The serious shortage of child and adolescent psychiatrists in the United States is well-described (1, 2). The rate of growth in the workforce does not appear to match the projected need. Kim et al. (1) estimated the needed number of child and adolescent psychiatrists in 2020 as over 12,000; given the current growth rate, the estimated supply would be 8,000. Data regarding entrance into child and adolescent psychiatry (CAP) fellowships suggests some improvement, but concern regarding future need continues (3). A review of 2006 National Resident Match Program data for CAP fellowships revealed that just 61% of matching residents were U.S. medical school graduates (4). The number of postgraduate year 1 (PGY-1) general psychiatry positions filled by U.S. medical school graduates has remained stable over the past 2 years, rising from 641 to 643 (5). The number of U.S. graduates matching into the limited number of combined pediatrics/psychiatry/CAP programs has hovered between 16 and 18 over the past 4 years (5). While recruitment into general psychiatry programs has increased some, efforts made by U.S. medical schools to promote recruitment into general psychiatry, and ultimately into CAP, are clearly needed.
Several authors have examined areas related to choosing a career path in general psychiatry. Sierles et al. (6) noted that the most useful predictor of a school’s proportion of students matching into psychiatry was the proportion of students matched the year prior, emphasizing that intrinsic factors (e.g., quality and characteristics of the psychiatry department and the medical education program at each school) seemed most important to the process. The psychiatry clerkship has been identified as a highly important variable that could impact residency recruitment into general psychiatry (7). However, data regarding the direction and extent of this impact appear to be mixed (8–11), pointing to the high degree of variability in psychiatry clerkship experiences. Studies of medical student attitudes regarding the field of psychiatry generally show improved attitudes after the clerkship, but little effect on career decision (12, 13).
Despite the vital need for recruitment, there is less research examining attitudes and career choice in CAP, particularly during the psychiatry clerkship. Martin et al. (14) studied perceptions toward aspects of psychiatry and CAP using pre- and postclerkship surveys. Student views toward positive features of both general psychiatry and CAP did not change, and the perception that CAP was stressful decreased. However, perception of poor family or community support for children as a negative feature of CAP increased. The effect of the clerkship (which included clinical exposure to CAP for 55% of the students) on student interest in specializing in psychiatry was small, increasing by two students (14).
Our study was designed to identify clinical exposure to CAP during the psychiatry clerkship and its impact on the interest in pursuing child psychiatry as a career. Further, we looked at elements of clinical experiences as a whole and their overall value in increasing interest in psychiatry and in CAP.
Participants were medical students completing third-year psychiatry clerkships at the University of North Carolina (UNC) School of Medicine. For 91 students, the entire clerkship was at UNC Hospitals during the 2005–2006 academic year. Twenty-five completed a clerkship during the first rotation period of 2006–2007 with a revised curriculum which included other sites. The 2006–2007 students were added to increase statistical power after determining that their responses did not differ significantly from those of the students from 2005–2006 (see Results section).
Students completed a 6-week clerkship divided into two 3-week blocks. Students were assigned to two of six UNC Hospital psychiatry inpatient units or to the consultation and liaison service. The 2006–2007 students took three 2-week blocks, which included at least one of the UNC services. Efforts were made to honor student service requests. On inpatient services, students participated in all clinical activities related to the evaluation and treatment of their patients. Students also took part in the evaluation of patients with acute psychiatric presentations in the walk-in clinic and the emergency department. Students worked with outpatients in clinics with an upper-level resident tutor providing weekly supervision. Seminars and case conferences rounded out the core curriculum for the clerkship.
For approximately two-thirds of the students, exposure to child and adolescent psychiatry (CAP) was limited to two lectures (one of those lectures usually included a patient interview) and variable experience on call in the emergency department and the walk-in clinic. About one-third of students completed a rotation with the CAP inpatient unit.
The Child and Adolescent Psychiatry Experiences Questionnaire
We constructed the Child and Adolescent Psychiatry Experiences Questionnaire (CAPE-Q) to give to all students at the end of the psychiatry clerkship. Completion of the survey was voluntary and anonymous; a coversheet explaining this was attached to each survey. Our materials and study plan were approved by the UNC Office of Human Research Ethics.
The CAPE-Q consists of questions regarding student demographics, rotation dates, and estimation of clinical exposure to CAP in various settings (e.g., inpatient, outpatient, emergency department, residential treatment, and partial hospital). Given the clinical opportunities in child and adolescent psychiatry (CAP) available to this sample, students could easily be classified into high CAP exposure and low CAP exposure based on whether or not they did part of their clerkship with a CAP inpatient unit. The survey queries students’ consideration of general psychiatry and CAP as career choices both before and at the end of the clerkship with yes/no questions. Following this is a series of 12-item Likert scale questions designed to assess student attitudes regarding aspects of the clerkship—with a focus on CAP experiences—and their relation to interest in psychiatry and CAP. A yes/no question regarding the influence of the CAP experience on the student’s decision to pursue or not pursue CAP was included. Several open-ended questions regarding the CAP exposure (or lack thereof) and its influence on career decision were also included.
Statistical analyses were conducted using SPSS, Version 14.0 (Chicago, 2005). The 12-item Likert scale portion of the CAPE-Q was assessed for internal consistency and factor structure. All independent and dependent variables were assessed for frequencies and nonparametric correlations (Kendall’s tau-b). Primary independent variables included cohort (3-week or 2-week rotation period) to assess the effects on responses due to curriculum time, rotation changes, level of exposure to CAP during the rotation, impact of rotation on the student’s CAP interest, and pre/post change in interest in general psychiatry or in CAP careers. Age and gender also served as independent variables. Dependent variables included having children, level of training, and the 12 categorically scored items of the CAPE-Q. Group comparisons of independent variables on dependent variables involved chi-square analyses. For a sample size of N=98, a G*Power analysis and consultation of Cohen’s statistical power tables for nonparametric analyses obtained an estimated power of 0.84 to detect a medium effect size of 0.3 SD units, using a two-tailed type I (alpha) error rate of 0.05 (15, 16). Due to multiple dependent variables in chi-square analyses, Bonferroni’s adjustment in the type I error rate was made, resulting in chi-square analysis alpha =0.005.
Overall there was an 83.6% response rate, with 97.8% of the 2005–2006 students and 44.0% of the 2006–2007 students completing the survey. The 98 students were 52.1% women and 7.7% of the students had children. Age ranges were 20–25 years (57.3% of sample), 26–30 years (32.3%), 31–35 years (8.3%), and 36–40 years (2.1%). Because the surveys were completed anonymously, demographic information regarding the group of students who did not return the surveys was not available.
We assessed the 12 items of the CAPE-Q for internal consistency (reliability and factor structure). Psychometric properties of the 12 items, each scored on a 5-point Likert scale, are provided in Table 1
. Most items intercorrelated significantly at p≤0.01. Item 2 (“My child psychiatry inpatient experience encouraged me to go into psychiatry”) significantly correlated with every other item, usually at the p<0.01 level. Reliability analysis on the entire 12-item scale yielded an excellent Cronbach alpha at 0.894. A maximum likelihood factor analysis with oblique factor rotation yielded one-to-two factors (Eigenvalues =6.352, 1.764) that explained 52.9%–67.6% of scale variance in scores (R2=0.529–0.676). These preliminary results indicate that the CAPE-Q represents a strong instrument for assessing students’ experiences in a rotation with CAP experiences and the impact of these experiences upon their decision to pursue a career in child psychiatry. There were low response rates to inpatient CAP related items (Table 1
, items 2, 7, 8); the majority of students did not complete an inpatient CAP rotation.
Additionally, there were no significant differences between the two genders on the CAPE-Q items or between the age groups on these items. Kendall’s tau-b correlations based on gender ranged from −0.043 (p=0.655) to 0.185 (p=0.154), and correlations based on age ranged from −0.150 (p=0.234) to 0.214 (p=0.084).
We were specifically interested in whether having an inpatient child and adolescent psychiatry (CAP) experience affected attitudes or interests in student rotation experience and subsequent careers. Based on student estimations, students were grouped as high CAP exposure (required inpatient experience, n=36) or low CAP exposure (n=62). Level of exposure to children was not significantly associated with its co-independent variables (i.e., gender or age). However, it was significantly (p≤0.005) associated with the following dependent variables: consideration of child psychiatry after the rotation (“Has your child psychiatry experience influenced your decision to pursue/not pursue child psychiatry?”), and CAPE-Q items 1, 2, 4, and 12 (Table 2
). The high CAP exposure group was thus more likely to agree with the following: they had enough exposure to CAP to make a career decision; their CAP inpatient experience encouraged them to pursue the field; they would consider going into CAP if they did not have to do a residency in general psychiatry; and participating in evaluations of youth increased their interest in child psychiatry.
We also looked for differences among students who reported an increase, decrease, or no change in their child psychiatry career interest based on their clerkship experience. Students were asked about their interest in CAP before and after the clerkship on a scale of 0–2 (No/Maybe/Yes). Of 94 students responding to both the pre- and postrotation CAP interest levels, 17 students (17.9%) exhibited an increased interest in CAP, 76 students (80.0%) showed no change, and only two students (2.1%) showed decreased interest. The high CAP exposure group had almost triple the percentage with increased CAP interest (27.8%) compared to the low CAP exposure group (9.7%, Figure 1
). Furthermore, overall change in CAP interest correlated significantly (r=0.510, p<0.01, n=66) with the question, “Has your child psychiatry experience influenced your decision to pursue/not pursue child psychiatry?” (Table 2
). The Pearson chi square for this association was likewise significant (χ2=24.6, p=0.000, n=66). There were also significant associations with CAPE-Q items 2, 4, 5, and 6 (Table 2
We explored differences among students who reported an increase, decrease, or no change in their general psychiatry career interest based on their CAP rotation experience. Students were asked about their interest in general psychiatry before and after their rotation on a scale of 0–2 (No/Maybe/Yes). Of 96 students responding to both the pre- and postrotation general psychiatry interest levels, 19 students (19.6%) exhibited increased interest, 73 students (75.3%) showed no change, and five students (5.1%) showed decreased interest. The pre/post change in interest was significantly associated with CAPE-Q item 3 (Table 2
Finally, we explored any differences on the survey items between students who reported that their CAP rotation experience influenced their decision to pursue child psychiatry and those who did not. Students were asked about how their CAP rotation experience influenced their decision to pursue child psychiatry on a scale of 0–1 (No/Yes). Of 68 students responding to this question, 18 students (26.5%) answered affirmatively. Responses to this question were significantly associated with consideration of child psychiatry prior to rotation, consideration of child psychiatry after rotation, and pre/post CAP interest change (Table 2
). Responses to this question were also significantly associated with consideration of psychiatry prior to rotation and consideration of psychiatry after rotation (Table 2
), but not for pre/post change in interest in psychiatry (χ2=3.12, p=0.210, n=67). CAPE-Q items 2, 4, and 12 were also significant (Table 2
As noted, we determined that there were no significant differences between the 2-week and 3-week rotation experiences on the dependent variables. A nonparametric correlation analysis between the dichotomously scored 2-week versus 3-week rotation students and each of the other categorical study variables yielded no significant correlations. Correlation coefficients ranged from −0.206 (p=0.063) to 0.118 (p=0.254), and corresponding chi-square analyses were not significant (χ1–42=0.129–6.55, p=0.164–0.934). Therefore, we included the eleven 2-week cohort students in all subsequent analyses to strengthen statistical power (15, 16).
In our sample, the CAPE-Q demonstrated strong internal consistency/reliability and content validity for the measurement of student-reported learning experiences in child and adolescent psychiatry (CAP) and their effect on interest in the field and on career decisions (17, 18). The factor analyses indicate that the CAPE-Q may be either uni- or bi-dimensional; further evaluations of the instrument’s factor structure will be necessary to fully evaluate its construct validity. Students without structured clinical experience in CAP did not believe that they had enough exposure to the field in order to make a career decision. While the final decision to pursue CAP is made several years after the clerkship, a critical intermediary decision—to pursue residency in general psychiatry—must be made in a matter of months. Therefore, early clinical exposure makes sense.
Here, CAP experience in the psychiatry clerkship exerted a positive influence on students’ decisions to pursue CAP at the end of the clerkship. A structured clinical experience in CAP was associated with this positive influence and with increased interest in the field. This is encouraging, particularly given that in this case the clinical experience was largely an inpatient one. Inpatient CAP can be challenging in a number of ways. In 1997, Beresin (19) discussed the impact of health care economics on CAP programs and education; 10 years later, average lengths of stay remain shortened, posing a challenge to providing meaningful clinical exposure to students. Other potentially concerning inpatient issues include managing aggression, seclusion/restraint, and distressing psychosocial situations. Despite these concerns, students found clinical interactions—particularly in evaluations of children—valuable in increasing their interest in CAP. The value of direct clinical contact in a clerkship CAP experience has been introduced (20); our larger sample underscores this finding.
Our findings have some limitations. The lack of a prerotation survey may make responses prone to a recall bias. We decided to administer only the postrotation survey due to concerns regarding guaranteeing anonymity, limited resources, and attempts to optimize the number of responses and statistical power. Moreover, the results reported may not generalize to all psychiatry clerkship programs. As has been noted for medical student interest in general psychiatry (6), intrinsic factors within departments and medical student programs could play a larger role than, for example, the presence of a CAP inpatient site. Here, intrinsic factors such as the quality and quantity of attending and resident teaching, the characteristics of different cases, and staff attitudes, particularly on inpatient services, could play a role. Also, given the few current opportunities for student exposure to child psychiatry, our high CAP exposure group may be somewhat biased as it is a selective experience. However, not all students in that group ranked CAP as their first choice. In the spirit of early identification of students who may develop an interest in CAP, we deemed this bias unavoidable. Further, while we note no significant differences between the group of students from the 2005–2006 entire academic year and the much smaller group from the first rotation of the 2006–2007 academic year, the response rate from the 2006–2007 group was much lower than that from the 2005–2006 academic year. One possible explanation for this was the change to a new curriculum, in which many students split their time between UNC and outside sites, with administrative adaptation of evaluation processes to the new curriculum.
Does a positive interest in child psychiatry at the end of the clerkship predict the eventual decision to enter a CAP residency 4 years later? The aforementioned clerkship study by Clardy et al. (9) noted that students indicating an interest in psychiatry careers at the end of the clerkship were much more likely to ultimately pursue residency training in psychiatry. Yet, the construct of our survey does not measure this, given our focus on ensuring anonymity for candid reporting and return rates.
Kirkpatrick’s (21) triangular-shaped hierarchy of levels of evaluation places our measure and its findings into additional context (22). With the most complex “highest” level of the triangle being improved psychiatric care for children, evaluation of student attitudes toward CAP and their interest in pursuing the field are at much broader, less complex levels. Assessment at all levels is important; in this case our measure could be employed locally to identify students for whom electives in general and child psychiatry could be arranged early in the fourth year. The results could be applied toward the development of mentoring relationships, noted by graduating medical students as a highly influential factor in determining specialty choice (23). It would also be valuable to identify students who may choose psychiatry and eventually CAP residencies.
Our findings support a compelling argument for including a structured CAP component in the third year psychiatry clerkship. In 1993, Wagner and Pollard (24) estimated that only 12% of medical students had CAP experience. To this end, the CAPE-Q could be utilized in individual programs to identify the strengths or weaknesses of the clerkship and could be used together with other measures of skills and knowledge in child psychiatry. Given the continued workforce need and overall limited clinical exposure to CAP in medical education, recommendations for high-quality, longitudinal experiences throughout the curriculum should be considered.
FIGURE 1. Change in Interest in Child and Adolescent Psychiatry Before and After the Psychiatry Clerkship
No change-Yes=student indicated interest in CAP both at the start of and at the conclusion of the clerkship; No change-No=student indicated no interest in CAP both at the start of and at the conclusion of the clerkship
The authors wish to acknowledge Myra Daniel, M.A., for her assistance in the preparation, distribution, and collection of the surveys. They also acknowledge Thomas Anders, M.D., and Elizabeth Armstrong, Ph.D., for their support in the development of this project.