Editorials |

The Definition of Survival FREE TO VIEW

Robert L. Thurer, MD, FCCP (Boston, MA)
Author and Funding Information

Dr. Thurer is affiliated with the Beth Israel Deaconess Medical Center and Harvard Medical School.

Correspondence to: Robert L. Thurer, MD, FCCP, Deaconess Medical Center, 330 Brookline Ave, Boston, MA 02212-5400; e-mail: rthurer@bidmc.harvard.edu

Chest. 1999;116(3):593-594. doi:10.1378/chest.116.3.593
Text Size: A A A
Published online

Physicians, surgeons, and other health professionals often use survival data to guide the treatment of patients with cancer. Such data (especially when stratified by stage and cell type) allow the evaluation of competing therapies and help patients gain a perspective on their illness. In most instances, a statistical discussion between patients and caregivers occurs around the time of diagnosis, and cumulative survival data are cited. However, as the course of a patient’s disease unfolds, information related to the patient’s prognosis following an initial period of survival becomes more appropriate. In this issue of CHEST (see page 697), Merrill and colleagues provide us with such“ conditional” survival data for patients with lung cancer. From their analysis of data from the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute, one can predict, for example, the likelihood that a patient who presents for 2-year follow-up after lobectomy for early stage lung cancer will survive an additional 5 years. Clearly, such information allows for a more enlightened evaluation and discussion. It may also help determine how aggressively to intervene in other health problems.

Unfortunately, the manner in which the SEER data are collected does not provide as much information as we might like. Instead of using the International Staging System, patients are divided into localized, regional, distant, and unstaged disease. Whereas localized disease translates into stages IA and IB, patients with regional disease include those in stages IIA, IIB, IIIA, and IIIB, a heterogeneous group with differing cumulative survivals. In addition, observed survival (from all causes) rather than survival relative to the overall population is reported. This makes it seem that the disease causes more deaths than is actually the case.

Aside from the obvious scientific utility of this information, it is interesting to speculate on the use of survival data in discussing treatment options with an individual cancer patient. Patients with cancer often ask for survival statistics, and their doctors are more than willing to provide that information. In contrast, patients with equally serious benign illnesses (such as diabetic ketoacidosis or dilated cardiomyopathy, for example) are less likely to engage in these detailed statistical discussions. I believe that there are reasons for this discrepancy and that our patients’ welfare may be jeopardized by our complicity with their wish for numerical prognostications.

Statistical description may represent an attempt to gain a sense of control over illness when our therapeutic success is limited. Despite improvements in surgical therapy, better chemotherapy, and modern radiation oncology, the likelihood of successfully treating patients with lung cancer has changed little over the past decades. This frustration and seeming lack of ability to impact on the disease process may, I believe, lead to an emphasis on the statistical description of the illness. Concern over these often-negative statistics may serve to mask the positive interventions that can prolong life and ease suffering. A preoccupation with data may dehumanize rather than help patients and their loved ones.

These survival curves can also be destructive when they create (inevitably) a perception that cancer is an “all or none” disease. Patients and their doctors forget that many patients who are not“ cured” can in fact live healthy and productive lives with cancer. Newer therapeutic strategies such as antiangiogenic metalloproteinase inhibitors may help improve the lives of patients with metastatic disease.1 Treatment does not have to be curative to be effective. Neither coronary artery bypass surgery nor angioplasty nor medical therapies cure coronary artery disease. Insulin does not cure diabetes. However, there is no doubt that these therapies are worthwhile. They decrease suffering and prolong healthful lives.

When faced with nonmalignant chronic illness, caregivers are typically concerned with prolonging good health (diabetics are not only given insulin but are encouraged to lose weight, stop smoking, and exercise). However, when the chronic illness is cancer, it is often “survive or die.” Framing the disease in this way not only robs patients of the hope of living well with either persistent or recurrent disease, it may lead to inappropriate risk taking in an attempt at cure. In addition, interventions not destined to “cure” may be rejected even though they have a substantial chance of improving survival. The needs of patients who cannot be cured should be given more rather than less of our attention.

Most patients do not understand the implications of statistical analyses on their particular situation. Unlikely events may be thought of as impossible. A risk of 1%, for example, is usually discounted completely. When a low chance of survival is cited, the patients may see their situation as hopeless. They do not realize that the people who survive are 100% alive. When using numbers, we must be careful to relate them to a patient’s particular situation in an understandable fashion.2 We should discuss not only survival but also quality of life. Patients should be given the opportunity to contribute their perspectives as well.

Studies of large groups of patients such as this one by Merrill and colleagues, as well as detailed studies of smaller groups, are invaluable for improving the care of patients with cancer. For the individual, these data can be a helpful guide in facing life’s inevitable choices. However, we must be careful to interpret this information for our patients and to keep their overall best interests foremost in mind. Cancer, like life, is not an “all or none” phenomenon. As physicians, we have the privilege of helping patients deal with life’s complexity. Let’s not oversimplify the task.


ACKNOWLEDGMENT: The author would like to thank Drs. Daniel D. Karp and Robert G. Johnson for their thoughtful review of this submission.

Brown, PD, Giavazzi, R (1995) Matrix metalloproteinase inhibition: a review of anti-tumour activity.Ann Oncol6,967-974. [PubMed]
McNeil, BJ, Pauker, SG, Sox, HC, Jr, et al On the elicitation of preferences for alternative therapies.N Engl J Med1982;306,1259-1262. [PubMed] [CrossRef]




Brown, PD, Giavazzi, R (1995) Matrix metalloproteinase inhibition: a review of anti-tumour activity.Ann Oncol6,967-974. [PubMed]
McNeil, BJ, Pauker, SG, Sox, HC, Jr, et al On the elicitation of preferences for alternative therapies.N Engl J Med1982;306,1259-1262. [PubMed] [CrossRef]
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543