Assessing the completeness of infant and childhood immunizations within a provincial registry populated by parental reporting: a study using linked databases in Ontario, Canada
Wilson SE, Wilton AS, Young J, Candido E, Bunko A, Buchan SA, Crowcroft NS, Deeks SL, Guttmann A, Halperin SA, Kwong JC, Wilson K, Tu K. Vaccine. 2020; 38(33):5223-30. Epub 2020 Jun 20. DOI: https://doi.org/10.1016/j.vaccine.2020.06.003
Introduction — In Ontario, Canada, parents have the responsibility to report their child’s routine infant and childhood vaccines to the provincial immunization registry (the Digital Health Immunization Repository; DHIR) without healthcare provider validation. Despite its use in routine immunization coverage monitoring, no study has previously examined the completeness of immunization data within the DHIR.
Methods — We assessed the completeness of DHIR immunizations, as compared to immunizations within the Electronic Medical Records—Primary Care (EMRPC) database, also known as EMRALD, a network of family physician electronic medical records (EMRs). We linked client records from the DHIR and EMRPC to a centralized population file. To create the study cohort, we examined children born during 2005–2008 and further defined the cohort based on those rostered to an EMRPC physician, visit criteria to ensure ongoing care by an EMRPC provider, and school attendance in Ontario at age 7. We calculated up-to-date (UTD) immunization coverage at age 7 for individual vaccines and overall using data from the DHIR and EMRPC separately, and compared the estimates.
Results — The analytic cohort to assess DHIR data completeness included 2,657 children. Overall UTD coverage (all vaccines assessed) was 82.0% in the DHIR and 67.6% in EMRPC. UTD coverage was higher in the DHIR for all vaccines assessed individually, with the exception of meningococcal C conjugate vaccine (difference = 0.3%). After excluding two EMRPC sites with irregularities in immunization data, the difference in overall UTD coverage between systems decreased from 14.4% to 6.6%.
Interpretation — These results validate the use of DHIR for coverage assessment but also suggest that bidirectional exchange of immunization information has the potential to increase immunization data completeness in both systems.
View full text