|
Abstract:
|
Test-negative case-control studies have become common in the evaluation of rotavirus and influenza vaccine effectiveness (VE). Test-negative designs enroll eligible-consented individuals with a certain clinical diagnosis, e.g., acute diarrhea and then perform laboratory tests to determine if an individual has the disease of interest or not, i.e., a case or control. The case-control design is both relatively easy and feasible to implement and enrolls a cohort of individuals without knowledge of the outcome, here rotavirus diarrhea. Case-control designs are more efficient soon after vaccine introduction and not when vaccination coverage is high. We conducted a test-negative case-control study to evaluate vaccine effectiveness of rotavirus vaccine after routine vaccine introduction in South African children. Using the data collected in this case-control study, we applied propensity score methods to develop a matched subset. Here we compare VE estimates of one and two doses of rotavirus vaccine (vs zero doses) against hospitalization for rotavirus diarrhea obtain using the propensity score matched subset to those obtained using the test-negative case-control approach.
|
