BENZODIAZEPINES AND ZOLPIDEM FOR CHRONIC INSOMNIA - A METAANALYSIS OFTREATMENT EFFICACY

Objective.-To evaluate the efficacy of benzodiazepines and zolpidem ta rtrate in chronic insomnia based on a quantitative review of literatur e. Data Sources.-Articles from 1966 to 1998 were identified using MEDL INE, by a manual review of relevant journals, and from bibliographies of identified articles. Study Selection.-Studies using randomized, dou ble-blind, placebo-controlled, parallel or crossover designs with benz odiazepines or zolpidem in adults younger than 65 years with chronic i nsomnia (modified Diagnostic and Statistical/Manual of Mental Disorder s, Fourth Edition criteria for primary insomnia) were selected for rev iew. Self-report and polysomnographic measures of sleep-onset latency, total sleep time, number of awakenings, and sleep quality were select ed as outcomes. Data Extraction.-Twenty-two studies met the selection criteria. A combined test of P values was performed, pooling broadly f rom the 22 studies to determine whether medication was superior to pla cebo. A combined test of effect sizes was performed on the subset of s tudies that reported effect size information to determine the magnitud e of medication effect. Data Synthesis.-A homogeneous sample of studie s summarized 1894 patients treated for a median duration of 7 days. Th e combined test of P values demonstrated that medication was superior to placebo in all 4 outcome measures. Treatment response was moderate in magnitude by the combined test of effect sizes. Conclusions.-Benzod iazepines and zolpidem produced reliable improvements in commonly meas ured parameters of sleep in patients with chronic insomnia. Relative t o the chronic and recurring course of insomnia, both the limited durat ion of treatments studied and the lack of follow-up data from controll ed trials represent challenges for developing evidence-based guideline s for the use of hypnotics in the management of chronic insomnia.