Registration in the Danish Hip Arthroplasty Registry - Completeness of total hip arthroplasties and positive predictive value of registered diagnosis and postoperative complications

Background There are few publications regarding the validity of data in hip arthroplasty registers. The Danish Hip Arthroplasty Registry (DHR) is a nationwide clinical database of THAs and revisions in Denmark. Patients and methods We assessed the completeness of registration of primary total hip arthroplasties (THAs) and revisions in the DHR from 1995-2000. In addition, the positive predictive value (PPV) of registered data for diagnoses in patients undergoing primary THA and postoperative complications was analyzed. Completeness was assessed using the Danish National Registry of Patients (NRP) as a reference, which is a nationwide and population-based registry of all somatic hospital admissions since 1977. The positive predictive value of registered data was assessed by review of medical records and preoperative radiographs from samples of randomly selected patients registered in the DHR using a standardized form. Results The overall completeness of registration for primary THAs and/or revisions was 94% (26 129 patients registered in both NRP and DHR and 27 757 patients registered in the NRP). There was a lower degree of completeness for revisions than for primary THAs (81% versus 94%). The completeness did not vary substantially according to gender and age. Completeness of registration was lower for university hospitals than for other hospitals (91% versus 95%), and for low-volume hospitals (87%) relative to those with medium (92%) and high volumes (96%). Overall, the diagnoses in patients undergoing primary THA could be confirmed in 84% (387/459) of the patients who were reviewed. The diagnosis of fresh fracture of the proximal femur was confirmed in only one third (22/73) of the cases. Postoperative complications were confirmed in two-thirds (36/54) of the patients reviewed. The specificity of the registration of postoperative complications was, however, only one-third (8/26). Interpretation We conclude that the Danish Hip Arthroplasty Registry is a potentially valuable tool for quality improvement and research due to the high degree of completeness of registrations regarding THA, and its moderate-to-high positive predictive value of registration for diagnoses in patients undergoing primary THA. However, information on several diagnoses for primary THA and on postoperative complications should be used with caution.