Abstract
Objectives:
Training future clinician-researchers remains a challenge faced by Canadian psychiatry departments. Our objectives were to determine the prevalence of residents interested in pursuing research and other career options as part of their practice, and to identify the factors associated with interest in research.
Method:
Data from a national online survey of 207 Canadian psychiatry residents from a total of 853 (24.3% response rate) were examined. The main outcome was interest in research as part of residents’ future psychiatrist practice. Bivariate and multivariate analyses were performed to identify demographic and vocational variables associated with research interest.
Results:
Interest in research decreases by 76% between the first and fifth year of psychiatry residency (OR 0.76 per year, 95% CI 0.60 to 0.97). Training in a department with a residency research track did not correlate with increased research interest (χ2 = 0.007, df = 1, P = 0.93).
Conclusions:
Exposing and engaging psychiatry residents in research as early as possible in residency training appears key to promoting future research interest. Psychiatry residency programs and research tracks could consider emphasizing research training initiatives and protected research time early in residency.
Introduction
Research improves our understanding of mental illness, enhances patient care, and guides novel psychiatric health services. Nonetheless, the number of physician-scientists have declined in the United States 1 –6 and Canada 7 despite growing needs for translational research, 8,9 and the drop is even more significant in psychiatry. 10,11 Research tracks (RTs) were recently created across Canada to help promote residents’ interest in research. However, research training during residency remains challenging due to the wide range of skills that need to be acquired often concurrently with clinical training, 12 including study design, data collection and analysis, writing manuscripts and grants, training students, and running a laboratory.
To sustain a steady pipeline of future psychiatrist-researchers, it will be of major importance to better understand what factors promote residents’ research interest. Previous studies have shown the critical importance of early exposure to research 13 in the first years of residency training 14 and even before, 1,15 as well as good mentoring. 8,10,13,16 In contrast, high level of trainee’s debts 6,8,13 may impede research interest.
In our paper, we aimed to describe psychiatry residents’ interest in research as part of their future psychiatrist practice, among other career options. We then attempted to identify potential correlates of research interest. We hypothesized that research interest would correlate with factors previously identified, particularly mentorship and early exposure to research. We also hypothesized that training in a research-focused university with a residency research track (RT) would be associated with interest in research.
Methods
We used data from a cross-sectional Internet survey sent to all Canadian psychiatry residents from postgraduate year PGY-1 to PGY-5 among the 17 medical schools with postgraduate psychiatry training programs. The parent study was promoted as an assessment of residents’ needs and interests in geriatric psychiatry by the Canadian Academy of Geriatric Psychiatry. 17 An online survey website was used (https://surveymonkey.com) and the design and implementation of this survey were based on Dillman’s Tailored Design Method. 18
Recruitment
Between September and October 2013, email invitations to participate were sent twice to all Canadian psychiatry residency program directors through the Coordinators of Psychiatric Education (COPE). Program directors then forwarded invitations to residents, who were offered a chance to win a $400 Amazon gift certificate if they participated. Ethics approval was obtained from University of Saskatchewan and McGill University and all participants gave online written consent.
Outcomes
Residents’ interests in research and 18 other fields of psychiatry and practice settings were assessed using their answer to the question, “What fields or practice settings are you interested in as part of your future psychiatrist practice?” on a 5-point Likert scale (1 = not at all interested, 3 = neutral, and 5 = very interested). To facilitate interpretation, a rating of interested (4 or 5) or very interested (5) were defined as interest in this field of psychiatry or practice setting, while ratings of neutral or below (1 to 3) were considered not interested.
Variables Included in the Survey
Demographic information were gathered: age, 1,13 sex, 19 ethnicity, 15 first language, marital status 13 ; as well as information on their residency training: the year of residency, 19 geographic location or university, and the presence of an RT in the university. 6,13 Residents also rated the importance of the following vocational variables: mentors, 6,8,10 annual conferences in a field of interest, intellectual stimulation, 13 subjective self-rated availability of research opportunities or job opportunities, 13 the importance of incomes, 13,20 lifestyle, 8,20 prestige, and of the length of training. Using a similar 5-point Likert scale ranging from not important at all (1 point) to very important (5 points), the score corresponds to residents’ perception or opinion of the importance of those factors.
Statistical Analysis
The prevalence of interest in research and other practice fields or settings were assessed. Our sample’s demographics were compared with all Canadian psychiatry residents using aggregated data from the Canadian Post-MD Education Registry (CAPER). Potential bivariate correlates of interest in research as a field of psychiatry were ascertained using chi-square, Student t test, and Mann–Whitney U tests, as appropriate. Then, logistic regression analyses, including all variables with significant bivariate correlations (2-tailed P < 0.05), were conducted to assess independent correlates of research interest. Interaction terms (for example, year of residency × lifestyle ) had been assessed as part of the logistic regression modelling.
Results
Respondent Characteristics
The survey was completed by 207 Canadian psychiatry residents from a total of 853 (24.0%), with 205 responding to items about career interests. Respondents were mostly aged 26 to 30 years (51.2%), female (70.0%), Caucasian (68.6%), and in long-term relationships (66.7%) (Table 1). There were 45 residents who reported to be in PGY-1, 36 in PGY-2, 53 in PGY-3, 38 in PGY-4, and 33 in PGY-5. Compared with all Canadian residents (n = 853), survey respondents did not differ significantly in age, sex, year of residency, or geographic location.
Association of residents’ characteristics with interest in research as part of your future psychiatrist practice, n = 205.
The prevalence of residents’ interest in a number of fields or settings of practice is described (Figure 1). Relatively high rates of residents were interested in adults aged 18 to 64 years (89.8%), outpatient (80%), psychosis (69.3%), and large urban settings (69.3%), while less residents were interested in personality disorders (42.0%) and rural settings (28.8%).
Residents’ responses to the question: “In what fields or settings of psychiatry are you interested in practicing?” (n = 205). Note, a 5-point Likert scale was used. Interested was defined by a rating of 4 or 5. Very interested corresponded to a rating of 5.
Prevalence and Bivariate Correlates of Interest in Research as a Part of Future Psychiatry Practice
Residents (30.2% and 9.3%) were interested and very interested in research as part of their psychiatric practice, respectively. Potential correlates of this interest were then assessed using bivariate analyses (Table 1). Year of residency was inversely correlated with interest in research (Mann–Whitney U = 3319.0, P = 0.006). PGY-5 residents were 76% less likely to be interested in research than those in PGY-1 (OR 0.24, 95% CI 0.08 to 0.70), with interest decreasing from 48.9% (n = 22/45) to 18.8% (n = 6/32). Other demographic variables were not significantly correlated. Compared with people aged 25 to 35 years, extreme age groups (aged 20 to 25 or >35 years) were more interested in research (44.7%, compared with 25.9%, χ2 = 6.02, df = 1, P = 0.014); however, this was an exploratory post hoc finding.
Additionally, a few vocational factors were associated with research interest: mentors in my field of interest (χ2 = 5.29, df = 1, P = 0.02) and annual conferences in my field of interest (χ2 = 5.82, df = 1, P = 0.02). Meanwhile lifestyle (χ2 = 5.91, df = 1, P = 0.02) showed a negative association: residents interested in research were less likely to find lifestyle (shorter working hours) important.
Lastly, compared with programs without an RT, research interest was no different in the research-oriented psychiatry programs with RTs 12 (the University of Toronto, the University of British Columbia, and McGill University) (χ2 = 0.007, df = 1, P = 0.93). Training programs with an RT had somewhat lower response rates 18.2% (62/340), compared with 27.9% (143/513) (χ2 = 10.4, df = 1, P = 0.001). In non-RT universities, 41.2% of residents interested in an academic career were also interested in research, compared with 6.5% of those striving for nonacademic careers (χ2 = 17.9, df = 1, P < 0.001). This was similar to findings in RT universities (41.3%, compared with 0%, χ2 = 9.53, df = 1, P = 0.002). Respondents from a program with an RT were not different from non-RT program respondents in terms of their year of residency.
Independent Correlates of Interest in Research as a Part of Future Psychiatry Practice
In logistic regression using the 4 variables found to be correlates of interest in research with a P value of less than 0.05, 2 variables were independently and inversely associated with research interest: year of residency (OR 0.76 per additional year of residency, 95% CI 0.60 to 0.97) and considering lifestyle (shorter working hours) to be important (OR 0.46, 95% CI 0.29 to 0.73). In a second logistic regression including only these 2 variables, year of residency (OR 0.76 per year, 95% CI 0.60 to 0.95) and considering lifestyle to be important (OR 0.35, 95% CI 0.14 to 0.86) remained independent correlates of research interest.
Discussion
Our study was the first nation-wide survey of Canadian residents assessing interest in various fields or settings of psychiatry, including research. We also examined a broad range of potential predictors of research interest.
In our survey, 30.2% of residents were interested in research as part of their future psychiatric practice, which is comparable with a survey in 33 large American university programs where 29% of psychiatry residents had a moderate or high interest in research. 13
In keeping with the literature, we found that increasing age and marital status were not predictably associated with research interest. 13 Also, we found that lifestyle (shorter working hours) are not as important among residents with research interest, which may suggest that a balanced work–life might not be as important for research-interested residents as previously considered. 8 Also in contrast with previous studies, 13,19,21 sex was not associated with research interest. Meanwhile mentoring and annual conferences may be contributory to research interest, 9,17,22 even if they may not independently predict interest. These findings support current initiatives put in place by the Canadian Psychiatric Association (CPA), such as the CPA mentoring program 23 and making conferences accessible to trainees.
The most striking finding was that rates of research interest decreased markedly during residency. Compared with PGY-1 residents, those in PGY-5 were 76% less likely to be interested in research (OR 0.24, 95% CI 0.08 to 0.70) (Figure 2). Further, increased year of residency was found to independently predict lower research interest (OR 0.76 per year, 95% CI 0.60 to 0.97). Previous authors found that early developmental factors were critical in research, 1,13 for example showing that over 86% of residents interested in a research career considered research prior to residency. 13
Proportion of psychiatry residents interested in research, by year of residency.
Ours is the first paper to suggest that residents’ research interest may decline as their training progresses. This could be due to the various challenges faced by residents who wish to pursue a research career, such as the lack of continuity in research time during potentially productive years, competing for grants with graduating PhD students from other disciplines, as well as a relative lack of research mentoring. 1
Also striking is the discrepancy between the percentage of residents interested in working in an academic setting (69.8%) and the number of residents actually interested in doing research (30.2%). In both RT and non-RT universities alike, only 41.2% of residents with academic aspirations were interested in research. One interpretation may be that residents underestimate research as intrinsically meaningful and as an often necessary component of career promotion in academic settings.
Interestingly, completing a residency program in a psychiatry department with an RT did not predict higher research interest (χ2 = 0.007, df = 1, P = 0.93). This was notable as the residency programs with an RT were situated in the most productive psychiatry research departments in Canada (University of Toronto, McGill University, and the University of British Columbia). Perhaps the most motivated residents will pursue research interests in spite of difficulties obtaining support locally: that is, that researchers can potentially bloom from any garden.
Limitations
A limitation of this secondary-analysis study was that the parent study 17 was promoted as a survey of interest in becoming a geriatric psychiatrist (although interest in other areas had also been assessed), which probably skewed the results toward a higher prevalence of geriatric psychiatry interest. Nonetheless, this did not appear to impact the correlation between research and potential predictors, as including geriatric psychiatry interest in the multivariate analysis did not modify our main findings. Certain potential predictor variables include the level of trainees’ debts, 6,8,13 being a parent, 13 exposure to clinical research, 8 or the presence of a mandatory research experience in the curriculum. 24 Academic productivity tended to increase in internal medicine or surgery programs that included obligatory scholarly activity, 25 –27 but this information had not been collected in the parent study. 17 Further, information about research experience at any stage of the training including prior to residency was not part of the study, even though it may correlate with research interest in psychiatry residents 13 and other recently graduated physicians. 28 As well, causality is difficult to infer from a cross-sectional study, for example, to confirm whether research interest decreases during residency, one could follow a cohort of residents. Also, multiple comparisons may have contributed to alpha inflation. However, our most robust findings still approach significance even if a Bonferroni correction is applied. Also, our relatively large survey was able to test a broad range of hypotheses previously identified in smaller quantitative and qualitative studies. Future qualitative studies could help tease out why residents lose interest in research over the course of training. Finally, there was only a 24.3% response rate, although this was still quite reasonable for a demographically-representative web survey of residents, 29 which often have dismal 10% response rates. 18
Conclusion
Research interest diminishes markedly with additional years of psychiatry residency training: a 76% decrease over 5-years of residency. Additionally, attending a residency program in a research-intensive university with an RT does not seem to predict increased interest in research. These findings have important implications for the recruitment and training of future clinician-researchers in psychiatry. Training programs, even those with an RT, 12 have tended to place almost exclusive importance on clinical training in the first 2 years of residency, with little dedicated time available for research activities until PGY-4. 12
Training future clinician-researchers remains a priority for Canadian psychiatry departments and universities have invested extensive resources into RTs. 12 Residency programs and their RTs could consider emphasizing research training initiatives and protected research time as early as possible in residency (for example, in PGY-1 and PGY-2), and perhaps even in medical school. 13
Footnotes
Acknowledgements
We thank the CAPER, and the COPE for their generous help in implementing this survey. We also thank Dr Dallas Seitz for his support and advice with study methodology.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article. This work was supported by the Canadian Academy of Geriatric Psychiatry (CAGP). Dr Rej has received Master’s Training Awards from the Canadian Institutes of Health Research and the Fonds de Recherche en Santé du Quebec. Dr Rapoport is the president of the CAGP, while Dr Rej and Dr Davidson are CAGP board members. Both Dr Rapoport and Dr Rej are on the CAGP research committee.