The following is a summary of “Investigating sources of non-response bias in a population-based seroprevalence study of vaccine-preventable diseases in the Netherlands,” published in the February 2024 issue of Infectious Diseases by Postema et al.
Researchers conducted a retrospective study to evaluate the representativeness of PIENTER 3 (P3) and identify non-response bias sources and trends in declining participation rates across all PIENTER studies.
They categorized P3 invitees into four survey response types, Full Participants (FP), Questionnaire Only (QO), Non-Response Questionnaire (NRQ), and Absolute Non-Responders (ANR). FP’s demographic and health indicator data were discussed with Dutch national statistics and intra-group comparisons among response types. Response type prediction was performed using random forest algorithms. Subsequently, they conducted comparative analyses across FPs from all three PIENTER surveys to elucidate the evolving profile of survey participants over time.
The results showed that P3 FPs tended to exhibit better health, younger age, and higher educational attainment than the general Dutch population. The random forest model struggled to distinguish between FPs and ANRs, although it revealed indications of a healthy-responder bias when differentiating FPs from NRQs. Analysis across the three PIENTER surveys indicated participant similarity over time, facilitating comparability. In alignment with national trends, P3 participants displayed a reduced inclination towards vaccination compared to previous cohorts.
Investigators concluded that the PIENTER biobank effectively tracks population immunity to vaccine-preventable diseases (VPDs) in the Netherlands, but future studies should improve recruitment from underrepresented groups and consider mixed survey modes to minimize bias, especially in studies with diverse vaccination/disease exposures.
Source: bmcinfectdis.biomedcentral.com/articles/10.1186/s12879-024-09095-5
