Well-Being in Graduate Medical Education: A Call for Action

Job burnout is highly prevalent in graduate medical trainees. Numerous demands and stressors drive the development of burnout in this population, leading to significant and potentially tragic consequences, not only for trainees but also for the patients and communities they serve. The literature on interventions to reduce resident burnout is limited but suggests that both individual- and system-level approaches are effective. Work hours limitations and mindfulness training are each likely to have modest benefit. Despite concerns that physician trainee wellness programs might be costly, attention to physician wellness may lead to important benefits such as greater patient satisfaction, long-term physician satisfaction, and increased physician productivity. A collaborative of medical educators, academic leaders, and researchers recently formed with the goal of improving trainee well-being and mitigating burnout. Its first task is outlining this framework of initial recommendations in a call to action. These recommendations are made at the national, hospital, program, and nonwork levels and are meant to inform stakeholders who have taken up the charge to address trainee well-being. Regulatory bodies and health care systems need to be accountable for the well-being of trainees under their supervision and drive an enforceable mandate to programs under their charge. Programs and individuals should develop and engage in a "menu" of wellness options to reach a variety of learners and standardize the effort to ameliorate burnout. The impact of these multilevel changes will promote a culture where trainees can learn in settings that will sustain them over the course of their careers.

Publication Rates of Abstracts Presented at the Society of General Internal Medicine Annual Meeting

BACKGROUND

Abstracts accepted at scientific meetings are often not subsequently published. Data on publication rates are largely from subspecialty and surgical studies.

OBJECTIVE

The aims of this study were to 1) determine publication rates of abstracts presented at a general internal medicine meeting; 2) describe research activity among academic general internists; 3) identify factors associated with publication and with the impact factor of the journal of publication; and 4) evaluate for publication bias.

DESIGN

Retrospective cohort study.

PARTICIPANTS

All scientific abstracts presented at the Society of General Internal Medicine 2009 Annual Meeting.

MAIN MEASURES

Publication rates were determined by searching for full-text publications in MEDLINE. Data were abstracted regarding authors' institution, research topic category, number of study sites, sample size, study design, statistical significance (p value and confidence interval) in abstract and publication, journal of publication, publication date, and journal impact factor.

KEY RESULTS

Of the 578 abstracts analyzed, 274 (47.4%) were subsequently published as a full article in a peer-reviewed journal indexed in MEDLINE. In a multivariable model adjusting for institution site, research topic, number of study sites, study design, sample size, and abstract results, publication rates for academic general internists were highest in the areas of medical education (52.5%, OR 5.05, 95% CI 1.57-17.25, reference group Veterans Affairs (VA)-based research, publication rate 36.7%), mental health/substance use (67.7%, OR 4.16, 95% CI 1.39-13.06), and aging/geriatrics/end of life (65.7%, OR 3.31, 95% CI 1.15-9.94, p = 0.01 across topics). Publication rates were higher for multicenter studies than single-institution studies (52.4% vs. 40.4%, OR 1.66, 95% CI 1.10-2.52, p = 0.04 across categories). Randomized controlled trials had higher publication rates than other study designs (66.7% vs. 45.9%, OR 2.72, 95% CI 1.30-5.94, p = 0.03 across study designs). Studies with positive results did not predict higher publication rates than negative studies (OR 0.89, 95% CI 0.6-1.31, p = 0.21).

CONCLUSIONS

This study demonstrated that 47.4% of abstracts presented at a general internal medicine national conference were subsequently published in a peer-reviewed journal indexed in MEDLINE.

Factors Influencing Physicians' Selection of Continuous Professional Development Activities: A Cross-Specialty National Survey

INTRODUCTION

We sought to understand what influences physicians' decisions about participation in continuous professional development (CPD) activities, and how often physicians engage in specific CPD activities.

METHODS

From September 2015 to April 2016, we administered a survey to 4648 randomly sampled licensed US physicians. Survey items addressed perceived barriers to CPD, factors that might influence participation in four prototypical CPD activities (reading an article, or completing a local activity, online course, or far-away course), and frequency of CPD engagement.

RESULTS

Nine hundred eighty-eight (21.6%) physicians responded. The most important barriers were time (mean [SD] 3.5 [1.3], 1 = not important, 5 = extremely important) and cost (2.9 [1.3]). In prioritizing factors influencing participation in four prototypical CPD activities, topical relevance consistently had the highest average rank. Quality of content and time to complete the activity were also frequently selected. Over the past 3 years, most physicians reported having participated in patient-focused learning and self-directed learning on a weekly basis; quality improvement and local continuing medical education (CME) activities several times per year; online learning, on-site courses, and national board-related activities a few times per year; and interprofessional learning less than once per year. Physicians believed that they ought to engage more often in all of these activities except board-related activities. They would like CME credit for these activities much more often than currently obtained.

DISCUSSION

The reasons physicians select a given CPD activity vary by activity, but invariably include topic and quality of content. Physicians want CME credit for the CPD activities they are already doing.

Interventions to prevent and reduce physician burnout: a systematic review and meta-analysis

BACKGROUND

Physician burnout has reached epidemic levels, as documented in national studies of both physicians in training and practising physicians. The consequences are negative effects on patient care, professionalism, physicians' own care and safety, and the viability of health-care systems. A more complete understanding than at present of the quality and outcomes of the literature on approaches to prevent and reduce burnout is necessary.

METHODS

In this systematic review and meta-analysis, we searched MEDLINE, Embase, PsycINFO, Scopus, Web of Science, and the Education Resources Information Center from inception to Jan 15, 2016, for studies of interventions to prevent and reduce physician burnout, including single-arm pre-post comparison studies. We required studies to provide physician-specific burnout data using burnout measures with validity support from commonly accepted sources of evidence. We excluded studies of medical students and non-physician health-care providers. We considered potential eligibility of the abstracts and extracted data from eligible studies using a standardised form. Outcomes were changes in overall burnout, emotional exhaustion score (and high emotional exhaustion), and depersonalisation score (and high depersonalisation). We used random-effects models to calculate pooled mean difference estimates for changes in each outcome.

FINDINGS

We identified 2617 articles, of which 15 randomised trials including 716 physicians and 37 cohort studies including 2914 physicians met inclusion criteria. Overall burnout decreased from 54% to 44% (difference 10% [95% CI 5-14]; p<0·0001; I²=15%; 14 studies), emotional exhaustion score decreased from 23·82 points to 21·17 points (2·65 points [1·67-3·64]; p<0·0001; I²=82%; 40 studies), and depersonalisation score decreased from 9·05 to 8·41 (0·64 points [0·15-1·14]; p=0·01; I²=58%; 36 studies). High emotional exhaustion decreased from 38% to 24% (14% [11-18]; p<0·0001; I²=0%; 21 studies) and high depersonalisation decreased from 38% to 34% (4% [0-8]; p=0·04; I²=0%; 16 studies).

INTERPRETATION

The literature indicates that both individual-focused and structural or organisational strategies can result in clinically meaningful reductions in burnout among physicians. Further research is needed to establish which interventions are most effective in specific populations, as well as how individual and organisational solutions might be combined to deliver even greater improvements in physician wellbeing than those achieved with individual solutions.

FUNDING

Arnold P Gold Foundation Research Institute.

Physician Attitudes About Maintenance of Certification: A Cross-Specialty National Survey

OBJECTIVES

To determine physicians' perceptions of current maintenance of certification (MOC) activities and to explore how perceptions vary across specialties, practice characteristics, and physician characteristics, including burnout.

PATIENTS AND METHODS

We conducted an Internet and paper survey among a national cross-specialty random sample of licensed US physicians from September 23, 2015, through April 18, 2016. The questionnaire included 13 MOC items, 2 burnout items, and demographic variables.

RESULTS

Of 4583 potential respondents, we received 988 responses (response rate 21.6%) closely reflecting the distribution of US physician specialties. Twenty-four percent of physicians (200 of 842) agreed that MOC activities are relevant to their patients, and 15% (122 of 824) felt they are worth the time and effort. Although 27% (223 of 834) perceived adequate support for MOC activities, only 12% (101 of 832) perceived that they are well-integrated in their daily routine and 81% (673 of 835) believed they are a burden. Nine percent (76 of 834) believed that patients care about their MOC status. Forty percent or fewer agreed that various MOC activities contribute to their professional development. Attitudes varied statistically significantly (P<.001) across specialties, but reflected low perceived relevance and value in nearly all specialties. Thirty-eight percent of respondents met criteria for being burned out. We found no association of attitudes toward MOC with burnout, certification status, practice size, rural or urban practice location, compensation model, or time since completion of training.

CONCLUSION

Dissatisfaction with current MOC programs is pervasive and not localized to specific sectors or specialties. Unresolved negative perceptions will impede optimal physician engagement in MOC.

Incentive and Reminder Strategies to Improve Response Rate for Internet-Based Physician Surveys: A Randomized Experiment

Background

Most research on how to enhance response rates in physician surveys has been done using paper surveys. Uncertainties remain regarding how to enhance response rates in Internet-based surveys.

Objective

To evaluate the impact of a low-cost nonmonetary incentive and paper mail reminders (formal letter and postcard) on response rates in Internet-based physician surveys.

Methods

We executed a factorial-design randomized experiment while conducting a nationally representative Internet-based physician survey. We invited 3966 physicians (randomly selected from a commercial database of all licensed US physicians) via email to complete an Internet-based survey. We used 2 randomly assigned email messages: one message offered a book upon survey completion, whereas the other did not mention the book but was otherwise identical. All nonrespondents received several email reminders. Some physicians were further assigned at random to receive 1 reminder via paper mail (either a postcard or a letter) or no paper reminder. The primary outcome of this study was the survey response rate.

Results

Of the 3966 physicians who were invited, 451 (11.4%) responded to at least one survey question and 336 (8.5%) completed the entire survey. Of those who were offered a book, 345/2973 (11.6%) responded compared with 106/993 (10.7%) who were not offered a book (odds ratio 1.10, 95% CI 0.87-1.38, P=.42). Regarding the paper mail reminder, 168/1572 (10.7%) letter recipients, 148/1561 (9.5%) postcard recipients, and 69/767 (9.0%) email-only recipients responded (P=.35). The response rate for those receiving letters or postcards was similar (odds ratio 1.14, 95% CI 0.91-1.44, P=.26).

Conclusions

Offering a modest nonmonetary incentive and sending a paper reminder did not improve survey response rate. Further research on how to enhance response rates in Internet-based physician surveys is needed.

Parental satisfaction of U.S. physicians: associated factors and comparison with the general U.S. working population

BACKGROUND

Physicians work considerably longer hours and are less satisfied with work-life balance than U.S. workers in other fields. There is, however, minimal data on physicians' parental satisfaction.

METHODS

To evaluate differences in parental satisfaction among physicians and workers in other fields, we surveyed U.S. physicians as well as a probability-based sample of the general U.S. working population between August 2014-October 2014. Parental satisfaction and the perceived impact of career on relationships with children were evaluated.

RESULTS

Among 6880 responding physicians (cooperation rate 19.2 %), 5582 (81.1 %) had children. Overall, physicians were satisfied in their relationships with their children, with 4782 (85.9 %) indicating that they were either very satisfied [n = 2738; (49.2 %)] or satisfied [n = 2044 (36.7 %)]. In contrast, less than half believed their career had made either a major [n = 1212; (21.8 %)] or minor positive [n = 1260; (22.7 %)] impact on their relationship with their children, with a slightly larger proportion indicating a major (n = 2071 [37.2 %]) or minor (n = 501 [9 %]) negative impact. Women physicians were less likely to believe their career had made a positive impact as were younger physicians. Hours worked/week inversely correlated with the belief that career had made a positive impact on relationships with children. Both men (OR: 2.75; p < 0.0001) and women (OR: 4.33; p < 0.0001) physicians were significantly more likely to report that their career had a negative impact on relationships with their children than the sex-matched U.S. working population.

CONCLUSIONS

U.S. physicians report generally high satisfaction in their relationships with their children. Despite their high satisfaction, physicians have a more negative perception of the impact of their career on relationships with their children than U.S. workers in general.

Relationship Between Clerical Burden and Characteristics of the Electronic Environment With Physician Burnout and Professional Satisfaction

OBJECTIVE

To evaluate associations between the electronic environment, clerical burden, and burnout in US physicians.

PARTICIPANTS AND METHODS

Physicians across all specialties in the United States were surveyed between August and October 2014. Physicians provided information regarding use of electronic health records (EHRs), computerized physician order entry (CPOE), and electronic patient portals. Burnout was measured using validated metrics.

RESULTS

Of 6375 responding physicians in active practice, 5389 (84.5%) reported that they used EHRs. Of 5892 physicians who indicated that CPOE was relevant to their specialty, 4858 (82.5%) reported using CPOE. Physicians who used EHRs and CPOE had lower satisfaction with the amount of time spent on clerical tasks and higher rates of burnout on univariate analysis. On multivariable analysis, physicians who used EHRs (odds ratio [OR]=0.67; 95% CI, 0.57-0.79; P<.001) or CPOE (OR=0.72; 95% CI, 0.62-0.84; P<.001) were less likely to be satisfied with the amount of time spent on clerical tasks after adjusting for age, sex, specialty, practice setting, and hours worked per week. Use of CPOE was also associated with a higher risk of burnout after adjusting for these same factors (OR=1.29; 95% CI, 1.12-1.48; P<.001). Use of EHRs was not associated with burnout in adjusted models controlling for CPOE and other factors.

CONCLUSION

In this large national study, physicians' satisfaction with their EHRs and CPOE was generally low. Physicians who used EHRs and CPOE were less satisfied with the amount of time spent on clerical tasks and were at higher risk for professional burnout.

Associations between resident physicians' publications and clinical performance during residency training

BACKGROUND

Both research and clinical medicine requires similar attributes of efficiency, diligence and effective teamwork. Furthermore, residents must succeed at scholarship and patient care to be competitive for fellowship training. It is unknown whether research productivity among residents is related to broad measures of clinical achievement. Our goal was to examine associations between the quantity of internal medicine residents' publications and validated measures of their knowledge, skills and multi-source evaluations of performance.

METHODS

This was a longitudinal study of 308 residents graduating from Mayo Clinic from 2006 to 2012. We identified peer-reviewed articles in Ovid MEDLINE between July of each resident's match year and the end of their graduation. Outcomes included American Board of Internal Medicine (ABIM) certification examination scores, mini clinical examination (mini-CEX) scores, and validated assessments of clinical performance by resident-peers, faculty and non-physicians. Performance assessments were averaged to form an overall score ranging from 1 to 5. Associations between quantity of resident publications--and ABIM, mini-CEX and performance assessment scores--were determined using multivariate linear regression.

RESULTS

The residents published 642 papers, of which 443 (69.0%) were research papers, 198 (30.8%) were case reports, and 380 (59.2%) were first-authored. On adjusted analysis, multi-source clinical performance evaluations were significantly associated (beta; 99% CI; p-value) with the numbers of research articles (0.012; 0.001-0.024; 0.007), and overall publications (0.012; 0.002-0.022; 0.002).

CONCLUSIONS

To our knowledge, this is the first study to demonstrate that scholarly productivity based on journal publication is associated with clinical performance during residency training. Our findings suggest that residents who invest substantial efforts in research are not compromised in their abilities to learn medicine and care for patients.

Changes in Burnout and Satisfaction With Work-Life Balance in Physicians and the General US Working Population Between 2011 and 2014

OBJECTIVE

To evaluate the prevalence of burnout and satisfaction with work-life balance in physicians and US workers in 2014 relative to 2011.

PATIENTS AND METHODS

From August 28, 2014, to October 6, 2014, we surveyed both US physicians and a probability-based sample of the general US population using the methods and measures used in our 2011 study. Burnout was measured using validated metrics, and satisfaction with work-life balance was assessed using standard tools.

RESULTS

Of the 35,922 physicians who received an invitation to participate, 6880 (19.2%) completed surveys. When assessed using the Maslach Burnout Inventory, 54.4% (n=3680) of the physicians reported at least 1 symptom of burnout in 2014 compared with 45.5% (n=3310) in 2011 (P<.001). Satisfaction with work-life balance also declined in physicians between 2011 and 2014 (48.5% vs 40.9%; P<.001). Substantial differences in rates of burnout and satisfaction with work-life balance were observed by specialty. In contrast to the trends in physicians, minimal changes in burnout or satisfaction with work-life balance were observed between 2011 and 2014 in probability-based samples of working US adults, resulting in an increasing disparity in burnout and satisfaction with work-life balance in physicians relative to the general US working population. After pooled multivariate analysis adjusting for age, sex, relationship status, and hours worked per week, physicians remained at an increased risk of burnout (odds ratio, 1.97; 95% CI, 1.80-2.16; P<.001) and were less likely to be satisfied with work-life balance (odds ratio, 0.68; 95% CI, 0.62-0.75; P<.001).

CONCLUSION

Burnout and satisfaction with work-life balance in US physicians worsened from 2011 to 2014. More than half of US physicians are now experiencing professional burnout.

Digital health interventions for the prevention of cardiovascular disease: a systematic review and meta-analysis

OBJECTIVE

To assess the potential benefit of digital health interventions (DHIs) on cardiovascular disease (CVD) outcomes (CVD events, all-cause mortality, hospitalizations) and risk factors compared with non-DHIs.

PATIENTS AND METHODS

We conducted a systematic search of PubMed, MEDLINE, EMBASE, Web of Science, Ovid, CINHAL, ERIC, PsychINFO, Cochrane, and Cochrane Central Register of Controlled Trials for articles published from January 1, 1990, through January 21, 2014. Included studies examined any element of DHI (telemedicine, Web-based strategies, e-mail, mobile phones, mobile applications, text messaging, and monitoring sensors) and CVD outcomes or risk factors. Two reviewers independently evaluated study quality utilizing a modified version of the Cochrane Collaboration risk assessment tool. Authors extracted CVD outcomes and risk factors for CVD such as weight, body mass index, blood pressure, and lipid levels from 51 full-text articles that met validity and inclusion criteria.

RESULTS

Digital health interventions significantly reduced CVD outcomes (relative risk, 0.61; 95% CI, 0.46-0.80; P<.001; I(2)=22%). Concomitant reductions in weight (-2.77 lb [95% CI, -4.49 to -1.05 lb]; P<.002; I(2)=97%) and body mass index (-0.17 kg/m(2) [95% CI, -0.32 kg/m(2) to -0.01 kg/m(2)]; P=.03; I(2)=97%) but not blood pressure (-1.18 mm Hg [95% CI, -2.93 mm Hg to 0.57 mm Hg]; P=.19; I(2)=100%) were found in these DHI trials compared with usual care. In the 6 studies reporting Framingham risk score, 10-year risk percentages were also significantly improved (-1.24%; 95% CI, -1.73% to -0.76%; P<.001; I(2)=94%). Results were limited by heterogeneity not fully explained by study population (primary or secondary prevention) or DHI modality.

CONCLUSION

Overall, these aggregations of data provide evidence that DHIs can reduce CVD outcomes and have a positive impact on risk factors for CVD.

A national study of medical students' attitudes toward self-prescribing and responsibility to report impaired colleagues

PURPOSE

The authors investigated medical students' attitudes about appropriate prescribing behaviors, their personal responsibility to report impaired colleagues, and what factors may influence these beliefs.

METHOD

The authors conducted a cross-sectional study of U.S. medical students in 2012 to assess attitudes about appropriate prescribing behaviors and responsibility to report impaired colleagues, and to explore relationships between prescribing beliefs and burnout, depression, and alcohol abuse/dependence. Chi-square test and multivariate logistic regression were performed.

RESULTS

Of 12,500 medical students invited to participate, 4,402 (35%) completed surveys. Believing it is appropriate to prescribe an antidepressant to self or spouse was rare (<10%) in comparison with believing it is appropriate to prescribe an antibiotic for oneself (34.5%) or a spouse (57.7%).In multivariate analysis, students with burnout were more likely to agree that each of the inappropriate prescribing behaviors was acceptable (ORs 1.15-1.51). Students with burnout were less likely to believe they had a personal responsibility to report colleagues with impairment due to alcohol or substance use (OR 0.87). Students personally experiencing symptoms of depression were less likely to believe medical students should report colleagues impaired by mental health problems (OR 0.72). Similarly, students with alcohol abuse/dependence were less likely to believe they had a duty to report colleagues impaired by alcohol/substance use (OR 0.55).

CONCLUSIONS

Suboptimal attitudes about prescribing and personal responsibility to report impaired colleagues are common among medical students. Suboptimal attitudes are associated with personal distress, further evidence of a link between personal distress and professionalism.

The effects of fatigue and dissatisfaction on how physicians perceive their social responsibilities

OBJECTIVE

To examine how fatigue and dissatisfaction with practicing medicine relate to US physicians' perceptions of their professional responsibilities in a time of upheaval in health care.

METHODS

From May 30, 2012, through September 1, 2012, we mailed an 8-page paper survey, including measures of perceived social responsibility, fatigue, and satisfaction, to a random sample of 3897 physicians selected from the American Medical Association Physician Masterfile. We performed bivariate tests and multivariable logistic regression to examine associations between key predictors (fatigue and satisfaction) and 3 dependent variables: obligations to care for the uninsured and underinsured, obligations to address policy issues, and agreement with cost-containment.

RESULTS

A total of 2556 physicians (65.6%) responded to the survey. Nearly half of physicians (1160 [45.4%]) reported high levels of fatigue, whereas most (1810 [70.8%]) expressed satisfaction with practicing medicine. Dissatisfaction in practicing medicine proved to be a significant predictor in how physicians perceive their professional responsibilities and in medical decision-making. Overall, physicians who rated themselves as very dissatisfied had a significantly reduced likelihood of favoring limiting reimbursement to expand basic coverage (odds ratio, 0.5; 95% CI, 0.3-0.7), recognizing an obligation to care for the uninsured (odds ratio, 0.5; 95% CI, 0.3-0.7), and addressing societal health policy issues (OR, 0.7; 95% CI, 0.5-0.9).

CONCLUSION

Although fatigue and satisfaction are strongly associated, only satisfaction appears to correlate with physicians' perceived social responsibilities. Physicians who are dissatisfied with their profession may be less inclined to address health policy issues, embrace charity care, or practice cost containment.

The prognostic importance of weight loss in coronary artery disease: a systematic review and meta-analysis

OBJECTIVE

To assess the prognostic impact of weight loss on clinical outcomes in patients with coronary artery disease (CAD).

METHODS

We performed a systematic review and meta-analysis of the prognostic effects of weight loss in patients with CAD on a composite outcome of all-cause mortality, cardiovascular mortality, and major adverse cardiac events considering studies published between January 1, 1964, and August 8, 2013. We considered weight loss "intentional" when it occurred in the presence of programmed therapeutic lifestyle changes and "observational" when no such intervention was specified.

RESULTS

We searched 1218 abstracts, of which 12 studies with 14 cohorts met the inclusion criteria. A total of 35,335 patients (mean age, 64 years; 72% male; body mass index [BMI], 30; 3.2 years of follow-up) were included. Overall, weight loss was associated with a greater risk of the composite outcome (relative risk [RR], 1.30; 95% CI, 1.00-1.69; P=.05). However, heterogeneity was high (I(2)=90%) and was substantially explained by weight loss intentionality. Presumed intentional weight loss (4 cohorts) was associated with improved outcomes (RR, 0.67; 95% CI, 0.56-0.80; P<.001), whereas observational weight loss (10 cohorts) was associated with worsened outcomes (RR, 1.62; 95% CI, 1.26-2.08; P<.001; interaction P<.001).

CONCLUSION

Whereas observational weight loss is associated with increased adverse cardiovascular events, intentional weight loss is associated with lower clinical events. These results suggest that the underlying mechanism of weight loss (ie, intentional or unintentional) affects its impact on subsequent risk in persons with known CAD.

Hospitalist career decisions among internal medicine residents

BACKGROUND

Hospital medicine is a rapidly growing field of internal medicine. However, little is known about internal medicine residents' decisions to pursue careers in hospital medicine (HM).

OBJECTIVE

To identify which internal medicine residents choose a career in HM, and describe changes in this career choice over the course of their residency education.

DESIGN

Observational cohort using data collected from the annual Internal Medicine In-Training Examination (IM-ITE) survey.

PARTICIPANTS

16,781 postgraduate year 3 (PGY-3) North American internal medicine residents who completed the annual IM-ITE survey in 2009-2011, 9,501 of whom completed the survey in all 3 years of residency.

MAIN MEASURES

Self-reported career plans for individual residents during their postgraduate year 1 (PGY-1), postgraduate year 2 (PGY-2) and PGY-3.

KEY RESULTS

Of the 16,781 graduating PGY-3 residents, 1,552 (9.3 %) reported HM as their ultimate career choice. Of the 951 PGY-3 residents planning a HM career among the 9,501 residents responding in all 3 years, 128 (13.5 %) originally made this decision in PGY-1, 192 (20.2 %) in PGY-2, and 631 (66.4 %) in PGY-3. Only 87 (9.1 %) of these 951 residents maintained a career decision of HM during all three years of residency education.

CONCLUSIONS

Hospital medicine is a reported career choice for an important proportion of graduating internal medicine residents. However, the majority of residents do not finalize this decision until their final year.

Statins use and risk of depression: a systematic review and meta-analysis

IMPORTANCE

Statin use has been associated with depression; however studies of the association between statin use and depression have yielded mixed results.

OBJECTIVE

To determine whether statin use is associated with depression and to evaluate the evidence supporting this association.

DATA SOURCES

Ovid MEDLINE In-Process & Other Non-Indexed Citations, Ovid MEDLINE, EMBASE, PsycInfo, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and Scopus were searched through December 28, 2012.

STUDY SELECTION

We included studies that evaluated exposure to statins, reported the development of depression, and relative risks or odds ratios (ORs) or provided data for their estimation. Two reviewers screened 981 abstracts independently using a standardized form, reviewed full text of 59 selected articles, and included 7 studies in this metaanalysis.

DATA EXTRACTION AND SYNTHESIS

Study design, statin exposure, development of depression, and study quality were extracted by 2 independent reviewers. A pooled OR with 95% confidence interval (CI) was estimated using the random-effects model and heterogeneity was assessed using Cochran's Q test and the I(2) statistic.

RESULTS

Seven observational studies (4 cohort, 2 nested case-control, and 1 cross-sectional) from 5 countries enrolling 9187 patients were included. Statin users were 32% less likely to develop depression than nonusers (adjusted OR, 0.68; 95% CI, 0.52-0.89). Modest heterogeneity was observed between the studies (I(2)=55%, P=0.01), which could be accounted for by one study, exclusion of which removed the heterogeneity (P=0.40, I(2)=2%) and further strengthened the antidepressant effect of statin (adjusted OR, 0.63; 95% CI, 0.43-0.93). Heterogeneity could not be explained by study design or study population. The quality of supporting evidence was fair.

CONCLUSIONS AND RELEVANCE

This systematic review and meta-analysis suggests that statin use is associated with lower risk for depression. However, higher-quality studies are needed to confirm the magnitude of this association.