Cancer Types


Site Map


Survival Predictions

Is it possible to make accurate survival predictions for individual patients with cancer? As discussed below (hospice Medicare benefit) it may be necessary to try.

Most patients do not want to hear these numbers, but the health care decision makers need to have some realistic idea about survival, which is the purpose of these tables. Studies show that doctors tend be overly optimistic in making survival predictions. Overall cancer statistics and survival by stage is here, most of the data below is for patients with advanced (metastatic or incurable) cancers. Read the general section about understanding statistics here.

  Karnofsky Score (performance score)
Karnofsky Score and Survival
Karnofsky Score Prior to Death for Cancer
Karnofsky Score Prior to Death in Hospice
Life Expectancy Tables
Medicare Hospice Study (Christakis)
  and survival curve and data by diagnoses
Survival Curve for Hospice Patients
Survival by KPS from NHS
Survival by Symptoms from NHS
Survival 1980 compared to 1990
Survival Expectations with Advanced Cancer
  Bone Metastases
  Brain (Glioma) class    and survival
  Brain Metastases
  Breast Cancer
  Liver Metastases
  Lung Cancer
A recent study from a clinic using palliative radiation developed tables to predict short term survival (go here.)
More data about survival in patients with advanced cancer in tables: #1 , #2#3

The Medicare Hospice benefit requires physicians to certify that the patient has only six months to live. This is of course impossible to predict accurately (and only the government would come up with something so stupid.) Many patients  get into hospice much later than might have been of benefit to them (and their family.) The GAO report from Sept. 2000, noted that in the 90's the average length of stay fell 20% from 74 days in 1992 to only 59 days in 1998. The government has responded to this by persecuting programs that have patients who "live too long" (since 10% of the patients live longer than 6 months.) It is useful for doctors to at least have some idea of average prognoses which is the purpose of the tables listed above.