Examining the “Halo Effect” of Surgical Care Within Health Systems

The “halo effect” suggests that improvements in health care delivery for one disease may improve outcomes in closely related diseases, but, to our knowledge, only a few studies have examined this concept in surgery.^1-4 We evaluated quality performance across 4 complex but common surgical procedures to determine whether high-quality care is consistently provided for multiple procedures at a single institution.

This is a retrospective review of data from the University HealthSystem Consortium. Data for 85 621 patients at 225 academic medical centers (AMCs) undergoing pancreaticoduodenectomy, colectomy (excluding low anterior resection, abdominal-perineal resection, and total colectomy), abdominal aortic aneurysm repair, or esophagectomy from October 2010 to June 2014 were reviewed. Hospital performance in each of 5 clinical variables independently associated with optimal outcomes (postoperative complication rate, length of stay, 30-day all-cause readmission rate, mortality rate, and hospital volume) was assessed and assigned a point value based on quintile performance (worst possible score of 1, best possible score of 5). Both mortality rate and length of stay were risk adjusted and presented as observed to expected ratios. Hospital volume was defined as the total number of the specified procedure performed during the study period.

A composite quality index score for each procedure was created by totaling the 5 individual outcome scores (worst possible score of 5, best possible score of 25) to identify the best- and worst-performing AMCs (Table). Hospitals in the top quintile were designated high performers; those in the bottom quintile were designated low performers. This was done for each procedure. We then determined whether AMCs that were high performers for 1 surgical procedure were also high performers in any of the other 3 procedures to evaluate whether optimal outcomes for 1 operation correlated with optimal outcomes for multiple surgical procedures. We performed a similar evaluation for low-performing hospitals. Statistical analysis was performed using χ² testing and stepwise backward regression. This research was deemed exempt by the UC Davis Institutional Review Board. Informed consent was not required owing to the retrospective nature of the study.

Of 229 AMCs, 33% achieved high performance in 1 procedure, but only 8% of AMCs achieved high performance in 2 or 3 procedures (Figure). No center achieved high performance in all 4 procedures. Similarly, 29% were low performers in 1 procedure and 20% were low performers in 2 or 3 procedures. Only 1 hospital (0.4%) was a low performer in all 4 procedures. The number of procedures in which a hospital was a high vs low performer was not randomly distributed on χ² analysis (P = .01).

Interestingly, 15% of AMCs were high performers in 1 procedure but low performers in another. One-fourth (25%) of all AMCs did not achieve high or low performance in any procedure.

When examining performance by procedure, high-performance AMCs usually achieved top performances in only 2 of 5 quality indicators, with fewer achieving top performances in 3 or 4 metrics of quality outcomes (data not shown). Only 1 center achieved top performance in all 5 outcomes, which was for patients undergoing colectomy. However, most high-performance AMCs were not low performers in any outcome and were never a low performer in more than 1 outcome. Low-performance AMCs rarely achieved top performance in any outcome but frequently had low performance in 2 or 3 quality outcomes. However, low-performance AMCs were rarely low performers in all 5 outcomes. These trends were consistent across all 4 procedures.

Stepwise backward regression predicting mortality among AMCs showed that both morbidity and hospital volume were always independently associated with mortality. This was consistent across all 4 procedures.

Excellence in 1 procedure did not transfer to excellence in other procedures within the same health system. In fact, 15% of AMCs were high performers for 1 procedure but low performers for another. Individual surgeons and policies within specific surgical divisions may be more responsible for optimal outcomes than processes of care delivery throughout an entire health system. High-quality centers achieved that designation with top performance in only 2 or 3 of 5 quality metrics. Even high-quality AMCs have room to improve.^5,6

Corresponding Author: Richard J. Bold, MD, Division of Surgical Oncology, Ste 3010, UC Davis Comprehensive Cancer Center, 4501 X St, Sacramento, CA 95817 (rjbold@ucdavis.edu).

Published Online: June 29, 2016. doi:10.1001/jamasurg.2016.1000.

Author Contributions: Drs Brown and Anderson had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Brown, Bold.

Acquisition, analysis, or interpretation of data: Anderson, Burgess, Bold.

Drafting of the manuscript: Brown, Anderson.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Anderson.

Administrative, technical, or material support: Burgess.

Study supervision: Brown, Bold.

Conflict of Interest Disclosures: None reported.