LAUSR

INTRODUCTION Home-based records (HBRs) (also known as vaccination cards) and caregivers' recall are the main means to ascertain vaccination status; however, data on the quality of HBRs and the validity of recall vaccination data compared to HBRs is scarce. This manuscript presents results from two analyses related to HBRs, one on HBR pictures taken during a vaccination coverage survey, including an assessment of the HBR quality and legibility, and an evaluation of the agreement between caregivers' recall and the vaccination information in the HBRs. METHODS Using pictures from 500 randomly selected HBRs collected during the 2016 district-based immunization coverage evaluation survey in Lebanon, two independent researchers assessed the quality of the picture and then of the HBR itself against a pre-defined set of criteria. HBRs were classified into three types: private, public and all others. In addition, caregivers' recall was compared to data found in vaccination HBRs to assess measures of vaccination status agreement for 5713 children for whom both sources of data were available. RESULTS Over 90% of the 500 HBR pictures reviewed were considered adequate to assess the HBR quality. In the sample, most cards were type 1 (41%), followed by type 2 (34%). Most HBRs met the set criteria for quality in terms of physical condition and legibility, while, among the 28 different types of cards, vaccination cards' content and design met a moderate level of quality. Concordance, sensitivity, specificity, positive and negative predictive values, and the Kappa statistic showed diverse levels of agreement for vaccination status per vaccine dose between caregivers' recall and vaccination HBRs. CONCLUSION This study illustrates that taking pictures of HBRs in a coverage survey is feasible and useful to conduct secondary analyses related to HBRs, such as assessing their quality and comparing recall with HBRs when both sources of data are available.