Cancer care quality measures: diagnosis and treatment of colorectal cancer.

OBJECTIVES To identify measures that are currently available to assess the quality of care provided to patients with colorectal cancer (CRC), and to assess the extent to which these measures have been developed and tested. DATA SOURCES Published and unpublished measures identified through a computerized search of English-language citations in MEDLINE (1966-January 2005), the Cochrane Database of Systematic Reviews, and the National Guideline Clearinghouse; through review of reference lists contained in seed articles, all included articles, and relevant review articles; and through searches of the grey literature (institutional or government reports, professional society documents, research papers, and other literature, in print or electronic format, not controlled by commercial publishing interests). Sources for grey literature included professional organization websites and the Internet. REVIEW METHODS Measures were selected by reviewers according to standardized criteria relating to each question, and were then rated according to their importance and usability, scientific acceptability, and extent of testing; each domain was rated from 1 (poor) to 5 (ideal). RESULTS We identified a number of well-developed and well-tested CRC-related quality-of-care measures, both general process-of-care measures (on a broader scale) and technical measures (pertaining to specific details of a procedure). At least some process measures are available for diagnostic imaging, staging, surgical therapy, adjuvant chemotherapy, adjuvant radiation therapy, and colonoscopic surveillance. Various technical measures were identified for quality of colonoscopy (e.g., cecal intubation rate, complications) and staging (adequate lymph node retrieval and evaluation). These technical measures were guideline-based and well developed, but less well tested, and the linkage between them and patient outcomes, although intuitive, was not always explicitly provided. For some elements of the care pathway, such as operative reports and chemotherapy reports, no technical measures were found. CONCLUSIONS Some general process measures have a stronger evidence base than others. Those based on guidelines have the strongest evidence base; those derived from basic first principles supported by some research findings are relatively weaker, but are often sufficient for the task at hand. A consistent source of tension is the distinction between the clinically derived fine-tuning of the definition of a quality measure and the limitations of available data sources (which often do not contain sufficient information to act on such distinctions). Although some excellent technical measures were found, the overall development of technical measures seems less advanced than that of the general process measures.