Several papers in the March issue of Health Affairs expose some of the challenges with reporting information about health care quality to the public, including the shortcomings of hospital reporting, the importance of framing quality information in ways consumers can understand and apply to real-world decisions, and the need for more consumer-relevant measures. Each paper offers useful research findings for elements of our national public reporting strategy.
But policymakers and payers shouldn’t draw the wrong conclusion from looking at these snapshots taken through the rear-view mirror. If we extrapolate too much from studies that are based on flawed measures deployed in flawed settings, we won’t get much insight on how to create a flow of information that provides value to the health care system we’re trying to encourage. Those responsible for shaping accountable care organizations, value-based payment programs, or the meaningful use of electronic health records won’t learn much from process measures layered on a fragmented, fee-for-service system with no incentive or accountability for the goals of the Triple Aim.
In fact, most of the findings reported in the March issue have been known for decades and been generally ignored in the design of the large scale reporting programs we see today. (See Note 1) The question is, why?
Let’s consider three commonly cited barriers to progress:
- We don’t have relevant measures, particularly for outcomes
- We don’t know how to communicate effectively to consumers
- Providers don’t have a ‘business case’ for reporting on quality.
1. We don’t have the measures
There is an abundant, scientific, commonly used library of measures that address symptoms, functioning, and outcomes that matter to patients. Work done by John Ware and many of the PORT grantees, the Picker Institute in the US and Europe, the Foundation for Accountability, and the PROMIS team has generated dozens of instruments and measures of quality. The Swedish approach to orthopedics quality provides one illustration. For every hospital and county in Sweden, they report annually on:
- Reported health gain (EQ-5D index gain after one year).
- Patient satisfaction one year after total hip arthroplasty.
- Short-term complications two years after total hip arthroplasty.
- Ten-year implant survival following total hip arthroplasty.
- One-year implant survival following hemi-arthroplasty.
None of these measures are in general use in the United States, and physicians often protest that accessing patient outcomes data is infeasible. Yet the one-year rate of patient follow-up after surgery in Sweden – where patients report on their pain level, functioning, and satisfaction – is 90 percent. In the United Kingdom, where measurement of orthopedics outcomes is only two years old, 80 percent of patients are reporting on their post-operative outcomes and the results are available by hospital. These data reveal significant variations in patient outcomes after surgery. Comparisons among Swedish regions and hospitals show a 21 percent variation in patient satisfaction results and a 41 percent variation in quality-of-life improvement.
Certainly if we wanted to measure and report on orthopedics outcomes in the United States, the measures and methods to do so exist. And the same could be said for cardiology, gastroenterology, ophthalmology, asthma, breast cancer, diabetes, and many other prevalent conditions.
2. We don’t know how to communicate with consumers
Oddly, Americans are famous for knowing how to influence consumer decisions. Our economy and political campaigns hinge on ever more sophisticated methods. It’s not an exaggeration to say that US hospitals’ advertising and marketing budgets would fund, by a factor of 10, the cost of collecting and distributing quality information about clinical outcomes and patient experience rather than the vacuous self-promotion that dominates urban billboards. The many earnest experiments at publishing quality data – ranging from the CMS Hospital and Physician Compare websites to the Healthgrades commercial products – attract miniscule numbers of consumer users, while Yelp is now in front of 60 million Americans every month looking for restaurant and dry cleaners’ ratings. What do Yelp, Angie’s List, and Urbanspoon know that CMS doesn’t?
3. Providers don’t have a business case
Thanks to Jack Wennberg and others, we have ample evidence that there is enormous variation in the performance of hospitals and doctors. In our California data, for example, the range of performance among primary care doctors on longstanding measures like colorectal cancer screening range from 28 to 78 percent (10th to 90th percentile range). No patient knows that the likelihood of receiving basic preventative screenings and chronic illness monitoring depends on which doctor’s office door they enter.
Payers, purchasers, and regulators have failed to require disclosure of performance data from those with whom they contract, failed to reward positive performance in any significant way, and allowed professionals and facilities to profit from unjustified service volume. We are caught in an unproductive loop: payers will not adopt measures that lack an evidence base and endorsed methodology; providers are not motivated to collect data for measures that are not likely to be used for payment; and the evidence base stagnates since no systems are in place to collect the necessary data.
Let’s consider the positives. Many measures and instruments exist to assess outcomes of interest to patients, but they are not in general or standardized use. American marketers – on the web and in print – know how to push information into consumers’ hands to influence decisions. Public agencies, employers, and insurance plans are moving steadily towards value-based purchasing. Rapid adoption of electronic health records and registries will provide widespread access to reliable clinical data. The research that shows low public interest in measures of low consumer value presented through inaccessible channels should not keep us from giving people the information they want and need to make good, personal health care decisions.
We will need to follow a few key principles:
- Implement measures that consumers say they want – regardless of the inherited and limited data infrastructure. Start building the necessary information systems now, leveraging electronic health records, registries, and standards-based health information exchange.
- Recognize that consumers want information relevant to them – not necessarily what experts think they should want. The Consumer-Purchaser Disclosure Project, a leading coalition of purchasers and consumers has some great insights on this.
- Meet consumers where they are; don’t try to lure them to a stand-alone website or publication. When shopping for a car or refrigerator, the performance information is on the sticker in the showroom; you don’t have to go to a government website to find it. Let’s integrate accountability information with health plan directories, Yelp, Castlight and similar shopping services, and search engines and kill off the attempt to create central scorecard websites. The search engine Bing’s mash-up of CAHPS and Hospital Compare data was a great illustration of this approach.
- Don’t try to make QI and accountability measures the same. The most vibrant QI work uses external measures to identify opportunities for improvement, but turns to statistical process control, run charts, and other tools to actually alter workflows and care processes. In general, public policy should express what the accountability framework is and then private parties can measure whatever they find useful to drive improvements towards those public objectives. The vast majority of PCPI and PQRS measures do not need to be reported to the public and do not need tight specification; individual societies and enterprises should articulate them however is useful to them.
We are at a critical juncture in the evolution of our health care system. Growing government interest in performance measurement and reporting, the adoption of value purchasing models keyed to measured performance, implementation of standards-based EHRs, disease registries, and health information exchange, and greater consumer exposure to health care decisions – including through the new insurance Exchanges – will all rest upon a foundation of publicly reported quality measures. Let’s not continue the madness of over-specified, highly granular process measures that burden providers with unnecessary cost and frustration yet fail to speak to patients and families.
We can’t let this recent flurry of studies and articles divert our attention and slow us down. Policymakers and private payers must understand and commit to these principles, and use their strength to put us on the right path.
Note 1. A few similar studies from 1993-1996: Hibbard JH, Jewett JJ. What type of quality information do consumers want in a health care report card? Med Care Res Rev. 1996 Mar;53(1):28-47; Edgman-Levitan S, Cleary PD. What information do consumers want and need? Health Aff (Millwood). 1996 Winter;15(4):42-56; Lansky D. A patient-focused approach to measuring the quality of cardiac care. Qual Lett Healthc Lead. 1993 Nov;5(9):18-23; McCarthy MJ Jr, Shroyer AL, Sethi GK et al. Self-report measures for assessing treatment outcomes in cardiac surgery patients. Med Care. 1995 Oct;33(10 Suppl):OS76-85; Ware, JE. What Information Do Consumers Want and How Will They Use It? Medical Care. 33(1):JS25, January 1995.