Statistical analysis of failure times in total hip replacement
Prof. Dr. M. Schumacher, FDM, Institute of Medical Biometry and Medical Informatics
Dr. Th. Maurer, Department of Orthopedics, Kantonsspital Liestal, Switzerland
Prof. Dr. P. E. Ochsner, Department of Orthopedics, Kantonsspital Liestal, Switzerland
- Summary of the project
Since August 1984, data on primary total hip arthroplasty has been documented prospectively in a specialized hospital in Liestal, Switzerland. Data is collected in standardized questionaires with information on time to revision, time to death and various prosthesis- and patient-specific covariates. Examinations are scheduled at 1, 2, 5 and 10 years, postoperative.
Aim of the statistical analysis was the identification of potential risk factors responsible for hip prosthesis failure due to aseptic loosening of the stem. Various prosthesis- and patient-specific covariates were considered in the analysis. Primary interest was focused on the effect of stem type because an accumulation of failures was observed for a specific stem type. In a primary analysis (1,2), we considered only the failure time of the first implanted total hip prosthesis during the observation period to ensure independence of observations. Cox proportional hazard models for cause-specific hazard functions were used to determine the influence of potential risk factors on hip prosthesis failure and death before revision, respectively.
In a subsequent analysis (2), shared gamma frailty models were
used in order to account for potential dependencies in the failure
times of bilaterally treated patients. Results indicate that
failure times of bilateral hip prostheses could be treated as if
they behave independent when prognostic relevant factors are
considered in the analysis. However, this conclusion is presently
based on a relative small data set with few bilateral failures and
has thus to be validated in a much larger data set.
- Publications
(1) Maurer TB, Ochsner PE, Schwarzer G, Schumacher M
Increased loosening of cemented straight stem prostheses made from
titanium alloys. An analysis and comparison with prostheses made of
cobalt-chromium-nickel alloy,
International Orthopaedics, 25, 2001, 77-80
(2) Schwarzer G, Schumacher M, Maurer TB, Ochsner PE
Statistical analysis of failure times in total joint replacement,
Journal of Clinical Epidemiology, 54, 2001, 997-1003