Concordance of patient-reported and routinely collected perinatal outcome data: A systematic review

Background: Patient-reported perinatal outcome data offer some distinct advantages over routinely collected data yet concerns about data quality and validity persist. Objectives: This review sought to answer two questions: how concordant are patient-reported and routinely collected data on perinatal outcomes, and under what conditions is concordance higher or lower? Data sources: PubMed, CINAHL, and PsycINFO databases. Study selection and data extraction: We included original research that reported on the concordance of patient-reported and routinely collected data for 11 International Consortium of Health Outcomes Measurement (ICHOM) Pregnancy and Childbirth Standard Set outcomes and met other criteria. We used double independent title and abstract screening, full text review, and data extraction. Synthesis: We conducted a narrative synthesis, summarising study results both overall and by perinatal outcome, recall period for patient-reported data collection, data collection mode, and study timeframe. Results: Fourteen studies reported on 78 analyses of the concordance of patient-reported and routinely collected data across four outcomes: stillbirth, maternal length of stay, preterm birth, and oxygen dependence. They used a range of different metrics of concordance (e.g., percentage agreement, sensitivity, specificity) and reported perfect concordance of patient-reported and routinely collected data in some contexts but poor concordance in others. Conclusions: In some contexts, there may be greater methodological equivalence of patient-reported and routinely collected perinatal outcome data than is typically acknowledged. Further research is needed both on outcomes not yet studied and to systematically examine the methods, processes, and conditions that optimise accuracy and concordance of patient-reported and routinely collected perinatal outcome data.