Page 3 - Kansas Journal of Medicine, Volume 10 Issue 3
P. 3

KANSAS JOURNAL of M E D I C I N E                                      from the Kansas City campus, 91% (61/67) from the Wichita re-
                                                                       gional campus, and 100% (8/8) from the Salina regional cam-
STUDENT EXPENSES IN RESIDENCY TRAINING                                 pus. The mean respondent age was 28 (range 24 - 55) years and
continued.                                                             130 (80%) were white (Table 1). Of the 32 non-responders, 26
                                                                       were male (81%), and 26 were from the main campus (81%).
The primary care mission of the regional campuses was expected to      	 Seventy-six students (47%) applied to primary care programs.
lower costs, but this could be countered by the increased distances    The percentage of primary care applicants was higher for women
from major cities and generally higher air fares from regional sites.  (52%) than men (42%) but not statistically significant (χ2(1, N =
                                                                       163) = 0.18, p = .21). Similarly, the percentage of regional campus
METHODS                                                                students applying to primary care (55%) was not significantly
	 All fourth-year students of the University of Kansas School of       higher than the main campus (43%; χ2(1, N = 163) = 0.11, p = .15).
Medicine (KUSM) who participated in NRMP during 2016 were              	 Volume of interviewing. Students applied to an average
surveyed immediately following announcement of NRMP re-                of 38 programs (range 1 - 124), received 16 interview invita-
sults. The survey questionnaire was distributed by e-mail weekly       tions (range 1 - 54), and completed eleven interviews (range
for four weeks. Class leaders sent social media reminders two to       1 - 28; Figure 1). One hundred and fifty-eight students (98%)
three times weekly encouraging students to complete the ques-          interviewed out-of-state, covering 42 states, including Alaska.
tionnaire. As an incentive, a donation proportional to the response    	 A MANOVA to determine the effect of gender, campus (main
rate was offered to each campus graduation celebration fund.           or regional), and primary care on the five dependent variables
	 The 33-item questionnaire was based on a 2015 study con-             related to the volume of interviewing (i.e., the numbers of ap-
ducted on the KUSM Wichita campus,8 literature reviews,1-3,5-12        plications, interview invitations and completions; and cost and
and input from faculty, residents, and students. The question-         time as limiting factors in interview decisions) found no signifi-
naire addressed the number, specialty, and location of programs,       cant difference for gender (Wilks’s Λ = .98, F(5,147) = .73, p =
variables influencing interview choices, cost and time of inter-       .60, η2 = .02), but significant differences between regional and
viewing, sources of funding of interviews, and any costs covered       main campuses (Wilks’s Λ = .89, F(5,147) = 3.75, p = .003, η2
by programs. The questionnaire included opportunities for nar-         = .11) and between primary care and non-primary care appli-
rative comments on specific items and the overall interviewing         cants (Wilks’s Λ = .75, F(5,147) = 9.8, p < .001, η2 = .25; Table 2).
process. The instrument was pilot-tested by eight students who
participated in early match processes. Minor changes were made         Table 1. Study participants.
to four questions to clarify meaning and avoid potential ambiguity.
	 Descriptive analyses provided details about the students and                           Respondents (%)
their survey responses. Chi-square tests were used to determine                              (N = 163)
if there were any statistical differences by specialty choice (pri-
mary care versus non- primary care), gender (male and female), as      Sex
well as campus location (main and regional). This test was chosen
because it is used to compare observed frequencies to expected              Female       79 (49)
frequencies. T-tests were used to compare the average costs of in-
terviewing by specialty choice (primary care versus non-primary              Male        84 (52)
care) and campus location (main and regional). This test was se-
lected for these variables because they only have two levels, and      Race
the differences between the two levels were of interest. Multivari-
ate analysis of variance (MANOVA) tests were used for simul-                 White       130 (80)
taneous comparisons between students by gender, campus, and
application to primary care (defined as all family medicine, in-             Asian       19 (12)
ternal medicine, pediatrics, and medicine/pediatrics programs).
Analyses of variance (ANOVA) on the dependent variables were                 Black       6 (4)
conducted as follow-up tests to all significant MANOVA results.
Using the Bonferroni Method, each ANOVA was tested at the              Other or missing  8 (5)
.025 level. MANOVAs were used because they are able to con-
trol for any correlations between dependent variables, while test-     Campus
ing for significance between multiple groups. This study was ap-
proved by the University of Kansas Institutional Review Board.               Main        94 (58)

RESULTS                                                                     Regional     69 (42)
	 Participants. Of 195 eligible students, 163 (84%) complet-
ed the questionnaire. The response rates were 78% (94/120)                   Speciality of Application

                                                                       Primary Care      76 (47)          31 (19%) family medicine
                                                                                                          28 (17%) internal medicine
                                                                                                          13 (8%) pediatrics
                                                                                                          4 (2.5%) medicine-pediatrics

                                                                       Non-Primary       87 (53)          35 (22%) surgical specialties
                                                                            Care                          10 (6%) anesthesiology
                                                                                                          9 (6%) obstetrics/gynecology
                                                                                                          7 (4%) radiology, emergency
                                                                                                          6 (4%) psychiatry
                                                                                                          3 (2%) dermatology, neurology,
                                                                                                          1 (0.6%) preventive medicine,
                                                                                                          pediatric neurology, other

   1   2   3   4   5   6   7   8