Accuracy of surgical abortion data capture in Canadian Government Health Administration Databases

Abstract

Objectives: Utilization of single-payer health administrative data to estimate clinical events in a population is limited by the potential for data entry and coding errors. Birth events are accurately captured in health administrative data, but the accuracy of such data on abortion events is unknown. We conducted a randomized controlled clinical trial at the time of an index surgical abortion and used health administrative data to follow subsequent events. If administrative data capture is accurate for surgical abortion, this resource can be used to detect subsequent abortion events among consenting women. For all enrolled participants, we confirmed an index clinical event via clinical records. Our objective was to use the clinical records of these events to assess the accuracy of data capture in the health administrative databases.