Related Experiment Videos

A Canadian comparison of data sources for coronary artery bypass surgery outcome "report cards"

W A Ghali1, D M Rothwell, H Quan

  • 1Departments of Medicine and Community Health Sciences, University of Calgary, Calgary, Alberta, Canada. wghali@ucalgary.ca

American Heart Journal
|August 31, 2000
PubMed

Insights

Comparing coronary artery bypass graft (CABG) surgery report cards in Ontario, Canada, found general similarities between administrative and clinical data. However, clinical data are essential for public scrutiny of individual hospital outcomes.

Area of Science:

  • Health Services Research
  • Cardiovascular Surgery Outcomes Analysis

Background:

  • Previous studies on administrative vs. clinical data for coronary artery bypass graft (CABG) surgery report cards, primarily from the US, have yielded conflicting validity conclusions.
  • This study addresses this gap by comparing two CABG surgery outcome report cards from Ontario, Canada: one using clinical data (Cardiac Care Network of Ontario) and one using administrative data (Canadian Institute for Health Information).

Purpose of the Study:

  • To compare the validity and consistency of administrative versus clinical data sources for generating coronary artery bypass graft (CABG) surgery outcome report cards in Ontario, Canada.
  • To assess the correlation and agreement between risk-adjusted mortality rates and hospital rankings derived from distinct data sources.

Main Methods:

  • Utilized data from four fiscal years (1992-93 to 1995-96) for 9 hospitals performing CABG surgery in Ontario.
  • Employed logistic regression models to calculate risk-adjusted death rates, incorporating clinical variables (e.g., left ventricular ejection fraction) and administrative data for comorbidities and in-hospital mortality.
  • Linked Cardiac Care Network clinical data with Canadian Institute for Health Information administrative data.

Main Results:

  • Risk-adjusted death rates showed substantial similarity between the two data sources for 7 out of 9 hospitals.
  • Absolute differences in adjusted death rates of 0.58% and 0.64% were observed for two hospitals.
  • A strong positive correlation was found for both risk-adjusted hospital death rates (intraclass correlation coefficient = 0.927) and rankings (Spearman correlation coefficient = 0.828).

Conclusions:

  • Outcome report cards for CABG surgery in Ontario demonstrate general comparability between administrative and clinical data sources.
  • Despite similarities, clinical data remain crucial for detailed public scrutiny of individual hospital performance in outcome reporting.
Abstract

Related Concept Videos