This review tried to determine the optimal dose of haloperidol decanoate for preventing relapse in schizophrenia. The authors concluded that the optimal dose seems to be 50 to 100 mg every 4 weeks, with little gained at doses above 100 mg every 4 weeks. The considerable limitations to this review, such as the use of mean data in dose-response curves, mean that these conclusions may not be reliable.

To establish the range of doses of haloperidol decanoate (HD) that are effective in preventing relapse in schizophrenia.

MEDLINE, EMBASE, PsycLIT, and the Cochrane Library were searched; the search terms were reported. A Cochrane review of HD was also searched.

Study designs of evaluations included in the review

There were no inclusion criteria relating to the study design.

Specific interventions included in the review

Studies describing the use of specific doses of HD for at least 6 months were eligible for inclusion. The included studies used doses ranging from 15 to 1,800 mg every 4 weeks.

Participants included in the review

Studies of people with schizophrenia were eligible for inclusion. The studies included in- and out-patients.

Outcomes assessed in the review

Studies reporting on relapse of schizophrenia were eligible for inclusion.

How were decisions on the relevance of primary studies made?

The author did not state how the studies were selected for the review, or how many reviewers performed the selection.

The author did not state that they assessed study quality.

The author did not state how the data were extracted for the review, or how many reviewers performed the data extraction. Relapse rate and the mean dose of HD were extracted.

How were the studies combined?

The author plotted dose-response curves (mean or fixed dose, and log dose of HD against percentage remaining well).

How were differences between studies investigated?

Some study details were tabulated.

Six studies (n=251) were included in the review. One study was a double-blind, randomised placebo-controlled trial, four were specified as double-blind trials, and one was an open evaluation. The duration of the studies ranged from 8 months to 1 year.

The relapse rate reported in the included studies varied between 0% (1 study using 385 mg HD/4 weeks) and 60% (1 study using 25 mg HD/4 weeks). The dose-response curves showed HD as having little effect on relapse rate below 25 mg/4 weeks. The relapse rate decreased when the dose was increased to between 25 and 100 mg, then levelled off between 100 and 400 mg. The author stated that the effect on relapse rate may be maximal at 50 mg/4 weeks.

The recommended dose range for HD (50 to 300 mg/4 weeks) is not supported by this review. Little can be gained by increasing the dose of HD to above 100 mg/4 weeks.

The research question was clear in terms of the participants, intervention and outcomes, but no inclusion criteria were specified for the study design. Several relevant sources were searched, although it was unclear whether language restrictions were applied and the potential for publication bias was not investigated. There was no information relating to the review methodology, therefore it was not possible to determine whether any efforts were made to reduce error and bias during the review process. There was no systematic assessment of the quality of the included studies, so the results from these studies and any synthesis may not be reliable. It was not reported whether the data were analysed on an intention-to-treat basis. In addition, there was limited information about the participants and co-interventions, which means that the comparability of study populations cannot be judged. The dose-response curves were produced using known fixed doses along with mean doses plotted as single points, which may produce misleading results. The author acknowledged that the analysis did not take account of effect of the duration of follow-up on relapse rate, and this too might have affected the reliability of the results. The potential for error and bias, the inclusion of mean data in the production of the dose-response curves, and other limitations discussed above, mean that the author's conclusions may not be reliable.

Practice: The author stated that optimal efficacy seems to be achieved at 50 to 100 mg/4 weeks, and that the administration of doses above 100 mg/4 weeks is not supported by the data available. Further, clinicians need to decide whether doses higher than this are justified given the lack of increased efficacy, as well as the increased cost and adverse events (neither of which were assessed in the review).

Research: The authors did not state any implications for future research.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.