1 Despite their conceptual appeal and potential benefits, the results of many COE initiatives have been largely disappointing. For example, analyses by Mehrotra et al.19 found no evidence that the Blue Distinction Centers had better outcomes than nondesignated hospitals with spine surgery and total joint replacement. In another study examining the effects of CMS’ 2006 policy restricting bariatric surgery to COEs, Dimick et al.20 found no evidence that COEs had better performance than non-COEs, or that the CMS policy had improved outcomes overall. As a result of that study and others reaching similar conclusions,12,21 CMS decided to rescind its policy restricting bariatric surgery to COEs in 2013. These examples highlight problems with current approaches to assessing the comparative quality of hospitals and surgeons for COE purposes. These challenges include using administrative data to assess outcomes other than mortality, putting too much weight on hospital attributes and processes of care not tightly linked to outcomes, and relying on self-reported data.22
Measurement challenges aside, COE programs have other downsides. Strategies that displace patients from their usual site of care and regular physicians may interfere with coordination of care. They tend to be highly polarizing, dividing hospitals and surgeons into winners and losers. In alienating the latter, a price of COE programs may be lost opportunities for engaging physicians in other types of quality improvement efforts. And finally, such programs improve outcomes exclusively to the extent that they steer care away from poor performers. It provides no mechanism for helping non-COE hospitals and surgeons improve their outcomes.
PAY FOR PERFORMANCE
Another approach to improving surgical outcomes is to encourage hospitals and surgeons to provide higher-quality care with financial incentives. Both public and private payers have active pay-for-performance programs involving surgery. CMS has both a Value-Based Purchasing program aimed at hospitals and a more recent effort targeting physicians – the so-called Physician Value-Based Payment Modifier program.23 Although such programs are moving toward specialty-specific outcome-based measures (as described later), pay-for-performance programs have until now focused on adherence to evidence-based practices related to perioperative care. For example, many private payers link hospital reimbursement to process measures established by the Surgical Care Improvement Project (SCIP),24 including practices aimed at reducing rates of surgical site infection, postoperative cardiac events, venous thromboembolism, and ventilator-associated pneumonia.
Among their strengths, pay-for-performance initiatives aimed at improving compliance with specific processes of care are considerably less polarizing than selective referral. In theory, anyone can “win.” To the extent that surgeons can “play to the quiz,” process-based pay-for-performance programs have the potential to achieve rapid and significant improvements in many aspects in perioperative care. Studies in primary care suggest that pay-for-performance programs may be particularly effective in improving process compliance among poor performers and thus reducing overall variation.13,25 In surgery, evidence to date suggests similar improvements in compliance with SCIP measures.26
2 Nonetheless, the extent to which improvements in evidence-based practice compliance translate to improvements in surgical outcomes remains uncertain. Studies to date have generally failed to demonstrate a strong correlation between provider performance with specific process measures and patient outcomes. For example, hospitals successful in examining at least 12 lymph nodes from the surgical specimens of patients undergoing colectomy for colon cancer, one popular measure, did not have better 5-year survival rates than those scoring worse on this quality indicator.27 Even when hospital performance is measured across multiple, evidence-based practices related to a specific postoperative complication, process compliance is not well correlated with patient outcomes. In examining national data compiled by CMS’ Hospital Compare program, for example, Nicholas et al.28 found that hospital compliance with the “bundles” of SCIP process measures related to surgical site infection and venous thromboembolism varied widely. Moreover, hospital compliance rates were not related to the incidence of those two postoperative complications.
A more recent study by Shih et al.29 provides the most direct evaluation of the overall effects of surgical pay-for-performance programs. Focusing on coronary bypass surgery and total joint replacement, Medicare’s Premier Hospital Quality Incentive Demonstration (HQID) was intended to incentivize quality and quality improvement by increasing reimbursements to both top performing hospitals and those demonstrating the greatest degree of improvement over time.30 While mortality and complication rates for cardiac and orthopedic patients decreased significantly over time in premier hospitals, improvements of similar magnitude were observed in nonparticipating hospitals.29
3 Together, these studies do not invalidate the importance of timely antibiotics and other well-tested prophylactic measures against surgical complications. Rather, they underscore the complexity of surgical care and the limitations of focusing exclusively on a small number of discrete steps in managing a surgical patient. As we describe later, these studies also highlight that surgical outcomes also depend on aspects of care not traditionally addressed in pay-for-performance programs, like surgeons’ technical skill and their proficiency in managing complications after they have occurred.
OUTCOME FEEDBACK
In contrast to strategies led by payers and policy makers, surgeons and their professional organizations are focusing primarily on registry-based quality improvement programs.14,15,31 Their goal is to provide hospitals and surgeons with rigorous feedback about their outcomes relative to those of their peers. Some of these programs have centralized approaches to coordinating quality improvement activities, while others focus primarily on performance feedback, leaving improvement work at the local level.
The Scientific Registry of Transplant Recipients and the Adult Cardiac Surgery Database of the Society of Thoracic Surgeons were among the earliest and most recognized surgical outcome registries. The large majority of US hospitals involved in those two specialties now contribute to those databases.32,33
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
