Strengthening the Business Case for Patient Safety
After more than a decade in the national spotlight, the problem of patient safety remains a serious concern in United States' hospitals. Although there have been some high profile successes (1), progress has been slow.(2) Well-recognized barriers include limited evidence about the effectiveness of interventions intended to improve patient safety and even less understanding of the contextual factors and strategies required to translate new knowledge into routine practice.(3,4)
An additional barrier has been the historically weak "business case" for patient safety, which can be traced to the original design of Medicare's Inpatient Prospective Payment System (IPPS). Not only did the IPPS fail to pay a premium for higher quality care, but it has traditionally rewarded hospitals for poor outcomes by assigning patients who develop a complication to a higher paying diagnosis related group (DRG), whether or not the complication was preventable. Consequently, most savings realized through patient safety initiatives have tended to accrue to payers, in the form of reduced outlays, instead of to the organization making the investment.(5,6)
However, this is changing. The Deficit Reduction Act of 2005 was, in part, a legislative attempt to do away with these perverse incentives, and in 2008 Medicare implemented the Hospital-Acquired Conditions policy, eliminating additional DRG payments for 10 types of complications felt to be preventable through the adoption of evidence-based guidelines, including catheter-associated urinary tract infections and falls. While Medicare's 2008 decision to penalize hospitals for poor quality and safety was one type of policy change to promote improvement, another direction has been to support capacity building. For example, despite strong evidence that computerized provider order entry can reduce the frequency of medication errors and adverse events (7), the high upfront acquisition and implementation costs of these systems, together with an uncertain return on investment, had hindered adoption. The EHR Incentive Program of the HITECH Act of 2009, which will provide more than $20 billion to hospitals and physicians for implementing information technology systems that meet certain standards, is thus another example of federal efforts to strengthen the business case for patient safety.
The implementation of Medicare's Value-Based Purchasing and Readmissions Reduction Programs in October 2012 is the latest, and most aggressive, attempt to address the weaknesses of existing payment policy. The policy, a product of the Affordable Care Act of 2010, represents a defining moment for pay-for-performance programs in US hospitals. It also marks the first time that pay-for-performance has been applied directly to the problem of patient safety. Under the program, 1% of base hospital DRG payments are withheld, with value-based incentive payments awarded later to institutions that have demonstrated high levels of performance or improvement. In its first iteration the program includes 13 measures taken from Medicare's Inpatient Quality Reporting (IQR) program, including process measures focused on the prevention of surgical infection, postoperative venous thromboembolism, and postoperative cardiovascular complications. In 2014, the program will expand to include the prevention of catheter-associated urinary tract infection, and more significantly, will begin to incorporate outcome measures, specifically hospital 30-day mortality rates for patients with acute myocardial infarction, heart failure, and pneumonia. By 2014, 45% of incentive payments will be weighted based on hospital performance on process measures, 30% on measures of patient experience and 25% based on patient outcomes. The hospital withhold is set to increase over the first 5 years of the program, from 1% in FY 2013 to 2% in FY 2017. In addition to Value-Based Purchasing, the Readmission Reductions Program penalizes institutions with higher than expected readmission rates. In 2012, more than 2000 hospitals were affected; collectively, they paid approximately $280 million in penalties.
Given the range and scale of these initiatives, those who have longed for a "business case for safety and quality" are undoubtedly cheered. Yet, against this backdrop, pay-for-performance has faced significant criticism. Opponents argue that the programs don't work, raise concerns about their impact on access to care and disparities, and suggest that financial incentives will corrode the professionalism and sap the intrinsic motivation upon which health care depends.(8-10) In addition, there are technical concerns about the adequacy of risk-adjustment and the limited statistical power to detect differences between institutions (a particular problem for some rare patient safety events). Moreover, there is a fear that hospitals may attempt to game the system by focusing more on documentation and coding than actual changes in care.(11,12) Even advocates argue that the current approach requires retooling, needing better measures and larger incentives applied in different ways.(13)
In weighing these arguments it is important to keep in mind that what we know about the benefits and risks of these programs is based on a small n. Within the US, the Hospital Quality Incentive Demonstration (HQID), which initially paid up to 2% in bonuses to hospitals with high relative performance and served as the pilot for today's Value-Based Purchasing program, is now viewed as a negative trial. While rates of adherence to process measures for patients with acute myocardial infarction, heart failure, and pneumonia were modestly higher after 2 years at participating hospitals, the difference between hospitals receiving pay-for-performance payments and control hospitals (which experienced public reporting and assorted other pressures to improve) faded over time, and more frustratingly, did not translate into better survival.(14-16) A similar story emerged from CMS's Hospital-Acquired Conditions program. Using data from the National Healthcare Safety Network, a recent analysis found that while catheter-associated bloodstream infections and catheter-associated urinary tract infections had been on the decline, the introduction of financial incentives (in the form of withholding additional payment for preventable complications) did not accelerate the rate of change.(17) On the other hand, an evaluation of a natural experiment with pay-for-performance in the United Kingdom, one modeled on the HQID, but with a 4% bonus instead of 2%, yielded positive results. Within this program, English hospitals exposed to financial incentives tied to performance on the same set of process measures as in the HQID demonstrated a small but significant reduction in 30-day mortality.(18)
Although the track record is spotty at best, these pay-for-performance programs have strong appeal—in no small measure because traditional models of hospital payment are so problematic and can be viewed as at least partially to blame for our underinvestment in quality and safety. If quality expert Paul Batalden was right and "every system is perfectly designed to get the results it gets," then whatever successes the patient safety field has achieved over the past decade are all the more impressive given the headwinds posed by traditional fee-for-service payment system. Turning a battleship the size of the US Medicare program, with its large number of stakeholders and huge dollars at stake (approximately 5% of the US economy), is no trivial undertaking. The architects and developers of the Hospital-Acquired Conditions and Value-Based Purchasing programs should therefore be applauded for their willingness to experiment with novel payment approaches that begin to address the weaknesses of the legacy system they inherited. Did they get everything right? Not by a long shot. Bonuses of 1% to 2% seem puny by the standards used in business and may be insufficient to drive the desired behavior change. And it was clear from the start that most patients who developed a hospital-acquired condition had other complications and comorbidities that would still allow for a higher DRG payment. Moreover a reasonable argument can be made that incentive payments (and penalties) should focus exclusively on outcomes, lest the hospitals worry about the steps needed to get there—especially when current process measures explain only a small fraction of observed variations in outcomes.(19) But, fears of reduced access have thus far not panned out (20), and while these new financial incentives are likely to affect hospital (and ultimately physician) behavior and image, our current approach to financing, which favors volume over value, may well be responsible for even less desirable actions.
In the end, whether or not pay-for-performance is successful will depend on the willingness of Medicare (and other payers) to embrace the tenets of any learning health care system. At this point, the goals we are trying to achieve are clearer than the best methods to reach them. Therefore, it will be vital that we invest in, and then use, research about the outcomes of these initiatives (both intended and otherwise), to guide future policy making. The safety of the more than 36 million individuals who find themselves patients in US hospitals each year depends on it.
Peter K. Lindenauer MD, MScDirector, Center for Quality of Care Research, Baystate Medical CenterAssociate Professor of Medicine, Tufts University School of Medicine
6. Mello MM, Studdert DM, Thomas EJ, Yoon CS, Brennan TA. Who pays for medical errors? An analysis of adverse event costs, the medical liability system, and incentives for patient safety improvement. J Empir Leg Stud. 2007;4:835-860. [Available at]
12. Meddings JA, Reichert H, Rogers MAM, Saint S, Stephansky J, McMahon LF. Effect of nonpayment for hospital-acquired, catheter-associated urinary tract infection: a statewide analysis. Ann Intern Med. 2012;157:305-312. [go to PubMed]
19. Bradley EH, Herrin J, Elbel B, et al. Hospital quality for acute myocardial infarction: correlation among process measures and relationship with short-term mortality. JAMA. 2006;296:72-78. [go to PubMed]