Inclusive data practices: disaggregating race and assessing comorbidities among American Indian and Alaska Native individuals in PRAMS.

Abstract

Importance: To assess potential misclassification or exclusion of American Indian and Alaska Native (AI/AN) individuals within the Pregnancy Risk Assessment Monitoring System (PRAMS) Phase 8, we compared differences between aggregated and self-reported race variables and their impact on maternal comorbidities.

Methods/design: We utilized several CDC-provided ethnoracial identity variables alongside a disaggregated variable we created. We then estimated comorbidity prevalences between these groupings to determine the impact of these methodological differences.

Results: PRAMS variables, MRACE_AMI and MAT_RACE_PU, included 13,341 (no distinction between AI and AN) and 7,494 AI (excluded AN altogether), respectively. Our constructed variable (n = 13,383) included 19 ethnoracial-subgroups and 42 tribal members not selecting AI/AN race. We found significant differences in the prevalence of comorbidities by these variables. For instance, the prevalence for diabetes with MAT_RACE_PU was 4.93%, with MRACE_AMI was 4.04%, but our subgroup AI (alone) was 5.46%, and AN (alone) was 1.37%.

Conclusion: Our results highlight significant disparities in maternal comorbidities among AI/AN women when different racial classification strategies are employed. Disaggregating these data revealed differences that are crucial for understanding the unique health challenges faced by various subgroups.