CORR Insights®: Defining Patient-relevant Thresholds and Change Scores for the HOOS JR and KOOS JR Anchored on the Patient-acceptable Symptom State Question

Where Are We Now?

Historically, users of patient-reported outcome measures (PROMs) have focused on effect sizes, as measured using tools like the minimum clinically important difference (MCID), and substantial benefit or threshold benefits, such as the patient-acceptable symptom state (PASS), which focus on the final score or outcome [3]. Meaningful changes or thresholds in PROMS using these tools have been evaluated by anchor-based or distribution-based approaches [1]. When anchor-based approaches are used, PROM scores may be compared at a single timepoint between different groups. The determination of whether the between-group differences are large enough to care about can be made using one or more patient-assessment questions (anchor questions) that allow the patient to characterize whether a level of improvement is large enough to be clinically important. Alternatively, PROM scores can be compared longitudinally in a single group over time with the use of a patient-assessment anchor to evaluate the size of an observed change. By contrast, distribution-based approaches depend upon on statistical criteria to quantify the size of an intervention’s clinical impact [8].

There are limitations to each method. Anchor-based approaches can be subjective and depend on the validity of the anchor question itself. Distribution-based approaches are simple to use but often a smaller effect size is needed to reach the MCID. Thus, distribution-based methods have the potential to underestimate the MCID [2]. To date, there is no consensus on the best way to use these metrics to determine the efficacy of the operations we perform [5, 7]. In other words, we lack universal acceptance on a score change or final value that defines success after hip and knee arthroplasty. Likewise, the method of collection or number of time points necessary to detect an MCID is unclear. Could a single-item PASS question be an adequate PROM after hip and knee arthroplasty?

In this issue of Clinical Orthopaedics and Related Research®, Dekhne et al. [3] raise the question as to whether simply collecting a postarthroplasty PASS is a more efficient method of measuring success than collecting pre- and postoperative PROMs and then calculating MCIDs and substantial clinical benefit metrics. They found that the single-item PASS performed well compared with previously established MCIDs for the HOOS JR and KOOS JR [3]. Typically, again, the MCID is defined as the minimum change in PROM scores that patients perceive as beneficial or clinically meaningful [4]. The discovery that a single-item instrument could be comparable to traditional PROM collection may help to simplify measurement and reporting of clinical success after hip and knee arthroplasty. Based on this study, we learn that single-item anchors or questionnaires may be appropriate to verify success after hip and knee arthroplasty with a minimum collection burden to patients and providers.

Where Do We Need To Go?

We need to devise some sensible, data-driven national guidelines about how and when we need to use PROMs in patient care outside of research settings, if at all. Researchers can debate how and when to use them in the studies they perform, but in everyday practice, PROMs have real limitations: Their collection is time-consuming, and given the high proportion of patients who don’t report back after arthroplasty, loss to follow-up should cause us to question any claims based on aggregated PROM results where the length and completeness of follow-up is not reported.

Although the results from Dekhne et al. [3] affirm commonly held beliefs about hip and knee outcomes, there is no consensus on what PROM scores are needed to indicate surgery or validate clinical results. Furthermore, for patients in whom surgery is still indicated but because of the patients’ comorbidities the procedure may be associated with a high risk that the patient won’t improve, the thoughtless application of PROMs could result in stalled preauthorization processes, decreased payments, or other penalties tied to those PROMs. Patients with diagnoses associated with a high likelihood of poor improvement or death after surgery may be unable to find a capable surgeon or institution if PROMs dominate the decision-making process. For these reasons, we need to define whether PROMs are necessary for routine arthroplasty procedures outside of clinical research setting in the United States. If so, acceptance of which instrument or anchor and how we define success beyond the MCID or PASS are necessary for each procedure.

If PROMs are necessary for clinical research, minimum values on the numbers of measurements and collection points must be defined, but the burden of collection on patients and institutions should be minimized. In complex procedures such as revision arthroplasty, oncologic settings, or posttraumatic reconstructions, a clear difference with respect to PROMs in primary cases is expected for institutions and surgeons.

Measurable PROMs are leading outcome measures for clinical success and value after hip and knee arthroplasty despite their limitations. Nevertheless, we have yet to define the best PROM or collection method after hip and knee arthroplasty. Like in Dekhne et al. [3], previous research from the Veteran’s Administration has used a method to estimate MCIDs for HOOS/KOOS PROMs [6], but rather than using the PASS single-item survey instrument, the Self-Administered Patient Satisfaction Scale (SAPS) for primary hip or knee arthroplasty functioned as the anchor questionnaire. As we move forward, we need to clarify or reach agreement upon the most appropriate anchor question or battery of questions for research and or practice.

How Do We Get There?

The present study raises the novel concept that a single-point outcome measurement after surgery may be comparable to classic PROM collection and therefore may be applicable to most arthroplasty procedures in the United States that are not performed as part of a clinical research project. This offers an alternative with the least burden to patients and providers regarding data collection.

To use PROMs appropriately at the national level, we need to set expectations for PROM use with respect to research, clinical surveillance, and authorization of payment. In terms of research, we need to agree upon a best PROM (HOOS versus HOOS JR, for example) for most arthroplasty outcome studies and employ it across platforms. Thus, measurements will be comparable across institutions, registries, and medical publications. Likewise, further agreement on the best anchor instrument will help decrease the burden of collection and reporting. With respect to clinical surveillance, the same universal data could be used to follow the performance of an institution overall or that of a procedure across multiple hospitals, registries, or locations. Finally, PROM collection could be misused with regard to authorization of payment for hip and knee arthroplasty. PROM collection should not be a prerequisite for surgery or a hoop for providers to jump through before elective arthroplasty. Rather, institutions who collect PROMs and report their results beyond the threshold of minimal improvement should be recognized and rewarded.