IMPORTANCE: Health-related social needs are drivers of worse health and high health care spending. In North Carolina Medicaid's 1115 waiver, the Healthy Opportunities Pilots (HOP) program allows for nonmedical services to address health-related social needs, such as healthy food boxes and housing navigation, financed by Medicaid, in 3 regions of North Carolina. OBJECTIVE: To determine whether individual participation in the HOP program is associated with lower Medicaid spending. DESIGN, SETTING, AND PARTICIPANTS: Comparative interrupted time series evaluation using North Carolina Medicaid data from March 2021 through November 2023. The 13 227 HOP enrollees were compared with 73 469 Medicaid beneficiaries reporting a health-related social need (a HOP eligibility criterion), but who were ineligible for HOP due to county of residence. EXPOSURE: Participation in HOP. MAIN OUTCOMES AND MEASURES: The primary outcome was the sum of spending by Medicaid on medical and HOP services per beneficiary per month. Secondary outcomes included health care use (emergency department visits, inpatient admissions, and outpatient visits). RESULTS: Among 13 227 HOP and 73 469 comparison group members, the mean age was 23.4 (SD, 18.8) years, 63.9% were female, 42.4% were Black, and 57.4% were White (participants could report >
1 race). Mean monthly preindex spending was 28 (SD, 105). Of HOP enrollees, 89% received at least 1 HOP service, and 85% of services were food services. The comparative interrupted time series analysis estimated that spending increased at the month of HOP enrollment (change in level, 87
95% CI, 20-54). However, the trend in spending for HOP participants was differentially lower (change in trend, -5 per beneficiary per month
95% CI, -22 to -8
P <
.001) relative to a counterfactual scenario without HOP. Taken together, these findings indicated that monthly spending was equivalent to estimated counterfactual expenditures by month 8 after starting HOP and lower thereafter. Emergency department visit trend was differentially lower for HOP participants (-6 per 1000 person-months
95% CI, -8 to -4). There was no statistically significant differential trend in hospitalizations (-1 per 1000 person-months
95% CI, -2 to 0) or outpatient visits (1 per 1000 person-months
95% CI, -7 to 9). CONCLUSIONS AND RELEVANCE: Results of this study suggest that the HOP program was associated with increased spending at enrollment, followed by a subsequently lower spending trend.