gms | German Medical Science

Objectives: The HOOS-12 is a recently published shortform of the widely used Hip Osteoarthritis Outcome Score (HOOS). It comprises of 12 questions from the original 40-item HOOS and provides Pain, Function and Quality of Life (QoL) scales as well as a summary hip impact score. While the HOOS-12 has been validated by its originators, to our knowledge it has not yet been externally validated. Furthermore, the originators proposed validation of the score in a non-US cohort. We therefore aimed to evaluate responsiveness, convergent construct validity, internal consistency and floor and ceiling effects of the HOOS-12.

Methods: Patients who received THA in one of three german hospitals as part of the multi-center prospective PROMISE study following ERAS principles were eligible for the study. Mean changes between baseline and one year postop. were determined to assess responsiveness. Internal consistency was assessed using Cronbach's alpha. Pearson correlations between HOOS-12 scales and full length HOOS and the EQ5D-5L were calculated for the assessment of construct validity. Percentages of patients with the maximum and minimum score at baseline and at one year postoperatively were calculated for the assessment of floor and ceiling effects.

Results: A total of 935 patients who underwent THA between May 2018 and March 2020 were included in the study (413 male, 522 female; mean age 66.0 (SD 10.6); mean BMI 28.3 kg/m2 (SD 5.5)).

Internal consistency of all three HOOS-12 subscales was high (>0.7).

Floor effects (percent with the lowest (worst) possible score) for all measures were very low (<2%) pre- and post-THA. Ceiling effects (percent with the highest (best) possible score) were negligible at baseline. Post-THA, there were notable ceiling effects for all subscales (Pain 52.8%; Function 40.3%; QoL 23.2%).

Tests of construct validity supported convergent and discriminant validity of HOOS-12 at baseline and one year postop. The correlation of HOOS-12 Pain and HOOS Pain was very high (r=0.90), indicating that all reliable variance in the full-length HOOS Pain scale was captured by the HOOS-12 Pain scale. The HOOS-12 Function scale also had high correlations with HOOS ADL (r=0.89) and Sport/Recreation (r=0.74). Moderate correlation was found between HOOS-12 Pain and EQ-5D Pain (r=-0.58) and HOOS-12 Function and EQ-5D ADL (r=-0.59).

Conclusions: The HOOS-12 showed good convergent construct validity and is responsive to changes in pain, function, QoL and knee impact between preoperatively and 1 year postoperatively. While it does not exhibit a floor effect, there is a substantial ceiling effect for all subscales at 1 year postoperatively.