A positive association for one postoperative outcome should not be generalized to all endpoints.
The retrospective cohort included 4,307 hip or knee arthroplasty patients. Preoperative benzodiazepine prescribing was associated with higher adjusted readmission odds. It was not clearly associated with respiratory failure or longer stay, and some other complications were too uncommon for analysis.
Treatment was not randomized. Anxiety, withdrawal, drug interactions and underlying disease could affect readmission. Prescription duration and actual exposure were incompletely characterized. A nonsignificant respiratory result does not demonstrate absence of risk.
For a postoperative opinion, identify the specific reason and timing of readmission. Keep unmeasured or unanalyzable endpoints separate from negative findings. The paper supports a particular observational signal while leaving both its mechanism and the cause of an individual patient’s complication unresolved.
Source: Publication PMC8131602. Applications to expert review are interpretive analysis.
Related reading: Arthroplasty Cohort Studied New Prescription Overlap Rather Than Overdose.