Background: Pregnant women have increased rates of hospitalization, morbidity, and mortality from seasonal influenza. Poor outcomes in pregnant women have also been reported from previous influenza pandemics, and in the initial data from the current novel influenza A (H1N1) pandemic. Nevertheless, many barriers exist to vaccine protection of pregnant women. Vaccines may not be available or may offer incomplete protection, and pregnant women may refuse or not have access to immunization. Antiviral medications may provide useful postexposure prophylaxis. Lee and colleagues developed a computer simulation model to predict the potential economic effect of antiviral medications for influenza postexposure prophylaxis in pregnant women.

The Study: The authors used epidemiologic data to construct a decision analysis model representing the decision to administer antiviral medication to a pregnant woman who may have been exposed to influenza through a close contact. The model incorporated the degree of exposure risk, antiviral drug effectiveness and cost, and the probability of potential outcomes such as hospitalization, preterm delivery, and mortality. Costs were estimated for the duration of the pregnancy. The antivirals considered were oseltamivir (Tamiflu), zanamivir (Relenza), amantadine (Symmetrel), and rimantadine (Flumadine). Each simulation was based on 1,000 healthy pregnant women with an average age of 27 years.

Results: For seasonal influenza, the routine use of antiviral medications in pregnant women was not cost-effective. Assuming usual attack and virulence rates, more women had adverse effects than experienced benefit. Conversely, antivirals were cost-effective in a range of H1N1 scenarios. The degree of benefit was strongly influenced by the attack rate, probability of preterm birth in infected women, outcomes for preterm neonates, and probability of influenza-related hospitalization. The outcomes were not influenced substantially by maternal age, but did change notably with gestational age. Before 20 weeks of gestation, the model favored not using prophylaxis, even with high attack rates (up to 30 percent). After 20 weeks of gestation, antiviral use was beneficial in all cases when the attack rate was 20 percent or higher. The model suggested that medications remained cost-effective until viral resistance caused effectiveness to drop to less than 30 percent.

Conclusion: The authors conclude that, among pregnant women, routine postexposure prophylaxis with antivirals is not recommended for seasonal influenza, but is beneficial in pandemic situations.