The association of baseline psychiatric comorbidity with healthcare utilization, status epilepticus, and all-cause mortality in adults with epilepsy: an exploratory cohort study

Psychiatric comorbidity is common among adults with epilepsy and may influence subsequent healthcare use and clinical outcomes. We examined its association with recorded mortality, status epilepticus, and healthcare utilization during the first year after a first recorded qualifying epilepsy event. We conducted a retrospective cohort study using the TriNetX Global Collaborative Network (171 contributing healthcare organizations), from January 1, 2015, to January 1, 2025. Adults aged ≥ 18 years with a first recorded qualifying epilepsy event (ICD-10-CM G40) followed by a recorded antiseizure medication on or after the index event and within nine months were classified according to the presence or absence of selected baseline psychiatric disorders (ICD-10 F20–F48) recorded during the five years preceding or on the index date, and 1:1 propensity-score matched on measured demographic, comorbidity, procedure, and antiseizure-medication covariates. A single baseline matched cohort was analyzed with Cox proportional-hazards models across an initial (days 1–90) and a subsequent landmark (days 91–365) post-index window, reporting hazard ratios (HRs) with 95% confidence intervals (CIs). The study was exploratory, without adjustment for multiple comparisons. After propensity-score matching, 75,895 patients were included in each cohort. Baseline psychiatric comorbidity was associated with higher recorded emergency and inpatient utilization in both windows (emergency HR 1.353 [95% CI 1.271–1.441] initial, 1.446 [1.362–1.536] subsequent; inpatient HR 1.162 [1.107–1.219] initial, 1.131 [1.061–1.205] subsequent). No statistically significant association was detected with status epilepticus (initial HR 0.746 [0.471–1.182]; subsequent 0.811 [0.608–1.080]), although event counts were low and estimates imprecise. Recorded mortality was lower in the initial window (HR 0.834 [0.793–0.877]) and attenuated toward the null in the subsequent window (HR 0.939 [0.884–0.998]); this inverse association is most consistent with selection, observability, ascertainment, and residual-confounding limitations rather than a protective effect. Baseline psychiatric comorbidity was associated with greater recorded acute healthcare utilization, particularly emergency care among adults with a first recorded qualifying epilepsy event. The status-epilepticus findings were inconclusive, and the inverse recorded-mortality association should not be interpreted as protective. These exploratory findings are hypothesis-generating; replication using validated incident-epilepsy definitions, aligned eligibility and follow-up, equivalent pre-index observability, and more complete mortality ascertainment is warranted.