History of Nonsteroidal Anti-inflammatory Drug Use and Functional Outcomes After Spontaneous Intracerebral Hemorrhage

Background and Purpose Preclinical and clinical studies have suggested a potential benefit from COX-2 inhibition on secondary injury activation after spontaneous intracerebral hemorrhage (ICH). The aim of this study was to investigate the effect of pre-admission NSAID use on functional recovery in spontaneous ICH patients. Methods Consecutive adult ICH patients enrolled in the Intracerebral Hemorrhage Outcomes Project (2009–2018) with available 90-day follow-up data were included. Patients were categorized as NSAID (daily COX inhibitor use ≤ 7 days prior to ICH) and non-NSAID users (no daily COX inhibitor use ≤ 7 days prior to ICH). Primary outcome was the ordinal 90-day modified Rankin Scale (mRS) score. Outcomes were compared between cohorts using multivariable regression and propensity score-matched analyses. A secondary analysis excluding aspirin users was performed. Results The NSAID and non-NSAID cohorts comprised 228 and 361 patients, respectively. After 1:1 matching, the matched cohorts each comprised 140 patients. The 90-day mRS were comparable between the NSAID and non-NSAID cohorts in both the unmatched (aOR = 0.914 [0.626–1.336], p = 0.644) and matched (aOR = 0.650 [0.392–1.080], p = 0.097) analyses. The likelihood of recurrent ICH at 90 days was also comparable between the NSAID and non-NSAID cohorts in both the unmatched (aOR = 0.845 [0.359–1.992], p = 0.701) and matched analyses (aOR = 0.732 [0.241–2.220], p = 0.581). In the secondary analysis, the non-aspirin NSAID and non-NSAID cohorts comprised 38 and 361 patients, respectively. After 1:1 matching, the matched cohorts each comprised 38 patients. The 90-day mRS were comparable between the non-aspirin NSAID and non-NSAID cohorts in both the unmatched (aOR = 0.615 [0.343–1.101], p = 0.102) and matched (aOR = 0.525 [0.219–1.254], p = 0.147) analyses. The likelihood of recurrent ICH at 90 days was also comparable between the non-aspirin NSAID and non-NSAID cohorts in both the unmatched (aOR = 2.644 [0.258–27.091], p = 0.413) and matched (aOR = 2.586 [0.228–29.309], p = 0.443) analyses. After the exclusion of patients with DNR or withdrawal of care status, NSAID use was associated with lower mRS at 90 days (aOR = 0.379 [0.212–0.679], p = 0.001), lower mRS at hospital discharge (aOR = 0.505 [0.278–0.919], p = 0.025) and lower 90-day mortality rates (aOR = 0.309 [0.108–0.877], p = 0.027). Conclusions History of nonselective COX inhibition may affect functional outcomes in ICH patients.