Precise answers to the wrong question: prospective clinical trials and the meta-analyses of pneumococcal vaccine in elderly and high-risk adults.

Ten prospective clinical trials conducted in elderly and high-risk adults have failed to show that pneumococcal vaccine prevents pneumococcal bacteraemia and all pneumonia. Several of these trials focused on unrepresentative populations and most had serious methodological problems. Few adequately considered sample size requirements in pre-trial planning. Retrospective sample size calculations based on the findings of the individual trials showed that none was large enough to rule out false negative results. Five published meta-analyses have attempted to determine the efficacy of pneumococcal vaccine by pooling the results of the individual clinical trials. The resulting study populations often were not representative of the populations of elderly and high-risk adults for whom vaccination is recommended. The meta-analysts often omitted clinical trials that should have been evaluated, included other trials that should have been omitted and miscounted the numbers of subjects and outcome events in the individual trials. Retrospective sample size calculations showed that none of the meta-analyses included an adequate number of person years of observation to rule out false negative results. The prospective clinical trials and meta-analyses of pneumococcal vaccine in elderly and high-risk adults have been inconclusive, but they should not be regarded as negative studies. The clinical effectiveness of vaccination in preventing pneumococcal bacteraemia in elderly and high-risk adults has been demonstrated in observational studies, and vaccination is cost-effective. This evidence is sufficient to justify wider use of pneumococcal vaccine.