It Matters What Is Measured

It Is Time to Eliminate the Review of Systems

It is time to eliminate the review of systems (ROS). Long revered as a central element of good patient care,¹ it may do more harm than good. There are 3 good reasons to rethink this time honored ritual.

First, the ROS is an undirected “fishing trip” for information that we hope will improve the diagnostic process. It is ironic that we scorn diagnostic testing, particularly laboratory testing and imaging, that randomly—rather than systematically—searches for information. Health care payers limit the number and types of laboratory tests that can be grouped into “panels” to be ordered with a single keystroke, and many imaging procedures must pass rigorous “necessity” criteria relevant to the clinical situation. While it may be important to ask specific questions to shed light on an evolving differential diagnosis or to clarify the status of a diagnosis, each question asked that does not support these efforts is a screening test—and one without a target. A complete ROS is the verbal equivalent of a whole-body computed tomography scan without the radiation.

Second, the ROS takes time that might be used more productively in other ways. To do “a complete 14-point ROS,” a physician must usually talk more than listen. Perhaps we should not be surprised that studies have shown that the time initially spent listening to our patients before interrupting is measured in seconds.² We should be talking less, and listening more.

The third and perhaps most important reason to reconsider the ROS is that payers are assigning value to it, when its value has not been demonstrated. Why? One reason may be that those submitting and those reviewing charges for services can hire individuals with little insight into the intricacies of the clinical encounter and train them to look at patient visit documentation and count the number of systems reviewed. But “not everything that can be counted counts, and not everything that counts can be counted.”³ The number of body systems reviewed in the context of a patient encounter simply does not count—for anything. But we adhere to the ritual because it is a reimbursement factor.

Evaluation and Management Documentation and Coding Need to Change

Health care measurement has dramatically increased recently, along with the time and attention required of physicians. But third-party payment for health care has long required measurement of the service provided. While procedural services are relatively easy to measure, cognitive services are not. Evaluation and management (E&M) coding for reimbursement is arcane, and documentation to allow coding requires an inordinate amount of time and effort that does not support good patient care. It can be argued that a focus on reimbursement has driven the evolution of the electronic health record to an extent that a focus on improving patient care is an opportunity lost thus far. We are being forced to record information that can be counted, not information that counts.

Primary Care Payment Reform Is Essential

Payment and documentation reform are urgently needed if family medicine as a discipline is to survive and thrive to meet the primary health care needs of the public. Current efforts at payment reform keep in place the existing E&M coding for most income. This system stifles innovation, limits care models, and inflates the cost of care. To the extent that reimbursement continues to be encounter-based, a more meaningful measurement system is urgently needed. Perhaps the elements that reflect care provided lie in the simple constructs of diagnosis or symptom; new or continued; and stable, improving, or worsening.

But all encounter-based reimbursement has significant limitations. Comprehensive primary care payment is essential to enable better care for patients, better health of the public, and greater value for the resources invested.

Measuring Value Is Limited to Counting What Can Be Counted

The focus on counting is now changing, at least in a small way, as we move toward value-based reimbursement. As a society we should be able to buy more health for the money we spend on health care, and payment systems are developing that purport to reward the value of care provided, not just the quantity of services provided. An industry has developed to support the dramatic expansion of measurement, but it is targeted almost exclusively at measuring health care processes and intermediate outcomes, not health. It is difficult to measure the impact of health care services on health. Not everything that can be counted counts, and not everything that counts can be counted.

Adding a component for value is admirable, but the continued expansion of disease-focused process and intermediate outcome measures threatens to replicate the mistakes of the past. In the recently released list of quality measures for family medicine to be used in the Medicare Merit-based Incentive Payment System, only 2 of 55 measure a health outcome. And aside from patient satisfaction, none attempt to measure the value of the basic pillars of primary care that are widely viewed as essential to maintaining and improving health: (1) first-contact care; (2) longitudinal continuity over time; (3) comprehensiveness, with the capacity to provide care for the majority of health problems; and (4) coordination of care with other parts of the health care system.

Research into Measurement Is Essential

In this issue of the journal, Etz et al⁴ report on the results of an open-ended, electronic survey of primary care clinicians addressing 2 questions: (1) How do you know good primary care when you see it? (2) What questions would you ask a practice to learn the extent to which it is helping to deliver health and wellness? Only 57.5% of responses could be categorized as consistent with currently used measures, and I argue that many of those are not commonly used. Among clinician responses, 42% could not be assigned measure-based codes, but rather reflected concepts considered important to the health of our patients that are currently not part of the measurement landscape. Primary care physicians understand that the value provided to a patient cannot be easily reduced to disease-focused processes and intermediate outcome measures. To support a fragile primary care infrastructure, rather than a relentless pursuit of ever more measures created in the existing paradigm, scholars and payers should be looking for fewer measurements that clearly reflect the care processes needed to support health. As articulated by Etz et al, we need less and more in primary care measurement. The current measures of the care we provide—and particularly the value of that care—are woefully inadequate. There is an urgent need to drastically change course.

It is time to abandon the ROS. It simply does not count and it should not be counted. Let us use our time in ways that actually might help the patient. Perhaps we could listen more without interrupting.