Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Pneumococcal vaccination update


Bandolier 72 related a systematic review [1] concluding that pneumococcal vaccines were ineffective for most outcomes in most people, especially those more likely to have them in western countries. Other reviews confined to randomised trials [2] have largely confirmed this.

A problem, though, is that some outcomes, like bacteraemia or death from pneumococcal pneumonia, are rare (usually affecting less than 1%), so even though large studies have been done, there are few actual events. Of course, if the events are that rare or the difference between vaccine and placebo that small, the chance of a difference in clinically relevant outcomes will also be small. But the trials may have missed something important, so a revisit to a large epidemiological study [3] makes sense.


This was a retrospective cohort study. The population was a health organisation in Washington State, in people at least 65 years old. They were followed from March 1998 to 2001 (three years). Records were examined for evidence of vaccination with pneumococcal polysaccharide vaccine, and for other information about health status, other diseases, and smoking status.

The primary outcome was hospital admission for community-acquired pneumonia, pneumonia where patients were not admitted (outpatient pneumonia), and pneumococcal bacteraemia. Analysis included assessment of various possible co-variates, like time since vaccination, smoking status, nursing home residence, or other conditions.


There were 47,000 people with 127,000 years of evaluation, with two-thirds of this following pneumococcal vaccination. Hospital admission for confirmed community-acquired pneumonia occurred in 1,428 people, of whom 61 had pneumococcal bacteraemia. There were 3,061 cases of pneumonia not admitted, and 5,690 deaths from all causes.

There was no significant association between vaccination and the risk of hospital admission for community-acquired pneumonia, outpatient pneumonia, or death from any cause when the results were adjusted for age, sex, nursing home residence, influenza vaccination, smoking status and co-morbid conditions (Table 1). Time since vaccination made no difference to the results.

Table 1: Pneumococcal vaccination and outcomes


Hospital admission for pneumonia
Outpatient pneumonia
Pneumococcal bacteraemia
Death from any cause
Unvaccinated rate per 1000 patient years (%)
Vaccinated rate per 1000 patient years (%)
Fully-adjusted risk
1.1 (0.98 to 1.3)
1.0 (0.94 to 1.1)
0.53 (0.31 to 0.93)
0.94 (0.87 to 1.01)

The presence of pneumococcal bacteraemia was lower in vaccinated persons based on adjusted and unadjusted rates (Table 1). The unadjusted difference was 30 persons for every 100,000 vaccinated.

In immunocompromised persons (9,158), defined by presence of cancer, use of immunosuppressive medication, and chronic liver or renal disease, pneumococcal vaccination made no significant difference for any outcome. Neither was there any difference for immunocompetent persons with chronic lung disease.

Also reported was results with influenza vaccination. Influenza vaccination was associated with a significant reduction in the risk of hospital admission with community-acquired pneumonia (0.78, 95% confidence interval 0.65 to 0.95) and risk of death (0.68; 0.62 to 0.76) during the influenza season.


These results are broadly consistent with results of several meta-analyses of randomised trials, none of which showed any effect of pneumococcal vaccination on pneumonia. Meta-analysis did not show any effect on pneumococcal bacteraemia in elderly or at risk subjects [2], but based on many fewer subjects than in this cohort study, and with non-significantly reduced risk. So for this outcome there is again broad agreement.

What would be the impact of vaccinating older people? About 20% of the UK population is aged 65 or older. In a typical population of 100,000, universal vaccination of those aged 65 or older would prevent about six cases of pneumococcal bacteraemia over a period of a few year.

The authors concluded that their study supports the use of pneumococcal vaccine to prevent bacteraemic disease in adults aged 65 or older. Their conclusion was based mainly on a health economic study [4], whose main conclusion was that vaccination saved $8.27 (about £5) and gained 1.2 quality-adjusted days of life per person vaccinated. The most unfavourable assumption was that the vaccination cost per quality-adjusted life-year ranged from $35,822 for ages 65 to 74 years to $598,487 for ages 85 years and older.

What is good is the agreement between a large and detailed cohort study and meta-analyses of randomised trials. Not everyone will be convinced by the argument that vaccination is worthwhile, and it is a topic that needs looking at with a cold and fishy eye, especially when newer, and hopefully better, vaccines come along. In the meantime, only very large cohort studies are likely to provide sufficient events to give better information on the effectiveness of the pneumococcal vaccines we have now.


  1. BG Hutchison et al. Clinical effectiveness of pneumococcal vaccine: meta-analysis. Canadian Family Physician 1999 45: 2392-93.
  2. RA Moore et al. Are the pneumococcal polysaccharide vaccines effective? Meta-analysis of the prospective trials. BMC Family Practice 2000 1:1 (
  3. LA Jackson et al. Effectiveness of pneumococcal polysaccharide vaccine in older adults. New England Journal of Medicine 2003: 348: 1747-1755.
  4. JE Sisk et al. Cost-effectiveness of vaccination against pneumococcal bacteremia among elderly people. JAMA 1997 278: 1333-1339.

previous or next story