Keep the Customer Satisfied

Every so often I find this song playing in the back of my mind as I go about my work in the emergency department. Written by Paul Simon, it was recorded by Simon & Garfunkel and appears on the 1970 album Bridge Over Troubled Water. I was in junior high school when that album was released, so it didn't resonate with me the way it does now. In fact, it would be quite a few years before that phrase took on the meaning it now has for so many of us.

When I was in medical school and residency training there was, as far as I knew, no such thing as this notion of customer satisfaction in the practice of medicine. We were taught to be skilled and thorough in gathering data - medical history and physical examination - and to select the right diagnostic tests as we sought to discern the cause of a patient's symptoms. The goal was to arrive at the correct diagnosis and prescribe treatment that was safe and effective. Ideally, it should also be cost-effective. If, in addition, a doctor happened to have a good "bedside manner," that was icing on the cake.

Doctors didn't think too much about whether their patients found them likable. We all have different personality styles. A patient who doesn't care for one physician's personality style will choose a different doctor - and promptly be replaced by someone who does like the first doctor. It all evens out. Some patients really don't care whether a physician even has much in the way of bedside manner, as long as he or she is diligent, thorough, and competent. Frankly, that's the way I look at it when I'm a patient: if my doctor has the requisite knowledge and skill to figure out what's wrong with me and what to do about it, I don't care if he has the personality of a cigar store Indian.


Times have changed. Everyone is keenly focused on whether patients are satisfied as customers in the business of health care, and everyone wants to measure their satisfaction with the "patient experience." In many health care enterprises, doctors are financially rewarded for getting high scores on patient satisfaction surveys. And there's nothing like money to influence behavior.



This isn't really a new phenomenon, though. In 1990 I wrote an article for the trade publication Medical Economics entitled "Some Days I Feel Like I Work at K-Mart: How Consumerism Has Affected the Doctor-Patient Relationship." Recently I re-read it and was a little surprised that I had already begun to perceive these same challenges more than two decades ago.

Take a look at the photo to the left. Doctor and patient, right? The doctor is smiling and engaged in the interaction. It looks promising for patient satisfaction. Actually, this photo comes from the website of a financial services firm, the doctor is the customer, and the business wants to provide the capital she needs to run her practice. But this notion of customer satisfaction is now pervasive in medical practice, and if you want your doctor to be focused on the medicine and not worrying about getting her patients to think she is kind and caring, forget it. That's the old paradigm. (Forgive me for using that word, but send me a scolding e-mail if I ever - and I do mean ever - use the phrase "paradigm shift." That would be unforgivable.)

There can't be anything wrong with doctors being nice and patients liking them, can there? Of course not. So you can stop reading right now. Except for one thing. I think you've heard of it. It's called the Law of Unintended Consequences. Because of this law, good intentions sometimes fail to produce good results.


When you start trying to measure customer satisfaction and give doctors a financial incentive for getting high marks from their patients, you'd better pay attention to what you're measuring and what the doctors think they have to do to improve their scores.


I believe the biggest problem facing us in this realm is that in many other spheres of their lives doctors are customers, and so they have a sense of customer satisfaction from that side of the transaction. What does it take to create a satisfied customer? It's quite simple. Figure out what the customer wants and give it to him.

So the patient has cold symptoms or a sore throat. The doctor thinks the patient must want an antibiotic. Why else would he be consulting a physician? To be told it's a cold and advised to go to the drugstore and buy an off-the-shelf medicine to treat the symptoms? The patient could have just watched TV commercials or asked the druggist at his local pharmacy if he wanted that. No, he must want something by prescription. So he gets an antibiotic for his head cold, sore throat, or chest cold, some illness caused by a virus. Antibiotics don't work against viruses, only bacteria. So there is no potential for benefit - only harm. But the illness gets better on its own, because that's how these things go, and the patient, who did not go on to have eternal bronchitis, figures the antibiotic must have helped. So what will he do every time he gets a viral respiratory infection henceforth and forever? Right. And all those unnecessary antibiotic prescriptions will help promote the development of resistant bacteria.


The doctor who wants to practice good medicine and still get high patient satisfaction scores can just take some extra time and explain to the patient why antibiotics are not indicated, how they are more likely to harm than help, and how they contribute to the problem of resistant bacteria. And then he can hope that his scores will be just as high as those of his colleague in the next office who just writes the antibiotic prescriptions, taking a fraction of the time and mental effort. Good luck with that.

And then there is the challenge of diagnostic testing. Patients like tests. They believe in tests. They have no idea that a smart doctor can often figure out what's wrong with them by eliciting a detailed account of their symptoms and performing a focused physical examination, without ordering any tests at all. Most of the time, test results serve primarily to confirm the diagnosis that was already established in the physician's mind.

For a recent study, investigators queried patients visiting an emergency department with abdominal pain about their confidence in the doctor's diagnosis and their overall satisfaction with the "patient experience." Confidence and satisfaction were low (shockingly low, I thought) when the doctor took a history and examined the patient but ordered no tests, which surely reflects the physician's belief that the diagnosis was simple and straightforward. Both confidence and satisfaction rose steadily as the number of tests increased, with the highest levels reached when the evaluation included a CAT scan.



What this study demonstrated, however, was something emergency physicians already knew intuitively. Abdominal pain is the single most common reason for visits to the emergency department. Many CAT scans are performed. How many of them are really necessary? How many are performed in the pursuit of diagnostic certainty? (And how useful is it to raise diagnostic certainty from 93% to 97%?) How many are ordered because the doctor is afraid of being sued if she misses something? How many are ordered because it will improve customer satisfaction? How would we know? What we do know, however, is that CAT scans are expensive and expose patients to radiation and sometimes to intravenous contrast agents that occasionally cause harmful reactions.

I don't know about you, but I want my doctor to order a CAT scan only if it is really necessary, and certainly not because he thinks I will have more confidence in his diagnosis and a higher level of satisfaction with the "patient experience" if he orders more tests.

Doctors now live and work in a world in which health care managers worship at the altar of customer satisfaction. There are serious problems with how patients' satisfaction is measured and how the results are interpreted. We are just beginning to understand these problems.

And yet, at a time when we should still be very concerned about our ability to define and measure patient satisfaction, to figure out what to do with the results when we do measure it, and to prevent unintended consequences, the federal government is already implementing a system that will financially punish hospitals that don't get high scores.

None of us should be surprised that public policy is being formulated and implemented without good science to support it. But we should be worried.