Know Your Chances

About the Charts

The numbers in the charts come from Federal government vital statistics records, the best available data for cancer and non-cancer deaths. The specific sources are the Surveillance, Epidemiology and End Results Program (SEER, National Cancer Institute) and the National Center for Health Statistics (Centers for Disease Control).

The charts are adapted from work originally published in the Journal of the National Cancer Institute:

Woloshin S, Schwartz LM, Welch HG. The risk of death by age, sex, and smoking status in the United States: putting health risks in context. J Natl Cancer Inst. 2008 Jun 18;100(12):845-53.

The website was created in collaboration with Lisa M. Schwartz, MD, MS and Steven Woloshin, MD, MS from the Dartmouth Institute for Health Policy and Clinical Practice, Lebanon, NH.

Return to Top

About the Numbers

Most of us aren't used to seeing numbers about the chance of dying. A 60-year-old man has a 4 out of 1000 chance of dying from colon cancer in the next 10 years. Is that a big or small chance? To decide, it helps to get perspective by comparing this chance with other chances. Comparing the chance of dying from different diseases helps you to appreciate which are the biggest threats you face (the ones you might want to do something about) and which threats are less worrisome. Compared to his chance of colon cancer, the 60 year old man's chance of dying from coronary heart disease in the next 10 years is much bigger (30 out of 1000); his chances of dying from prostate cancer (3 out of 1000) is about the same, while his chance of dying from rabies is much lower (less than 1 out of 1000).

Another important way to get perspective is to consider the chance of dying from anything (called "all causes" in the charts). It allows you to see how each individual cause of death contributes to the total. Because there are so many causes of death, you can't simply add up the numbers in each column of the charts. So we have a row across the top of the big picture charts with the chance of dying from all causes at each age.

The big picture charts are all about the chance of dying over the next 10 years. While the 10-year time frame is arbitrary, it makes sense to us. It's not too long (it's easy to imagine) and it's not too short (so the chances aren't forced to look too small). And it allows time to do things like change your lifestyle (most importantly not smoking) or consider proven screening tests - to lower your chances.

Changing the time frame matters a lot. The longer the time, the larger the chance. To see how chances change over time, you can create custom charts which provide the chances over the next 10 years, 20 years or for over a person's remaining life expectancy (how much longer someone that age is expected to live).

The your chances option lets you see the top causes of death for a person based on the sex, race and exact age you select. To show you how things change over time, we also show the top causes for the same person when they were 10 years younger and when they will be 10 years older.

And the special cancer tables give you a different kind of perspective: they let you compare the risk of diagnosis and death for a variety of cancers. For most diseases, getting the disease is much more likely than dying from it. Although many people believe cancer is a death sentence, this is far from the case. By comparing the risk of being diagnosed with a cancer to your risk of death from the cancer you can get a sense of how deadly the cancer really is.

Finally, the numbers in the charts are averages. In health statistics, age and sex are generally the most important predictors of what will happen to you. That's why the charts break down the chances by age and sex. The charts do not account for other important predictors - like a strong family history of a certain disease or various behaviors (like smoking) or exposures (working with asbestos) - which affect your chances too. When reliable methods for generating more personalized estimates become available, we will incorporate them into the charts.

If you want to read more about how to make sense of messages about chance, you can download the book "Know your chances: Understanding health statistic - How to see through the hype in the medical news, ads and public service announcements" (University of California Press) from PubMed Health's Understanding Research Results.

Return to Top

About the Cause of Death Categories

In most cases, we used standard National Center for Health Statistics' 113 most common causes of death categories as defined in the Deaths: Leading Causes reports available from the CDC's National Vital Statistics Reports. We modified standard categories, however, in several instances to enhance clinical meaning. For example, the NCHS 113 list includes the category "ischemic heart disease" which consists of deaths from acute myocardial infarction, other acute and chronic ischemic heart disease, and atherosclerotic cardiovascular disease. We modified the list to create a category called "coronary heart disease," which includes all the foregoing "ischemic heart disease" deaths plus deaths from associated complications (eg, congestive heart failure and arrhythmia). In addition, some causes of death have been removed from our list (e.g., "Certain other intestinal infections"). Refer to the Cause of Death Definitions for the ICD-10 codes used to define the causes of death included in this website.

Return to Top

About the Problems With Death Certificate Data

The accuracy of the charts ultimately depends on accurate attribution of death on death certificates and the methods used by the National Center for Health Statistics to assign the underlying cause of death. Assigning an underlying cause of death for individuals with multiple medical problems is particularly challenging since there is often uncertainty in the chain of events leading to death. It is likely that both under- and over-counting of deaths occurs. Diabetes and hypertension deaths, for example, are probably under-counted: they are often reported as contributing factors rather than as the underlying cause of death.

Return to Top

Methodology

DevCan- Probability of Developing and Dying from Cancer

Statistical models are used to compute the probability of being diagnosed or dying of cancer from birth or conditional on a certain age. DevCan statistical package is used to compute these probabilities. DevCan takes cross-sectional counts of incident cases from the standard areas of the Surveillance, Epidemiology, and End Results (SEER) Program conducted by the National Cancer Institute, and mortality counts for the same areas from data collected by the National Center for Health Statistics, and uses them to calculate incidence and mortality rates using population estimates from census data for these areas. These rates are converted to the probabilities of developing or dying from cancer for a hypothetical population. Please note that when the program refers to cancer or incidence, it is referring only to the cancer site that you requested.

To obtain additional information about the methodology, please access the following link: http://surveillance.cancer.gov/devcan/.

Return to Top

Data Source

Incidence data used in these charts were obtained from the 18 registries of the Surveillance, Epidemiology, and End Results (SEER) program for the special cancer section. While mortality data for the same areas collected by the National Center for Health Statistics (NCHS) were used for cancer death chart in the special cancer section. Three most recent years of diagnosis & death (2010-2012) were used for all the calculations.

Return to Top

Cause of Death Definitions

Return to Top