Chuck Klosterman, author of Sex, Drugs, and Cocoa Puffs, will tell you that you absolutely can compare apples to oranges. However, he’s never worked in health care. Sure, both apples and oranges are fruits—just like all patients are people. But picture this: two PT clinics—both alike in dignity, in fair USA, where we lay our scene. One clinic averages 10 visits per case; the other, 20. If that’s all you have to go by, then you probably assume the first clinic is much more efficient—and effective—at patient care.
Not so fast. After all, case and patient mix are critical factors. Perhaps the second clinic treats patients with more severe injuries or complicated pre-existing conditions. Thus, comparing clinics based solely on case-load efficacy is unfair. To truly measure the quality of care provided, one must examine the patient outcomes achieved.
But again, we run into an apples-to-oranges problem. Considering scores of regularly completed outcome measurement tools (OMTs) alone, one clinic may demonstrate faster functional improvement compared to other clinics. However, certain patients heal better than others. Let’s use two ACL reconstruction rehab patients as an example: one takes 10 visits longer to achieve the same level of function as the other. But that's not necessarily a reflection of the quality of the care each patient received, because one patient is 30 years older than the other and also suffers from knee osteoarthritis. Both patients receive excellent care, and both patients get better. But without taking into account these critical patient case details, all you have on paper is a visit count—and that’s not the sole way practitioners want to be judged.
Thus, if clinicians don’t take complicating factors into account, then the quality of their data, the justification of their care, and ultimately, the legitimacy of PT in general are all invalid. And that hurts bottom lines, as this study from the Journal of the American Physical Therapy Association explains: “assessments of provider performance that are tied to public reporting or financial incentives that are based on unadjusted outcomes may penalize providers treating the sickest patients who fail to show enough improvement or require more visits in a treatment episode.”
So, how do practitioners account for complicating factors consistently and at scale? Enter risk adjustment. According to this overview published on CMS’s website, “The purpose of risk adjustment when comparing outcome rates (e.g., hospitalization rates) for two different patient samples is to statistically compensate (or adjust) for risk factor differences in the two samples so that the outcome rates can be compared legitimately despite the differences in risk factors.”
Whereas unadjusted outcomes can lead to poor payment or penalties, risk-adjusted outcomes—and the valid data they provide—help level the data playing field. According to this risk adjustment primer by the Veterans Health Administration (VHA), “risk adjustment facilitates more rational resource allocation based on need. Most capitated payment methods use some type of risk adjustment to come up with payment amounts that take into account providers’ resource requirements for treating their patients.”
Furthermore, risk-adjusted OMTs allow healthcare practitioners and payers to establish baselines for tracking quality and efficacy over time, both internally and across facilities, regions, and the country. Similarly, such measures allow practitioners to prove the effectiveness of specific treatments or interventions. “Without adequate risk adjustment, it is impossible to say whether perceived improvements in patient outcomes reflect better treatment, healthier patients or other factors,” states the VHA primer.
To circle back to our two ACL patients, let’s assume the therapists regularly use the lower extremity functional scale (LEFS) to assess improvement throughout both episodes of care. LEFS accounts for age and other complicating factors. Thus, the therapists are able to better—and more fairly—assess and prove the efficacy of their care.
For all of the above reasons, OMTs must be risk-adjusted, which means test results take into account differing levels of patient complexity, such as age, weight, litigation, diabetes, cancer, and heart disease. (For a list of some of the most popular risk-adjusted OMTs, check out this post.) Patient complexity doesn’t end with the aforementioned comorbidities, though. According to the VHA primer:
Geographic factors, population densities and distances between facilities may have profound effects on practice patterns, access to care and patient outcomes. Coverage limits that differ across state medical aid programs may influence veterans’ illness severity, as well as the spectrum of services those veterans seek from VHA. In addition, recent research has shown that higher patient income is positively associated with better clinical outcomes.
And as the previously cited CMS overview states, “In general, risk factors for an outcome are chosen first by conceptually and clinically specifying the potential risk factors, and then assessing which ones are empirically related to the outcome.” (For a deeper dive into the methodologies and processes for developing those models, I recommend this CMS overview. While it focuses specifically on home health, its discussion of risk adjustment is very informative.)
Payment reform is happening, and the key to ensuring not only proper payment, but also physical therapy’s rightful place in the care continuum, is outcomes data. But it has to be the right outcomes data. And as Heidi stated in her recent founder letter, that’s data that is “measureable, comparable, actionable, and perhaps most importantly, meaningful to all stakeholders.” Risk-adjustment ensures the comparability of data, which moves PTs one step closer to proving their worth and effecting change with payers. To frame it Matt Damon-style, it empowers you to ask, “How you like dem apples?”