Clinically Relevant Thresholds in Patient-Reported Outcomes: Do Patients' Expectations Evolve Over Long-term Follow-up? Response
Author(s):Walsh EG, Domb BG.
DOI Link: 10.1177/03635465251407254
Background
This correspondence addresses methodological considerations regarding the calculation of clinically relevant outcome thresholds—minimal clinically important difference (MCID), patient acceptable symptom state (PASS), and substantial clinical benefit (SCB)—in hip arthroscopy outcomes research, particularly their evolution over time.
Methods
The authors clarify their use of distribution-based MCID calculations in a longitudinal cohort study evaluating outcomes at 2-, 5-, and 10-year follow-up. They discuss methodological alternatives, including anchor-based and change-distribution approaches, and the associated statistical requirements for validity (e.g., ROC AUC performance).
Key Findings
Anchor-based MCID calculation was not utilized due to inadequate ROC curve performance (AUC < 0.7), limiting reliability of this method in the study cohort.
A distribution-based method (½ standard deviation of baseline scores, per Norman et al.) was used, consistent with common practice in the literature.
The authors acknowledge that different MCID methodologies (baseline distribution vs change distribution vs anchor-based) can yield variable threshold values.
They note that ~78% of prior studies in a referenced systematic review utilized a similar distribution-based approach.
The change-from-baseline distribution method may better capture temporal changes and time-dependent effects, but was not applied due to methodological familiarity and standard practice at the time of analysis.
Conclusion
The authors acknowledge limitations in MCID methodology selection and support further investigation comparing distribution-based and anchor-based approaches to better standardize clinically relevant threshold calculations in hip preservation research.
What Does This Mean for Providers?
Clinically relevant thresholds such as MCID are highly method-dependent and may vary substantially depending on statistical approach. Providers interpreting outcomes literature should recognize that MCID, PASS, and SCB values are not universal constants and may differ based on calculation method and patient population. Standardization of methodology remains an important area for future research to improve comparability across studies.