Clinical innovations - With Innovation, Timing is Everything

I ought to know. My own career makes me an expert on innovation, bad timing, and failure.

As I described in Innovation-Driven Health Care: 34 Key Concepts for Transformation, I’ve been involved in more than one failed innovation. That’s innovation. You start. You fail. You start again. You fail again. And you keep going until you get it right.

Here’s a short list of my failed innovations;

• In the early 1970s, Doctor Russell Hobbie, professor of physics at the University of Minnesota, and I developed software for producing a differential diagnosis based on a patient’s age, gender, and abnormal laboratory tests. Eventually the program spanned 600 lab tests. Hobbie wrote the program in FORTRAN. He ran it on an early version of the Internet. The Department of Defense created the Internet. It placed it in a dozen or so university computing centers as a communications network fallback in case of a Soviet nuclear attack.. In our laboratory report, UNIPORT (Unified Presentation of Relevant Tests), we listed five diagnostic possibilities. The report included the right diagnosis 80% of the time – not great, but not bad, considering our limited database.

Our timing was bad for two reasons: 1) we had no other clinical information other than age, gender, and lab tests; 2) we overlooked medical legal and privacy issues. Nevertheless, our laboratory sent out 6 million of these reports before the program was abandoned after a national drug company acquired our lab. Today, our differential diagnosis program would fall into the “diagnostic support” category, would include clinical data extracted from claims data, would be available at the doctors’ fingertips at the point of care,, and would contain the “right” diagnoses nearly 100% of the time.

• In the late 70s and early 80s, I put forth the idea you could write software to measure a patient’s Health IQ. The health IQ would be like a person’s mental IQ, which, of course, was, on average, 100. Why not, I reasoned, compare a person’s normal values for blood pressure, body mass index based on height and weight, girth in inches, laboratory tests (sugar and blood lipids and lipid ratios), and personal and family history of stroke or heart attack, and then assign an HQ value. To arrive at the HQ, you would simply add or subtract from 100 and weight the value depending on whether it was suboptimal or greater than “normal.” If your HQ was less than 70, you were in retarded health, if more than 130, you were Olympic material. If it were less than 50, you ought to see a doctor immediately. To each report, we attached a personal letter explaining to people where they stood in terms of health and what they could do to improve their health.

The HQ succeeded in spotting patients with high blood pressure, obesity, diabetes, those with faulty lipids, those at imminent risk of a heart attack, and those who needed to see a doctor. In Oklahoma, we ran the HQ on more than 4200 state employees and found an unhealthy population – overweight, many pre-diabetic or diabetic, and nearly 9% at high risk for a heart attack. At least a half dozen patients suffered predictable heart attacks while the study was going on. Word quickly spread among employees that if your total cholesterol/HDL exceeded 13.5, you ought to see your doctors pronto .Furthermore, the HQ was cheap at $35 a patient. Some people who received the report came to me, carrying change purses, and offering me money to do an HQ on their relatives, presumably because their own health report meant so much to them. But alas, the timing was wrong. Industry was not yet in a mood to pay for broad scale health screening for employees, and in those screening programs that existed, mostly based on questionnaires, participants were more interested in social interaction and swapping health-giving advice than in hard data. Information privacy may have also been an issue.

• In the early 1990s, a hospital executive and I created a series of more than 150 bundled bills for common hospital procedures. The bills “bundled” fees for the hospital, physicians, and required doctor consultants. We backed the fees with re-insurance if something went wrong, and the bundled bill exceeded budget. We thought it was a great idea. It was clear and transparent, told patients in advance exactly what the bill would be, and we knew it was a model that could be used to compare prices between hospitals. But it failed. It turns out health plans’ data systems could not handle such a “radical” idea. The old approach, fee-for-service for every service under the sun on both sides of the hospital physician aisle, was more profitable. Besides, the health plans preferred to negotiate separately with hospitals and doctors. A “divide and conquer” strategy worked better for health plans. This failure, in short, was about perverse incentives.

Today, in the world of consumer-driven care, our bundled bill concept might be considered a revelation. It shows the power of competitive pricing, transparency, and bundling. I was thinking of this as I read an article by John Goodman, president of the National Center for Policy Analysis in Dallas (“Perverse Incentives in Health Care, April 5, 2007). Here are a few things Goodman had to say, “In a normal market, entrepreneurs in search of profit would solve problem (lack of efficient, high quality, low price care) by repackaging and repricing their services in order to make customer-pleasing adjustments... Take cosmetic and Lasik surgery, for example, in both markets, patients pay with their own money. They also have no problem finding what is virtually impossible to find for other types of surgery – a packaged price covering all aspects of the procedure. People can compare prices, and in some cases, quality.” Goodman goes on to say, the price for cosmetic surgery has declined, the price of Lasik surgery fell by 30%.

When it comes to innovation, don’t get too far ahead of the curve, but keep on trying anyway.