"How Yelp Reviews Can Help Improve Patient Care" - The New York Times
By Aaron E. Carroll
Hospitals and many insurance carriers care about patient satisfaction. It especially matters to hospitals because insurance payments can be influenced by how patients rate the care they receive, as well as by the health of the patient, which hospitals usually report.
Many people in the health care profession are put off by this. They argue that patient satisfaction scores aren’t necessarily aligned with outcomes. Moreover, they say that trying to improve satisfaction is a waste of time.
It’s possible, however, that patient satisfaction is being rewarded already, and that the efforts we are making to highlight it aren’t helping as much as we think.
Almost every study on patient satisfaction uses the Hospital Consumer Assessment of Healthcare Providers and Systems (H.C.A.H.P.S.) survey. Studies show it is correlated with clinical measures of quality, although some other studies dispute this.
Collecting such information is costly, but there may be other sources for quality assessments that don’t require investment from the health care system. In 2012, researchers in the Journal of General Internal Medicine examined online reviews from RateMDs.com and Yelp. They found that a majority of reviews were positive. They noted, however, that patients reported on aspects of care that extended beyond the patient-physician encounter. They were concerned about staff, access to the hospital and convenience. They also cared greatly about the bedside manner of the doctors they encountered.
Dr. Naomi Bardach, associate professor of pediatrics and health policy at U.C.S.F. Benioff Children’s Hospital San Francisco, and her colleagues looked at Yelp to determine how consumer ratings compared with those of the hospital consumer assessment survey. Of the almost 3,800 hospitals with survey and other data, about 25 percent also had ratings on Yelp. The correlation between Yelp and the survey was quite strong. Moreover, high ratings at Yelp were correlated with lower mortality for myocardial infarctions and pneumonia — and fewer readmissions for those problems as well as heart failure.
A recent study in Health Affairs expanded on this work. Researchers compared the content of Yelp narrative reviews with the factors considered important in the hospital consumer assessment survey. They found that Yelp reviews did cover most things the survey tallied. But they also covered an additional 12 criteria not in the survey.
These included the cost of a visit; insurance and billing; scheduling; compassion of staff; family member care; and the quality of many staff members. All of these things were important to patients, and all might be correlated with outcomes. More important, nine of the 15 most prevalent criteria in reviews were not included in the survey.
The use of metrics like the hospital consumer assessment survey assumes that those in the health care system have figured out how best to measure patient satisfaction. They also assume that all the information we need should come from patients or from the medical record.
That’s not how the real world works. In a more recent study, Dr. Bardach found that the perceptions of other family members matter and can also be powerful because they focus on safety in a way that patients may not be able to do. The same is true of those who help care for patients at home and help make medical decisions for them. Their opinions are ignored by the survey.
Those publicly available data may even be more comprehensive than that gathered at the behest of payers. The next question is whether the health care system needs to measure satisfaction — maybe the publicly available data, like at Yelp, is sufficient. Could those data be used alone to incentivize providers?
A couple of weeks ago, my colleague Austin Frakt wrote a column on hospital quality and market share. He argued that people could improve their health by choosing a hospital that has a higher quality rating. He also underscored that patient satisfaction scores are often aligned with quality and with better outcomes.
He highlighted a paper published in The American Economic Review that looked at how performance of hospitals was related to market share. Conventional wisdom holds that patients lack information on quality and that they cannot tell, or favor, providers who seem better. But researchers found that hospitals that performed better, on both outcomes and process-based measures, tended to have greater market share, and experienced greater market growth.
Further, they found that as patients shifted to hospitals with higher performance, that change alone drove a significant amount in the improvements seen in overall survival rates for a number of conditions. Overall survival improved, in part, just because patients shifted from hospitals with lower quality to higher quality. If patients had the ability to choose between hospitals, they tended to gravitate to those with higher performance.
In other words, this may be an area of health care where the free market is working. When allowed to choose, patients seem able to discern quality — as they define it — and gravitate toward it. It’s not clear that we need to be forcing the issue with measurement and reimbursement.
It’s important to recognize, though, that this was a study of Medicarepatients, all of whom arguably have more flexibility in terms of hospitals and doctors than those with private insurance. Those with private insurance often are restricted by “narrow networks” and directed to a few facilities and offices.
Americans cry out for more choice in their health care. This sometimes gets translated to mean a choice of insurance companies. But Americans with government-provided Medicare, who have the least choice of insurers, have the most choices when it comes to providers. And they seem to use that freedom to choose providers who perform better.
I asked Dr. Bardach about this. “It’s still unclear how people are getting the information to choose the hospitals, but the power of stories is likely an important part of why Yelp and other online reviews are compelling,” she said. “Stories add nuance and context to the otherwise somewhat sterile numbers that the H.C.A.H.P.S. produces.”
Research shows that patients reward quality on their own — when they can. We might just need to make it easier for more of them to do so.