[Skip to Content]
[Skip to Content Landing]
June 19, 1987

The Use of Claims Data to Evaluate Health Care

Author Affiliations

University cf Alberta Edmonton

University cf Alberta Edmonton

JAMA. 1987;257(23):3226. doi:10.1001/jama.1987.03390230062013

To the Editor.—  The article entitled "Use of Claims Data Systems to Evaluate Health Care Outcomes" by Wennberg et al1 illustrated the use of claims data in the evaluation of health care outcomes by determining mortality and reoperation rates following prostatectomy. Although they have shown that these types of data are useful in the evaluation of the "hard" outcomes of mortality and reoperation rate, we urge caution in the use and interpretation of claims data for other "softer" health care outcomes. We believe that claims data obtained from the various Canadian provincial health care plans can contain important biases that effectively limit their usefulness in studying many health care processes and outcomes.Reasons for the unreliability of claims data include physician choice of diagnostic options to yield the highest financial return; the coding of only those diagnoses used for billing and not the complete diagnostic spectrum of the patient