Does a multivalent pneumococcal vaccine given to infants prevent otitis media?

BACKGROUND: An effective pneumococcal vaccine could reduce the incidence of otitis media caused by the most common bacterial pathogen, Streptococcus pneumoniae.

POPULATION STUDIED: Researchers from the Finnish Otitis Media Vaccine Trial recruited infants from prenatal and child health centers in 3 communities in Finland. The 2497 infants enrolled in the study from December 1995 through April 1997 represented 55% of the eligible children in the study communities and were therefore similar to a primary care population. This analysis studied 1662 of the enrolled children; 831 infants were assigned to receive the pneumococcal vaccine and another 831 to receive the hepatitis B vaccine as a control. Totals of 786 (94.6%) in the pneumococcal vaccine group and 794 (95.5%) in the control vaccine group completed the study. The remaining 835 were assigned to receive a different pneumococcal vaccine and were not included in this analysis.

STUDY DESIGN AND VALIDITY: The study was a prospective randomized double-blind cohort study. Allocation concealment was not specified. Infants were randomized to receive either the heptavalent pneumococcal-CRM197 conjugate vaccine or the control hepatitis B vaccine at the age of 2, 4, 6, and 12 months. They were also given a combined diphtheria-tetanus-pertussis and Haemophilus influenzae type b (DTP-Hib) vaccine at 2, 4, and 6 months, inactivated polio vaccine at 7 and 12 months, and measles-mumps-rubella vaccine at 18 months. Subjects were monitored until they were aged 2 years, and their parents were urged to record adverse reactions and bring their children to the study clinics for symptoms suggestive of respiratory infection or otitis media. Acute otitis media (AOM) was diagnosed according to predefined clinical criteria and confirmed by culture of middle ear fluid obtained by myringotomy with identification of specific bacterial serotypes.

OUTCOMES MEASURED: Researchers measured adverse reactions to the vaccines and episodes of AOM in several overlapping categories: all episodes meeting clinical criteria; culture-confirmed pathogen-specific episodes; episodes due to serotypes included in the vaccine, serotypes which cross-react with those serotypes, and other pneumococcal serotypes and groups; episodes due to H influenzae; and episodes due to Moraxella catarrhalis. They followed strict definitions for separate episodes in the same study subject and for recurrent otitis media.

RESULTS: Among the 1662 subjects, there were 2596 episodes of clinical AOM during the follow-up. The vaccine did not reduce the overall number of episodes of AOM. The vaccine reduced the number of all culture-confirmed pneumococcal episodes by 34% (95% confidence interval [CI], 21%-45%; number needed to treat=6). It reduced episodes due to serotypes contained in the vaccine and to serotypes that cross-react with those in the vaccine by 57% and 51%, respectively. Because 95% of enrolled subjects completed the study, intention-to-treat analysis produced similar results.