Please note: In 2003, the CTF updated its Grades of Recommendations to include an "I Recommendation" for situations where insufficient evidence exists to allow a recommendation to be made.  (Formerly, these situations were captured under a "C Recommendation".)  This change is not retroactive, and all "C Recommendations" made prior to 2003 have not been reevaluated in light of the new "I" recommendation grade.  For a discussion of these recommendation grades, please link to the 2003 article in the Canadian Medical Association Journal here.

Overview
In its 1979 report the Canadian Task Force on the Periodic Health Examination recommended antipneumococcal vaccination for those with chronic debilitating illness, sickle cell anemia and anatomic or functional asplenia but not for the general population. In 199 1 more detailed recommendations were made, based on the evidence for specific risk groups using a stringent measure of protective efficacy – the development of illness after exposure to the organism.

Burden of Suffering
The annual incidence of pneumococcal pneumonia in the United States has been estimated to be between one and five episodes per 1,000 subjects. One population-based study showed that pneumococcal bacteria were found in 18.7 subjects per 100,000 population, with an incidence of 53 per 100,000 among those 65 years or more. The incidence of pneumococcal pneumonia has been estimated to be three to five times this value. In Canada 7,069 people were admitted to hospital because of pneumococcal pneumonia during 1983-84; this represented 86,323 hospital days. The rate of death among such patients is approximately 15% to 20% despite appropriate antibiotic therapy.

S. pneumoniae is a major cause of otitis media, which has been estimated to occur in two-thirds of the general population by 3 years of age; about half will have had three or more episodes. It was responsible for 73 (37%) of the reported cases of bacterial meningitis in Canada in 1986.

Those at high risk for pneumococcal infections include: patients with anatomic or functional asplenia, those with Hodgkin’s disease or nephrotic syndrome, and those receiving immunosuppressive therapy. Other groups in which such infections have been associated with increased rates of illness and death include those with chronic lung or cardiac disease or diabetes mellitus, those in institutions and over age 55. However, some authors conclude that there is no evidence to support an unusual frequency of pneumococcal infections in people with chronic cardiac, renal or hepatic disease or diabetes mellitus.

Maneuver
Pneumococcal vaccine, originally an octavalent vaccine, is now available as a 23-valent vaccine containing the strains of S. pneumoniae that have been isolated most often from patients with invasive disease. There have been no comparative studies of the old and new vaccines.
Effectiveness of Prevention and Treatment

Specific Patient Populations
The efficacy of pneumococcal vaccine was initially demonstrated in randomized controlled trials involving people working or living in close proximity to each other. Among South African miners<1,2> pneumococcal pneumonia or bacteremia developed in 17 of 1,493 vaccine recipients, as compared with 83 of 1,480 placebo recipients, for a vaccine efficacy rate of 78.5%.<1> Smit and collaborators<2> found an identical protective efficacy in a trial involving 983 recipients and 2,036 control subjects. Among 8,586 vaccine recipients in a study of U.S. military recruits<3> 4 had the pneumonia against which they were immunized, and 56 had other types of pneumonia; the corresponding numbers among the 8,449 saline recipients were 26 and 59.

Vaccine efficacy has also been confirmed in a cohort study involving patients with sickle cell disease and those who had undergone splenectomy.<4> Patients with Hodgkin’s disease, on the other hand, have demonstrated poor antibody responses to the vaccine.<5>

General Pediatric Populations
In a randomized controlled trial in Australia involving 1,273 healthy children aged 6 to 54 months, there was no reduction in the number of respiratory and otic complaints among the vaccine recipients.<6> In addition, there were no differences in the rates of general hospital admission or admission because of respiratory tract infection. In a subgroup of patients, no difference was found between the recipients and the control subjects in nasal carriage rate of the S. pneumoniae serotypes included and not included in the vaccine.<7> In a multicentre study in Massachussetts 124 children aged 5 to 21 months were given the same vaccines.<8> The group that received the vaccine with the "target" serotypes had fewer infections than the other group but an equal number of ear infections.

For 179 infants in Alabama aged 6 to 21 months who had had at least one episode of otitis media before 10 months of age, no difference was observed in the overall infection rate between two groups randomly assigned to receive vaccines containing "target" serotypes (those accounting for 75% of cases of otitis media) and "nontarget" serotypes.<9> The incidence of subsequent infections in black children 6 to 11 months of age was found to be decreased by half<10> but this may have been a chance occurrence resulting from multiple subgroup analyses. Makela and colleagues<11> examined the efficacy of vaccine in 827 Finnish children aged 3 months to 6 years. The subjects received either a 14-valent pneumococcal vaccine or Hemophilus influenzae type b vaccine in a double-blind manner. Although there was a significant reduction in the number of ear infections within the first 6 months after vaccination, the difference was lost with longer follow-up.

General Adult Populations
Shapiro and Clemens<12> conducted a case-control study of the efficacy of vaccine in patients at high risk for serious pneumococcal infections. In the 90 case-control pairs, 7% of the case subjects had received vaccine, as compared with 18% of the control subjects. The protective efficacy of the vaccine was estimated to be 67% (95% confidence interval (CI): 13% to 87%). This was mainly because of a reduction in the number of infections in the group at moderate risk; minimal efficacy was observed in the immunocompromised group. In an extension of this study, with 1,054 case-control pairs,<13> 13% of the case subjects and 20% of the control subjects had previously received vaccine, for a protective efficacy rate against vaccine serotypes of 56% (95% CI: 42% to 67%) for immunocompetent patients. The vaccine was not shown to be efficacious against infections caused by serotypes not represented in the vaccine (protective efficacy –73%; p=0.15) or for immunocompromised patients (protective efficacy 21%; p=0.48).

Using the same study design but excluding immunocompromised patients, Sims and co-workers<14> found that 8% of 122 infected patients had been vaccinated, as compared with 21% of 244 control subjects, for a vaccine efficacy rate of 70% (95% CI: 36% to 86%).

A case-control study involving 89 patients in hospital because of pneumococcal bacteremia showed no difference between the two groups in the vaccination rate or the distribution of responsible serotypes.<15> However, 30% of the subjects had hospital-acquired rather than community-acquired infection. The inclusion of people at high-risk who have previously been shown to respond poorly to vaccination and the small sample limit the significance of the findings.

In a two-year cohort study<16> involving elderly patients at two hospitals with long-term care facilities in Rochester, NY, the vaccine coverage was initially 84% (it was reduced to 65% by the end of the study). A small group of patients who refused to be vaccinated and a larger group of patients who were inadvertently not offered the vaccine acted as control subjects. The results of the first year revealed no significant difference between the two groups in the number of cases of pneumonia or deaths per 1,000 patient-years. This study is susceptible to volunteer bias and only a two-fold or greater difference in incidence would have been detected given the sample size. It has also been suggested that the etiologic serotypes in the elderly patients differed from the vaccine serotypes, which had been selected because they most frequently produced bacteremic disease in patients in general hospitals. However, there was no difference between the two groups when the vaccine serotypes were differentiated from nonvaccine serotypes.

In a cross-sectional analytic study<17> and among patients over 10 years of age, the proportions of isolates of vaccine serotypes (from blood and CSF samples) in vaccinated and nonvaccinated groups were 50% and 66% respectively, for an efficacy rate of 60%. The vaccine was not efficacious in a high-risk group with various immunologic or splenic disorders. In a subsequent study, involving more patients, efficacy was estimated to be 47%.<18> In addition, Bolan and associates<19> observed that the incidence of infection with pneumococcal serotypes found in the vaccine was higher among the nonvaccinated subjects (67%) than among the vaccinated subjects (49%). The efficacy rate was estimated to be 55% (95% CI: 2% to 82%) among patients with chronic cardiovascular disease, pulmonary disease or diabetes mellitus and 61% (95% CI: 3% to 99%) among those aged 61 years or more.

In a small randomized controlled trial of 1,300 mentally ill elderly patients randomly assigned to receive two hexavalent pneumococcal vaccines or saline placebo, there was no statistical difference in death rate or recovery rate of pneumococcal isolates (vaccine strains or all strains).<20>

A study involving 13,600 subjects aged 45 years or more who were not immunocompromised also demonstrated no difference in the incidence of radiologically diagnosed pneumonia or isolation of pneumococci from sputum samples.<20> This study had low statistical power because the incidence of pneumococcal pneumonia was low.

A trial involving 189 patients with severe chronic obstructive pulmonary disease<21> and the Veterans Administration Cooperative Study, involving 2,295 patients at high-risk,<22> also failed to detect differences in the rates of illness and death among vaccine recipients and control subjects. In the latter study the rate of death from lower respiratory tract infection was actually 15% greater among the recipients. The maximum achievable difference between the two groups would have been only 10%. The methodology of this study has been criticized regarding the lack of separation of colonization from infection, choice of endpoint, the possible failure of randomization and the high risk of ß error.

Kaufman and collaborators<23> conducted a 6-year study involving almost 11,000 inhospital patients over 40 years of age, about half of whom had received a divalent or trivalent vaccine. There were 99 cases of pneumonia in 5,750 vaccinated people, as compared with 227 cases in 5,153 control subjects (p<0.001). There were fewer deaths in the vaccinated group than in the control group (40 vs. 98, p<0.001).

In a randomized controlled trial, the vaccine was found to be efficacious in patients in geriatric hospitals and homes for the aged in France.<24> Patients with multisystem disease, and those who were immunocompromised were excluded. There were 1,234 patients aged 55 to 85 years without special risk factors, and 452 had diabetes or chronic cardiac, renal or pulmonary disease. Pneumonia occurred in 31 control subjects and 9 vaccine recipients, for a vaccine efficacy rate of 77% (95% CI: 51% to 89%). This study may be criticized because the cause of pneumonia was not confirmed and the pneumonia was not confirmed radiologically. It also differs from the U.S. studies because a higher frequency of pneumonia was observed in the control group.

Because of an absence of clinical studies controversy exists about recommendations for widespread vaccination in the general population. The efficacy of vaccine among people with chronic lung disease remains to be proven. Recent studies have shown that the proportion of people with serologic responses to the vaccine and the upper limit of the antibody concentration achieved were lower among elderly subjects with chronic lung disease than among young, healthy control subjects. These observations differ from those in previous studies in which some of the measured antibody was reactive to substances other than capsular polysaccharide.

Recommendations of Others
In 1989, the National Advisory Committee on Immunization recommended that a single dose of pneumococcal vaccine be given to all persons 65 or more years of age and to patients with chronic cardiorespiratory disease, cirrhosis, alcoholism, chronic renal disease, diabetes mellitus (adults), HIV infection, conditions associated with immunosuppression, asplenia, splenic disorders or sickle cell disease.

Conclusions and Recommendations
There is as yet no role for universal immunization with pneumococcal vaccine. Results of randomized trials have supported its use in studies of military recruits and South African miners under epidemic conditions but did not confirm a benefit among infants and young children. Therefore, there is good evidence to exclude pneumococcal vaccination from the periodic health examination of the general population and of infants and young children (E Recommendation).

The results of studies involving adults have been contradictory because of differences in study design, populations and end-points. In a cohort study the vaccine was effective in people with sickle cell disease and those who had undergone splenectomy. The efficacy rate of the vaccine in case-control studies has been from 60% to 70% among patients over 55 years of age, particularly if they have chronic pulmonary, cardiac or renal disease or are alcohol dependent. In cohort studies the rate has been 50% to 60%. Two randomized trials involving people in institutions have shown the vaccine to be efficacious. The data fit the model of increased exposure and incidence of infection observed among people living under crowded situations. Thus, there is good evidence to include vaccination in the periodic health examination of patients who have undergone splenectomy, those with sickle cell disease, immunocompetent elderly people living in institutions and adults living or working in crowded conditions (A Recommendation). Before the final decision to vaccinate is made, however, the physician must consider the overall condition of the person.

Immunocompromised patients do not seem to respond reliably to the vaccine, and the benefit in this group remains to be proven. There is fair evidence to exclude vaccination from the periodic health assessment of immunocompromised patients (D Recommendation).

Randomized trials in non-institutional settings have shown negative results, but all had limited power because of the low frequency of pneumonia as an outcome. If one assumes that 2.5 cases of pneumococcal pneumonia occur per 1,000 nonvaccinated people at high risk and a vaccine efficacy rate of 70% (the highest reported), then 571 subjects need to be vaccinated to prevent one case of pneumonia. The trend toward increased rates of death from lower respiratory tract infections among vaccine recipients in the Veterans Administration Cooperative Study prevents a stronger recommendation for vaccination considering its efficacy in the institutional setting. There is insufficient evidence to include vaccination in or exclude it from the periodic health examination of people over age 55 living independently (C Recommendation).

Unanswered Questions (Research Agenda)

1. Should pneumococcal vaccine be used in the ambulatory elderly population? We suggest conducting further trials, perhaps in countries where universal vaccination of those over 55 years of age has not been accepted, to replicate the Veterans Administration Cooperative Study.
2. Conjugate or other pneumococcal vaccines need to be developed that are more immunogenic in order to improve the efficacy rate among immunocompromised people.

This review was initiated in March 1988 and the recommen-dations were finalized by the Task Force in September 1990. A report with a full reference list was published in March 1991.<25>

Full Citation
Wang, E.E.L. Administration of pneumococcal vaccine. In: Canadian Task Force on the Periodic Health Examination. Canadian Guide to Clinical Preventive Health Care. Ottawa: Health Canada, 1994; 386-395.