I had the opportunity to interview one of the nation's foremost experts on pay-for-performance and health care quality measurement, Harvard professor Ashish K. Jha, MD, MPH. His entertaining and insightful blog "An Ounce of Evidence" tops my bookmarks. He's known in the business community for his forceful candor on the need for much more transparency and better payment systems in health care. He also serves as a volunteer member of my organization's Blue Ribbon Panel on the Hospital Safety Score.
Health Care Improvement: "Excruciatingly Slow"
Leah Binder: Patent safety has been a big topic of conversation in health policy circles, especially since 2000 when the IOM issued its report To Err Is Human, estimating as many as 100,000 Americans were dying of preventable hospital errors every year. Have we made progress?
Ashish K. Jha: There has been some progress, particularly in two areas. First, there has been a culture shift over the past couple of decades. 15 years ago every error was considered an individual's fault, and the notion that systems were responsible was completely novel and even strange to many people. Over the past 15 years, we have come to understand that systems fail patients far more often than people fail patients. People make mistakes because we're human and the job of good health systems is to catch those mistakes and prevent them from hurting patients.
Second, we have seen some progress in certain areas of safety. For instance, infection rates are generally down from where they were 10 years ago.
But I find the pace of improvement excruciatingly slow, and I think the biggest thing that hinders us is the lack of clear and strong incentives for making patient care safer. You can be a hospital with a mortality rate that's twice as high as your neighbor's down the street, and the financial consequences of that are pretty trivial. And while we continue to tolerate that kind of failure, we shouldn't be surprised that the progress is going to be slow.
The Key: Keeping Hospital CEOs Awake at Night
Binder: So should we reform the payment system to impose more financial consequences on doctors and nurses when there are problems with patient safety?
Jha: As a physician, I don't need to be paid more to provide safe care for my patients. I can only do so much as a physician; patient safety depends largely on the organization in which I practice in. And organizational investments in safety are very much driven by financial incentives than individual incentives. If an organization felt an existential threat from providing unsafe care, you can bet that organization would start making investments in patient safety.
I actually don't think pay-for-performance for individual doctors and nurses is going to make a big difference. If you came to me and said, "Ashish, 20 percent of your salary is going to be tied to the infection rates of your patients." I'd respond by trying to redouble my efforts to wash my hands, etc. But, there's only so much I can do personally. On the other hand, if you went to the hospital and said,"Twenty percent of the entire reimburse is going to be tied to your infection rate,"-- that's a make-or-break proposition for that hospital, because if they lose 20 percent they have to shut down. And, they would make the kinds of investments we need to make care safer.
I haven't heard of any hospital that went out of business because its care was unsafe. I also haven't heard about any CEO who got fired because the hospital's infection rate was too high. It doesn't happen, and that's telling.
Right now, when I talk to hospitals and hospital executives, what I hear is that safety is on the checklist of things that chief quality officers have to do. It's not on the top priority list for CEOs. It's not what keeps CEOs awake at night. And until we get CEOs losing sleep about unsafe care, we're not going to make a big dent in the failures of our health care system.
The Problem with Obamacare's Focus on Readmissions
Binder: Creating financial incentives for hospitals to improve is a goal in the Affordable Care Act. You've expressed some ambivalence about tying those payments to readmissions rates. Why?
Jha: I don't find that readmission rates are a very patient-centered measure. If you said to someone you have three choices: Go to a hospital with high death rate, go to a hospital with high infection rate, or go to a hospital with high readmission rate, nobody would say, "God, I'll take the high death rate or the high infection rate- but please don't let me be readmitted!" I don't think most patients would say readmission is a fate worse than death. And yet, that's how we have prioritized it: When you look at the set of metrics that the ACA uses to reward or penalize hospitals, the one that gets the most dollars tied to it is readmissions, and that has never made much sense to me.
Readmissions is not a measure of quality, it is a surrogate marker for quality. Readmission is a utilization measure, measuring whether a patient uses hospital care within 30 days of discharge. Some readmissions are clearly tied to bad quality care, but others are tied to the fact that some patients are just really sick. A lot of readmissions occur because the patient doesn't have good social support at home or can't access primary care, and so when they get sick they have to come back to the hospital. And when they do, I don't want the hospital to feel pressure not to admit them if that's what the patient needs. We need to make better investments in the strength of the primary care system.
Binder: Are you suggesting that readmissions are always a bad measure to use for pay-for-performance?
Jha: It isn't necessarily bad, though readmissions wouldn't be my top priority. And this is where I think there is some amount of nuance is important because we actually have empirical data.
We find that for medical conditions -- heart attack, pneumonia, heart failure -- much of what drives readmissions is how sick patients are, or what kind of social support they have at home. These are factors that have less to do with the quality of hospital care. These patients are often older, chronically ill, with major medical problems. Twenty years ago many of these patients might have died in the hospital, but today, we are able to keep them alive. However, when they go home, even small issues can land them back in the hospital. There is no doubt we have to figure out how to improve care for these chronically ill people. It's just not clear to me that focusing on readmissions is the best approach.
The story is very different for surgery. All of the empirical data suggests that surgical readmissions are much more about quality. They seem to be more closely related to what happened during and after surgery, and as a clinician this makes a lot of sense.
Usually, it's complications of the procedure or of the hospital stay that bring surgical patients back to the hospital. So, if you want to lower surgical readmissions, you have to focus on improving surgical and post-operative care and asking hospitals to do that is reasonable.
Time to Take the Long View:
Binder: What would you like to see instead of readmission rates as an incentive for hospital improvement?
Jha: As I've said in a couple of my blogs, I'd like to see much more long-term thinking by Medicare, things like bundled payments tied to real quality measures that take a long view. Not 30 days but 90 days, or even longer. But, we aren't going to get there until we have a much better ability to measure the value of that long-term care. We need major investments in better understanding post-acute and longer-term care. But when we do, and when we ask providers to be accountable for the long-term outcomes of patients, we are much more likely to drive the kind of change we want to see than simply focusing on short-term readmissions rates.
The piece first appeared on Forbes