expected cluster size of 16 pharmacies in each group, and using a type I error rate of 0.05 and 80% power to detect the aforementioned difference, the resulting sample size was determined to be 256 patients. To account for 15% attrition during the follow-up period, we planned to enroll a total of 300 patients (150 patients in each group). Based on a previous report indicating that around 50% of individuals with diabetes receiving pharmacological treatment visit ophthalmologists22), we planned to enroll around 10 participants per pharmacy at 32 sites and targeted pharmacies with at least 20 patients receiving diabetes medications annually. Ultimately, there were only 278 eligible participants who met the criteria among the 32 pharmacies included in the study; however, due to the high follow-up rate of 93.7%, the study included more than the required sample size of 256 cases, which allowed us to conclude recruitment.
The analysis was based on the intention-to-treat principle. Descriptive statistics were presented as means and standard deviations or medians and interquartile ranges for continuous variables, and as frequencies and percentages for categorical variables. The risk difference and risk ratio for the primary outcome were calculated using the generalized estimating equations model23),24) to account for the cluster units. A complete case analysis was performed with two adjusted models: model 1, adjusted for age and sex; and model 2, another adjusted for age, sex, and factors associated with diabetic retinopathy, such as duration since diagnosis of diabetes, HbA1c levels, use of insulin, history of diabetic retinopathy, and family history of diabetic retinopathy. These pre-specified adjusted variables were selected based on a known or clinically anticipated strong association between baseline characteristics and the primary outcome22). Moreover, a post-hoc analysis was conducted using Model 2, with further adjustments for variables related to socioeconomic status and access to ophthalmic care. For secondary outcomes, the mean changes from baseline to 6 months in both groups were calculated, and the mean difference between these changes was assessed.
All analyses were conducted using two-sided tests, with statistical significance set at p-value of less than 0.05. Analyses were performed using Stata (version 18.0; StataCorp, College Station, TX, USA).
A total of 278 patients were enrolled from 32 pharmacies between February 2023 and April 2024. Of these, 10 (3.6%) participants were excluded because they did not meet the following inclusion criteria: one was ≥75 years and nine had visited an ophthalmologist within the past year. Ultimately, 268 patients were included and randomized to the intervention (n = 133) and control (n = 135) groups (Fig. 1). Of the participants allocated to each group, data for the 6-month outcome was unavailable for 8 individuals in the intervention group (5 loss to follow up, 1 protocol deviation, and 2 declined to the answer the questionnaire), and 9 individuals in the control group also had unavailable data (9 loss to follow up).
Table 1 shows the baseline characteristics of the participants in each group. Totally, the mean age of the participants was 60.1 (10.7) years, and 35.8% (96/268) of the participants were women. The median duration since the diagnosis of diabetes was 5.2 (2.4–10.8) years, and the mean HbA1c level was 7.5 (1.6), with 11.2% (30/268) being insulin users. A total of 3.7% (10/268) of patients with a history of retinopathy had not received ophthalmological care for more than a year and were included in this study.
| Characteristic | Intervention Group (n = 133) |
Control Group (n = 135) |
ASB |
|---|---|---|---|
| Patients | |||
| Age, mean (SD), year | 60 (10.4) | 60 (11.0) | <0.01 |
| Sex, Female | 44 (33.1) | 52 (38.5) | 0.11 |
| Body weight, kg | 70.9 (13.1) | 70.0 (15.3) | 0.06 |
| Duration of diabetes, median (IQR), year | 4.9 (2.4–11.1) | 5.2 (2.3–10.3) | 0.05 |
| HbA1c, mean (SD) | 7.4 (1.7) | 7.5 (1.6) | 0.07 |
| Insulin usage | 15 (11.3) | 15 (11.2) | <0.01 |
| History of diabetic retinopathy | 4 (3.0) | 6 (4.8) | 0.09 |
| Family history of diabetic retinopathy | 9 (6.9) | 6 (4.6) | 0.10 |
| Comorbidities other than diabetes | |||
| Ocular diseases other than diabetic retinopathy | 1 (0.8) | 3 (2.2) | 0.12 |
| Hypertension | 72 (54.1) | 79 (58.5) | 0.09 |
| Dyslipidemia | 40 (30.1) | 49 (36.3) | 0.13 |
| Cardiovascular disease | 14 (10.5) | 14 (10.4) | <0.01 |
| Cerebrovascular disease | 10 (7.5) | 7 (5.2) | 0.10 |
| Kidney disease | 2 (1.5) | 3 (2.2) | 0.05 |
| Liver disease | 5 (3.8) | 4 (3.0) | 0.04 |
| Cancer | 7 (5.3) | 5 (3.7) | 0.08 |
| Others | 17 (12.8) | 17 (12.6) | <0.01 |
| Adherence to diabetes medication, mean (SD)b) | 9.3 (1.5) | 9.1 (1.9) | 0.07 |
| Time required to reach the ophthalmologist | |||
| <15min | 72 (62.6) | 61 (49.6) | 0.26 |
| 15min to 29min | 30 (26.1) | 48 (39.0) | 0.28 |
| 30min to 59min | 10 (8.7) | 11 (8.9) | <0.01 |
| 60min≤ | 3 (2.6) | 3 (2.4) | <0.01 |
| Educational attainment | |||
| Junior high school graduation | 9 (7.0) | 15 (11.3) | 0.15 |
| High school graduation | 77 (60.2) | 90 (67.7) | 0.16 |
| Associate degree | 14 (10.9) | 5 (3.8) | 0.28 |
| Bachelor’s degree or over | 28 (21.9) | 23 (17.3) | 0.12 |
| Annual household income | |||
| <¥3,000,000 | 41 (43.6) | 45 (39.8) | 0.08 |
| ¥3,000,000 to ¥4,999,999 | 30 (31.9) | 42 (37.2) | 0.11 |
| ¥5,000,000 to ¥6,999,999 | 11 (11.7) | 17 (15.0) | 0.10 |
| ¥7,000,000 to ¥9,999,999 | 9 (9.6) | 9 (8.0) | 0.06 |
| ¥10,000,000≤ | 3 (3.2) | 0 (0) | 0.26 |
| Employment situation – Employed | 91 (73.4) | 100 (78.7) | 0.13 |
| Lives alone | 27 (20.9) | 25 (18.7) | 0.06 |
| Pharmacies | |||
| Prescriptions/day <100 | 13 (81.3) | 12 (75.0) | 0.15 |
| Pharmacists’ years of experience, mean (SD) | 12.3 (8.7) | 10.3 (10.1) | 0.23 |
| Diabetes certification among pharmacists | 49 (55.7) | 29 (60.4) | 0.07 |
Note: Missing data on body weight in 6 (2.2%) participants, duration of diabetes in 48 (17.9%), HbA1c in 60 (22.4%) participants, insulin usage in 1 (0.4%) participants, history of diabetic retinopathy in 11 (4.1%) participants, family history of diabetic retinopathy in 8 (3.0%) participants, adherence to diabetes medication in 23 (8.6%) participants, time required to reach the ophthalmologist in 30 (11.2) participants, educational attainment in 6 (2.2%) participants, annual household income in 14 (5.2%) participants, employment situation in 17 (6.3%) participants, lives alone in 5 (1.9%) participants.
Abbreviations: ASB, Absolute Standardized Bias; IQR, interquartile range; SD, standard deviation.
a) Data are reported as number (percentage) of participants unless otherwise indicated.
b) Adherence was measured using a Visual Analog Scale.
In total, 19.8% (53/268) of the participants received ophthalmological care during follow-up period, with 18.8% (25/133) in the intervention group and 20.7% (28/135) in the control group (unadjusted risk difference, −0.03 [95% confidence interval [CI], −0.14 to 0.08]; unadjusted risk ratio, 0.88 [95% CI, 0.53 to 1.44]). In the multivariable model adjusted for age and sex, the adjusted risk difference and risk ratio were −0.03 (95% CI, −0.14 to 0.08) and 0.88 (95% CI, 0.54 to 1.45), respectively, with no significant difference observed (Table 2). Similarly, in the multivariable model adjusted for age, sex, and factors associated with diabetic retinopathy, no difference was found between the two groups (adjusted risk difference, −0.07 [95% CI, −0.21 to 0.08]; adjusted risk ratio, 0.73 [95% CI, 0.41 to 1.30]) (Table 2). Consistently, a post-hoc analysis—adjusted for variables related to socioeconomic status and access to ophthalmic care—also showed no difference between the two groups in terms of receiving ophthalmological care during the follow-up period (Supplement 3).
| Intervention Group (n = 133) |
Control Group (n = 135) |
|
|---|---|---|
| Primary outcome, n (%) | 25 (18.8) | 28 (20.7) |
| Crude | ||
| Risk Difference (95% CI) | −0.03 (−0.14 to 0.08) | reference |
| Risk Ratio (95% CI) | 0.88 (0.53 to 1.44) | reference |
| Adjusted model 1a) | ||
| Risk Difference (95% CI) | −0.03 (−0.14 to 0.08) | reference |
| Risk Ratio (95% CI) | 0.88 (0.54 to 1.45) | reference |
| Adjusted model 2b) | ||
| Risk Difference (95% CI) | −0.07 (−0.21 to 0.08) | reference |
| Risk Ratio (95% CI) | 0.73 (0.41 to 1.30) | reference |
Abbreviations: CI, confidence interval.
a) Model 1 adjusted for age and sex.
b) Model 2 adjusted for age, gender, and diabetic retinopathy-related factors (disease duration, HbA1c, insulin use, history and family history of diabetic retinopathy
The mean differences in change from baseline to 6 months in behavior for ophthalmic visits, dietary therapy, and exercise therapy across both groups were 0.07 (−0.23 to 0.37), −0.09 (−0.30 to 0.13), and 0.07 (−0.17 to 0.32), respectively, with no significant differences between the groups (Table 3). Similarly, the mean differences in change from baseline to 6 months in adherence, HbA1c levels, and body weight were 0.24 (−0.14 to 0.63), −0.28 (−0.76 to 0.20), and 0.17 (−0.83 to 1.83), respectively, with no significant differences between the groups (Table 3).
| Intervention Group (n = 133) |
Control Group (n = 135) |
Difference between Groups (95% CI) |
|
|---|---|---|---|
| Behavior changes regarding ophthalmic visits | |||
| Baseline | 0.49 (0.66) | 0.56 (0.66) | — |
| Change from baseline to 6 months | 0.19 (1.10) | 0.12 (1.15) | 0.07 (−0.23 to 0.37) |
| Behavior changes regarding dietary therapy | |||
| Baseline | 1.76 (0.94) | 1.65 (1.06) | — |
| Change from baseline to 6 months | −0.07 (0.77) | 0.02 (0.94) | −0.09 (−0.30 to 0.13) |
| Behavior changes regarding exercise therapy | |||
| Baseline | 1.35 (1.01) | 1.48 (1.05) | — |
| Change from baseline to 6 months | 0.02 (0.94) | −0.05 (1.00) | 0.07 (−0.17 to 0.32) |
| Adherence to diabetes medication | |||
| Baseline | 9.33 (1.39) | 9.36 (1.33) | — |
| Change from baseline to 6 months | 0.13 (1.15) | −0.11 (1.69) | 0.24 (−0.14 to 0.63) |
| HbA1c, % | |||
| Baseline | 7.47 (1.87) | 7.39 (1.14) | — |
| Change from baseline to 6 months | −0.33 (1.62) | −0.05 (1.57) | −0.28 (−0.76 to 0.20) |
| Body weight | |||
| Baseline | 70.7 (13.4) | 70.5 (15.4) | — |
| Change from baseline to 6 months | −0.53 (4.79) | −0.70 (2.84) | 0.17 (−0.83 to 1.18) |
a) Data are reported as mean (standard deviation) unless otherwise indicated.
This cluster RCT, involving 32 pharmacies, assessed the efficacy of community pharmacist-led educational interventions for individuals with diabetes who did not regularly visit ophthalmologists. Pharmacists who had completed standardized training on the basic knowledge of diabetic retinopathy conducted an active intervention using a guidebook on diabetic retinopathy to encourage patients to visit an ophthalmologist. However, this intervention did not result in an increase in the proportion of patients who visited ophthalmologists, nor did it lead to changes in behavior for ophthalmic visits or improvements in diabetes-related outcomes, such as HbA1c levels.
Worsening eye prognosis among individuals with diabetes due to improper visits to ophthalmologists has become a global concern, requiring the effective method to ensure timely ophthalmologic visit25–27). Several interventional studies have been conducted to encourage ophthalmology visits among individuals with diabetes who have not visited ophthalmologists appropriately. In a multicenter RCT conducted in the United States15), monthly educational sessions and cognitive behavioral therapy for diabetic retinopathy were implemented over 6 months for individuals with diabetes aged 65 years and older, which significantly increased the proportion of individuals undergoing retinal examination for diabetic retinopathy after 6 months (85.7% vs. 51.1%; risk difference 0.346 [0.20–0.46]). However, applying this intervention in real-world clinical settings is difficult and highlights the need for more practical and effective intervention strategies. Other RCTs examined whether distributing an educational pamphlet on diabetic retinopathy to individuals with diabetes could improve ophthalmology visit rates and found that pamphlet distribution alone did not effectively motivate patients to visit eye care28),29). In the current study, we evaluated the efficacy of a practical intervention, where community pharmacists used a detailed guidebook on diabetic retinopathy to conduct a brief, single-session educational intervention aimed at raising awareness of diabetic retinopathy. However, this pharmacist-led intervention did not lead to an increase in timely ophthalmology visits among individuals with diabetes who had not previously visited an ophthalmologist for over a year.
In this study, the educational intervention conducted by community pharmacists using a detailed guidebook on diabetic retinopathy for individuals with diabetes did not lead to an increase in timely ophthalmology visits or an improvement in diabetes-related outcomes. There are several possible explanations for these findings. First, our cluster-level randomization resulted in some imbalances in patient background factors related to financial, social, and geographical barriers. For example, although employment status and low income are key factors associated with decreased participation in eye screening30), the proportion of unemployed individuals and those with an annual household income of less than ¥3,000,000 (the lowest income category in our study) was higher in the intervention group in our study, potentially biasing the results toward an underestimation of the intervention effect. The intervention group had a higher proportion of participants with greater educational attainment than the control group, and a higher proportion who could reach an ophthalmologist in less than 15 minutes, both of which may have potentially led to an overestimation of the intervention effect. Although the post-hoc analysis adjusting for these factors related to socioeconomic status and access to ophthalmic care showed no significant intervention effect, these imbalances may have influenced the study results. Second, we prioritized the practicality of the education strategy, and the time allocated for the explanation was kept to a minimum to avoid significant deviations from the usual patient care in the community pharmacies. This may have resulted in the importance of timely ophthalmology visits not being effectively communicated within such a short period. Third, our study did not include follow-up with patients after the intervention. Previous studies demonstrating the effectiveness of interventions to improve ophthalmology visits among individuals with diabetes have shown that multiple reminders via phone calls delivered by bilingual interventionists16) or repeated face-to-face educational interventions conducted by community health workers15) are necessary. There is also a report that repeating brief three-minute educational interventions by pharmacists could impact on improving blood glucose control in individuals with diabetes31). Since our educational intervention was also conducted for three minutes, our educational intervention, which did not involve repeated education, may not have been sufficient to induce behavioral changes in ophthalmology visits or to improve diabetes-related outcomes. Furthermore, in Japan, visits to ophthalmologists are generally based on recommendations from physicians, and this aspect of the medical system and cultural norms may have contributed to the limited efficacy observed in our study. Given that individuals with diabetes regularly visit pharmacies for medications, it may be crucial to provide repeated education during these visits and to track their ophthalmology visit status while coordinating with their primary care physicians. Finally, since the effect was lower than expected, the potential lack of statistical power due to sample size limitations could be another reason for the null findings.
There were some limitations in this study. First, due to the nature of the study design, it was not possible to fully mask the patients and the pharmacists delivering the intervention. However, since we conducted randomization at the pharmacy level, neither the pharmacists nor the patients were aware of the detailed interventions assigned to the other groups. Second, as the outcomes were measured based on patient self-reports via questionnaires, the possibility of measurement errors in diabetes-related outcomes, such as the HbA1c level and body weight, cannot be ruled out. Third, there is a potential bias in the sample representativeness, as our study was conducted within a specific pharmacy group. However, this group covers a wide range of typical hospital types in Japan and handles prescriptions from all medical departments.
In this cluster RCT targeting individuals with diabetes who had not visited ophthalmologists, a single community pharmacist-led educational intervention for diabetic retinopathy prevention did not result in an increase in the proportion of ophthalmology visits, nor did it lead to changes in behavior for ophthalmic visits or improvements in diabetes-related outcomes. Practical and effective educational strategies to encourage ophthalmology visits for diabetic retinopathy are required.
AO is an employee of Total Medical Services Corporation. The other authors have no conflict of interest to disclose related to this research.
This work was supported by a research fund from the Total Medical Service Co., Ltd.
The authors would like to thank Tsukasa Kamitani and Shunichi Fukuhara for their suggestions and comments on this study.
Concept and design: all authors.
Acquisition, analysis, or interpretation of data: AO, SO, YY.
Drafting the work or revising: all authors.
Final approval of the manuscript: all authors.
Yosuke Yamamoto is one of the Editorial Board members of Annals of Clinical Epidemiology (ACE). This author was not involved in the peer-review or decision-making process for this paper.
