Evaluating Doctors on Their Performance with Patients

The Centers for Medicare and Medicaid Systems (CMS) and private insurers have identified numerous pay-for-performance measures, confirmed in large part by evidence-based medicine, and implemented them in hospitals, clinics, and doctors’ offices.

For example, there are process measures for hypertension and heart disease where, for example, periodic readings of patient’ blood pressure and blood sugar levels are reported. But testing does not tell you whether the patient is bringing these diseases under control. So intermediate outcome measures that focus on patients’ actual blood pressure and blood sugar levels are noted to indicate whether the patient is improving, stable, or deteriorating.

Then there are final outcome metrics that demonstrate what happened to patients who received treatment in hospitals, went to another facility, or returned home. Complications after surgery such as infections, strong reactions to chemotherapy that requires re-admission into hospitals for further treatment, or death are examples of such outcome measures.  CMS and private insurers identified scores of such measures as a basis for allocating or withholding payments to hospitals, groups of physicians, and individual doctors ( pay for performance of doctors)

As one would expect when attaching high stakes to metrics in a helping profession such as medicine where there are many stakeholders (e.g., insurers, employers, doctors, medical staff, patients) views on pay-for-performance measures diverge, especially since insurers have published “report cards” displaying rankings, percentages, and results of these different measures for organizations and individual doctors. Divergent views of performance measures and “report cards” are inevitable when one examines the complex terrain that physicians inhabit and the predicaments they inherently face: expertise is never enough, making decisions amid uncertainty is common, and dependence upon the patient for improvement is essential. No surprise, then, these metrics and their outcomes, thus far, have generated mixed reactions. See here, here, and here.

Many policy makers, administrators, and doctors are satisfied that the measures are consistent with findings derived from evidence-based medicine and their experiences with patients. They welcome efforts to raise the quality of care and reduce costs. While many primary care doctors do agree with the policy initiative, they still question the measures because they know that these metrics—even when evidence-based medicine endorse the measures–seldom pick up individual differences among patients who have breast cancer, heart disease, or diabetes.

Consider a pediatrician whose practice includes adolescent girls. He tells the story of what happened to him when one of his insurer’s measures of quality care is a requirement to test all sexually active girls for chlamydia, a sexually transmitted disease. Since insurers do not and cannot read every single medical chart, they use a proxy measure to determine whether a girl is sexually active. They check to see if patients take birth control pills. That proxy complicates matters greatly because the pediatrician’s patients don’t take those pills for contraception, but for acne and menstrual pain.

The pediatrician asks a colleague: “So do I skip the testing for Chlamydia and fail my quality standards?” he asked. “Or do I order a test that the patient doesn’t need and that will probably not be covered by her insurance?” He ended his story by saying: “I’m all for quality. I just don’t think this is quality.”

For another primary care physician who has been highly ranked in the past on these measures from the local health insurer, her current “report card” left her dismayed.

“The quarterly “report card” sits on my desk. Only 33% of my patients with diabetes have glycated hemoglobin levels that are at goal. Only 44% have cholesterol levels at goal. All my grades are well below my institution’s targets. It’s hard not to feel like a failure when the numbers are so abysmal.”

She pondered the report card and told a colleague: “[They] focus on diabetes in pristine isolation [when] my patients inconveniently carry at least five other diagnoses and routinely have medication lists in the double digits.”

Moreover, according to the doctor, one of her patients has diabetes that can be controlled but has failed to come into the office regularly even though staff had contacted her many times. The doctor told her colleague that the patient “just can’t afford to take that much time off from work.”  “Does that,” the physician asked the colleague, “make me a worse doctor?”

These recent policy efforts to create accountability through establishing performance measures and dispensing monetary rewards to those who reach the benchmarks end up steering clinical practice toward standardization. These physicians I described are caught in multiple predicaments. They are involved, unknowingly, in the  struggle that has gone back and forth in medical circles for decades over the worth of doctors’ intuitive judgments vs. empirical evidence when uncertainty reigns and the abiding quandary of being dependent upon patients for success when the metrics make the doctor wholly responsible for outcomes. And now there is a full-court press that policymakers, saturated with data, have mobilized to standardize medical decision-making to increase efficiency and effectiveness.

---