The validity of self-report as a technique for measuring short-term complications after total hip arthroplasty in a joint replacement registry.

This study evaluated concordance between self-reports and surgeon assessments of short-term complications. A total of 3976 primary total hip arthroplasty patients consented for an institutional registry (5/2007-12/2008); 3186 (80.1%) completed a 6-month survey; 137 (4.4%) reported deep venous thrombosis, pulmonary embolism, major bleeding, fracture, or dislocation. Patients reporting complications were called. Positive predictive values and 95% confidence intervals (95% CI) for patient self-report were measured, using surgeon assessment for comparison: pulmonary embolism, 88.9% (95% CI, 78.4%-99.4%); dislocation, 81.1% (95% CI, 75.9%-86.5%); fracture, 73.7% (95% CI, 63.8%-83.5%); deep venous thrombosis, 69.7% (95% CI, 61.9%-77.5%); major bleeding, 32.0% (95% CI, 19.4%-44.5%); any bleeding, 88.0% (95% CI, 75.3%-99.9%). Of 97 confirmed complications, 64.95% presented to outside institutions. Registry data on self-reported complications may overcome limitations of traditional methods, but data should be interpreted cautiously. Concordance was high for PE and dislocation but low for major bleeding.