You know that an iPhone 8 is better than an iPhone 5 by comparing specs for the two phones, says Harvard Business School Professor Robert Huckman. How can patients and providers do that in health care? How should quality measures be used — or not used?
“As a patient, how do we engage in that iPhone 5 verses iPhone 8 comparison when it comes down to whether we should have surgery done by Provider A or Provider B or whether we should have surgery done at all?” asks Huckman. “For providers themselves, how do I know, relative to the industry, if I’m an iPhone 5 or an iPhone 8 when it comes to serving my patients in a particular area?”
Quality measurement has some good uses, one of which is to describe a problem, answers Harvard Medical School Professor Michael McWilliams. Many seminal articles on deficient quality in health care relied on measures. Another use for quality measurement is evaluation.
“But we do run into problems when we tie public reporting or financial incentives to them,” says McWilliams. “There are some serious costs.” Some of these include inadequate risk adjustment leading to disparities, detracting resources away from things that can’t measure well but that are important for quality, gaming behavior, wasteful behavior like risk selection, and the reporting burden on providers.
That said, McWilliams agrees that health care is missing out on quality comparisons like the iPhone example. He acknowledges that “you are an astute observer of what you like; you know what you like in an iPhone.”
“To some extent, we should rely on what patients think, but they don’t have MDs so they can’t appreciate a lot of clinical complexities,” says McWilliams. “The only people left over — doctors, nurses, health care professionals — are trained to help us know who gives better quality care.” If you need surgery or an emergency hospital visit, for example, and asked a physician for his or her opinion on different surgeons or hospitals, you would “get a very strong opinion,” he says. A lot could be done to expose physicians to more information so that they can improve those impressions and better guide referrals, and this would help guide quality improvement.
What if, asks Huckman, underlying that clinician opinion is a reputational concern, not necessarily an understanding of the quality measure? How do we know that we’re getting the clinician’s opinion, not the reputational one?
How provider markets are structured plays a part in that, says McWilliams. “A big requirement for this to start working would be more competitive provider markets. For example, if I am not happy with the way my practice is providing primary care in this market where we have a lot of dominant health systems, I just don’t have that many choices as to where I want to practice.” More competition means more choices and would go a long way toward exposing physicians to variability in quality that is not measure dependent. “That being said, there is clearly data you can gather to help that process,” he says.
From the NEJM Catalyst event Navigating Payment Reform for Providers, Payers, and Pharma, held at Harvard Business School, November 2, 2017.