A number-needed-to-treat analysis of the use of respiratory syncytial virus immune globulin to prevent hospitalization

Abstract

Objectives: To estimate how many infants in selected high-risk subgroups would require treatment with respiratory syncytial virus immune globulin (RSV-IG) to avoid 1 hospital admission and to determine whether this is economically justified. Design: Cost-benefit analysis. Data from 3 randomized controlled trials of RSV-IG are used to estimate the number needed to treat to prevent 1 hospital admission for respiratory syncytial virus infection. The threshold number needed to treat is computed according to a formula incorporating costs and benefits of RSV-IG prophylaxis. Estimates of the willingness to pay were obtained from a sample of 39 health care providers (35 physicians and 4 nurses). Main Outcome Measures: The number needed to treat to prevent 1 hospital admission for respiratory syncytial virus infection. The threshold number needed to treat that would balance costs with benefits. Results: More than 16 (95% confidence interval, 12.5-23.8) infants would need to be treated with RSV-IG to avoid 1 hospital admission for respiratory syncytial virus infection, ranging from 63 for premature infants without chronic lung disease to 12 (confidence interval, 6.3-100.0) for infants with bronchopulmonary dysplasia. A sensitivity analysis of the costs and values of hospital admission for respiratory syncytial virus infection and RSV-IG treatment resulted in a weak recommendation against the treatment of infants with bronchopulmonary dysplasia and strong recommendations that the costs and risks of RSV-IG treatment outweigh the benefits for the combined sample of infants and premature infants without lung disease. Conclusions: The number-needed-to-treat procedures offer a method to assess evidence of treatment effects and decision rules for whether to accept treatment recommendations. Under plausible assumptions, treatment with RSV- IG is not recommended for infants without lung disease. Institutions can examine cost and benefit assumptions that best fit their own practice setting.