Nearly 2 million children worldwide who are younger than five years die each year from acute respiratory illness, many related to Streptococcus pneumoniae (pneumococcus). In the United States, high rates of invasive pneumococcal disease occur in White Mountain Apache and Navajo children, who have rates of 1,820 and 537 deaths, respectively, per 100,000 children younger than two years. Because young children respond poorly to T-cell–independent antigens, current pneumococcal polysaccharide immunizations offer little protection in these vulnerable populations. A seven-valent pneumococcal vaccine conjugate to CRM₁₉₇ (PnCRM7) already has been proved effective against invasive pneumococcal disease in children younger than two years in a northern California population and against pneumococcal otitis media in young children in Finland. O’Brien and colleagues tested this vaccine in a randomized, double-blind study of White Mountain Apache and Navajo children.

They recruited 8,292 children between six weeks and 24 months of age living on or close to the reservations. Children with contraindications to the vaccine, hypersensitivity to any vaccine components, and severe medical conditions were excluded. The study randomized communities to intervention or control groups. In intervention communities, every child for whom consent was obtained was given the vaccine. Children in control communities were given a control vaccine for Neisseria meningitidis group C. The randomization was designed to ensure comparable numbers of children in each treatment group and comparability between the groups of communities in important variables such as number of daycare centers, social mixing, and sociodemographic factors. The children received two to four doses of vaccine, depending on their age at entry to the study, at two-month intervals. All children were followed by the Indian Health Service for serious adverse events such as hospital admission, seizure, allergic event, serious illness, and death. Health professionals were instructed to obtain blood cultures for all seriously ill infants to detect invasive pneumococcal illness. The study was designed to continue until 48 cases of invasive pneumococcal illness had been detected, but it was stopped early because of evidence of significant public health benefit in the intervention communities.

The 4,142 children who received PnCRM7 were similar to the 3,903 control children in all significant variables. In three years of observation, eight cases of invasive pneumococcal disease were detected in control children compared with two in children immunized with PnCRM7. The authors calculate the per-protocol primary efficacy of PnCRM7 as 76.8 percent and the intention-to-treat efficacy as 82.6 percent. During the study, 1,867 children (approximately 22 percent in each group) were admitted at least once to the hospital, but only four children had illness judged to be vaccine-related. Twenty-two children died during the study, but none of the deaths was vaccine-related. Adverse events occurred within 72 hours of vaccination in 15 children who did not require hospital admission.

The authors conclude that PnCRM7 has high efficacy against invasive pneumococcal disease in vulnerable populations, and they argue that it should be included in routine childhood immunizations in other countries and in regions of the United States with high rates of invasive pneumococcal disease.