Clinical investigation
Palliation
Accuracy of survival prediction by palliative radiation oncologists

https://doi.org/10.1016/j.ijrobp.2004.07.697Get rights and content

Purpose

To examine the accuracy of survival prediction by palliative radiation oncologists.

Methods and materials

After consultation of cancer patients with metastatic disease for referral of palliative radiotherapy, radiation oncologists estimated the survival of the patients. These were compared with the actual dates of death obtained from the Cancer Death Registry. The time to death from all causes was the outcome. The survival times were measured from the date of the first consultation at the palliative radiotherapy clinics.

Results

Six radiation oncologists provided estimates for 739 patients. Of the 739 patients, 396 were men and 343 were women (median age, 69 years). The median survival of all patients was 15.9 weeks. The mean difference between the actual survival (AS) and the clinician predicted survival (i.e., actual survival minus clinician predicted survival) was −12.3 weeks (95% confidence interval, −15.0 to −9.5) for the entire population. The mean difference was −21.9 weeks when the actual survival was ≤12 weeks, −19.2 weeks when the AS was 13–26 weeks, −9.7 weeks when the AS was 27–52 weeks, and +23.0 weeks when the AS was >52 weeks.

Conclusion

In this study, the prediction of survival by radiation oncologists was inaccurate and tended to be overly optimistic.

Introduction

Patients with advanced cancer and their families often request an estimate of the life expectancy to plan end-of-life issues and remaining time together. Medical professionals also rely on such estimates to guide appropriate clinical decisions, plan supportive care, and allocate resource use (1).

Clinicians are often overly optimistic in the survival prediction of terminally ill cancer patients (2, 3, 4). This may result in delaying referral to hospice and other end-of-life care (5, 6). Moreover, the choice of dose fractionation in palliative radiotherapy may be dependent on the survival prediction. No study has yet been done addressing the accuracy of survival prediction by palliative radiation oncologists.

The primary purpose of this study was to compare the clinician predicted survival (CPS) with the actual survival (AS) of patients with advanced cancer. The secondary objectives were to examine whether the CPS improved with clinician experience and the impending death of the patient.

Section snippets

Methods and materials

The Rapid Response Radiotherapy Program at Toronto Sunnybrook Regional Cancer Centre provides quick access to radiotherapy to relieve suffering and improve the quality of life of patients with advanced cancer. Patients referred have a pathologic diagnosis of cancer and documentation of metastatic disease by pathologic confirmation, clinical examination, or imaging studies. The Rapid Response Radiotherapy Program is staffed by six palliative radiation oncologists. After the consultation for

Results

A total of 739 patients (396 men and 343 women) were included in this study between January 1, 1999 and December 31, 2001. The median patient age was 69 years. The lung, breast, and prostate were the most common primary cancer sites. Close to 40% of patients had had significant weight loss (at least 10% during the past 6 months). The median Karnofsky performance score was 60 (Table 1).

Six radiation oncologists provided a total of 739 survival estimates. Their years of certification in Radiation

Discussion

Our study is the first to specifically evaluate the accuracy of the survival prediction by radiation oncologists. The radiation oncologists' survival estimates tended to be inaccurate in an optimistic direction. For patients with a short life span (≤12 weeks), the CPS estimates from the radiation oncologists were, on average, at least double the AS. For patients with a life span of 13–26 weeks, the predicted survival estimates were again overly optimistic. For an AS between 27 and 52 weeks, the

Conclusion

In our study, the prediction of survival by radiation oncologists was inaccurate and tended to be overly optimistic.

Acknowledgments

We thank Drs. Andrea Bezjak, Padraig Warde, and Ian Tannock for constructive comments and Ms. Melissa Mulder and Ms. Rama Thiagarajan for secretarial assistance.

References (21)

There are more references available in the full text version of this article.

Cited by (0)

Supported in part by the Michael and Karen Goldstein Cancer Research Fund, Department of Radiation Oncology, University of Toronto, and Toronto Sunnybrook Regional Cancer Center Radiation Program Fund.

View full text