Can Health IT Do More Harm Than Good?

Neuroscientists say the brain is a pattern recognition machine, which no doubt explains why experienced diagnosticians can often spot a disease in a patient even before a complete lab workup is done.

Unfortunately, the brain can sometimes be fooled into seeing a disease pattern when none really exists. And by extension, so can a computerized diagnostic program. Right now, it seems physicians, computers, and the general public may be unknowingly falling into this trap.

Medicalization? Says Who?

Medicalization refers to the mistake of turning normal variations in human physiology into disease. An age-related decline in male sexual drive, for instance, becomes erectile dysfunction (ED). Normal menopause-induced declines in estrogen levels and the subsequent change in libido become female sexual dysfunction (FSD). Of course, some patients' hormone levels are so low that they genuinely do constitute a disease. But Americans have blurred the lines between normalcy and pathology, which in turn leads to needless treatment.

What does this tendency have to do with health IT? Take osteoporosis and osteopenia as examples. Osteoporosis is clearly a condition that needs treatment because it causes spontaneous fracture; but osteopenia, sometimes referred to as pre-osteoporosis, is not. This relatively small drop in bone density may or may not eventually lead to symptomatic disease. And in a patient who has no other risk factors to suggest the eventual onset of osteoporosis, there's little need for drug therapy.

In Overdiagnosed, Making People Sick in the Pursuit of Health, Dr. H Gilbert Welch and his colleagues write about a healthy woman with osteopenia who was put on estrogen/progesterone therapy--because it increases bone density--only to discover down the line that the hormones probably increase the risk of breast cancer and stroke. A switch to a different medication brought on esophageal ulcers.

So what happens when a physician orders a bone densitometry test and the T score comes back at -1.5, indicating osteopenia? If he plugs that number into his clinical decision support software, depending on how sophisticated the program is, the readout may say: "Below population norm, consider bisphosphonate therapy."

And given the eight-minute window he has before seeing his next patient, there's a good possibility the doctor will take the easiest, least time-consuming approach and write a prescription for Fosamax or one of the other bisphosphonate drugs. Unfortunately, since there's no hard evidence to show that the benefits of these drugs outweigh the risks--at least in patients with osteopenia and no other risk factors--many experts recommend against this approach.

This same scenario is playing out over and over in hospitals and doctors' offices nationwide. We now have disorders like "prediabetes" and "prehypertension" in the medical literature which, if not carefully analyzed, will likewise encourage the needless prescription of medications that offer more risks than benefits.

If a patient's fasting blood glucose is above 125 mg/dl, he likely has diabetes. If it's below 100 mg/dl, he's considered normal. But if his reading comes back between 100 and 125 and the clinician plugs that number into the electronic health record--which in turn is linked to a clinical decision support system and a set of practice guidelines--the system may come back with a conclusion such as "Patient has impaired glucose tolerance, or 'prediabetes.' Prediabetes increases the risk of cardiovascular disease by 50%. Consider the following treatment options ..." When that does happen, you have to wonder: Are the IT tools you've put in place encouraging your doctors to diagnose a disease that will never happen?

That's not to suggest that some of these pre-diseases should be ignored. There are changes in diet and lifestyle that have the potential to slow down the onset of real diseases. But the operative word here is potential. It's far less dangerous to treat a potential threat of disease with diet, exercise, or stress management than to prescribe a drug known to cause real adverse effects.

And since only a very small percentage of patients with pre-disease ever become symptomatic, is drug therapy a wise decision--or just one more excuse to medicalize the normal aging process? If that's the case, health IT executives have to sit down with their medical colleagues and rethink the diagnostic cutoff points they're programming into the health care system.

The Healthcare IT Leadership Forum is a day-long venue where senior IT leaders in healthcare come together to discuss how they're using technology to improve clinical care. It happens in New York City on July 12. Find out more.