Critical appraisal — structured clinical viva

Validity (strong relative to many psychopharmacology trials). Central concealment, double-blind design with blinded assessors, ITT, and multi-centre recruitment support internal validity on RoB 2 domains for randomisation, deviations, and outcome measurement — still probe missing data rates and selective reporting of secondaries.^[2] Industry funding is a conflict to name, not an automatic veto; look for independent analysis and pre-registration.

Results. Statistically significant mean PANSS difference with CI excluding zero; magnitude (−4 points) may be below some proposed MCIDs — discuss clinical importance separately from p-value. Prefer dichotomous outcomes (response/remission) and absolute differences if available for NNT framing.^[3] EPS benefit is clinically meaningful if scales and rates are robust; similar discontinuation is not proof of equal overall tolerability — absence of difference ≠ equivalence without a proper non-inferiority design and powering.^[6]

Applicability. First-episode patients may differ from chronic multi-episode samples; duration 6 weeks is acute, not long-term metabolic risk; shared decision-making must include metabolic monitoring, patient values (EPS vs weight), and cost. Place this trial against the broader meta-analytic landscape of SGA vs FGA trade-offs rather than a single abstract.^[5][1]

GRADE language. Moderate certainty if only minor concerns remain (industry, short duration, continuous primary); strong recommendation for this specific drug over haloperidol would require larger absolute benefits, longer outcomes, and consistency across studies.^[4]