“Strive for five”—not just for the food pyramid anymore Why a flawed bell curve is skewing patient satisfaction scores by Erik DeLue, MD
Published in the September 2008 issue of Today's Hospitalist
IN THIS ERA OF INCREASING PHYSICIAN ACCOUNTABILITY—which, to my mind, is a good thing—it is rewarding to find that your patients view you as a “good” doctor.
Or so I thought. A quick lesson in Press Ganey, the most-used ranking system for patient satisfaction surveys, demonstrates how “good” can be so bad. On the survey, it turns out that a “very good” ranking translates to an absolute score of 100%, while a “good” corresponds to an absolute score of 80%—and puts a physician with that relative ranking in the bottom 1%. Myself,
“Give us a four, and we will all be poor.”
well, there have been times where I am sure the likes of Dr. Kevorkian would outscore me, this even with an absolute score that would make my parents proud.
A flawed bell curve
As many of us know, hospital CEOs and the like aren’t too interested in the argument that absolute scores matter more than relative scores. And it can be difficult to make a counter argument. Even if we all perform very well, it is unreasonable to believe we should all get A grades.
Thankfully, the converse is true as well—and if my organic chemistry professor hadn’t graded on a curve, I wouldn’t be writing this commentary today. Yet if we are to have real information that we can compare about our patients’ perceptions of physician performance, it can’t be flawed information in, flawed information out.
Let me explain what I mean. The “flawed information in” is the Press Ganey bell curve. In looking at physician scores on these patient satisfaction surveys, what stands out is how narrow the curve is that generates these scores.
You wouldn’t think that the relative difference between absolute scores of 82% and 85% would be statistically significant in terms of the overall quality of a doctor’s bedside manner. Yet on Press Ganey’s narrow absolute scale, that small difference—between a B and a high B minus, for those of us who kept score in high school—turns into the yawning gap between an absolute ranking in the bottom 10th percentile and one in the top 25th.
Then there’s the issue of “flawed information out”: No matter how many doctors care for the patient, the patient survey will affect the ranking of only the attending of record.
The car salesman’s approach
Which brings me to one of the new catch phrases in health care: “Strive for five.” That’s a reference to the five-point scale that patients use to grade our attentiveness and care.
I was first exposed to this phrase by way of a public service announcement targeting the nation’s obesity epidemic, which stressed the importance of eating five servings each of fruits and vegetables every day. But healthy eating has very little to do, literally or metaphorically, with this new “phrase that pays.”
All cynicism aside, I advocate for the consumer-driven model of health care as long as it leads to improvements in quality and effectiveness. Indeed, patients should have a voice in assessing their treatment, and their feedback can help bring about change for the greater health care good. And it may well be that Press Ganey is simply the best system available for letting patients’ voices be heard.
Yet with a bell curve this narrow, physician behavior modification may be less likely to result in better scores than in an attempt to work the system. Improving my group’s attentiveness to patients’ needs remains a major initiative, but working the system requires taking the same general tack used by a car dealer who wants his customers to know that scoring him a paltry 4 (good) on the inevitable customer satisfaction survey condemns him to the most thankless of car sales jobs—say, selling Hummers with gas at $5 a gallon. Hence, the five phone calls you get after you leave the dealership to ensure that you fork over the five-star ranking.
So we, like many hospitals and hospitalist programs, will continue to conduct this “strive for five”campaign. Our patient information flyer reads as follows: “If you believe that the care you have received has met your standards, we hope you will feel comfortable rewarding us with a five on the survey.”
The downside, of course, is that all our collective attempts to work with the system may eventually drive the bell curve even narrower, thus magnifying the damage of that rogue patient who believes his or her care was worth only a four. (I’ve actually enjoyed imagining anti-four slogans for use in our promotional material: “Give us four and we will all be poor.” Or “Giving me four is the same as a boot out the door.”)
What is the upshot of the new common ground that doctors and car salespeople share? We want feedback in the hope that it will make the patient, er, I mean consumer, more comfortable during the stressful period that is hospitalization. I actually believe this may be the end result, even if the means we use to get there do not do justice to the desired and extremely important end: improving the patient-physician relationship.
Erik DeLue, MD, is medical director of the hospitalist program at Virtua Memorial Hospital in Mt. Holly, N.J. Check out Dr. DeLue’s blog and others on the Today’s Hospitalist Web site at www.todayshospitalist.com.