Ambiguous findings concerning potential advantages of depot antipsychotics

Abstract

We review the recent literature on the comparative efficacy/effectiveness of antipsychotic long-acting injections (LAIs) vs. oral antipsychotics. Randomized controlled trials (RCTs), mirror-image studies and cohort studies are reviewed, and the strength and weaknesses of each design discussed. The data are conflicting and differ by trial design. RCTs tend to show equivalence for the two treatments, whereas mirror-image studies and cohort studies are more likely to show benefits for LAIs, although a prescribing/channelling bias in cohort studies suggests that LAIs are prescribed to more severely ill patients. Two recent, large RCTs, in a more real-world treatment setting, found a longer time to relapse or treatment failure with a LAI versus oral antipsychotics. Each of the three trial designs has strengths and weaknesses. The traditional RCT does not seem to be the gold standard to compare the effectiveness of oral antipsychotics and LAIs because of selective recruitment and altered ecology of the treatment. Conversely, weaknesses of observational studies include lack of blinding and difficulty disentangling the effect of treatment from prescribing bias and confounders. The best approach to determining comparative effectiveness may be a pragmatic RCT with broad inclusion criteria and minimal exclusion criteria and minimal assessment following baseline assessment and randomization.