Competing events influence estimated survival probability: when is Kaplan-Meier analysis appropriate?

Competing risks survival analysis applied to data from the Australian Orthopaedic Association National Joint Replacement Registry

BACKGROUND AND PURPOSE

The Kaplan-Meier (KM) method is often used in the analysis of arthroplasty registry data to estimate the probability of revision after a primary procedure. In the presence of a competing risk such as death, KM is known to overestimate the probability of revision. We investigated the degree to which the risk of revision is overestimated in registry data.

PATIENTS AND METHODS

We compared KM estimates of risk of revision with the cumulative incidence function (CIF), which takes account of death as a competing risk. We considered revision by (1) prosthesis type in subjects aged 75–84 years with fractured neck of femur (FNOF), (2) cement use in monoblock prostheses for FNOF, and (3) age group in patients undergoing total hip arthroplasty (THA) for osteoarthritis (OA).

RESULTS

In 5,802 subjects aged 75–84 years with a monoblock prosthesis for FNOF, the estimated risk of revision at 5 years was 6.3% by KM and 4.3% by CIF, a relative difference (RD) of 46%. In 9,821 subjects of all ages receiving an Austin Moore (non-cemented) prosthesis for FNOF, the RD at 5 years was 52% and for 3,116 subjects with a Thompson (cemented) prosthesis, the RD was 79%. In 44,365 subjects with a THA for OA who were less than 70 years old, the RD was just 1.4%; for 47,430 subjects > 70 years of age, the RD was 4.6% at 5 years.

INTERPRETATION

The Kaplan-Meier method substantially overestimated the risk of revision compared to estimates using competing risk methods when the risk of death was high. The bias increased with time as the incidence of the competing risk of death increased. Registries should adopt methods of analysis appropriate to the nature of their data.

Estimating implant survival in the presence of competing risks

In medical research, commonly, one is interested in the time to the occurrence of a particular event, such as the revision of an implant, and the analysis of these data is referred to as survival analysis. However, for some patients, the event is not observed and their observations are censored. These censored observations are particular to survival data and require specific methods for estimation. The Kaplan and Meier method is a popular method to estimate the probability of being free of the event over time and it is now widely applied in orthopaedics such as to report implant survival. However, one of the assumptions underlying the Kaplan-Meier estimator implies that patients whose observations are censored have the same risk of occurrence of the event than patients remaining in the study. However, because the revision of an implant cannot occur after a patient dies, and that dead patients have their observations censored in the Kaplan-Meier method, another setting must be considered. In the sequel we will demonstrate the inadequacy of the Kaplan-Meier method to estimate implant survival and detail the cumulative incidence estimator.

Results of 32 allograft-prosthesis composite reconstructions of the proximal femur

The use of allograft-prosthesis composites for reconstruction after bone tumor resection at the proximal femur has generated considerable interest since the mid1980s on the basis that their use would improve function and survival, and restore bone stock. Although functional improvement has been documented, it is unknown whether these composites survive long periods and whether they restore bone stock. We therefore determined long-term allograft-prosthesis composite survival, identified major complications that led to revision, and determined whether allograft bone stock could be spared at the time of revision. We also compared the radiographic appearance of allografts sterilized by gamma radiation and fresh-frozen allografts. We retrospectively reviewed 32 patients with bone malignancy in the proximal femur who underwent reconstruction with a cemented allograft-prosthesis composite. The allograft-prosthesis composite was a primary reconstruction for 23 patients and a revision procedure for nine. The minimum followup was 2 months (median, 68 months; range, 2-232 months). The cumulative incidence of revision for any reason was 14% at 5 years (95% confidence interval, 1%-28%) and 19% at 10 years (95% confidence interval, 3%-34%). Nine patients (28%) had revision of the reconstruction during followup; four of these patients had revision surgery for infection. Allografts sterilized by gamma radiation showed worse resorption than fresh-frozen allografts. Based on reported results, allograft-composite prostheses do not appear to improve survival compared with megaprostheses.

LEVEL OF EVIDENCE

Level IV, therapeutic study. See Guidelines for Authors for a complete description of levels of evidence.

Statistical analysis of arthroplasty register data

Data from arthroplasty registers are often analyzed using survival methods. Several methodological problems exist, for example relating to competing events, non-random censoring, non-proportional hazards and dependent observations. League tables and ranking of specific survival results leds to further methodological difficulties. Most of these problems are, however, well known and a number of methods for dealing successfully with the problems have been developed. These methods are usually accessible in commercially available statistical software packages. The statistical analysis and reporting of data from arthroplasty registers can thus be improved. Development of arthroplasty register guidelines for statistical analysis could play an important role in making these registers even more useful.

Ipsilateral femoral autograft reconstruction after resection of a pelvic tumor

BACKGROUND

Reconstruction of bone after the resection of a pelvic tumor is challenging. The purpose of the present study was to evaluate the use of the ipsilateral femur as the graft material for reconstruction.

METHODS

We performed a retrospective review of thirteen patients with a malignant pelvic lesion who underwent resection followed by reconstruction with an ipsilateral femoral autograft and insertion of a total hip replacement. The study group included nine men and four women with a median age of fifty-one years at the time of the reconstruction. The diagnosis was chondrosarcoma in eight patients, metastasis in three, and myeloma and radiation-induced malignant disease in one each. The surviving patients were assessed functionally and radiographically; the cumulative probability of revision was estimated while taking into account competing risks.

RESULTS

The median duration of follow-up was forty-nine months. At the time of the latest follow-up, seven patients were alive and disease-free and six had died from metastatic disease. Four patients had had revision of the reconstruction, two for the treatment of mechanical complications and two for the treatment of infection. Three other patients had mechanical complications but had not had a revision. The cumulative probability of revision of the reconstruction for mechanical failure was 8% (95% confidence interval, 0% to 23%), 8% (95% confidence interval, 0% to 23%), and 16% (95% confidence interval, 0% to 39%) at one, two, and four years, respectively.

CONCLUSIONS

Although it has attendant complications consistent with pelvic tumor surgery, an ipsilateral femoral autograft reconstruction may be an option for reconstruction of pelvic discontinuity in a subgroup of patients following tumor resection. This innovative procedure requires longer-term follow-up studies.

Predictive accuracy and usefulness of calibration of the ESC SCORE in Switzerland

BACKGROUND

Prevention of cardiovascular disease (CVD) at the individual level should rely on the assessment of absolute risk using population-specific risk tables.

OBJECTIVE

To compare the predictive accuracy of the original and the calibrated SCORE functions regarding 10-year cardiovascular risk in Switzerland.

DESIGN

Cross-sectional, population-based study (5773 participants aged 35-74 years).

METHODS

The SCORE equation for low-risk countries was calibrated based on the Swiss CVD mortality rates and on the CVD risk factor levels from the study sample. The predicted number of CVD deaths after a 10-year period was computed from the original and the calibrated equations and from the observed cardiovascular mortality for 2003.

RESULTS

According to the original and calibrated functions, 16.3 and 15.8% of men and 8.2 and 8.9% of women, respectively, had a 10-year CVD risk > or =5%. Concordance correlation coefficient between the two functions was 0.951 for men and 0.948 for women, both P<0.001. Both risk functions adequately predicted the 10-year cumulative number of CVD deaths: in men, 71 (original) and 74 (calibrated) deaths for 73 deaths when using the CVD mortality rates; in women, 44 (original), 45 (calibrated) and 45 (CVD mortality rates), respectively. Compared to the original function, the calibrated function classified more women and fewer men at high-risk. Moreover, the calibrated function gave better risk estimates among participants aged over 65 years.

CONCLUSION

The original SCORE function adequately predicts CVD death in Switzerland, particularly for individuals aged less than 65 years. The calibrated function provides more reliable estimates for older individuals.

Function, disability, and health-related quality of life after allograft-prosthesis composite reconstructions of the proximal femur

BACKGROUND

Factors affecting function, disability and health-related quality of life after allograft-prosthesis composite reconstructions have been poorly studied.

METHODS

Retrospective study of eighteen patients who underwent reconstruction with proximal femoral allograft-prosthesis composites and answered questionnaires to assess hip function (Postel and Merle d'Aubigné, self-reported Harris Hip Score), disability (Toronto Extremity Salvage Score), and health-related quality of life (SF-36).

RESULTS

The median Toronto Extremity Salvage Score was 76 (IQR: 48-85), the median self-reported Harris Hip Score was 90 (IQR: 48-95) and the median Postel and Merle d'Aubigné score was 17 (IQR: 12-17). Older age, female gender, and presentation with a pathologic fracture were associated with increased disability and poorest function. The median Physical Component Summary score was 44 (IQR: 39-45) and Mental Component Summary scores were 49 (IQR: 46-56). Male gender and recurrence of disease were associated with poorer health-related quality of life. There was a high degree of correlation between function, disability scores and Physical Component Summary score.

CONCLUSIONS

Patients' characteristics at presentation such as age, gender, and occurrence of a pathologic fracture play an important role in determining disability, function, and health-related quality of life after allograft-prosthesis composite reconstruction of the proximal femur.