Antipsychotic Co-prescription and Pain Trajectories in Hospitalized Internal Medicine Patients: A Retrospective Cohort Study

Pain management in hospitalized internal medicine patients is challenging due to multimorbidity, polypharmacy, and behavioral symptoms. Antipsychotics are frequently co-prescribed with analgesics in this population alongside other central nervous system-active medications, yet their relationship with acute pain trajectories remains poorly understood. This retrospective cohort study examined whether antipsychotic co-prescription was associated with pain trajectories and rehospitalization among hospitalized medical patients receiving analgesic therapy, including an opioid-treated subgroup. Internal medicine inpatients receiving in-hospital analgesic therapy were included. Pain outcomes comprised mean, maximum, and last visual analog scale (VAS) scores and peak-to-last change in VAS (ΔVAS); all-cause rehospitalization was assessed after discharge. Propensity score matching accounted for demographic and clinical factors. Additional analyses examined timing of antipsychotic exposure and co-treatment with other central nervous system-active medications. Among 18,387 patients hospitalized for ≥ 48 h, 6166 with at least one VAS assessment within the first 2 days of hospitalization formed the analgesic-treated cohort, including 2388 receiving opioids. After matching, antipsychotic co-prescription was associated with lower mean and maximum VAS scores in the analgesic-treated cohort, particularly among older adults. The association was most evident among patients already receiving antipsychotics at admission, whereas findings among those initiated during hospitalization were limited. No consistent differences in pain outcomes were observed in the opioid subgroup, while associations in the broader analgesic cohort persisted across central nervous system (CNS) co-medication strata. Antipsychotic co-prescription was not associated with sustained increases in rehospitalization risk. Antipsychotic co-prescription was associated with lower peak pain intensity in the broader analgesic-treated cohort, particularly among older adults, but not in the opioid subgroup. This context-dependent pattern differed according to analgesic treatment and timing of antipsychotic exposure and may reflect differences in multidimensional pain processing. These findings highlight the importance of considering central nervous system-active co-medications in pain management and warrant prospective investigation.