Assessment of the vaccine efficacy in adults

vaccination is currently recommended

Fine and colleagues completed a meta-analysis of all randomized clinical trials of pneumococcal vaccine . A MEDLINE search from 1966 to 1991, manual review of article references, communications with pneumococcal vaccine investigators and vaccine manufacturers identified nine randomized controlled trials.

Studies involving children and those in which the vaccine contained four valences (or serogroups) or less were excluded in this meta-analysis. Summary odds ratios and risk differences were calculated for outcomes, but the latter is because heterogeneity across studies was observed.

Several subgroup analyses were planned, in particular, population risk of pneumococcal vaccine (high vs. low risk). The high risk population includes those aged greater than 55 years of age, those with predefined underlying chronic medical conditions, constituting that group for which vaccination is currently recommended.

Significant risk differences were observed in two pneumococcal infection-related outcomes: definitive pneumococcal pneumonia where the risk difference was 4/1000 and definitive pneumococcal pneumonia with vaccine types only of 8/1000. In other words, 250 subjects need to be vaccinated to prevent one case definitive pneumococcal pneumonia.

The differences in other outcomes including mortality, all cause pneumonia and bronchitis were not significant and not necessarily suggesting efficacy of vaccination. The subgroup analysis for population risk observed the benefit from vaccination to be accrued in low risk, but not high risk populations.

Quality analyses of the studies involving high risk populations had actually observed these to be of higher quality and this lack of benefit cannot be attributed to poor study quality. Sensitivity analyses stressed that exclusion of several studies materially affected risk differences for pneumococcal related study outcomes.

In summary, this excellent meta-analysis suggests efficacy of unconjugated pneumococcal vaccine in preventing definitive pneumococcal pneumonia. The number that need to be vaccinated to accrue such a benefit is arguably quite large. More importantly, the benefit appears to be predominantly in otherwise healthy subjects rather than the high risk adults for whom pneumococcal vaccination is currently recommended.


Review of new literature that appeared since the last update on pneumococcal vaccine does not justify any change in recommendation at this time. The meta-analysis specifically excluded studies other than randomized controlled studies, which were the major evidence in support of pneumococcal vaccination.

These studies were available at the time of the original Task Force recommendations. The lack of efficacy amongst the high risk group may be related to the inability of at least some members of this heterogeneous group to mount an adequate immune response to vaccine.

The efficacy of vaccination should be revisited with availability of results from randomized trials of conjugate pneumococcal vaccine. Such a vaccine may overcome the problems of poor immunogenicity in the subgroup of adults for whom a number of groups recommend vaccination.

Until that time, there appears no urgency in offering pneumococcal vaccination to the population at high risk from pneumococcal disease.