Controlled clinical trials最新文献

Background

The interpretation of clinical trials and efforts directed at reducing the worldwide burden of coronary disease must take regional differences into account. This study examined the regional differences in baseline characteristics, treatment, and outcome in patients presenting with ST elevation myocardial infarction (STEMI) who were entered into the Magnesium in Coronaries (MAGIC) trial.

Methods and results

MAGIC randomized 6213 patients to standard care with either placebo infusion or infusion of intravenous magnesium sulphate. There was no difference in mortality between these groups. For this analysis, three geographic regions were identified (Region 1=United States and Canada; Region 2=Bulgaria, Georgia, and Russia; Region 3=Austria, Belgium, Chile, Hungary, Israel, the Netherlands, New Zealand, and Venezuela) and compared with respect to baseline characteristics, treatment, and 30-day mortality.

Patients in Region 2 had the highest prevalence of adverse risk factors at entry, including history of prior myocardial infarction, heart failure, stroke, and hypertension; anterior location of index acute myocardial infarction; and presence of pulmonary congestion at presentation. Furthermore, Region 2 patients infrequently received reperfusion therapy compared with those in Region 1. Region 3 was intermediate in this regard. Mortality was highest in Region 2, least in Region 1, and intermediate in Region 3.

Conclusion

Geographic location, particularly, parts of Eastern Europe, is strongly and independently associated with mortality following STEMI. This geographic variation in mortality confirms prior reports, although adequate explanations continue to be elusive and are beyond the scope of this large simple trial. Future international trials must recognize this variation in design, analysis, and interpretation.

The selection of a trial design is an important issue in the planning of clinical trials. One of the most important considerations in trial design is the method of treatment allocation and appropriate analysis plan corresponding to the design.

In this article, we conducted computer simulations using the actual data from 2158 rectal cancer patients enrolled in the surgery-alone group from seven randomized controlled trials in Japan to compare the performance of allocation methods, simple randomization, stratified randomization and minimization in relatively small-scale trials (total number of two groups are 50, 100, 150 or 200 patients). The degree of imbalance in prognostic factors between groups was evaluated by changing the allocation probability of minimization from 1.00 to 0.70 by 0.05.

The simulation demonstrated that minimization provides the best performance to ensure balance in the number of patients between groups and prognostic factors. Moreover, to achieve the 1 percentile for the p-value of chi-square test around 0.50 with respect to balance in prognostic factors, the allocation probability of minimization was required to be set to 0.95 for 50, 0.80 for 100, 0.75 for 150 and 0.70 for 200 patients. When the sample size was larger, sufficient balance could be achieved even if reducing allocation probability. The simulation using actual data demonstrated that unadjusted tests for the allocation factors resulted in conservative type I errors when dynamic allocation, such as minimization, was used. In contrast, adjusted tests for allocation factors as covariates improved type I errors closer to the nominal significance level and they provided slightly higher power. In conclusion, both the statistical and clinical validity of minimization was demonstrated in our study.

Context

The term “patient-reported outcomes” (PROs) has evolved to include any endpoint derived from patient reports, whether collected in the clinic, in a diary, or by other means, including single-item outcome measures, event logs, symptom reports, formal instruments to measure health-related quality of life (HRQL), health status, adherence, and satisfaction with treatment. This term coincides with the explicit interest from drug development researchers and regulatory authorities in the appropriate utilization and reporting of treatment impact measures.

Objective

To determine the level and nature of use of PROs compared to other types of effectiveness endpoints in approved product labeling for new drugs recently approved in the United States.

Design and sources

Review and analysis of effectiveness endpoints as reported in clinical study descriptions in approved product labeling of new molecular entities (NMEs) approved in the United States from 1997 through 2002.

Main outcome measures

Effectiveness study endpoints reported in approved product labeling that fall into the following categories of measurement: PROs, clinician-reported outcomes (CROs), and laboratory test/device measurement endpoints.

Results

PROs were reported in 64 (30%) of the 215 product labels reviewed. Clinician-reported outcomes were reported most frequently (62%) followed by laboratory/device endpoints (50%). PROs were the only type of endpoint used in the FDA-approved label for 23 products. Formal multiitem PRO scales were cited 22 times. Use of PROs is most common in antiinflammatory, CNS, gastrointestinal, respiratory, allergic conjunctivitis, and urologic therapy areas. The frequency of reported PRO use over this period did not change.

Conclusion

PROs, although quite variable as a class of study endpoints, were found to have a significant role in the development and evaluation of new medicines. More formal guidance from the FDA about use of such measures along with continued collaboration by PRO researchers to develop and disseminate standards will enhance the appropriate use of PROs in future drug development and labeling.

Background

Patients may be asked to register a symptom daily in clinical trials. A problem associated with this kind of registration is that patients do not always fill in the diary at the appropriate time. As there is evidence showing that memory is unreliable, this undermines the entire purpose of collecting daily data on paper diaries. We aimed to compare accuracy, autocorrelations of consecutive entries, and responsiveness in paper diaries (P-Diaries) with electronic diaries (E-Diaries) and telephone diaries (T-Diaries).

Methods

In a multi-centre, open, and parallel trial, 177 patients were allocated at random to P-Diaries, E-Diaries, or T-Diaries for the registration of symptoms through 4 weeks of treatment of gastro-oesophageal reflux disease (GORD). The primary outcome measure was the diaries.

Findings

The proportion of patients completing all morning and all evening entries on time was low for both E-Diary and T-Diary groups. By accepting entries that were done half a day late, the proportion was increased to 19/57 (33%) for the E-Diary group, and to 9/61 (15%) for the T-Diary group. For P-Diary, where no control for time registration and entries was adapted, 37/59 (63%) of the patients completed all morning and all evening entries. A significant higher autocorrelation in P-Diaries was also found. Responsiveness was similar regardless of method of data capture.

Interpretation

The results are consistent with the suggestion that data in the P-Diaries are not filled in at the appropriate time. Use of E-Diaries or T-Diaries improves quality and is recommended in future clinical trials.

In the literature, different statistical methods to evaluate bronchodilator studies are used. These approaches are all based on the absence of residual heterogeneity and on baseline independency of the parameter under analysis. A database containing the lung function values of newly referred patients was used to assess these assumptions as function of the underlying diagnosis (asthma, bronchitis and emphysema) and to chart the characteristics of analysis of covariance, which (partly) deals with these drawbacks. Bronchodilator data of 709 asthmatics, 522 bronchitic and 126 emphysema patients were used. It was shown that, in asthma, for almost all lung function parameters, bronchodilation was indeed dependent on baseline values, which was less strong in bronchitis and even weaker in emphysema. A negative effect of age on bronchodilation was found, which is strong in asthma and almost absent in emphysema, rendering the use of bronchodilation as a diagnostic tool less useful. The conclusion is that analysis of covariance is a good way to evaluate bronchodilation studies in obstructive lung disease, particularly in asthma. For bronchitic or emphysema patients, difference-based approaches may suffice. The assumptions underlying the other methods were not met.

Objective

The effect of funding source on the outcome of randomized controlled trials has been investigated in several medical disciplines; however, psychiatry has been largely excluded from such analyses. In this article, randomized controlled trials of second generation antipsychotics in schizophrenia are reviewed and analyzed with respect to funding source (industry vs. non-industry funding).

Method

A literature search was conducted for randomized, double-blind trials in which at least one of the tested treatments was a second generation antipsychotic. In each study, design quality and study outcome were assessed quantitatively according to rating scales. Mean quality and outcome scores were compared in the industry-funded studies and non-industry-funded studies. An analysis of the primary author's affiliation with industry was similarly performed.

Results

Results of industry-funded studies significantly favored second generation over first generation antipsychotics when compared to non-industry-funded studies. Non-industry-funded studies showed a trend toward higher quality than industry-funded studies; however, the difference between the two was not significant. Also, within the industry-funded studies, outcomes of trials involving first authors employed by industry sponsors demonstrated a trend toward second generation over first generation antipsychotics to a greater degree than did trials involving first authors employed outside the industry (p=0.05).

Conclusions

While the retrospective design of the study limits the strength of the findings, the data suggest that industry bias may occur in randomized controlled trials in schizophrenia. There appears to be several sources by which bias may enter clinical research, including trial design, control of data analysis and multiplicity/redundancy of trials.

Much is known about how to split alpha between or among several comparisons, or how to preserve the nominal alpha level with an exact analysis, but the issue of how alpha is generated, or where it comes from, has not received a commensurate degree of attention. It would seem that there is little point in working out methods to allocate or conserve alpha if it is unlimited in supply. Moreover, there seems to be a logical inconsistency in requiring that a given amount of alpha, generally 0.05, be split among the primary comparisons performed by a given set of researchers, yet allowing other researchers to analyze the same data with a new 0.05 to work with. We will address these inconsistencies, and ask more generally where alpha comes from, how it can be generated, and under what conditions it should be one-tailed or two-tailed.