Psychiatric drug research. Sample size requirements for one vs two raters

Clinical trials in psychiatry: background and statistical perspective

There have been many hundreds of studies which have examined the efficacy of different forms of treatment of psychiatric disorders over the past 40 years. This paper presents some background to these studies for statisticians, illustrating and discussing some of the difficult problems which arise in this specialty of medicine. It also demonstrates a major requirement for statisticians to influence both the design and presentation of clinical trials and provides some suggestions about how this may be done.

Subjective response to neuroleptics and outcome in schizophrenia: a re-examination comparing two measures

The schizophrenic patient's early psychological response to neuroleptic treatment has been demonstrated to be a significant predictor of treatment response. The validity of the construct of subjective response is evaluated by comparison of two measures. Fifty-five recently admitted and unmedicated schizophrenic patients were randomly allocated to chlorpromazine or haloperidol. Subjective responses at 24 and 48 hours as assessed by two different scales, the Van Putten & May scale and the self-administered Drug Attitude Inventory, were strongly correlated with outcome at three weeks. Early emergence of extrapyramidal symptoms was not related to subjective response, but dysphoric patients had a greater incidence of EPS by the end of treatment than did non-dysphoric patients. The two measures showed high concordance in identification of early drug dysphoria.

Beta, or type II error in psychiatric controlled clinical trials

Fifty-two reportedly "negative" psychiatric controlled clinical trials from 1980-1984 were re-examined to note if large enough samples had been studied to give a high probability (greater than 0.90) of detecting a 20% or 50% improvement in therapeutic response, defined as a reduction in the non-response rate. Forty-four of the trials had a larger than 10% risk of missing a true 20% difference in non-response rates, and with the same risk, 25 of the trials could have missed a 50% difference between treatments. Based on estimates of 90% confidence intervals, a potential 20% difference was possible for 47, and 30 had a potential true difference of 50% in non-response rates between experimental and control conditions. Far more reports describing "non-significant" studies could not be re-examined, due to lack of reporting part or all of the relevant results. The possibility of missing considerable therapeutic improvement, or important clinical differences, because of small sample size demands more concern in psychiatry.

A double-blind comparison of melperone and thiothixene in psychotic women using a new rating scale, the CPRS

Eighty-one women with psychosis of schizophrenic and paranoid type were assigned to a double-blind study comparing the clinical effects of melperone (100 mg X 3) and thiothixene (10 mg X 3). The antipsychotic effect was evaluated by clinical rating according to the CPRS and the NOSIE-30 scales before and after 2 and 4 weeks of drug treatment. A satisfactory interrater reliability was obtained for the CPRS. Significant correlation was also found between the CPRS and NOSIE ratings. Treatment with both drugs was associated with significant reductions in morbidity as estimated by several measures of therapeutic effect from the CPRS, by the NOSIE scale and by global ratings. There were no marked differences at any rating time point between the drugs in this regard. There were more extrapyramidal side effects in the thiothixene group than in the melperone-treated patients. The results encourage the use and further evaluation of melpherone in the treatment of psychotic patients.

Intensive design in evaluating anxiolytic agents

The purposes of this study were: (1) to test the usefulness of intensive design in detecting the effects of an established antianxiety agent in a single patient studied for a period as brief as 8 weeks and (2) to explore the usefulness of combining intensive and extensive designs by jointly analyzing the results from several similarly treated patients. Fifteen primarily anxious, psychoneurotic patients aged 21-50 and scoring 17 or more on the Taylor Manifest Anxiety Scale were admitted to the study; and 11 completed the full treatment program. Medications were diazepam 5 mg t.i.d. and a matching placebo, administered under double-blind conditions. Patients were treated for 8 weeks, divided into 42-week blocks. In each block, the patient received diazepam 1 week and placebo the other, with the order in each block determined at random. The patient came weekly for evaluation, including, self-ratings on the Hopkins Symptom Checklist (SCL), global status, global change; reports of occupational and social function; resting pulse; reaction time; psychiatrist's ratings on the Hamilton Anxiety Scale, global status and global change. The patient also reported daily his mood on the Profile of Mood States (POMS). Mean deviations from the general trend for post-diazepam and postplacebo scores on each criterion were compared within patients. Diazepam-placebo differences on each criterion were analyzed between patients. Criteria that clearly recorded the anti-anxiety effect of diazepam as compared to placebo included the Hamilton Anxiety Scale, the psychiatrist's global status and global change ratings, the SCL Anxiety and Somatization Scales, and the POMS Anxiety Scale. Other criteria that showed a reliable diazepam effect included SCL Depression (decrease), POMS Vigor (increase), POMS Fatigue (decrease), SCL Anger (increase), and reaction time (increase). The most sensitive criteria distinguished diazepam from placebo even when results were considered only from the first 6 patients during their first 4 weeks of treatment- a total of 24 patient weeks of treatment. The factors contributing to the sensitivity of this design were investigated and discussed.