Effectiveness of a Smartphone App for Guiding Antidepressant Drug Selection Colin Man; Cathina Nguyen; Steven Lin, MD BACKGROUND AND OBJECTIVES: Major depression is a prevalent chronic disease in the United States. However, many physicians lack access to decision support tools at point of care to help choose antidepressants in a rational, evidence-based manner. A patient-centered treatment model that uses a symptom-based approach to selecting antidepressants was developed into a smartphone application to provide instant, evidence-based recommendations and drug monographs. The purpose of this study was to assess the impact of this mobile application on the confidence level of family physicians in treating depression. METHODS: The smartphone application was provided to 14 family medicine residents and attending physicians from the O Connor Family Medicine Residency Program in San Jose, CA. Participants were asked to use the software as drug reference and clinical decision support during patient care activities. Three surveys were administered over a 12-week period to assess provider characteristics, outcome measures (ie, confidence in managing depression and choosing an initial antidepressant based on patient symptoms, medical comorbidities, potential side effects, and drug interactions), and fund of antidepressant knowledge. RESULTS: The average confidence levels in managing depression, starting an antidepressant on a patient with depression, and choosing an initial antidepressant based on patient symptoms increased significantly within the period of smartphone application usage. The average scores on the antidepressant knowledge tests also improved. CONCLUSIONS: The smartphone application was an effective tool for both increasing confidence in depression treatment and educating physicians. Future studies to evaluate the effectiveness and impact of smartphone applications on medical education and postgraduate training are warranted. (Fam Med 2014;46(8):626-30.) Unipolar major depressive disorder (MDD) is a chronic disease that affects more than 30 million adults in the United States. 1 It is the second leading cause of years lived with disability 2 and the fifth leading cause of disabilityadjusted life years in the nation. 3 Over a dozen different antidepressants have been developed, many in the past 2 decades, to treat MDD. 4 Although general treatment guidelines for MDD exist, comparative effectiveness studies of antidepressants have produced mixed results. 5 Recently, some family physicians have proposed an individualized, patient-centered treatment model for MDD that uses a symptom-based approach to selecting antidepressants. 6 This model was developed into a smartphone application for the purpose of providing instant, evidencebased recommendations and detailed drug monographs to primary care physicians at the point of care. 7 The objective of this pilot study was to evaluate the effectiveness of this smartphone application in increasing the confidence level of residents and attending physicians when treating MDD in a family medicine residency program. Methods This prospective study was conducted over a 12-week period from February 2013 to May 2013. In total, 14 physicians (five attendings and nine residents) from the O Connor Family Medicine Residency Program in San Jose, CA, were recruited on a voluntary basis. The only inclusion criterion was that all participants must have an iphone. The O Connor Hospital Institutional Review Board From the School of Humanities and Sciences, Stanford University, Stanford, CA (Mr Man); Department of Public Health Sciences, University of California-Davis (Ms Nguyen); and Center for Education and Research in Family and Community Medicine, Division of General Medical Disciplines, Department of Medicine, Stanford University (Dr Lin). 626 SEPTEMBER 2014 VOL. 46, NO. 8 FAMILY MEDICINE
exempted this study from formal review. Baseline Survey A 15-item baseline survey was administered to assess provider characteristics (ie, years of clinical experience, frequency of contact with MDD patients, experience with smartphone applications), outcome measures (ie, confidence in managing MDD and choosing an initial antidepressant based on patient symptoms, medical comorbidities, potential side effects, and drug interactions), and overall fund of knowledge using five true or false questions about antidepressants (Table 1) developed for this purpose. After completion of the baseline survey, each participant downloaded a copy of the smartphone application (available for free online at www.linvivo. com). Participants were encouraged to use the software for drug reference and clinical decision support in whatever manner they chose to supplement their usual care during patient care activities. Follow-Up Surveys A halfway survey was administered after 6 weeks and a final survey after 12 weeks of real-world access to the smartphone application. These surveys contained identical questions to the baseline survey in terms of outcome measures (ie, confidence in managing MDD) and fund of knowledge. Of the 14 physicians who started the study, nine (64%) completed the halfway survey, and 11 (79%) completed the final survey. Statistical Analysis Statistical analysis was performed using Minitab 15 Statistical Software for Windows. Survey data was coded using a Likert scale of 1 to 7 and then analyzed for significant differences between baseline, halfway point (6 weeks), and final point (12 weeks) using non-parametric tests, which assumed non-normal distribution populations. Fund of knowledge assessments were checked for correctness and then coded based on percentage of correct answers. Data were analyzed for significant differences between baseline and final point using a two-sided t test with pooled standard deviation. Participant characteristics data were analyzed for significant differences between baseline, halfway point, and final point using two-sided t tests. All t tests were designed to have 95% confidence interval. Results Participant characteristics are summarized in Table 2. Overall, the average confidence ratings of the study participants in managing MDD, starting an antidepressant on a patient with newly diagnosed MDD, and choosing an antidepressant based on patient factors (ie, presenting symptoms, medical comorbidities, potential side effects, and drug interactions) all improved significantly after 12 weeks with use of the smartphone application (Table 3). The improvements occurred with longer access to the application over time. In addition, the average scores on the antidepressant knowledge assessment quizzes also increased among smartphone application users (Figure 1), although not significantly due to small sample size. No statistically significant differences in participant characteristics (ie, years of clinical experience, frequency of contact with MDD patients, experience with smartphone applications) were detected. Discussion The findings from this small pilot study suggest that a smartphone application designed to provide instant, evidence-based recommendations for antidepressant selection and drug monographs to primary care physicians at the point of care may increase confidence in managing MDD. Physicians are more confident when they are not relying solely on their training but also on readily available clinical decision support tools such as smartphone applications. While studies on computerized decision support tools have been conducted before, 8 none have assessed the impact of mobile applications on physician confidence. Increasing the confidence of family physicians in managing MDD is crucial when nearly three quarters of Americans who seek care for MDD go to a primary care physician rather than a mental health specialist. 9 Although our study was small, statistically significant improvements in confidence levels of the study participants in managing MDD, starting an antidepressant on a patient with newly diagnosed MDD, and choosing an antidepressant based on patient factors were detected. The increase in confidence over a relatively short Table 1: True or Questions in the Knowledge Assessment Surveys Question There are more than half a dozen classes of antidepressants. All antidepressants need at least 4 weeks to take effect. Selective serotonin reuptake inhibitors (SSRIs) can be sedating, activating, or neutral. No antidepressant has been shown to be safe for breast-feeding. Newer generation antidepressants are proven to be more effective than tricyclic antidepressants (TCAs) in inducing remission in patients with unipolar major depression. Correct Answer True True FAMILY MEDICINE VOL. 46, NO. 8 SEPTEMBER 2014 627
Table 2: Participant Characteristics Characteristics No. (n=14) Years of clinical experience PGY 1 4 PGY 2 2 PGY 3 3 <5 post-residency 2 5 10 post-residency 1 10 20 post-residency 1 >20 post-residency 1 Estimated % of clinics where provider manages at least one patient with major depression <10 2 10 25 4 25 50 4 50 75 1 75 90 2 >90 1 Estimated % of clinics where provider uses a smartphone application <10 1 10 25 0 25 50 1 50 75 3 75 90 4 >90 5 Table 3: Comparison of Mean Physician Confidence Levels Before and After Access to the Smartphone Application Questions* Baseline After 6 Weeks P Value After 12 Weeks P Value Confidence in managing outpatient adults with major depression 4.214 4.444.607 5.364.048 Confidence in starting an antidepressant for newly diagnosed major depression 4.286 4.778.397 5.636.018 Confidence in choosing an antidepressant based on patient factors 3.642 4.889.034 5.273.010 * All ratings are on a 7-point Likert scale with 1 being Not Very Confident, 4 being Neutral, and 7 being Very Confident. 628 SEPTEMBER 2014 VOL. 46, NO. 8 FAMILY MEDICINE
Figure 1: Comparison of Mean Percentage of Correct Answers in the Knowledge Assessment Surveys Before and After Access to the Smartphone Application There were no significant differences in the mean total scores, with P value.205. time period suggests that a larger study will likely produce more relevant findings. Overall, the lack of published literature on the use of smartphone applications to aid clinical decision making in primary care suggests a need for further research to evaluate effectiveness, test feasibility, and identify barriers to implementation. Our finding that smartphone applications may increase clinical confidence of residents and attending physicians is unlikely to be limited to MDD treatment. Smartphone applications can be programmed with any clinical topic to provide evidence-based recommendations at the point of care. As the entire medical community moves toward an era of mobile-based decision support applications, 10 primary care physicians will increasingly use these tools to enhance the quality of care they provide. Smartphone applications can also be a powerful tool for educating resident physicians, whose busy schedules often preclude reading medical journals and keeping current on the most up-to-date practice guidelines. The ability of smartphone applications to be updated automatically as soon as new information is released provides a method for evidence to be disseminated rapidly in a format that is readily accessible at the point of care. Future studies to evaluate the effectiveness and impact of smartphone applications on medical education are needed. CORRESPONDING AUTHOR: Address correspondence to Dr Lin, Stanford University, Department of Medicine, 211 Quarry Road, Suite 405, Palo Alto, CA 94304. 650-725-7966. Fax: 650-498-7750. stevenlin@stanford.edu. ACKNOWLEDGMENTS: Conflict disclosure: Colin Man and Steven Lin, MD, created the smartphone app in this study. The app is available free of charge online. There is no financial conflict of interest. We thank the residents and attending physicians of the San Jose-O Connor Family Medicine Residency Program for their participation in this study. Smartphone application disclosure: Antidepressant rankings were determined based on the most up-to-date comprehensive literature review. Although current evidence does not favor particular antidepressants on the basis of efficacy alone, clinically important differences in drug side effects were considered in the overall ranking and used to match medications with patient symptom clusters. Where evidence was unavailable, expert opinion and usual practice were considered. Drug rankings are only recommendations based on the available evidence and do not replace provider judgment and clinical experience. Monographs of drugs were adapted from PDR and Lexicomp, and drug prices were updated from US Consumer Reports. References 1. Kessler RC, Berglund P, Demler O, et al; National Comorbidity Survey Replication. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R). JAMA 2003;289(23):3095-105. 2. US Burden of Disease Collaborators. The state of US health, 1990-2010: burden of diseases, injuries, and risk factors. JAMA 2013;310(6):591-608. FAMILY MEDICINE VOL. 46, NO. 8 SEPTEMBER 2014 629
3. Murray CJ, Lopez AD. Measuring the global burden of disease. N Engl J Med 2013;369(5):448-57. 4. Adams SM, Miller KE, Zylstra RG. Pharmacologic management of adult depression. Am Fam Physician 2008;77(6):785-92. 5. Gartlehner G, Thaler K, Hill S, Hansen RA. How should primary care doctors select which antidepressants to administer? Curr Psychiatry Rep 2012;14(4):360-9. 6. Lin SY, Stevens MB. The symptom clusterbased approach to individualize patientcentered treatment for major depression. J Am Board Fam Med 2014;27:151-9. 7. Linvivo. L Allegro. Available at http://www. linvivo.com. Accessed September 1, 2013. 8. Trivedi MH, Daly EJ, Kern JK, Grannemann BD, Sunderajan P, Claassen CA. Barriers to implementation of a computerized decision support system for depression: an observational report on lessons learned in real world clinical settings. BMC Med Inform Decis Mak 2009;9:6. 9. Montano CB. Recognition and treatment of depression in a primary care setting. J Clin Psychiatry 1994;55:18-34. 10. Lippman H. How apps are changing family medicine. J Fam Pract 2013;62(7):362-7. 630 SEPTEMBER 2014 VOL. 46, NO. 8 FAMILY MEDICINE