Objective To determine if composite measures based on process indicators are consistent with short-term outcome indicators in surgical colorectal cancer care.

Design Longitudinal analysis of consistency between composite measures based on process indicators and outcome indicators for 85 Dutch hospitals.

Setting The Dutch Surgical Colorectal Audit database, the Netherlands.

Participants 4732 elective patients with colon carcinoma and 2239 with rectum carcinoma treated in 85 hospitals were included in the analyses.

Main outcome measures All available process indicators were aggregated into five different composite measures. The association of the different composite measures with risk-adjusted postoperative mortality and morbidity was analysed at the patient and hospital level.

Results At the patient level, only one of the composite measures was negatively associated with morbidity for rectum carcinoma. At the hospital level, a strong negative association was found between composite measures and hospital mortality and morbidity rates for rectum carcinoma (p<0.05), and hospital morbidity rates for colon carcinoma.

Conclusions For individual patients, a high score on the composite measures based on process indicators is not associated with better short-term outcome. However, at the hospital level, a good score on the composite measures based on process indicators was consistent with more favourable risk-adjusted short-term outcome rates.