Surveillance, violence, and assistance: Distinct typologies of police contact and prospective health outcomes among people who use drugs.

Policing is increasingly understood as a structural determinant of health, yet the social epidemiology of police contact remains limited by limitations to variable-centered measurement which may not capture the complexity of exposure. We aim to address this gap in a cohort of people who use drugs, a population facing disproportionate and consequential police contact, using person-centered methods to examine qualitatively distinct exposure profiles. We applied latent class analysis to 11 police contact indicators among a longitudinal cohort of people who use drugs in Baltimore, Maryland (N = 486) and estimated associations between baseline police-contact class and health outcomes at 6-month follow-up using BCH-weighted logistic regression with cluster-robust standard errors. We identified three classes: minimal contact (64%), routine enforcement (25%), and violent enforcement (11%), the latter characterized by high surveillance alongside violence, coercion, and paradoxically, police assistance. After adjustment, those in the violent enforcement class had increased odds of nonfatal overdose at follow-up (aOR 3.23, 95% CI 1.78-5.87), while those in the routine enforcement class had higher odds of injection drug use (aOR 2.85, 95% CI 1.56-5.23) and lower odds of HIV testing (aOR 0.58, 95% CI 0.39-0.85). These findings demonstrate that police contact comprises qualitatively distinct typologies associated with distinct downstream health consequences among people who use drugs, and point to a need for longitudinal research to elucidate the mechanisms through which each operates. Implications of the co-occurrence of severe enforcement and police assistance among the most impacted individuals for reforms which aim to mitigate harmful police contact on population health are discussed.