Background/Purpose: The KSS questionnaire has previously been shown to have face, content and construct validity and test-retest reliability. Our objective was to assess the Responsiveness and Predictive Ability of the KSS.

Methods: We used data from the Mayo Clinic Total Joint Registry to assess the validity of KSS questionnaire, by including patients who underwent primary total knee arthroplasty (TKA) between 1993-2005 and responded to the baseline and 2-year post-primary TKA KSS questionnaire.  KSS questionnaire generates three domain scores: KSS function score (walking, stairs and use of knee supports), range of motion and KSS score (pain, range of motion and stability) scores. Convergent/divergent validity was examined with the association of select demographics (age, gender, number of joints involved) at baseline with KSS scores using linear regression and correlation analyses.  Responsiveness was assessed with Minimally Clinically Important Difference (MCID) and Really Important Difference (RID) were calculated corresponding to “somewhat better now” and “much better now” patient responses, respectively to the question at 2-years- Compared to your condition before the surgery, how would you rate your knee now?  For discriminant ability, we calculated effect size by taking the change in respective score from baseline to 2-years and dividing the result by the standard deviation at baseline. Predictive ability was assessed by association of improvement in KSS function (categorized as <=10, 11-40 and >40) and KSS scores at 2-years (minus preoperative scores; categorized as <= 35, 36-63, and 64-95) with revision surgery at time after 2-years. 

Results: For primary TKA, there were 5,280 knees with both a baseline and a 2-year data. The sample consisted of 2,375 males (45%) and 2,905 females (55%). The mean age at surgery (SD) is 68 (10), median age was 70 (range, 17-93).  

MCID and RID thresholds were as follows: KSS function score, 7.0 and 29.3; KSS score, 27.7 and 51.0; range of motion, 11.7 and 20.4.  Respective effect sizes at 2-years were large, 1.37 for function score, 2.60 for knee score and 1.56 for range of motion; and consistent with effect sizes for other validated scales for arthroplasty. 

Improvement in KSS-function score was associated significantly with the risk of revision after 2-years (Table 1); total KSS score showed a trend towards significant association (p=0.07)

Conclusion: KSS and KSS-function scores are responsive in patients undergoing primary TKA. KSS scores can predict the risk of early revision after primary TKA.  These findings imply that a regular monitoring of patient-reported outcomes of pain and function may allow signal detection for early TKA revision.

Table 1. Predictive ability of KSS score for revision TKA

	Hazard ratio (95% CI)	p-value
Improvement in KSS-function (n=4270 with 179 events)
<= 10 vs. 41-100	1.87 (1.23,2.83)	0.003
11-40 vs.41-100	1.18 (0.78,1.78)	0.43
Improvement in KSS (n=785 with 13 events*)
<= 35 vs. 64-95	16.63 (2.07,2150)	0.07
36-63 vs. 64-95	4.95 (0.56,658)	0.31

*186 TKAs did not have data to calculate KSS scores

---

« Back to 2013 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/responsiveness-and-predictive-ability-of-the-knee-society-scale-kss-score/