Measurement of American Indian and Alaska Native Racial Identity Among Medical School Applicants, Matriculants, and Graduates, 1996-2017.

Importance: Accurate racial/ethnic identity measurement is needed to understand the effectiveness of outreach, recruitment, and programs to support American Indian and Alaska Native (AIAN) people becoming physicians. Objective: To examine how changes in race/ethnicity data collection by the American Medical College Application System are associated with trends in applicants, matriculants, and graduates self-reporting as AIAN. Design, Setting, and Participants: In this cohort study, interrupted time series regression was conducted using data from the American Medical College Application system identifying medical school applicants and graduates between January 1, 1996, and December 31, 2017, who identified as AIAN. The number of students identifying as AIAN was compared before and after the American Medical College Application System changed how it collected race/ethnicity data in 2002. Data analyses were conducted between December 2019 and May 2019. Exposures: Applicants could select only 1 racial identity from 1996 to 2001 and could select more than 1 racial identity from 2002 to 2017. Main Outcomes and Measures: Rates of AIAN groups before and after changing how race/ethnicity data were collected. Covariates were age, sex, and Medical College Admission Test scores. Results: The total number of individuals identifying as AIAN in the study was 8361; the mean (SD) number of applicants per year was 380.0 (89.9) overall: 257.3 (39.6) in 1996 to 2001, with a mean (SD) age of 26.6 (5.5) years and 830 (54.0%) male individuals, and 426.1 (50.1) in 2002 to 2017, with a mean (SD) age of 25.5 (5.6) years and 3441 (50.5%) female individuals. Before the change, there was a decrease of 5% per year (relative rate [RR] of 0.95; 95% CI, 0.91-0.98; P < .001) in the rate of AIAN applicants. In 2002, the change in data collection was associated with an immediate 78% relative increase in applicants (RR, 1.78; 95% CI, 1.55-2.06; P < .001). From 2002 to 2017 there was a 10% increase in applicants per year (RR, 1.10; 95% CI, 1.06-1.14; P < .001). For matriculants, yearly trends indicated a nonsignificant 3% decrease before the change, whereas the change was associated with an immediate 62% relative increase in matriculants (RR, 1.62; 95% CI, 1.35-1.95; P < .001), with no difference in trend after the change. For graduates, a nonsignificant yearly decrease of 2% was found in the mean number of graduates before the change, whereas the change was associated with an immediate 94% relative increase (RR, 1.94; 95% CI, 1.57-2.38; P < .001), followed by no change in trend after the modification. Conclusions and Relevance: Changing the method of race/ethnicity data collection captured more AIAN applicants, matriculants, and graduates. Yearly trends indicate concerning although nonsignificant differences after the change for AIAN graduates. These findings should inform diversity efforts.