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.     

Early stopping of prevention trials when multiple outcomes are of interest: a discussion

In large prevention and screening trials, we want to answer multiple questions simultaneously. When monitoring prevention trials with multiple disease outcomes, composite measures may be useful, such as a count of important events, or a comparison of expected mortality. There are advantages to investigating multiple interventions in the same prevention or screening trial, using all-versus-none, factorial, or reciprocal control designs. In these situations it may be important to answer some questions early while others are still being investigated.     

Antileishmanial chalcones: statistical design, synthesis, and three-dimensional quantitative structure-activity relationship analysis

A large number of substituted chalcones have been synthesized and tested for antileishmanial and lymphocyte-suppressing activities. A subset of the chalcones was designed by using statistical methods. 3D-QSAR analyses using 67 (antileishmanial activity) and 63 (lymphocyte-suppressing activity) of the compounds for the training sets and 9 compounds as an external validation set were performed by using the GRID/GOLPE methodology. The Smart Region Definition procedure with subsequent region selection as implemented in GOLPE reduced the number of variables to approximately 1300 yielding 3D-QSAR models of high quality (lymphocyte-suppressing model, R2 = 0. 90, Q2 = 0.80; antileishmanial model, R2 = 0.73, Q2 = 0.63). The coefficient plots indicate that steric interactions between the chalcones and the target are of major importance for the potencies of the compounds. A comparison of the coefficient plots for the antileishmanial effect and the lymphocyte-suppressing activity discloses significant differences which should make it possible to design chalcones having a high antileishmanial activity without suppressing the proliferation of lymphocytes.     

Comments on a commentary: statistical evaluation of split mouth caries trials

When results of split mouth caries trials are reported, the most frequently used outcome measures are "effectiveness" and "net gain". In a recent review by Riordan & FitzGerald (Community Dent Oral Epidemiol 1994;22:192-7) of the statistical analysis of data from such trials, the close connection between effectiveness and the traditional epidemiological outcome measure, relative risk, was pointed out and the latter measure was recommended. The confidence intervals for these parameters were, however, incorrect. This note provides valid confidence intervals for net gain, effectiveness and relative risk based on data from split mouth trials and presents examples of the calculations.     

Some extensions and applications of a Bayesian strategy for monitoring multiple outcomes in clinical trials

We present some practical extensions and applications of a strategy proposed by Thall, Simon and Estey for designing and monitoring single-arm clinical trials with multiple outcomes. We show by application how the strategy may be applied to construct designs for phase IIA activity trials and phase II equivalence trials. We also show how it may be extended to incorporate the use of mixture priors in settings where a Dirichlet distribution does not adequately quantify prior experience, randomized phase II selection trials involving two or more experimental treatments, and trials with group-sequential monitoring for applications involving multiple institutions.     

Methodological artifacts in moderated multiple regression and their effects on statistical power.

Monte Carlo simulations were conducted to examine the degree to which the statistical power of moderated multiple regression (MMR) to detect the effects of a dichotomous moderator variable was affected by the main and interactive effects of (a) predictor variable range restriction, (b) total sample size, (c) sample sizes for 2 moderator variable-based subgroups, (d) predictor variable intercorrelation, and (e) magnitude of the moderating effect. Results showed that the main and interactive influences of these variables may have profound effects on power. Thus, future attempts to detect moderating effects with MMR should consider the power implications of both the main and interactive effects of the variables assessed in the present study. Otherwise, even moderating effects of substantial magnitude may go undetected. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Constipation and obesity: a statistical analysis

Constipation frequency was investigated among 1897 subjects (874 males and 1023 females), ranging in age from 6 to 70 years, chosen at random between city inhabitants and farmers. Bowel habit was studied in 966 obese patients (390 males and 576 females). The authors followed for the diagnosis of constipation either the commonest and restrictive criterion of the weekly bowel actions or the clinical one which implies also others parameters, like hard or small stools, difficulties of expulsion or feeling of incomplete evacuation after defecation. The statistical analysis showed that constipation frequency is 8.3% in obese patients and 1.5% in normal-weighting, according to weekly bowel actions criterion: the difference is statistically significative (p less than 0.001).     

Clinical trials of antibacterial agents: a practical guide to design and analysis. Statisticians in the Pharmaceutical Industry Working Party

Guidelines on the conduct of clinical trials of antibacterial agents produced by the US Food and Drug Administration, the British Society for Antimicrobial Chemotherapy, the Infectious Diseases Society of America and a European Working Party have been reviewed. Although very informative, these guidelines provide limited practical guidance on the design and statistical aspects of phase III studies of antimicrobial agents. This paper describes the differences between antibacterial trials and clinical studies in other therapeutic areas with regard to subjective endpoints, dual clinical and bacteriological endpoints, frequent protocol violations and difficulty of using placebo controls. The importance of a detailed protocol and planned analysis strategy is emphasized. The choice of comparator agents, practical issues with the blinding of trial materials and the documentation of patients excluded from study entry are discussed. The use of different patient groups and different endpoints in analyses are described. The principles of equivalence and their application to trials of antibacterial agents are discussed, together with an approach to calculating sample size. A variety of statistical analyses of results are compared for different situations indicating some of the problems that can arise. Different methods of presentation of study data are included with emphasis on regulatory submissions rather than scientific publications. Some graphical presentations are recommended and issues regarding data across different studies are discussed.     

Incidence of pelvic inflammatory disease in clinical trials with Cu-7 (intrauterine copper contraceptive): a statistical analysis

Pelvic inflammatory disease (PID) was recorded for 365 of 10,760 women using the Cu-7 for up to 48 months during clinical investigations in the United States. This involved a total exposure of 242,169 woman-months, or some 20,000 woman-years. The overall incidence of PID was 18 cases/1000 woman-years. Incidence was inversely related to age and to duration of use. It was statistically significantly greater in parous subjects than in nulliparous, after stratifying for age. The estimated rates in this study do not differ markedly from those in most other prospective investigations.     

Clinical outcome measures and rating scales in multiple sclerosis trials

In this review, we analyzed clinical outcome measures used in multiple sclerosis (MS) clinical trials in which the primary goal is to slow or arrest progression of disease. In addition, we examined rating scales that quantify symptomatic complications of MS (for example, spasticity) and the current role of magnetic resonance imaging in MS treatment trials. Each proposed scale has advantages and deficiencies, and none meets all the criteria for an ideal outcome measure. The validity of trial design may be improved by using combinations of selected components of current scales as well as new instruments targeted to specific variables (such as motor strength). Symptom-specific rating scales are most appropriately used in trials of symptomatic therapeutic strategies for MS. Until serial magnetic resonance imaging changes are definitely known to predict long-term impairment and disability in patients with MS, clinical outcome measures will remain the primary means of assessing therapeutic efficacy in phase III clinical trials.     

Applied statistical power analysis and the interpretation of nonsignificant results by research consumers.

Contends that the problematic publication of nonsignificant research results is partly due to statistical tests not powerful enough to detect an effect of meaningful size. Authors reporting nonsignificant results should demonstrate the worth of the results by discussing the power of tests. If they do not assume this responsibility, then consumers of research should conduct their own power analyses. The use of power analysis for the interpretation of nonsignificant findings is demonstrated. (33 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

New treatments for multiple sclerosis: a clinical perspective

A comprehensive screening program was initiated in Saudi Arabia in 1982 to determine the frequency of the sickle cell (Hb S) gene in the different regions of Saudi Arabia. Over a period of 10 years 30,055 samples were collected from 36 areas in the five provinces of Saudi Arabia; these were screened using electrophoretic techniques in alkaline and acid pH for the presence of Hb S in heterozygous and homozygous states. The overall prevalence of Hb AS was 7.36% and Hb SS was 1.06%, giving an Hb S gene frequency of 0.047. The results from the different regions were separated and the prevalence of Hb AS and Hb SS and Hb S gene frequencies were calculated in each province and each region. The prevalence of Hb AS and Hb SS cases ranged from O to 25.88% and O to 5.27%, respectively. No case of Hb AS or Hb SS were identified in certain areas of the central province (Al-Russ, Al-Unaiza, Al-Mesnab and Bkaria) and northern province (Qurayat and Al-Jouf). In the southern province both Hb S homozygotes and heterozygotes were encountered in all regions except Farasan Island. In all other regions the Hb S gene was encountered, but at a variable frequency. Applying Hardy-Weinberg equilibrium it was observed that in the majority of the regions the observed Hb S homozygotes were significantly higher compared to the number expected (p < 0.0001). This was believed to be due to bias as the samples were collected in the hospital. The Hb S gene frequency in the different regions ranged from 0 to 0.17 when the frequency was calculated on the basis of both Hb AS and Hb. SS cases, and 0 to 0.13 if the gene frequency was calculated after eliminating the Hb SS cases. A close correlation was observed between the Hb S gene frequency and malaria endemicity. This study shows that the Hb S gene occurs frequently in several regions of Saudi Arabia and there is an urgent need to implement control and prevention programs to reduce the number of Hb S homozygous cases.     

Statistical analysis and optimal design for cluster randomized trials.

In many intervention studies, therapy outcome evaluations, and educational field trials, random treatment assignment of clusters rather than persons is desirable for political feasibility, logistics, or ecological validity. However, cluster randomized designs are widely regarded as lacking statistical precision. This article considers when and to what extent using a pretreatment covariate can increase experimental precision. To answer this question, the author first optimizes allocation of resources within and between clusters for the no-covariate case. Optimal sample sizes at each level depend on variation within and between clusters and on the cost of sampling at each level. Next, the author considers optimal allocation when a covariate is added. In this case, the explanatory power of the covariate at each level becomes highly relevant for choosing optimal sample sizes. A key conclusion is that statistical analysis that fully uses information about the covariate-outcome relationship can substantially increase the efficiency of the cluster randomized trial, especially when the cost of sampling clusters is high and the covariate accounts for substantial variation between clusters. Recent multilevel studies indicate that these conditions are common. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Clinical and economic analysis of allogeneic peripheral blood progenitor cell transplants: a Canadian perspective

Allogeneic peripheral blood progenitor cell (PBPC) transplants are an alternative to BMT, although G-CSF mobilization dose, timing of pheresis and risk of GVHD are not well defined. We compared harvest characteristics, donor and recipient outcomes and costs of two PBPC transplant strategies with historical controls who received BMT. Twenty donors mobilized with four daily s.c. G-CSF doses (5 microg/kg/day) (group 1) and 20 mobilized with 10 microg/kg/day G-CSF (group 2) were compared with 20 BM controls (group 3). G-CSF and phereses were well tolerated. Four of 40 PBPC donors required femoral catheter placement. At least 2.5 x 10(6) CD34+/kg recipient weight were collected with two phereses in 19/20 donors (group 1) and 18/20 donors (group 2). Time to neutrophil (18 vs 20 vs 22 days, P = 0.02) and platelet (21 vs 24 vs 27 days, P = 0.005) engraftment was shorter in the PBPC groups (group 2 vs group 1 vs group 3) but secondary engraftment outcomes were not different. The incidence of grade 2-4 aGVHD was higher in the low-dose G-CSF group (group 1) but there was no difference in cGVHD, 100-day or 1-year survival. The mean PBPC transplant cost (group 1) at first hospital discharge was less than BM (group 3) ($34,643 vs $37,354) but the mean overall cost for both groups was similar at 100 days ($46,334 vs $46,083). Allogeneic PBPC transplant with short course, low-dose G-CSF mobilization is safe, feasible and cost equivalent to allogeneic BMT.