Date of Award


Document Type

Open Access Thesis

Degree Name

Master of Medical Science (MMSc)

First Advisor

Daniel Weinberger, PhD

Second Advisor

Virginia Pitzer, ScD


BACKGROUND - Respiratory syncytial virus (RSV) causes seasonal respiratory infection with potentially serious complications in children, and leads to hospitalization rates as high as 50% in high-risk infants. Safe and effective immunoprophylaxis (palivizumab) is available, but costly. The American Academy of Pediatrics (AAP) recommends palivizumab only for infants at high risk of complications, and only during the RSV season. Although the current AAP guidelines acknowledge the existence of spatial and temporal variation in RSV incidence, they do not recommend spatial or temporal adjustments to immunoprophylaxis regimen outside of Florida. In this study, we investigate the value of using spatial and temporal variation in RSV incidence to adjust the RSV prophylaxis regimen in Connecticut.

METHODS - We describe a generalized additive model of RSV incidence using cubic cyclic penalized splines and apply that model to hospital admissions in Connecticut between July 1995 and June 2013. We use the model to estimate the fraction of all RSV cases in Connecticut occurring while immunoprophylaxis (administered according to the AAP guidelines) offers protection from RSV infection (“preventable fraction”). We also formulate several alternative immunoprophylaxis regimens, with the same net pharmaceutical cost as the AAP-recommended regimen, and similarly estimate their preventable fraction.

RESULTS - Using preventable fraction to assess different immunoprophylaxis regimens, we found that regimens adjusted for county-level variation in timing of RSV seasons are superior to the current AAP-recommended regimen. We also considered the effect of rounding the timing of the first dose of prophylaxis to pragmatic calendrical boundaries (weekly, biweekly, and monthly), and found that benefits of alternative regimens over the AAP-recommended regimen persisted with biweekly rounding, but not with monthly rounding. Our best-performing pragmatic alternative to the AAP guidelines was based on the regional RSV season midpoint with biweekly rounding. In comparison to the AAP recommendation, whose preventable fraction is 94.08% (95% CI: 93.71 – 94.42%), that alternative yielded improvement to preventable fraction of 95.07% (95% CI: 94.74 – 95.36%). We also found that alternative regimens adjusted for annual variation in RSV season are non-superior to spatially adjusted regimens.

CONCLUSION - Overall, we recommend county-level spatial analysis of RSV incidence as the starting point for RSV immunoprophylaxis optimization in Connecticut. However, the potential reduction in RSV hospitalizations should be weighed against the potential increase in implementation cost.

Open Access

This Article is Open Access