Friday, October 19, 2012

Using patient satisfaction as a performance metric

One of the biggest problem with the US medical system is cost:  every time you put a patient in front of a provider, the provider does stuff.  Sometimes it is what we want the provider to do; many times it is not.  See our earlier blog post:  What do tonsilectomies have in common with auto repair?

The problem, of course is the incentives:  our fee-for-service payment scheme rewards physicians and hospitals for doing stuff to patients, regardless of whether it is the cost-effective thing to do. 

To try and fix the problem, the government is evaluating hospital performance using patient satisfaction scores.  The Wall St. Journal has a funny piece about its obvious shortcomings:
"Donna Barnett, a senior nurse at Grady [Hospital], cites a patient who had a hemorrhagic stroke and recovered swiftly enough to walk out of the hospital about a week later. On the survey the patient complained that meals were served cold and gave Grady low scores. 'It makes you want to throw your hands up,' said Ms. Barnett."

When Vanderbilt asked patients what they wanted, it was easy and free parking, which explains the valet parking and unsightly parking structures all over campus.  Patients didn't seem to care too much about the quality of care, and not at all about the cost because other people (you and I) pay for their care. 

And lets not forget the placebo effect, which means that patients are not satisfied unless providers do something to them.  


  1. Do people really not care about their care because someone else is paying for it? Isn't the patient's incentive to be healthy - or as healthy as possible? The problem with patients' views on satisfaction is that they do not know how to rate or compare their care. Physicians have been held in such high esteem and patients, especially older ones, have been socialized to not question their doctor's care for them. Quality is assumed to be good when it should not be. Medical care is complicated, making it difficult for someone not in the field to judge.

    Quality measurement has exploded over the last several years. Interestingly, the quality being measured is not tied to cost. It is assumed improving quality will decrease costs, but what if what is being measured has little or no impact on costs? A great example is the Beer's criteria of inappropriate medications for the elderly. This list was developed by an expert panel and implemented into quality measures. Sounds good, right? Keep the elderly off of medications that would cause drowsiness, increase likelihood of falls, etc. Then someone actually takes a look at what is causing the elderly to be hospitalized from medications. Turns out to be primarily two kinds of medications - blood thinners and medications to keep help control blood sugar.

    Quality needs to be questioned and measured in correlation to costs with objective data to improve care.

    1. Separating the recipient of the service from the one who pays for the service lessens the incentive to become an informed consumer. It doesn't mean that patients don't care about quality. They just believe that ensuring quality is someone else's responsibility. I'll illustrate with an example. I was waiting for my appointment at the Apple Store last week and I witnessed several elderly people getting service/instruction on their computers. They were very informed consumers. They knew how to operate the machines and asked very informed and educated questions. Why would someone spend more time researching a $1500 computer than they would researching their colonoscopy or their blood pressure medicine? Could it be that they perceive the cost of the former to be greater than the cost of the later?

  2. nice comments: I would love to hear some stories about your experience with customer satisfaction as a performance evaluation metric.

    For professors, customer evaluations are the best metric we have of teaching quality.

    1. I practiced ENT for 10 years and since tonsillectomies are a discussion item, I'll share my experience. I had a very successful referral-based practice and did a lot of tonsillectomies (several hundred annually). According to Facebook postings and word of mouth, the reasons that my patients were satisfied related to the cleanliness of my office, ease of parking, and mine and my staff's friendliness. Seldom did I see comments that had anything to do with my skill as a surgeon and NEVER did I have anyone comment on the cost-effectiveness of the care I provided or my complication rate.
      So, what's my point? Patients view healthcare as a commodity. They assume that a tonsillectomy at one location or by one surgeon is essentially the same "product" that they would get elsewhere. This clearly benefits providers, and not patients, as it allows achievement of high satisfaction scores that are often based on things that have nothing to do with the quality or cost-effectiveness of care.
      I don't propose to have all the answers but I do suggest that, if patients had to negotiate fees and pay out of their pocket for services like tonsillectomy, facilities and surgeons would be more inclined to advertise and promote things like cost, average recovery time, complication rate, etc. Since none of that really matters, all providers need to do is give patients a nice parking spot or a steak dinner and ensure that the "experience" is viewed as satisfactory. Surely there isn't anyone out there who would try to game such a system!

  3. Patient satisfaction like movie/restaurant reviews are highly subjective. The issues with these scores as mentioned above is that most of the times it will have nothing to do with the kind of care (timely medications, procedures, emergency interventions etc) that drive the cost.

    Another way that researchers are looking to improve quality and control cost is outcomes-based fee, where the hospital re-admission is a factor. Example: If a patient has been discharged from the hospital and is readmitted within 30 days, some sort of penalty will be levied against physician/hospitals. The issues with this kind of fee system might include, hospitals denying service to patients with multiple co-morbidities who are prone to get sick often.

  4. agree that One of the biggest problem with the US medical system is cost.