Boehringer Ingelheim International GmbH.

Background Randomization is the foundation of any clinical trial involving treatment comparison. It helps mitigate selection bias, promotes similarity of treatment groups with respect to important known and unknown confounders, and contributes to the validity of statistical tests. Various restricted randomization procedures with different probabilistic structures and different statistical properties are available. The goal of this paper is to present a systematic roadmap for the choice and application of a restricted randomization procedure in a clinical trial. Methods We survey available restricted randomization procedures for sequential allocation of subjects in a randomized, comparative, parallel group clinical trial with equal (1:1) allocation. We explore statistical properties of these procedures, including balance/randomness tradeoff, type I error rate and power. We perform head-to-head comparisons of different procedures through simulation under various experimental scenarios, including cases when common model assumptions are violated. We also provide some real-life clinical trial examples to illustrate the thinking process for selecting a randomization procedure for implementation in practice. Results Restricted randomization procedures targeting 1:1 allocation vary in the degree of balance/randomness they induce, and more importantly, they vary in terms of validity and efficiency of statistical inference when common model assumptions are violated (e.g. when outcomes are affected by a linear time trend; measurement error distribution is misspecified; or selection bias is introduced in the experiment). Some procedures are more robust than others. Covariate-adjusted analysis may be essential to ensure validity of the results. Special considerations are required when selecting a randomization procedure for a clinical trial with very small sample size. Conclusions The choice of randomization design, data analytic technique (parametric or nonparametric), and analysis strategy (randomization-based or population model-based) are all very important considerations. Randomization-based tests are robust and valid alternatives to likelihood-based tests and should be considered more frequently by clinical investigators.

BackgroundAlthough COPD exacerbations are known to occur more frequently in winter, there is little information on hospitalizations and cause-specific mortality. This study aimed to examine seasonal variations in mortality and exacerbations in patients with COPD during the TIOtropium Safety and Performance In Respimat® (TIOSPIR®) trial.Patients and methodsTIOSPIR was a large-scale, multicenter trial, which assessed the safety and efficacy of tiotropium delivered via HandiHaler® (18 μg once daily) or Respimat® Soft Mist™ (2.5 or 5 μg once daily) inhaler in patients with COPD. Patients were aged ≥40 years, with a smoking history ≥10 pack-years, and post-bronchodilator forced expiratory volume in 1 second ≤70% and forced expiratory volume in 1 second/forced vital capacity ≤0.70. COPD exacerbations and deaths were monitored throughout the trial. The data were pooled to examine seasonal patterns. Southern hemisphere data were shifted by 6 months to align with northern hemisphere seasons.ResultsTIOSPIR was conducted in 43 northern (n=15,968) and 7 southern (n=1,148) hemisphere (n=1,148) countries. The median duration of treatment was 835 days, with a mean follow-up of 2.3 years. Among 19,494 exacerbations, there were clear seasonal differences (winter, 6,646 [34.1%]; spring, 4,515 [23.2%]; summer, 3,198 [16.4%]; autumn, 5,135 [26.3%]). Exacerbations peaked in early winter (December in the northern hemisphere and June in the southern hemisphere), respiratory hospitalizations in midwinter, and respiratory deaths in early spring.ConclusionAlthough winter poses a 2-fold hazard for COPD exacerbations vs summer, respiratory deaths peak in early spring. These data suggest that seasonal intensification of preventive treatments may impact COPD morbidity and mortality.Trial registration numberNCT01126437.

BackgroundThe aim of this study was to evaluate whether Global Initiative for Chronic Obstructive Lung Disease (GOLD) classification could predict mortality risk factors and whether baseline treatment intensity would relate to mortality within each group, using data from TIOSPIR®, the largest randomized clinical trial in COPD performed to date.MethodsA total of 17,135 patients from TIOSPIR® were pooled and grouped by GOLD grading (A–D) according to baseline Medical Research Council breathlessness score, exacerbation history, and spirometry. All-cause mortality and adjudicated cardiovascular (CV) and respiratory mortality were assessed.ResultsOf the 16,326 patients classified, 1,248 died on treatment. Group B patients received proportionally more CV treatment at baseline. CV mortality risk, but not all-cause mortality risk, was significantly higher in Group B than Group C patients (CV mortality – hazard ratio [HR] =1.74, P=0.004; all-cause mortality – HR =1.18, P=0.11). Group D patients had a higher incidence of all-cause mortality than Group B patients (10.9% vs 6.6%). Similar trends were observed regardless of respiratory or CV medication at baseline. In contrast, respiratory deaths increased consistently from Groups A–D (0.3%, 0.8%, 1.6%, and 4.2% of patients, respectively).ConclusionThe data obtained from the TIOSPIR® trial, supporting earlier studies, suggest that proportionally more CV medication and CV deaths occur in GOLD Group B COPD patients, although deaths attributed to respiratory causes are more prevalent in Groups C and D.

Background: Randomization is the foundation of any clinical trial involving treatment comparison. It helps mitigate selection bias, promotes similarity of treatment groups with respect to important known and unknown confounders, and contributes to the validity of statistical tests. Various restricted randomization procedures with different probabilistic structures and different statistical properties are available.Methods: We survey available restricted randomization procedures for sequential allocation of subjects in a randomized, comparative, parallel group clinical trial with equal (1:1) allocation. We explore statistical properties of these procedures, including balance/randomness tradeoff, type I error rate and power. We perform head-to-head comparisons of different procedures through simulation under various experimental scenarios, including cases when common model assumptions are violated. We also provide some real-life clinical trial examples to illustrate the thinking process for selecting a randomization procedure for implementation in practice. Results: Restricted randomization procedures targeting 1:1 allocation vary in the degree of balance/randomness they induce, and more importantly, they vary in terms of validity and efficiency of statistical inference when common model assumptions are violated (e.g. when outcomes are affected by a linear time trend; measurement error distribution is misspecified; or selection bias is introduced in the experiment). Some procedures are more robust than others. Covariate-adjusted analysis may be essential to ensure validity of the results. Special considerations are required when selecting a randomization procedure for a clinical trial with very small sample size.Conclusions: The choice of randomization design, data analytic technique (parametric or nonparametric), and analysis strategy (randomization-based or population model-based) are all very important considerations. Randomization-based tests are robust and valid alternatives to likelihood-based tests and should be considered more frequently by clinical investigators.