A Simple Way to Measure Health Care Outcomes - Harvard Business Review

Despite the current uncertainty surrounding the fate of the Affordable Care Act (ACA), health care leaders must not let debates over access detract from what needs to happen regardless of the legislation’s fate: Their organizations must improve the value of care they deliver. When the ACA was passed, in 2010, many observed that although the ACA expanded access to health care, it did less to address the equally critical issue of improving the delivery of that care. Regardless of whether the ACA is repealed, this challenge remains.

The success of efforts to improve care delivery hinges on developing clear approaches to performance measurement. A guiding principle offered by many has been to focus on improving the value of care delivered, as measured by the health outcomes achieved per dollar spent on care. Typically, value is expressed as a ratio: the quality of outcomes (adjusted to account for the severity of a patient’s condition) divided by the cost of treating that condition. Improving value therefore occurs by increasing quality relative to cost. For the most part, care organizations are better able to measure the denominator (cost) than the numerator (quality) of the value equation. As a result, they spend too much time focusing on costs and cost growth instead of value and its growth.

Our suggestion for a starting point to measure outcomes is simple, easy, and inexpensive: Ask the patient. For many years, health care providers have worked to collect so-called patient-reported outcomes measures, or PROMs, which are measures of function and health status reported by patients. The widespread and consistent use of PROMs, however, has proven to be elusive for a range of reasons, including the complexity of the measures tracked and the varying reliability of patient assessments on many measures.

Insight Center

As providers continue to refine their approaches to collecting PROMs, they should consider developing simpler approaches for capturing feedback from patients. We emphasize that these simpler methods would not displace PROMs, but rather serve as a complement to them. For example, Net Promoter Score (NPS) is a method of measuring customer loyalty or advocacy across wide array of industries, from hospitality to financial services to consumer technology. It uses a simple question of whether a consumer would recommend a product or service to a friend or colleague. NPS at the level of a unit or organization is then calculated as the percentage of respondents who are “promoters” less the percentage who are “detractors.”

In health care, NPS could also be calculated at the level of a given condition or provider within an institution. Asking even a single question to patients may help organizations understand the needs of each patient and identify opportunities for improvement that may not require much investment. In addition to being used for internal improvement, this data could inform consumers, employers, and insurers in their decisions to purchase care from particular providers.

“The Elements of Value,” an article in the September 2016 issue of Harvard Business Review, discusses the role that product or service quality plays in customer advocacy. The article states: “Across all the industries we studied, perceived quality affects customer advocacy more than any other element. Products and services must attain a certain minimum level, and no other elements can make up for a significant shortfall on this one.” What this suggests is that patient-reported satisfaction may be a reasonable proxy for quality. But what does it reliably tell us about clinical outcomes?

Another recent HBR article by authors from the Geisinger Health System, which offers a satisfaction guarantee to patients, acknowledges that “critics of programs that improve patient satisfaction will often imply that there is a false equivalence between efforts geared towards improving quality and those aimed at improving patient experience — that quality efforts are in some way superior.” The authors challenge this critique, noting, “[Recent] strong evidence [see this article and this one] suggests that improved patient satisfaction is in fact correlated with better health outcomes and quality: Increased satisfaction is associated with decreased length of hospital stay, lower readmission rates, reduced mortality, and fewer minor complications.”

We acknowledge that correlation does not imply causation. Rather, a high-quality outcome is the result of several factors and actions that, when properly aligned, result in patient satisfaction. In this case, satisfaction is meaningful not because it causes quality but because it may indicate that quality has been delivered.

For many types of care — especially care that is routine, highly standardized, and completed over a short duration — one may thus get significant mileage out of asking patients very simple questions such as, “At this point, how satisfied are you with the outcome of your care?” Consider, for example, a knee replacement or appendectomy. Given that a return to normal function is a key outcome for these treatments and an expectation for many patients, asking patients at predetermined intervals post-treatment to rate the quality of their outcomes may prove highly valuable.

The value of such patient feedback in primary care and the management of chronic diseases such as diabetes or hypertension could also be substantial. Beyond the value of the information itself, the process of collecting it, which could involve interacting with a health coach or even a mobile app, may have the additional benefit of helping patients with the often-elusive goal of better adhering to their disease management plans.

For other types of health care, it will be harder for patients to assess outcomes that may be subtle, technical, and reveal themselves over extended periods of time. The fact that such assessment is difficult, however, should not be used to dismiss efforts to achieve it. For example, asking a patient with lung cancer to rate their care may not capture the quality of the medical treatment received, as a significant portion of the relevant outcomes of that treatment emerge during the months and years following initial diagnosis. Asking that patient to rate care at set intervals, however, may help observe some of those emerging outcomes.

We often assume that health care is different from other consumer products or services and that this uniqueness prioritizes the value of clinical measures of performance over those related to patient experience. This assumption might cause one to measure smartphones primarily on battery life or cars on fuel economy. Yet other attributes of smartphones and cars, such a price and ease of convenience, matter too.

The view that health care is different misses the insight that it is different because it comes in a wide variety of forms — some routine, others quite complex. When it comes to measuring outcomes and value, not all health care falls into the “complex and hard to measure” bucket. Much of it is routine and predictable.

As such, the use of measures that we typically see in other consumer products and services may be relevant and helpful in measuring, and thus improving, value. In short, the fact that simple patient ratings of care may not be applicable for all types of care should not cause us to dismiss their use for any types of care.