gms | German Medical Science

Objectives/Interrogation: Predicting the outcome of a medical intervention can help both the patient and surgeon in the decision-making process for treatment. There is little knowledge on the predicting factors for subjective and clinical outcome, especially for proximal interphalangeal (PIP) joint replacement. Therefore, our study objective was to identify the preoperative predictors for subjective and clinical outcome in patients with primary osteoarthritis (OA) one year after PIP joint surface replacement surgery.

Methods: All PIP joint OA patients treated with a CapFlex-PIP implant (KLS Martin Group, Tuttlingen, Germany) were prospectively documented in a clinical registry. Sociodemographic, surgical and radiographic data were recorded at baseline. Patients completed the brief Michigan Hand Outcomes Questionnaire (brief MHQ) and rated their pain at rest and during activities of daily living on a 10-point numeric rating scale before and one year after PIP joint replacement. In addition, PIP joint range of motion (ROM) was measured. In order to identify predictors for the brief MHQ, pain at rest and PIP joint ROM one year after surgery, all baseline variables were entered into three multivariate linear regression models and described with R-square (R²) coefficients.

Results and Conclusions: One hundred and nineteen patients with 119 PIP joints and a mean (± standard deviation) age of 70 (±9) years were included in the analysis. Predictors of a higher brief MHQ score one year after surgery were a better baseline brief MHQ score, higher baseline pain level at rest, greater preoperative PIP joint ROM, less finger axis deviation, and if one of the ulnar sided fingers (i.e. ring and small finger) was affected (R² = 0.24). However, it was impossible to predict postoperative pain at rest (R² = 0.06) and PIP joint ROM (R² = 0.10) using baseline factors.

Subjective hand function measurements are more suitable for the prediction of postoperative outcomes after PIP joint surface replacement. Consequently, the decision for treatment should preferably be based on patient-reported hand function rather than clinical variables. The fit of our models was quite low. Further evaluation of additional predictors seems necessary in order to provide a robust predicting model to support the future decision-making process for PIP joint replacement.