Abstract / Summary
Butorphanol is sometimes added to multimodal intravenous patient-controlled analgesia (PCA), but its incremental analgesic benefit and adverse symptom burden when combined with an established sufentanil-based regimen remain uncertain. This single-center retrospective cohort study used routinely collected perioperative, postoperative PCA follow-up, and hospital billing data from July 2022 to June 2025. Patients who underwent gynecologic surgery and received one of four eligible base PCA regimens were included. Movement-evoked pain on postoperative day 1 was the primary effectiveness outcome, and any dizziness was the primary safety outcome. Propensity score overlap weighting was used for the primary analysis. The cohort included 1557 PCA treatment records, of which 239 contained butorphanol and 1318 did not. After overlap weighting, the adjusted mean differences were 0.01 points [95% confidence interval (CI) − 0.09 to 0.10; P = 0.935] for movement-evoked pain, 0.07 points (95% CI − 0.02 to 0.16; P = 0.139) for resting pain, and − 0.78 mL (95% CI − 3.17 to 1.60; P = 0.539) for PCA volume delivered. Dizziness occurred in 13.6% of butorphanol-containing records and 6.0% of controls [adjusted risk difference 7.7 percentage points, 95% CI 3.3–12.5; odds ratio (OR) 2.49, 95% CI 1.53–4.03; P < 0.001]. Higher odds were also observed for nausea (OR 1.76, 95% CI 1.22–2.47; P = 0.005) and co-occurring dizziness and nausea (OR 5.85, 95% CI 3.10–11.41; P < 0.001). In the measured overlap population, adding butorphanol did not materially improve early analgesic outcomes but was associated with more dizziness and nausea. Residual confounding precludes definitive causal conclusions.