Statistical Analysis
Sample-size and power considerations were based on the annual rate of exacerbations as the primary outcome. We estimated that the placebo group would have an average of 1.54 exacerbations per person-year and that simvastatin would reduce this rate by 15%. The estimated rate for the placebo group was based on the observed exacerbation rate in the placebo group of a previous trial of azithromycin conducted by the COPD Clinical Research Network.8
We based our sample-size calculations on a two-sided alpha level of 0.05, a study power of 90%, the expected number of exacerbations per person-year in the placebo group (1.54), the expected number of exacerbations per person-year in the simvastatin group (1.54×0.85=1.31), a mean duration of follow-up of 2 years, and an equal probability of assignment to the placebo or simvastatin group. We calculated that we would need to enroll 911 patients; assuming a 10% loss to follow-up, the total sample-size estimate was 1002 patients.
An important secondary outcome was the time to the first exacerbation. Assuming that simvastatin treatment would reduce the exacerbation rate by 15% and adjusting for the 10% loss to follow-up, we estimated that 1126 patients would be required in order to provide the study with 90% power for this secondary outcome. Following FDA recommendations and guidance from the data and safety monitoring board for the exclusion of patients who were receiving certain drugs and the exclusion of patients with diabetes, respectively, we projected that 74 additional participants would be needed, for a final study population of 1200 patients.
All the patients who underwent randomization were followed until the end of the study. The final analysis was performed on an intention-to-treat basis. COPD exacerbation rates in the two study groups were compared with the use of a rate ratio. The independence of individual exacerbations was ensured by considering participants to have had two separate exacerbations if the onset dates were at least 14 days apart. Exacerbation rates in each group and the between-group differences were analyzed with the use of negative binomial regression modeling and time-weighted intention-to-treat analyses with adjustments of confidence intervals for between-participant variation (overdispersion).
Secondary outcome measures were assessed at baseline and at 12, 24, and 36 months (and at the final visit if it did not coincide with an annual visit). Analyses of the time to the first exacerbation and the rate of death from any cause were performed with the use of Kaplan–Meier methods, with survival curves for the probability of remaining outcome-free in the two groups as a function of time from randomization. Curves were compared with the use of the log-rank test. Continuous outcome measures, including absolute and relative changes in FEV1 and FVC, dyspnea assessment, and SGRQ scores were computed as rates of change over time, from baseline to the date of the last observation.