Accuracy of self-reported diagnosis of hip replacement

To examine the accuracy and validity of self-report of hip replacement for osteoarthritis (OA).

We compared self-reported hip replacement and the reason for surgery in elderly, white women age > or =65 years from the Study for Osteoporotic Fractures cohort with medical records and pelvis radiographs. Women, followed for an average of 8 years at the fifth clinic visit, were asked about any hip replacements since baseline.

Among 7,421 women attending the fifth clinic visit, 347 reported 387 hip replacements. Radiographs and/or medical records were available for 316 self-reported hip replacements. Participants accurately reported that hip replacements were for arthritis or fracture, with 94.5% and 97.2% of these self-reported diagnoses, respectively, confirmed from medical records. However, 1 in 8 self-reported hip replacements were not attributed by participants to either arthritis or a fracture; of these, medical records indicated that 88% were for OA. Overall, 302 self-reported hip replacements (95.6%) were confirmed as hip replacements (for agreement with self-report, kappa = 0.95 [95% confidence interval 0.92-0.96]). Under-reporting of hip replacements compared with hip replacements seen on radiographs was minimal (0.28%).

Elderly women accurately reported hip replacements and whether the surgery was for arthritis or a hip fracture, although a small number of hip replacements for arthritis were not attributed to this diagnosis by the women. Because hip OA and hip fracture have very different determinants, epidemiologic studies that use self-reported hip replacement as an indicator for the presence of hip OA or as an outcome of hip OA should verify the underlying cause by asking about the reason for surgery.