Misclassification of hospital volume in surgical volume-outcome analyses of persons aged 65 years and older: potential limitations of medicare data

This study evaluated the agreement between hospital procedure volume calculated using unrestricted data and data restricted to those 65 years of age or older. It also assessed the influence of volume misclassification on volume-outcome relations.

The study used administrative data from Ontario Canada on 3 surgical procedures (excision of colon or rectum for cancer, major lung resection for cancer and repair of unruptured abdominal aortic aneurysm) done between 1994 and 1999.

Hospital procedure volume rankings were determined using data for all patients, and for patients aged 65 years or older. Concordance between unrestricted and age-restricted volume was measured. For those aged 65 or older, the association between procedure volume and outcome (death within 30 days of surgery) was evaluated using both hospital volume categorisation methods.

A high degree of concordance was found between age restricted and unrestricted estimation of hospital procedure volume for the 3 procedures (weighted Kappa, 0.92–0.96). However, hospital volume categorisation based on persons 65 years of age or older resulted in a statistically significant association between procedure volume and outcome for lung resection, whereas hospital volume categorisation based on all persons did not.

Small amounts of hospital volume misclassification may lead to substantial changes in the apparent association between hospital procedure volume and outcome.