Differential Changes in QTc Duration during In-Hospital Haloperidol Use

Le résumé fourni par la source

AIMS: To evaluate changes in QT duration during low-dose haloperidol use, and determine associations between clinical variables and potentially dangerous QT prolongation. METHODS: In a retrospective cohort study in a tertiary university teaching hospital in The Netherlands, all 1788 patients receiving haloperidol between 2005 and 2007 were studied; ninety-seven were suitable for final analysis. Rate-corrected QT duration (QTc) was measured before, during and after haloperidol use. Clinical variables before haloperidol use and at the time of each ECG recording were retrieved from hospital charts. Mixed model analysis was used to estimate changes in QT duration. Risk factors for potentially dangerous QT prolongation were estimated by logistic regression analysis. RESULTS: Patients with normal before-haloperidol QTc duration (male ≤430 ms, female ≤450 ms) had a significant increase in QTc duration of 23 ms during haloperidol use; twenty-three percent of patients rose to abnormal levels (male ≥450 ms, female ≥470 ms). In contrast, a significant decrease occurred in patients with borderline (male 430-450 ms, female 450-470 ms) or abnormal before-haloperidol QTc duration (15 ms and 46 ms, respectively); twenty-three percent of patients in the borderline group, and only 9% of patients in the abnormal group obtained abnormal levels. Potentially dangerous QTc prolongation was independently associated with surgery before haloperidol use (OR(adj) 34.9, p = 0.009) and before-haloperidol QTc duration (OR(adj) 0.94, p = 0.004). CONCLUSION: QTc duration during haloperidol use changes differentially, increasing in patients with normal before-haloperidol QTc duration, but decreasing in patients with prolonged before-haloperidol QTc duration. Shorter before-haloperidol QTc duration and surgery before haloperidol use predict potentially dangerous QTc prolongation.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.