Trenton Honda1, Trenton D Henry2, Ellen D Mandel3, Alicia Quella4, José E Rodríguez5, Shahpar Najmabadi2, Virginia L Valentin6. 1. Bouvé College of Health Sciences, Northeastern University, Boston, MA, USA. t.honda@northeastern.edu. 2. Department of Family and Preventive Medicine, University of Utah, Salt Lake City, UT, USA. 3. Physician Assistant Program, St. Elizabeth University, Morristown, NJ, USA. 4. Department of Physician Assistant Studies, Augsburg University, Minneapolis, MN, USA. 5. Office of the Associate Vice President for Health Equity, Diversity, and Inclusion, University of Utah Health, UT, Salt Lake City, USA. 6. Division of Physician Assistant Studies Department of Family and Preventive Medicine, University of Utah, UT, Salt Lake City, USA.
Abstract
BACKGROUND: Physician Assistants (PA) are important members of the medical team, and increasing diversity in healthcare professionals has been consistently associated with improved health outcomes for underrepresented minority patients. In this study of a national cohort of PA program applicants, we investigated whether the number of programs a student applied to (Application Number, AN) was significantly associated with increased likelihood of matriculation into a PA program. METHODS: We examined all applications (n = 27,282) to the 2017-2018 admissions cycle of the Central Application Service for Physician Assistants, which is utilized by over 90% of accredited PA programs in the US. As we a priori hypothesized that associations would be non-linear, we used natural cubic splines to estimate the associations between matriculation and AN, controlling for multiple metrics of academic achievement, experience, and applicant demographics. We subsequently used segmented regression analyses (modified poisson regression with robust error variance) to investigate log-linear associations above and below inflection points identified in the spline analyses. Additionally, we explored for effect modification by race/ethnicity. RESULTS: The strongest associations were observed between application number 2-7, and a threshold effect was observed at > 16 applications, beyond which there was no significant, incremental benefit in matriculation likelihood. Associations differed by race, particularly for application number 2-7, wherein the incremental benefit from each additional application was highest for Black applicants (Likelihood Ratio [LR]: 1.243, 95% CI: 1.136 to 1.360) vs non-Latinx White (LR: 1.098, 95% CI: 1.072 to 1.125), with no additional, incremental benefit beyond 7 program applications. For all other races, significant increased likelihoods of matriculation were observed until 16 program applications. CONCLUSIONS: These findings can help guide pre-PA advisors and PA programs, providing recommended thresholds to applicants on the most cost effective ways to increase their likelihood of admissions, and the PA profession as a whole by providing actionable information that can potentially increase Race/Ethnic diversity in the PA profession and, by extension, medical teams.
BACKGROUND: Physician Assistants (PA) are important members of the medical team, and increasing diversity in healthcare professionals has been consistently associated with improved health outcomes for underrepresented minority patients. In this study of a national cohort of PA program applicants, we investigated whether the number of programs a student applied to (Application Number, AN) was significantly associated with increased likelihood of matriculation into a PA program. METHODS: We examined all applications (n = 27,282) to the 2017-2018 admissions cycle of the Central Application Service for Physician Assistants, which is utilized by over 90% of accredited PA programs in the US. As we a priori hypothesized that associations would be non-linear, we used natural cubic splines to estimate the associations between matriculation and AN, controlling for multiple metrics of academic achievement, experience, and applicant demographics. We subsequently used segmented regression analyses (modified poisson regression with robust error variance) to investigate log-linear associations above and below inflection points identified in the spline analyses. Additionally, we explored for effect modification by race/ethnicity. RESULTS: The strongest associations were observed between application number 2-7, and a threshold effect was observed at > 16 applications, beyond which there was no significant, incremental benefit in matriculation likelihood. Associations differed by race, particularly for application number 2-7, wherein the incremental benefit from each additional application was highest for Black applicants (Likelihood Ratio [LR]: 1.243, 95% CI: 1.136 to 1.360) vs non-Latinx White (LR: 1.098, 95% CI: 1.072 to 1.125), with no additional, incremental benefit beyond 7 program applications. For all other races, significant increased likelihoods of matriculation were observed until 16 program applications. CONCLUSIONS: These findings can help guide pre-PA advisors and PA programs, providing recommended thresholds to applicants on the most cost effective ways to increase their likelihood of admissions, and the PA profession as a whole by providing actionable information that can potentially increase Race/Ethnic diversity in the PA profession and, by extension, medical teams.
Entities:
Keywords:
Matriculation; Medical education; Physician assistant/associate; Race/ethnicity; Underrepresented minority