New comprehensive score for predicting difficulties in revision total hip arthroplasty.

Background: Revision total hip arthroplasty (THA) can be complex, and assessing possible difficulties is important to predict the operative time. No simple score for predicting difficulties has been assessed prospectively. We therefore developed an original score for the pre-operative evaluation of extraction and reconstruction difficulties. The objectives of this prospective study were to (1) assess correlations between score values and operative time, (2) determine whether the score predicted the need for revision implants and/or filling material, (3) determine whether the score predicted intra-operative and post-operative complications, and (4) evaluate the inter-observer and intra-observer reproducibility of the score.

Hypothesis: The score is reproducible and correlates well with the operative time, thereby allowing prediction of this parameter before surgery.

Material and methods: A prospective study of 103 revision THA procedures performed between March 2018 and August 2023 was conducted. The primary outcome was operative time and the secondary outcomes were use of a revision implant, use of filling material, and intra-operative and post-operative complications. The score was determined by four observers to allow evaluation of inter-observer agreement. Intra-observer agreement was assessed by having one of the observers determine the score a second time after inclusion of the last patient. The score has a maximum value of 20 and allows classification of the procedure as very difficult, difficult, and moderately difficult.

Results: Mean operative time correlated with the score value: 136.0 ± 33.9 min in the very difficult group, 102.0 ± 34.8 min in the difficult group, and 75.4 ± 65.5 min in the moderately difficult group (p = 0.0002). The score predicted the use of a reinforcement ring (40 procedures: 12/17 [70%], 11/25 [44%], and 17/61 [28%] in the very difficult, difficult, and moderately difficult groups, respectively; p = 0.01) and of a long stem (20 procedures: 8/17 [47%], 7/25 [28%], and 5/61 [8%] patients in the very difficult, difficult, and moderately difficult groups, respectively; p < 0.001). The score did not predict the use of filling material (42 procedures: 10/17 [59%], 9/25 [36%], and 23/61 [37%] in the very difficult, difficult, and moderately difficult groups, respectively; p = 0.250). The score predicted both intra-operative complications (5/17 [29%], 4/25 [16%], and 4/61 [6%] procedures in the very difficult, difficult, and moderately difficult groups, respectively; p = 0.028) and post-operative complications (4/17 [23%], 0/25 [0%], and 6/61 [9%] in the very difficult, difficult, and moderately difficult groups, respectively; p = 0.15). Inter-observer agreement was strong according to Landis-Koch criteria, with kappa values ranging from 0.70 to 0.79 [0.57-0.90]. The kappa value for intra-observer agreement was 0.74 [0.63-0.85].

Discussion: This score predicts surgical difficulties by adding criteria to bone destruction, in contrast to widely used classifications for revision THA. Moreover, the score is reproducible and predicts the operative time, thus potentially playing an important role during pre-operative planning.

Level of evidence: IV; prospective observational non-comparative study.