Level 2: Mid-level evidence
A randomized trial in Bangladesh evaluated the efficacy of influenza vaccine in pregnant women (N Engl J Med. 2008;359:1555-1564). Women (N=340) in their third trimester of pregnancy were randomized to receive inactivated influenza vaccine or 23-valent pneumococcal vaccine. Primary outcome in infants was the first episode of laboratory-confirmed influenza before 24 weeks of age; the trial lacked intention-to-treat analysis and included 316 mother-infant pairs. Laboratory-confirmed influenza occurred in six of 159 (3.8%) infants in the influenza-vaccinated group vs. 16 of 157 (10.2%) infants in the pneumococcal vaccine group (NNT 16, 95% confidence interval 8-118). In addition, there were statistically significant reductions in both infant and maternal rates of respiratory illness with fever in the influenza-vaccinated group.