Validating Australian Government Medicare Benefits Schedule Data Against the National Joint Replacement Registry Data

• Main Authors: Nicole Pratt, Katherine Duszynski, Aarti Gulyani, Stephen E Graves
• Format: Article
• Language: English
• Published: Swansea University 2020-12-01
• Series: International Journal of Population Data Science
• Online Access: https://ijpds.org/article/view/1580

Introduction To enhance the value of the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) we recently linked these data with administrative datasets including, Medicare Benefits Schedule (MBS) and Pharmaceutical Benefits Scheme (PBS). Understanding the validity of administrative data is important in establishing the reliability of these data for informing clinical practice and policy. Objectives & Approach The study objective was to determine the validity of MBS data for capturing the occurrence of a joint replacement procedure. Using the AOANJRR procedures as the gold standard, we determined the sensitivity of the MBS data in correctly identifying hip joint replacement procedures. Results Of the 178,047 patients with a single primary total hip replacement occurring in a private hospital setting and recorded in the AOANJRR, 76% had a same-day MBS service claim indicative of that procedure, 2% had MBS procedures within +/- 7 days of the procedure while 18% had no MBS procedure codes indicative of a total hip joint replacement procedure. Of the procedures with no total hip MBS codes, 2% had MBS procedures codes indicating a total knee procedure, 1.7% had MBS procedure codes indicating a revision hip on that day and 13% of procedures had an in-hospital MBS hip anaesthetic administration code claimed on that day. Conclusion / Implications Given the increasing application of MBS data to describe health service use it is important to understand the validity of these data for identifying procedures undertaken in the private hospital setting. Using validated, gold standard data captured by the AOANJRR we identified that MBS data likely underestimate the occurrence of total hip replacement procedures. In addition, some MBS procedure codes are misattributed to other procedure types such as knee procedures and revision procedures.