Background: Prior to the development of the Canadian Institute for Health Information’s (CIHI) stroke special project 340, Ontario reported acute stroke care data on a biennial basis through the Ontario Stroke Registry’s (OSR) Ontario Stroke Audit (OSA). Funding for the OSR data collection ended as of fiscal year 2012. Beginning April 1, 2012, CIHI special project 340 became mandatory and Ontario hospitals were required to collect relevant acute stroke and transient ischemic attack (TIA) care data in the National Ambulatory Care Reporting System (NACRS) and the Discharge Abstract Database (DAD). CIHI 340 data are currently used for the awarding of stroke distinction accreditation and will be used to calculate several of the stroke Quality Based Procedure (QBP) performance indicators.

Objective: To evaluate the accuracy and completeness of the reporting of the seven new CIHI 340 data elements in hospital administrative data by comparison with the 2012/2013 OSROSA.

Methods: For the purposes of this study, the OSA is considered the gold standard. The OSA’s acute stroke and TIA cases for the period April 1, 2012 to March 31, 2013 were linked to the NACRS and DAD administrative data. Agreement between the administrative data and the OSA data was calculated for each of the new data elements by individual hospital and for various hospital aggregations including Regional Stroke Centres, District Stroke Centres, non-designated hospitals, telestroke hospitals and hospitals with Stroke Distinction accreditation. A value of at least 85% was considered good agreement.

Results: Good agreement ( ≥85%) between the administrative data and the gold standard was found for two of the data elements: patient received a CT or MRI scan within 24 hours of arrival at hospital and patient received thrombolytic (tPA) therapy. The date and time of stroke or TIA symptom onset reported in the administrative data had poor agreement (< 34%) while agreement on the remaining new data elements fell somewhere in between these high and low points.

Conclusion: The accuracy and completeness of the CIHI 340 information are less than optimal for those data elements that require date and time values. Going forward, CIHI 340 data will be used in the calculation of four performance indicators for the stroke QBP initiative. A plan for improving and auditing data quality is essential if hospital funding is to be linked to these QBP indicators.

Download the report here.