Non‐Steroidal Anti‐Inflammatory Drug Versus Acetaminophen Initiation at Influenza Diagnosis and 14‐Day Neuropsychiatric Diagnoses in Adults: A Target Trial Emulation Using Japanese Claims Data
Satoshi Kutsuna, Shungo Yamamoto, Yoshiki KusamaABSTRACT
Purpose
After case series in which diclofenac and mefenamic acid were associated with worse outcomes in children with established influenza‐associated encephalopathy, Japan recommended acetaminophen for influenza fever in 2001. We estimated the 14‐day risk of newly recorded neuropsychiatric diagnoses after initiating a non‐steroidal anti‐inflammatory drug (NSAID) versus acetaminophen at influenza diagnosis in adults, a related but distinct question.
Methods
Using the JMDC Claims Database (2018–2025), we emulated an active‐comparator, new‐prescription‐user target trial in adults aged 18–64 years. Time zero was an influenza diagnosis with a same‐day, same‐facility prescription of an oral or rectal NSAID or single‐agent acetaminophen; the initial prescription defined the group. The outcome was a broad neuropsychiatric composite newly recorded on Days 1–14 after a 365‐day component‐specific diagnosis‐free period. We estimated the average treatment effect in the propensity‐score overlap population using overlap weighting.
Results
Among 1 125 391 eligible adults, 140 877 (12.5%) initiated an NSAID (85.9% loxoprofen) and 984 514 initiated acetaminophen. The composite occurred in 112 and 699 episodes; overlap‐weighted 14‐day risks were 79.6 and 71.2 per 100 000 (risk ratio: 1.12, 95% CI: 0.92–1.36; risk difference 8.4 per 100 000, 95% CI: −7.2 to 24.0). Estimates were similar across outcome definitions and sensitivity analyses. Diclofenac and mefenamic acid were too rarely prescribed to evaluate.
Conclusions
In adults, NSAID initiation, predominantly loxoprofen, showed no clear difference in short‐term neuropsychiatric diagnoses compared with acetaminophen, although a relative increase of up to 36% and residual confounding by presenting symptoms were not excluded. The findings do not address diclofenac, mefenamic acid, children or established influenza‐associated encephalopathy.
Trials Registration
UMIN000062444.