key: cord-0971622-glu9b0ay
authors: Hussein, Rand; He, Zhoushanyue; Bareham, Julia; Patel, Tejal; Killeen, Rosemary; Grindrod, Kelly
title: Evaluating the Effectiveness of the Computer-Based Education Platform, Pharmacy5in5, on Pharmacists’ Knowledge of Anticholinergic Toxicity Using a Randomized Controlled Trial
date: 2022-01-01
journal: Pharmacy (Basel)
DOI: 10.3390/pharmacy10010008
sha: 7ea616bf0b8a8fbe8248b641b3b6ec9ec8e12f19
doc_id: 971622
cord_uid: glu9b0ay

Background: Computer-based education has been widely implemented in healthcare professional development education. However, there has been little examination of the potential for computer-based education to enhance pharmacists’ knowledge. This study aims to assess the effectiveness of computer-based education on improving pharmacists’ knowledge compared to printed education material. Methods: This study was a web-based randomized controlled trial. Participants were randomly allocated to either an intervention group where they had access to the computer-based education module on Pharmacy5in5.ca or to a control group where they had access to printed educational material. Knowledge gain was assessed using a pre- and post-knowledge test. Results: A total of 120 pharmacists were recruited and 101 completed the post-knowledge test (50/60 in the intervention group; 51/60 in the control group). Both groups showed a significant increase in knowledge gain (intervention group: pre-test mean score 19.35 ± 3.56, post-test mean score 22.42 ± 3.812, p value < 0.001; control group pre-test mean score 19.22 ± 3.45, post-test mean score 23.29 ± 3.087, p value < 0.001). However, the difference in knowledge change was not significant between the two groups (22.42 vs. 23.29, p value = 0.333). Conclusions: In this study, a computer-based education module enhanced pharmacists’ knowledge to a similar degree to printed education material. Efforts should be made to provide computer-based education as an option to support pharmacists’ professional development.

Many commonly prescribed and over-the-counter medications such as antihistamines and tricyclic antidepressants have anticholinergic properties, contributing to a wide range of side effects such as dry mouth, delirium, and urinary retention [1] . The aging population, which is growing in Canada, is particularly vulnerable to these side effects due to comorbidities, polypharmacy and potential drug-drug interactions [2] [3] [4] . Using one anticholinergic medication may result in problematic side effects, but using multiple anticholinergic medications can create a cumulative anticholinergic burden leading to anticholinergic toxicity [5, 6] . Further, a 2015 systematic review by Salahudeen et al. [7] highlighted that a higher anticholinergic burden leads to a decline in cognitive ability and physical function among older adults. Moreover, anticholinergic toxicity can lead to emergency department admissions [4, 5] .

Pharmacists, as highly accessible healthcare professionals, are in an ideal position to identify patients at high risk of developing anticholinergic toxicity, and to provide or 

All potential participants were asked via email to complete an online 26-question pre-test and demographic survey. Participants who completed the baseline assessment were assigned a unique study identifier and randomly allocated to one of the two study arms. Participants allocated to the intervention group were sent a second email with a link to the Pharmacy5in5 platform, with access granted to an Anticholinergic Toxicity module. Participants were given one week to complete the one-hour Anticholinergic Toxicity module on Pharmacy5in5.ca. This module was designed specifically for this trial and had not been released to other platform users (only users enrolled in the intervention arm had access to this new module to reduce the risk of contamination [27] ). Users had the freedom to access all module components in any order they chose. No additional content development or refinement was performed on the platform or modules while conducting the trial. The intervention group had continuous access to the Pharmacy5in5 platform, which is available on computers, tablets, and smartphones. Given that recruited users were already registered users of the Pharmacy5in5 platform, no training was provided before or during the intervention on how to use the platform. Participants allocated to the control group were instructed via a second email to review printed education material, which included a paper-based decision aid, over the same one-week time frame.

After one week of accessing the Pharmacy5in5.ca module or the print education material, all participants were invited to complete a post-test via email. In the email, participants in the intervention group were asked to complete the module quizzes before taking the post-test. The post-test quiz included the same questions as the pre-test, but in a different order to assess knowledge gain. To access the post-test quiz, participants were asked to enter their study ID and were instructed to answer the questions to the best of their ability and not to consult other materials while taking the quiz. Throughout the study, up to four reminder emails were sent to prompt participants to complete the post-knowledge test.

Pharmacy5in5 is a computer-based education intervention that is freely available to all pharmacy professionals in Canada. The platform aims to accelerate the adoption of best practices by pharmacy professionals. On registration, users provide their demographic information along with consent to use their non-identified data for research purposes. The platform was developed through three cycles of user testing with pharmacists, pharmacy technicians, and students, where users were asked to interact with the platform while explaining their impressions, actions, and any areas of confusion. Users also completed the Systems Usability Scale (SUS) as a validated measure of usability.

Pharmacy5in5 is designed to regularly release modules that cover five take home messages about a clinical or pharmacy practice topic. Each module is created by a team of experts to provide real-world examples of how evidence, guidelines and practice tools can be used in daily practice. The platform also provides much needed information about trending topics. For example, in the recent "COVID-19" module, users were provided with latest evidence regarding the use of COVID-19 vaccines during pregnancy and breastfeeding. Each module has the following design components: One fast facts quiz with immediate feedback; six case-based quizzes, with delayed feedback [28] ; peer comparison; self-reflection questions to self-report behaviours; and multimedia resources • Quizzes: each module consists of one "Fast Fact quiz" and six "Case-based quizzes" that are based on the module learning objectives. The "Fast Fact quiz" tests users' basic knowledge about the clinical topic, while "Case-based quizzes" test users' knowledge and behaviour through helping a fictional pharmacist make the optimal decision in cases based on real-life scenarios. Questions used are multiple choice or true/false. • Feedback: in each module, online quizzes are followed with feedback on users' performance. Each "Fast Fact quiz" provides users with immediate feedback after each question to let them know if they got the right answer. "Case-based quizzes" provide delayed feedback after the user finishes the whole quiz series, along with the correct answer, rationale, and links to useful resources underneath each question. • Peer comparison: this feature aims to compare a user's performance to an average user and provides feedback on user progress. Peer comparison also rewards users who outperform an average user, and motivates users who perform below average. • Self-reflection: after each case-based quiz, users are asked to answer a question about their past behaviour. Self-reflection questions are structured as follows: "in the past 3 months have you/did you . . . ". This feature allows users to self-report their behaviours, which can help detect any changes in users' behaviour before and after taking the module. • Multimedia resources: there is a set of specific resources created for each module, including: flashcards, infographics, and short videos. These multimedia resources highlight the key take home message for each clinical topic and can be shared on social media.

The platform was designed using a combination of the Theoretical Domains Framework (TDF), the COM-B (capability, opportunity, motivation, and behaviour) model, and the Behaviour Change Wheel (BCW), to understand factors that influence pharmacist behaviour. The TDF includes 14 domains for potential determinants of behaviour (e.g., knowledge, skills, beliefs about capabilities, etc) [28] . The COM-B model demonstrates how an individual's capability, opportunity, and motivation impact behaviour, and can be linked back to specific TDF domains [29] . Based on the COM-B, Pharmacy5in5 aims to change pharmacist behaviour by targeting pharmacist capability. This corresponds to TDF be-haviour change domains such as knowledge, skills, decision processes, and behavioral regulation [29] . The BCW shows how the COM-B domains can be targeted by a variety of behavioural interventions and policies [30] . Pharmacy5in5 uses three behavioural interventions, including modelling (learning through imitation), education (gaining knowledge), and training (gaining skills).

The printed education material used for this study was provided as an 11 pager PDF document and included two references: (1) RxFiles: Reference List of Drugs with Potential Anticholinergic Effects which is a paper-based decision aid [31] ; and (2) Anticholinergic Toxicity review by Broderick et al., 2020 [5] . These two references were used in the development of the Pharmacy5in5 online module content. The RxFiles drug charts are unique decision aids used by Canadian pharmacists and physicians in practice. The charts provide evidence-based information about drugs to help providers make therapy decisions [32] .

To create the Anticholinergic Toxicity module, the Pharmacy 5in5 content developers started by drafting learning objectives (Box 1). Next, the learning objectives were used to develop two educational infographics ( Figure 2 ), one short animated video (1:06 min long), and five flash cards ( Figure 3 ). The content developers also developed an introductory immediate feedback quiz and six case-based delayed feedback quizzes to address the learning objectives. Next, all the module components were reviewed through two rounds with a panel of three experts. In the first round, the content of the module was shared with the panel via email to gather comments and critiques for each question and all multimedia resources, and modifications were made accordingly. A total of seventeen questions were modified and the two infographics were clarified based on expert feedback. For the second round, the full module was shared after modifications and 12 questions were reworded slightly as per expert feedback. Before launching the study, the module was piloted with five practising pharmacists to assess any concerns with readability and clarity of wording.

The main outcome measure was the difference in the post-test quiz score between the intervention and control group using pre-post validated knowledge tests. A secondary outcome measure was the change in the intervention group's and control group's quiz scores, before and after the intervention.

With a power of 80% and a two-tailed alpha of 0.05, an estimated 45 users were needed for each group to show a 5% difference in knowledge between the two arms post-test scores (standard deviation 1.23 based on a previous pilot study, 2-sided test, p < 0.05, power of 80%). Taking into consideration the high dropout rate with internetbased interventions [24, 33] , we enrolled 120 participants to account for an expected 25% dropout rate.

After enrollment, eligible participants were randomised using a 1:1 allocation ratio via a computer-generated random number by an independent research member who was not involved in the data analysis. Given the nature of the study, it was not feasible to blind the participants. However, only the primary investigator was aware of the allocations. The study team was blinded to allocation until the analysis was completed. Recognize anticholinergic side effects when assessing patients, including anticholinergic toxicity.

Identify higher risk anticholinergic drugs, including additive effects.

Classify patients according to their risk for serious anticholinergic side effects.

Identify safer alternatives for indications where anticholinergic drugs are commonly used.

Implement a plan to switch an anticholinergic drug to a safer alternative, deprescribing, tapering, switching, and incorporating washout periods.

Average scores were computed for the primary outcome (knowledge score out of 29) and reported as mean ± SD. The normality of pre-and post-knowledge test scores were tested using Kolmogorov-Smirnov and Shapiro-Wilk test. As the data was found to be skewed, nonparametric tests were used to compare the outcomes between the control and intervention groups using Mann-Whitney U test and Wilcoxon Signed ranks tests to compare within groups at before and after the intervention [18] . Additional analyses included assessing differences in baseline sociodemographic factors between the control and intervention groups, and between respondents and non-respondents using Chi-square and Fisher's exact test. p < 0.05 was considered statistically significant. Statistical analysis was performed using SPSS version 27 statistical package (IBM Corp, New York, NY, USA) [34] .

To assess if the use of the platform results in knowledge gain, a knowledge test was developed as outlined by Case and Swanson [35] . The first draft of knowledge test included a total of 33 questions that were developed based on the module learning objectives. The knowledge test questions were different from the questions in the module quizzes and included a combination of clinical vignette questions and standard questions. The questions were constructed to address Blooms' Taxonomy lower and higher order thinking levels. First, the knowledge test was reviewed by two senior content developers who were asked to code each question to the corresponding learning objective and Bloom's level to ensure that there were at least three questions addressing each learning objective as well as an equal number of questions assessing Bloom's Taxonomy lower and higher order thinking levels.

The knowledge test was validated with geriatric pharmacotherapy experts in two rounds. In the first round, the experts were asked to rate each question in terms of relative importance to the stated learning objectives (1 = Not relevant; 2 = Somewhat relevant; 3 = Quite relevant; 4 = Very relevant). They were also instructed to provide their feedback and comments to improve the questions. Agreement among experts was used to calculate the content validity index for each item (item content validly index [i-CVI]), where items with i-CVI values equal or less than 0.79 required revisions, and i-CVI values less than 0.70 were eliminated [36, 37] . The first round of content validation was conducted with seven experts who reviewed the 33 questions. After revision, the second draft was composed of 26 questions (23 multiple choice questions; 3 open-ended questions). Seven questions were deleted because they were considered repetitive or did not fit with any of the learning objectives. A total of 12 questions were revised or reworded to improve clarity. Of the 33 questions, only six questions had a low level of agreement among experts (i-CVI = 71.4) (see Appendix A).

For the second round, the revised version of the knowledge test with 26 questions was shared with the experts to gather their feedback and collect any additional comments. Only three experts agreed to participate in the second round. Minor grammatical modifications and comments were recommended for seven questions.

The knowledge test internal consistency and item difficulty index were assessed using the first 50 responses [38] . Items with an item difficulty index less than 0.2 or higher than 0.8 were eliminated or revised [37] . The internal consistency of the 26 items, using Cronbach's alpha (α), was 0.73. In terms of the item difficulty index, four items had a difficulty index of 0.9 or above and were deleted to allow for differentiation in learning. Two items had a difficulty index less than 0.2, the two items were retained and one of them was revised for clarify (10.a). The final draft of the knowledge test included 22 items:19 multiple-choice questions (MCQs) and 3 open-ended questions (see Appendix B). The final score was calculated out of 29 as follows: The score of the MCQ questions (out of 19) was automatically generated via Qualtrics, and the score for the open-ended questions (out of 10) was calculated based on a pre-validated grading rubric.

All pharmacists practising in Ontario who were registered users of Pharmacy5in5.ca were invited to participate in the trial. A total of 120 users agreed to participate and were randomized to the intervention group (60 participants) and the control group (60 participants). No statistically significant differences were found between the demographic characteristics of the two study groups (Table 1) . Most respondents were female (78.3%), held a bachelor's degree (63.3%), received their training in Canada (72.5%), had more than 10 years of pharmacy practice experience (58.3%), and were practising in community pharmacies (70.8%). In terms of previous training, 96% of participants denied receiving any training related to anticholinergic toxicity in the past 12 months. 

Of the 120 pharmacists who were randomized, all completed the pre-test and 101 (84.2%) completed the post-test. In particular, the post-test was completed by 83.3% (50/60) of the intervention group, and 85% (51/60) of the control group. There was no significant difference in baseline demographics of respondents and nonrespondents in the intervention group. There was no significant difference in the pre-test scores between respondents and nonrespondents in the intervention group (pre-test mean score 19.64 ± 3.49; pre-test mean score 17.9 ± 3.66; p value = 0.159). Similarly, in the control group, there was no significant difference in baseline demographics of respondents and nonrespondents, and no significant difference in the pre-test scores between respondents and nonrespondents (pre-test mean score 19.22 ± 3.37; pre-test mean score 19.22 ± 4.08; p value = 0.699)

Out of the 60 pharmacists allocated to the intervention group, 41 (68.3%) accessed the module and completed all seven quizzes, one user completed six out of the seven quizzes, one user completed five out of the seven quizzes, and 15 (25%) did not access the module at all. Of the 15 who did not access the module, seven completed the post-test.

On average, pharmacists in the intervention group spent 40 min to complete the quizzes of the module in a single session. Response data also showed that 15/43 (34.8%) of pharmacists completed the knowledge post-test within one day of completing the module quizzes, 26/43 (60%) completed the knowledge post-test after one week of completing the module, and 2/43 (4.6%) completed the knowledge test after two weeks of completing the module quizzes.

As shown in Table 2 , the mean score for the knowledge pre-test was 19.35 ± 3.56 (67%) in the intervention group and 19.22 ± 3.45 (66%) in the control group. There was no significant difference in the knowledge pre-test mean scores between the two groups (Mann-Whitney U: Z value = −0.018, p = 0.985). The mean score of the knowledge post-test increased significantly to 22.42 ± 3.812 (77%) in the intervention group and 23.29 ± 3.087 (80%) in the control group (Wilcoxon Signed Rank: Intervention group: Z value = −4.690, p < 0.001; Control group: value = −5.180, p < 0.001). However, there was no significant difference between the two study groups in the knowledge post-test (Mann-Whitney U: Z value = −0.968, p = 0.333). In terms of individual questions, the two questions that had the lowest performance were those that addressed classifying patients according to their risk for serious anticholinergic side effects. In particular, only 16% of the participating pharmacists were able to identify that low dose doxepin (3 mg) is not a risk factor for serious anticholinergic side effects. Moreover, only 16.8% recognized that Parkinson's disease is not a risk factor for serious anticholinergic side effects. See Appendices C and D for subgroup analysis of post-test knowledge questions.

This randomized controlled trial assessed Ontario pharmacists' knowledge of anticholinergic toxicity after completion of a computer-based education module. One of the key findings is that the computer-based education module improved pharmacists' knowledge significantly (p < 0.001). This result is consistent with previous studies conducted among nurses and physicians [39] [40] [41] . Moreover, the study showed that there was no difference between a paper-based decision aid by RxFiles and a computer-based education module in improving pharmacists' knowledge of anticholinergic toxicity, as there was no significant difference in post-test scores between the study groups (p = 0.208). Surprisingly, the study found that the topic of anticholinergic toxicity is not commonly addressed in pharmacy professional trainings with only 4% of participants reporting having received training related to the topic in the past 12 months. Given that anticholinergic medications are problematic in an aging population, and that the primary users of medications are the elderly, pharmacists are required to have the ability to address any drug-related problems that arise with the use of these medications. The Pharmacy5in5 platform offers busy pharmacists the opportunity to easily access modules via an online platform and test their knowledge and receive feedback on their behaviours.

This study contributes to existing literature with few studies comparing the effect of computer-based education to printed educational material among pharmacists. In a recent scoping review of the effect of computer-based education on health care providers' knowledge skills and behaviour, a total of 14 studies assessed knowledge gain [19] . Only one study targeting pharmacists was identified. Nesterowicz et al. [18] compared computerbased education to a two-hour on-site session on improving pharmacists' knowledge of blood pressure monitoring. The authors assessed knowledge gain after six months. Similar to our study, the increase in knowledge in both groups was significant by 29% in the intervention group and 27.2% in the control group, but the groups did not differ significantly, indicating that online learning is as effective as on-site learning. Pharmacists in the study had a low pre-test scores indicating more room for improvement in their knowledge. Despite the difference in topic addressed, this indicates that computer-based education could promote knowledge retention among pharmacists and that there should be a move towards readily available, easily accessible, self-directed computer-based education, especially during the COVID-19 pandemic.

While this study indicated that computer-based education is an appropriate intervention to educate pharmacists on anticholinergic toxicity, the knowledge gained is not necessarily translated into clinical practice [42] . A successful behaviour change intervention should address barriers and facilitators to promote practice change [43] . As such, computer-based education could alter pharmacists' behaviour by addressing unique barriers to reducing anticholinergic burden. A 2021 systematic review by Stewart et al. [44] identified a number of barriers among pharmacists including lack of confidence in skills and uncertainty about the need for reducing anticholinergic burden. The Anticholinergic Toxicity module used in this study could help pharmacists overcome these barriers through the six case based quizzes to build confidence, and the animated video emphasizing the consequences of anticholinergic toxicity [45] . Therefore, computer-based education could be tailored to address barriers using different design components, making it a promising intervention to optimize the use of anticholinergic medications

A strength of this study was a low dropout rate of 15.8% despite being conducted during the COVID-19 pandemic, as compared to other studies among healthcare professionals that reported dropout rates of 31% [46] and 47% [47] . This suggests that there is high feasibility for conducting online interventions with pharmacists. The dropout rate in both the intervention and control groups was similar, which suggests that both the printed and computer-based education platform were desirable. A second strength of the study was the triangulation of the post-test knowledge scores with the usage of the platform, where 86% of the users who completed the post-test had completed the module. Another study strength was the low level of contamination bias as only 1 out of 101 users reported that they knew other pharmacists who participated in the study, and 9 users were unsure. However, all users reported that they did not discuss their group allocation with a colleague, nor knew whether they were assigned to a different group. A potential limitation is that the content of the education modules was focused on a single topic (anticholinergic toxicity), with a plan for future work to assess other pharmacy practice-related content to demonstrate the generalizability of the effect of computer-based education on pharmacists' knowledge. Another limitation, is the limited ability to control or monitor the use of additional resources while completing the post-tests. Moreover, there was limited ability to confirm whether to the control group users completed the paper-based material before completing the post-knowledge tests. Another potential limitation is selection bias, with the possibility that only pharmacists who were interested in the topic agreed to participate in the study [48] .

This study provided much needed information about the effectiveness of the computerbased education platform in improving pharmacists' knowledge. The results also act as a feasibility study for a future study of how the website supports behaviour change and practice outcomes. Pharmacy5in5 is anticipated to be a highly impactful platform with the potential to advance the field of professional continuing education for pharmacists.

This study evaluated the effectiveness of a computer-based education module compared to a paper-based education in improving pharmacists' knowledge of anticholinergic toxicity. The study highlighted that computer-based education can improve pharmacists' knowledge similarly to printed education material. Further research is needed to assess the potential of computer-based education in improving pharmacists' knowledge in nonclinical topics such as ethics and policies, as well as assessing if knowledge gained from computer-based education is retained and implemented in pharmacy practice. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. (Q9) M.K., a 52-year-old female patient, comes to the pharmacy to ask for your recommendation for the appropriate dose for loperamide. She has been experiencing mild diarrhea for the past two days. She is not taking any chronic medications or OTC drugs other than acetaminophen for an occasional headache. 

Cognitive risks of anticholinergics in the elderly

Drugs with anticholinergic properties: A current perspective on use and safety

Potentially inappropriate anticholinergic medication use in communitydwelling older adults: A national cross-sectional study

Anticholinergic toxicity, polypharmacy, and inappropriate prescribing

Anticholinergic Toxicity. StatPearls 2020

Drugs contributing to anticholinergic burden and risk of fall or fall-related injury among older adults with mild cognitive impairment, dementia and multiple chronic conditions: A retrospective cohort study

Anticholinergic burden quantified by anticholinergic risk scales and adverse outcomes in older people: A systematic review

Interventions to reduce anticholinergic burden in adults aged 65 and older: A systematic review

Optimizing medication appropriateness in older adults: A randomized clinical interventional trial to decrease anticholinergic burden. Alzheimer's Res

Education to reduce potentially harmful medication use among residents of assisted living facilities: A randomized controlled trial

Printed educational materials: Effects on professional practice and healthcare outcomes

Education and assessment of pharmacists on the use of the drug burden index in older adults using a continuing professional development education method

Effectiveness of E-learning in pharmacy education

Post-Registration Learning Trends of Community Pharmacists

Evaluation of the impact of a continuing professional development worksheet on sustained learning and implementing change after a continuing pharmacy education activity

Using technology to meet the challenges of medical education

Interprofessional online learning for primary healthcare: Findings from a scoping review

Learning in Continuing Pharmacy Education is effective and just as accepted as on-site learning

Effects of computer-based education on health professionals' knowledge, skills, and behavior: A scoping review

Evidence for simulation in pharmacy education

Doctor of pharmacy student preferences for computer-based vs. paper-and-pencil testing in a, social and administrative pharmacy course

Evaluation of Computer-based Simulation Learning on Knowledge, Learning Approaches and Motivation among Pharmacy Students

A Narrative Overview of the Effectiveness of E-learning in Pharmacy Education

Improving and standardizing evaluation reports of Web-based and mobile health interventions

Methods to increase response to postal and electronic questionnaires

How to obtain excellent response rates when surveying physicians

A scoping review of the problems and solutions associated with contamination in trials of complex interventions in mental health

Making psychological theory useful for implementing evidence based practice: A consensus approach

Validation of the theoretical domains framework for use in behaviour change and implementation research

The behaviour change wheel: A new method for characterising and designing behaviour change interventions

ANTICHOLINERGICS: Reference List of Drugs with Potential Anticholinergic Effects

E-learning for health professionals

IBM SPSS Statistics for Windows V, 27.0; IBM Corp: Armonk

Constructing Written Test Questions for the Basic and Clinical Sciences

Design and implementation content validity study: Development of an instrument for measuring patient-centered communication

Development and validation of a new tool to measure the facilitators, barriers and preferences to exercise in people with osteoporosis

Health Measurement Scales: A Practical Guide to Their Development and Use

An evaluation of general practice nurses' knowledge of chronic kidney disease risk factors and screening practices following completion of a case study-based asynchronous e-learning module

Can an e-learning course improve nursing care for older people at risk of delirium: A stepped wedge cluster randomised trial

Behavioural change in primary care professionals undertaking online education in dementia care in general practice

Moles, R. Educational interventions to improve prescribing competency: A systematic review

The role of behavioral science theory in development and implementation of public health interventions

Barriers and facilitators to reducing anticholinergic burden: A qualitative systematic review

Identifying barriers, facilitators and behaviour change techniques to the adoption of the full scope of pharmacy practice among pharmacy professionals: Using the Theoretical Domains Framework

Breakeven, cost benefit, cost effectiveness, and willingness to pay for web-based versus face-to-face education delivery for health professionals

Improving Care After Myocardial Infarction Using a 2-Year Internet-Delivered Intervention: The Department of Veterans Affairs Myocardial Infarction-Plus Cluster-Randomized Trial

Identifying and avoiding bias in research