266 - Racial/Ethnic (R/E) Disparities in Neonatal Abstinence Syndrome (NAS) Using Dual Exposure and Withdrawal ICD-10 Codes: Improved Sensitivity

Poster Number: 266
Publication Number: 266.2242

Background: We previously reported weighted sample prevalence of US NAS using ICD-10-CM P96.1, with rates per 1000 population, of 9.8 in 2016 and 2019 HCUP-KID data, and R/E disparities, of 14.7 in White compared to 6.0 in Black and 4.3 in Hispanic newborns. A recent CDC study, however, suggests combined diagnostic exposure (P04.49) and withdrawal (P96.1) codes may be more sensitive for identifying NAS than P96.1 alone.

Objective: We examine the sensitivity of NAS prevalence and R/E disparities using dual exposure and withdrawal ICD-10-CM codes.

Design/Methods: Cross sectional HCUP-KID data from 2016 and 2019 were used to examine NAS prevalence using two code diagnostic criteria (P96.1 + P04.49) among newborns ≥35 weeks gestational age. Iatrogenic cases (P96.2) were excluded. We estimated R/E NAS disparities, reported as cases per 1000, using multivariable models with predictive margins adjusting for: sex, payer (Medicaid/Private), income (ecologic quartiles), and hospital size (large/medium/small), type (urban non-teaching/urban teaching), and region (Northeast/Midwest/South/West).

Results: Table 1 shows a sensitivity analysis for the weighted prevalence of NAS comparing single vs. dual codes. With dual codes NAS prevalence more than doubled overall and across most factors, and continued to be significantly higher among Whites compared to Blacks and Hispanics; however, the rates for Blacks almost quadrupled. Medicaid payer type (40.18) and rural hospital location (36.5) had the highest observed prevalence. Effect modification was evident by payer type and income level for all R/E comparisons, and consistent with prior research, Whites on Medicaid and in the lowest income quartile had the highest NAS rates [88 (95% CI:82.5, 93.5)]. Furthermore, NAS prevalence was significantly higher among Blacks compared to Hispanics across most factors except rural hospital location which is a substantive change from prior results using P96.1 only.

Conclusion(s): Using dual exposure and withdrawal ICD-10 codes, all NAS rates increased. Additionally, significant differences between Black and Hispanic newborns were observed. The use of dual exposure and withdrawal codes should be considered in future research as it may better identify NAS. Under-representing NAS prevalence may in turn lead to insufficient resources allocated to local programs for substance use disorder and its sequelae, resulting in suboptimal management of affected infants.