The problem of 'noise' in monitoring patient-based outcomes: generic, disease-specific and site-specific instruments for total hip replacement.

OBJECTIVES: To compare the performance of three types of patient-based health status instrument--generic, disease-specific and site-specific--in assessing changes resulting from total hip replacement (THR).
METHODS: A two-stage prospective study of patients undergoing surgery for THR involving an assessment at a pre-surgical clinic and a follow-up clinic at 6 months. 173 patients with a diagnosis of arthritis and being admitted for unilateral THR were recruited in the outpatient departments of a specialist orthopaedic hospital and peripheral clinics within Oxfordshire. Patients' health status was assessed using the 12-item Oxford Hip Score, the Arthritis Impact Measurement Scales (AIMS) and SF-36 general health questionnaire together with their surgeons' assessment using Charnley hip score obtained before and 6 months after surgery.
RESULTS: Effect sizes, used to compare change scores, revealed that pain and function domains changed most following THR on both the AIMS and the SF-36. 71 patients (41%) were assessed as having symptoms or problems currently affecting lower limb joints other than the hip recently replaced. Change scores were compared between these patients and all other patients who reported no current problems with other joints. The Oxford Hip Score found no significant difference between change scores for these two groups of patients while both AIMS and SF-36 physical and pain dimensions recorded significant differences of similar magnitude (physical P < 0.01, pain P < 0.05). Likely reasons for this were apparent on closer inspection of the item content of each instrument.
CONCLUSIONS: Assessment of outcomes in THR is necessarily long-term. Within studies of this kind, a hip-specific instrument (Oxford Hip Score) is likely to be more able to distinguish between symptoms and functional impairment produced by the index joint, as compared with other joints and conditions, than either a disease-specific instrument (AIMS) or a generic health status measure (SF-36). This is important given the high probability of existing and subsequent co-morbidity affecting such populations of patients. This consideration is likely to be relevant to any long-term assessment programme following treatment for a condition which threatens bilateral expression over time.