The Pain That Results From Pain Measurement

At this point, most clinicians understand that the opioid epidemic is real and important. In the discussions of what happened and what to do about it, critics point at measurement of pain in the HCAHPS survey, and express concern that financial incentives for hospitals to improve patient experience are causing doctors to prescribe narcotics too liberally. Some critics of the Affordable Care Act extend this logic train to conclude that Obamacare is driving the opioid epidemic.

This blog posting is not about politics, however, nor is it really about financial incentives. It is about measurement itself — and the question of whether the very act of measuring anything in health care can produce potentially perverse consequences. The short answer to that question is “Yes.” The next question is what to do about it.

First, some comments on my potential conflicts of interest. Much of my career has involved measurement of quality and efficiency with the goal of improving performance — so I am pro-measurement in general, and familiar with the pushback that often follows. Currently, I work for a company, Press Ganey Associates, Inc., that does a great deal of measurement, including measurement of patient experience. Press Ganey is not the only company that measures HCAHPS for hospitals and its counterpart surveys for physicians and emergency departments, but Press Ganey is sufficiently well known that it is often blurred with HCAHPS. That assumption is wrong, but not important for this piece, since I am in favor of HCAHPS measuring pain control.

Why do I dismiss the importance of financial incentives for pain control? Quite simply, financial incentives for pain control don’t really exist for doctors, who are the ones who prescribe medications. There is no CMS incentive that gives doctors more money if their patients’ pain control is better, or that takes money away if it is worse. Nor do I know of any commercial insurers or provider organizations that tie physician compensation to pain control.

Hospitals do have a small amount of money tied directly to pain control for inpatients — but it is so small that it barely qualifies for the term “rounding error.” CMS ties 0.5% of payment to patient experience, of which there are nine components. Pain control thereby “drives” 0.055% of hospital payments. If any hospital passes along financial risk for pain control to physicians, it is spending more money on the accounting than is at stake.

Some physicians have incentives to improve patients’ likelihood to recommend their organization, so they might think that if they prescribe more narcotics, they might get better recommendations. However, research shows the opposite is true — physicians who give more narcotics actually tend to have slightly lower patient ratings. In any case, prescription data demonstrate that the opioid epidemic began in the late 1990s, long before HCAHPS began measuring pain in hospital patients (2006) and even longer before the first financial incentives for patient experience were introduced (2012). So the case that the HCAHPS or Obamacare are driving the opioid epidemic is not a good one.

The Paradox of Pain Measurement

Nevertheless, I don’t dispute that simply knowing that pain control is being measured can change the way physicians practice — and that this awareness can lead to better care in some patients and worse care in others. As one thoughtful (and angry) emergency department physician said to me, he knew that it was not true that physicians had a financial incentive to prescribe narcotics . . . “but,” he said, “it feels like it is true.”

The larger issue is that no one likes being measured, and therefore any measurement can produce perverse effects. Measuring surgical mortality can cause physicians to avoid surgery on high-risk patients, even if those patients’ outcomes might be worse with non-surgical care. Measuring my patients’ experience might nudge me to do tests and give medications that are marginally indicated. Measuring my efficiency and my antibiotic stewardship might nudge me in the opposite direction and cause me to withhold interventions that could be beneficial.

Name a performance measure, and I can give you a potentially perverse effect that might result from its use. Pain control is the topic of the day, but tomorrow it might be efficiency, or readmissions, or the percentage of patients who have surgery in the last week of life.

In a profession that believes “First, do no harm,” a logical conclusion is that measurement itself is the problem — at least for the gray-zone issues that are so common in medicine. These gray zones exist because there are so many competing values. It’s clear that giving every patient in pain a narcotic is a bad idea, but so is giving opioids to none of them. If there is no right answer, some might argue that society should trust physicians’ judgement, give them autonomy, and leave them alone. And stop measuring pain control and many other metrics.

Yet this approach would mean accepting that improvement is not possible — when we know that improvement is an imperative. There is tremendous variation in management of pain, and depressing evidence that non-white patients are less likely to receive adequate analgesia. Not measuring would mean accepting this variation.

There are unintended consequences with everything one does in life — including measuring pain and other issues important in health care. But there are unintended consequences with not measuring, too. When patients are suffering, physicians should respond. Sometimes that response will involve medications, including narcotics. But often that response should involve conversations with the patient to put their symptoms in perspective and ease the fears that can magnify their discomfort.

Should the financial incentives that exist for pain control be changed? Some advocates for the current financial incentives for pain control worry that no one will pay attention to this issue if there is no money tied to it. Incentives are always subject to tweaking, but in general, I favor financial incentives for financial issues (e.g., efficiency) and non-financial incentives (e.g., transparency) for quality — and that goes for pain, patient experience, and other outcomes that matter to patients.

On balance, we do not need less measurement in health care; we need more wisdom about what to do with the data.