TY - JOUR T1 - A composite index for total hip arthroplasty in patients with hip osteoarthritis. JF - The Journal of Rheumatology JO - J Rheumatol SP - 347 LP - 352 VL - 29 IS - 2 AU - Jean Francis Maillefert AU - Alice Gueguen AU - Minh Nguyen AU - Laurent Berdah AU - Michel Lequesne AU - Bernard Mazieres AU - Eric Vignon AU - Maxime Dougados Y1 - 2002/02/01 UR - http://www.jrheum.org/content/29/2/347.abstract N2 - OBJECTIVE: We propose a composite index for considering total hip arthroplasty (THA) in hip osteoarthritis (OA). METHODS: We carried out a 3 year longitudinal study of patients with painful hip OA from 137 centers. Clinical data were collected at baseline and every 3 months; radiographs were taken at entry and each year. The decision to have surgery was made by the patient, the rheumatologist, and the surgeon, with no reference to outcome measures. Statistical analysis included discrete Cox analysis with time dependent covariates, on 3 month interval grouped data. The dependent variable was THA during the 3 months following the evaluated visit. Time dependent covariates collated at each evaluated visit included radiological joint space width (JSW), percentage decrease in JSW during the year preceding, patient's global assessment. Lequesne index, pain, and nonsteroidal antiinflammatory drug and analgesic intake. We compared 2 analyses differing in symptomatic variables entered: values obtained at a single time point vs mean values between 2 visits at a 3 month interval. Selection of the index was based on the best combination of variables to predict occurrence of THA. RESULTS: Of the 508 patients recruited for study, 42 were excluded. During the 3 year followup, 75 patients underwent THA. Symptomatic, therapeutic, and radiological variables were included in the index resulting from the 2 analyses. Based on the selected cutoff, the positive and negative predictive values for occurrence of THA in the 2 years following were 54.3 and 90.6%, respectively (single point model), and 52.9 and 86.7%, respectively (model using mean values of symptomatic variables between 2 visits). CONCLUSION: The poor positive predictive value of the composite indices obtained in this study suggests that there are other unmeasured factors determining access to surgery. On the other hand, the high negative predictive values suggest that these composite measures should be used by clinicians to determine which patients should not be referred to THA. ER -