Implementation of a statewide quality assessment program for rectal cancer.

Abstract

4 Background: The quality of surgical care for rectal cancer (RC) has been extensively studied in Europe; however, efforts in the United States are limited by the lack of detail about surgical care in available registries. We describe a unique quality assessment program for RC designed within the Michigan Surgical Quality Collaborative (MSQC), a collaborative quality improvement organization that includes community and academic medical centers. Methods: 10 MSQC hospitals contributed to this retrospective cohort study (2007-2012). Experienced nurse reviewers were trained to abstract cancer surgery-specific data from hospital medical records and local tumor registries. Five RC surgery-specific quality measures were designed based on literature review and current guidelines: adequate lymph node (LN) procurement (>=12); use of mesorectal excision; rate of margin positivity; use of neoadjuvant therapy for clinical stage II/III; and sphincter preservation rate for mid/upper RC. Results: 353 RC cases were studied (333 radical surgery, 20 local surgery). Participating hospitals varied in size (5 <400 beds), teaching status (6 major teaching, 2 minor teaching, 2 non-teaching), and urban/rural location (9 urban). Challenges encountered in the data abstraction training process included overcoming technical jargon in pathology and operative reports. Regular conference calls, access to a specialist for questions, and modifications to definitions helped overcome difficulties; 9 of 10 abstractors scored >90% correct on test cases. Analysis reveals wide variation between hospitals on quality measures. 75% of cases had adequate LN procurement (range by hospital 64-100%); mesorectal excision was performed in 76% of cases (38-97%); 9% had positive margins (0-17%); 94% of clinical stage II/III cases had neoadjuvant therapy (67-100%); and 86% of eligible cases had sphincter preservation (62-100%). Conclusions: A program designed to provide salient feedback to surgeons regarding RC treatment is feasible, and reveals that high-quality processes of care are not consistently used. These data suggest opportunities for process improvement. The program will be disseminated statewide for prospective data feedback and quality assessment.