How accurate is the data provided to the Irish hip fracture database?

Abstract

National databases, such as the Irish Hip Fracture Database (IHFD), are known to contain inaccuracies. The coordination of services, national funding and future research rely on the integrity of the IHFD so as to avoid policy and budget planning being based on misrepresentative data. An audit was performed to assess the accuracy of the 2016 IHFD data collected in our trauma unit. The Hospital In-Patient Enquiry (HIPE)-recorded fracture classification, IHFD-recorded fracture classification and IHFD-recorded operation performed for each hip fracture patient was identified. Each variable was compared with the classifications and procedures derived by the authors. Two hundred fifty-two cases were identified. The HIPE-recorded fracture classification was accurate in 29% of cases, and the IHFD-recorded fracture classification in 61% (p < 0.001). The IHFD-recorded operation performed was accurate in 76% of cases. Thirty-six cases (14%) were omitted by HIPE, and eight (3%) from the local IHFD (p < 0.001). Errors resulted from poor documentation, in determining the presence of fracture displacement, prosthesis coating and intramedullary nail length. Diagnoses and procedures were identified from ward and theatre logbooks that the data coordinator was unable to record. The data coordinator in our unit doubled the accuracy of the data, and reduced the rate of omitted cases by a factor of 4.5. Accuracy levels would be significantly improved with clear, thorough documentation by the medical team following education of junior surgical trainees on hip fracture classification and procedural coding. Inaccurate data has a negative effect on hospital reimbursement and compromises the integrity of the IHFD.