What will it take to motivate hospitals to do what we know works to make health care safer? Of the three major approaches to improving patient safety—regulation/accreditation, financial incentives, and public reporting—the most promising is public reporting of performance information and feedback to providers. Transparency is an idea whose time has come and both hospitals and the public will be better off because of it.
Data from a large number of hospitals, gathered by several sources, show wide variations in the incidence of one of the most lethal hospital-acquired complications, central line–associated bloodstream infections (CLABSIs). Compared with the evidence on how to prevent other types of infections—and most other kinds of adverse events—the evidence on how to prevent CLABSIs is quite strong. Peter Pronovost demonstrated the potential for complete elimination of central line infections in his intensive care unit at Johns Hopkins Hospital seven years ago.1 In 2005, in a stunning display of generalizability, Pronovost and his team taught staff in over 100 Michigan hospitals to implement his protocol for central line insertion, and 68 hospitals completely eliminated CLABSIs for six months or more.2
Yet, we still have significant rates of CLABSI in most hospitals, and some are very high. What is going on? What is going on is that the vast majority of hospitals have not implemented the Pronovost protocol because they have not made a meaningful commitment to reducing preventable injuries, much less eliminating them. Despite an avalanche of data, exhortation from all kinds of experts, and impressive results by some, most hospitals have in place programs to implement only a few of the known safe practices, and none has a strategic plan to implement all of the 34 evidence-based safe practices endorsed by the National Quality Forum. The "chasm" between what we know and what we actually do to prevent complications of our treatment matches that described for all of quality by the Institute of Medicine in its 2001 report, Crossing the Quality Chasm.3
What will it take to motivate hospitals to do what we know works to make health care safer? Evidence is available on the effectiveness of three major approaches: regulation/accreditation, financial incentives, and public reporting of performance and feedback to providers.
Because regulation is a state function, and there is tremendous variation in state approaches to quality and safety, its use has been spotty. Information from reporting systems, for example, is seldom used by regulators to improve safety. Although licensing functions are usually supported with public funds, state departments of public health seldom have the resources to monitor hospital practices. Given cost constraints and inertia, this situation seems unlikely to change in the near future. The leading accreditor, the Joint Commission, has been the most effective force for change by requiring hospitals to implement its Patient Safety Goals. However, monitoring safe practices is only a small part of the Joint Commission's activities. It seems unlikely that either it or the states will be able to exert enough pressure to get health care systems to make the quantum changes necessary in hundreds of processes to make health care safe.
Using the reimbursement system to improve quality of care has been in vogue for a decade or more. Incentives are usually positive: payment of a bonus as a percentage of reimbursement—2 percent in the Centers for Medicare and Medicaid Services (CMS)/Premier Hospital Quality Incentive Demonstration—although rewards are sometimes packaged with penalties for underperformers. Rewards tend to be for process improvement, not outcomes. There is some evidence that financial incentives improve compliance with quality indicators (such as use of certain medications following acute myocardial infarction), but little or no evidence of improved outcomes.4 Programs tend to reward mostly providers who are already leaders in quality. As with standards to improve public education, many fear that providers will perform to the test rather than proactively seek to improve quality, but this is a concern with all measurement.
Financial incentives for improving safety, on the other hand, are relatively new. In contrast to those for improving quality, which are positive and process-oriented, incentives for safety have been negative and outcome-oriented: instead of receiving a bonus for adhering to a safe practice, providers are penalized for the consequences of not doing so. The focus has been on selected "never events," taken from the list of serious reportable events developed by the National Quality Forum.5 These are significant patient harms that hospitals know how to prevent—or should know. Such financial disincentives began in Minnesota but have been given national reach with CMS' recent decision not to reimburse hospitals for the additional costs attributable to eight serious reportable events.
Although the stakes for any hospital are small (these are, or should be, rare events), the pushback has been considerable. Hospitals claim, with some justification, that not all of these events are preventable. This concern seems particularly apt for at least two of the "no-pay" events on the CMS list: falls and urinary catheter–associated infections.6 Reliable and valid practices for preventing these adverse events have yet to be developed.
Evidence that not paying for serious reportable events improves safety is also lacking. The experience in Minnesota is not encouraging. For example, the number of wrong-site surgeries and retention of foreign bodies in Minnesota hospitals actually increased substantially from 2006 to 2009.7 Some of this increase is undoubtedly related to improved compliance with the reporting requirement, but nonetheless it does not indicate that the penalty is having a positive effect.
So far, the most powerful method for reducing preventable injuries has been to require physicians to provide data on their own performance and then provide them with comparisons of their risk-adjusted complication rates with those of their peers. The Veterans Administration (VA) pioneered this approach in the 1990s with its National Surgical Quality Improvement Program, which has since been adopted and promoted by the American College of Surgeons. Under this program, each hospital's surgical specialty department receives feedback on its risk-adjusted complication and mortality rates, together with a comparison with all of the other (unidentified) surgical departments in the VA system. In response to these reports, below-average units made substantial improvements, leading over several years to system-wide declines in both complication rates and mortality that significantly exceeded the secular trend.8
It is reasonable to assume, though as yet unproved, that public reporting of similar types of data would spur hospitals to make greater efforts to reduce adverse events. Hospitals—or the public—can choose the benchmark level they prefer: above average, top decile, or others. But it seems evident that performance reporting works best when all providers participate—as in the VA experience. Thus, reporting has to be mandatory. As Wachter emphasizes, it is essential that the events to be reported are a) clinically significant, b) easily measured, and c) largely if not completely preventable.9 Risk adjustment is essential for fair comparisons.
The "benchmark" in safety, of course, should be zero. If it is, then risk adjustment is irrelevant. The hope is that, as it becomes public knowledge that some hospitals are able to eliminate certain types of adverse events, others will be motivated to follow. While a major thrust of the patient safety movement has been to eliminate blaming and shaming of individuals when they make mistakes, for organizations public reporting may be an appropriate use of shaming.
The larger issue here is transparency. From an ethical standpoint, the argument in favor of transparency is straightforward: the public has a vital stake in the outcomes of health care, and therefore it has a right to know how we are doing. (The contrary argument that hospitals and doctors have a right to keep their results secret in order to protect those with bad results is patently untenable.)
From an economic standpoint, Porter and others regard consumer access to full information as a critical element of value-driven purchasing of health care10 They contend that consumers can make meaningful choices only if they have complete information. While this formulation is attractive to some economists and policymakers, repeated studies over more than 20 years—going back to the Pennsylvania cardiac surgical scorecards of the 1980s—show that few patients and even fewer doctors pay much attention to this type of information in deciding with whom and where they will receive their medical care.
From the standpoint of improving patient safety, however, transparency is crucial. It is the cornerstone of the cultural transformation that our health care organizations need to undergo to become safe. Transparency is essential within an institution if caregivers are to feel safe in reporting and talking about their mistakes. The free flow of information is essential for identifying and correcting the underlying systems failures. Transparency is also the key to successful—and ethical—responses to patients when things go wrong. It is the cover-ups that lead to lawsuits. And transparency is essential for accountability, to show the public that the hospital or system responds ethically to its failures. Internal transparency begets external transparency—and vice-versa.
Although most hospitals are still skeptical about being transparent, evidence from a few organizations that have gone public with their bad news shows that it is a win-win. First, transparency motivates caregivers to improve care. Lives are saved. In addition, openness shows that the hospital feels accountable and has nothing to hide, which increases public confidence. Transparency is an idea whose time has come and both hospitals and the public will be better off because of it.