Doctors Are Dinosaurs In a High Tech World
New York Newsday, July 1, 1995
It was far easier to be a priest when parishioners couldn't read and weren't allowed to have Bibles, and it was easier to be a physician in the days when patients didn't arrive for a visit with three newspaper articles about the latest tests and treatments, a photocopy of the relevant anatomy from a medical textbook a print-out of their Medline search, and advice from their on-line ailment discussion group (not to mention a directory of medical malpractice lawyers). Patients often hear the latest medical news before their physicians, and certainly before many of those physicians have been able to digest its implications for their practice.
This week, for instance, a study confirmed that post-menopausal women taking estrogen have higher rates of breast cancer, leading to renewed debate about the explosive growth of estrogen prescriptions. Estrogen is just one example of the complexity of treatment decisions today.
Estrogen relieves hot flashes, reduces osteoporosis, colon cancer and heart disease, and improves sex, skin tone and mental clarity. On the other hand, estrogen increases the risks of uterine and breast cancer, and that possible side effects of blood clots, bleeding, weight gain, headaches, gallstones and irritability. A woman and her physician must factor into the decision to prescribe estrogen a woman's personal and family medical history, her risk factors for all these conditions, and her personal values and feelings about these risks and benefits.
Doctors aren't always very good at discussing these decisions with patients. They generally don't have the time or training to elicit values and feelings. A 1990 study showed that women were taking estrogen primarily to relieve hot flashes, which most male doctors consider trivial compared to the other risks and benefits.
Most doctors are also not very good at making these calculations, since they're able to take into account only a few of the hundreds of relevant facts. But computers have this capability in spades. That's one reason why, for all the moaning and groaning about the depersonalization of medicine, the deterioration of the doctor-patient relationship, and the attack of the killer HMOs, physicians who want to practice good medicine in the future, and to give good advice to patients, will have to be integrated into healthcare organizations, and will have to accept computerized decision-making.
Even the crudest of our current computerized diagnosis and treatment programs are more consistent and accurate than the decision-making of individual doctors. For instance, in the first two months of using their computerized drug prescription system, Abbott-Northwestern Hospital in Minneapolis flagged 1,500 physician medication orders as inappropriate. LDS Hospital in Salt Lake City found its new computerized drug interaction system reduced the incidence of adverse reactions from 5 percent to 0.2 percent of cases and reduced the post-surgical infection rate from 1.8 percent to 0.4 percent by systematizing administration of post-surgical antibiotics; these two innovations alone were estimated to have saved the hospital a million dollars a year. Now the hospital is struggling with how to handle the accumulating evidence that some physicians have practice patterns that are not only ineffectual, but actionable by the state medical board.
Today, good medicine can be practiced only by organizations. On their own, doctors vary wildly in their practice of medicine, and most err in the direction of doing too much, sometimes dangerously so. After all, doctors have traditionally been paid for doing more rather than less.
Doctors today are like auto workers forced to re-engineer every automobile they touch in accordance with the latest consumer polls, product testin~ and federal regulations-they don't, and can't, do it well. An industrial revolution is occurring in medicine, and its name is managed care. It may well sever the traditional doctor-patient relationship, just as ceramic factories severed the relationship between the local potter and his customers. But the result of managed care, like that of the industrial revolution, can be cheaper products meeting a higher, proven standard of quality.
Unfortunately, managed care will achieve this goal only with some help from the government. The nub of managed care is to create incentives that reward doctors and hospitals for doing less. When doctors and hospitals have been doing too much, managed care tends to ratchet them down to what they should be doing. HMOs have little incentive, however, to measure whether they're doing too little. The potential cost of malpractice verdicts doesn't outweigh the benefits from cutting costs.
The president's original health-care reform proposal addressed this problem in several ways. Under Clinton-style reform, HMOs would have been graded annually on how satisfied and healthy its clients were compared to the clients of other HMOs. These report cards would have allowed patients to make informed choices between different styles of care. Fortunately, in the absence of health-care reform, some states and employer groups are encouraging industry-wide report cards, but they are coming very slowly.
The Clinton administration also briefly flirted with the idea of "enterprise liability," which would shift malpractice liability from individual doctors to HMOs. To continue the industry analogy, the current malpractice system encourages victims of dangerous products to sue the individual assembly-line worker rather than the manufacturer.
If the HMOs were made responsible for the quality of care through regular report cards and enterprise liability, they would have far more incentive to standardize and improve the erratic quality of physician services. HMOs might discover, for instance, that allowing physicians to spend only 16 minutes per patient reduces the quality not only of communication but of care. The less time doctors and patients talk, the less likely doctors are to understand patients' lives, and the less likely patients are to understand their diagnoses and treatments.
But if HMOs were held accountable for the quality of patient communication and the outcomes of health care, they might also discover that many doctors aren't as skilled at communication as some other health workers. Medical education selects students for scientific prowess, not interpersonal competence, and devotes just 5 percent of its curriculum to interpersonal skills. Even the fraction of physicians who choose family or general practice are often poorly prepared to be effective frontIine "health educators." And even if all primary-care physicians were competent communicators, they make up only about a quarter of all doctors-far too few to fill the role of readily available health care counselors.
Nurses, on the other hand, are available in' the numbers required, and many studies have shown that they match or excel doctors in communication skills. At the cost of one doctor at ten minutes a visit, HMOs can provide two nurses for twenty-minute visits. Nurses can be quickly trained to, use computerized diagnosis and treatment programs to help them decide when a sore throat needs to be referred to a specialist as a possible throat cancer, and when it needs salt water. In fact, nurses using computers turn out to be better diagnosticians than doctors, because they are less like likely to override the computer's advice on the basis of their own idiosyncratic experience (or arrogance).
My rosy scenario has one major weakness. In most cases, idiosyncratic, anecdotal data is the only data we have about the best way to diagnose and treat disease. Again, the Clinton administration has encouraged expanding health services research, in order to determine which tests and treatments are effective, and to issue guidelines for practice. The meager budget of the flagship of this effort, the Agency for Health Care Policy and Research (AHCPR), is now targeted for "zeroing out" by the current Congress. Without the outcome data generated by the AHCPR, dangerous ineffective technologies and methods will proliferate, and we will all be held hostage to the increasingly limited abilities of physicians to synthesize medical knowledge for us.
Reasonably priced, high-quality, science-based medicine is actually
possible. We won't get there through blind faith in the old-fashioned free
market or that other dinosaur, the doctor-patient relationship, but we
may with the right combination of federal regulation, federal research,
computers, managed care and nurses.