Estimating mortality is an inexact science, but a recent report from researchers at UCSF (Cruz et al.; JAMA. 2013;309(9):874-876) suggests that a simple 12 question index may be effective at predicting an individual’s 10 year mortality. The “mortality index” was created from tracking over 20,000 adults who filled out a national health survey, approximately 6000 of whom subsequently died in the 10 year followup period. Interested in figuring out where you fall on the scale? Tally your points in the following way:
1. If you are a man, you automatically start with 2 points. Women start with 0.
2. If you are < 60, add 0 points. Men and women ages 60 to 64 get 1 point; ages 65 to 69 get 2 points; ages 70 to 74 get 3 points; and 85 or over get 7 points.
3. Add 2 points for each of the following questions you answer YES to:
- Do you have a current or previous cancer diagnosis, excluding minor skin cancers?
- Do you have lung disease which limits activity or requires oxygen?
- Do you have heart failure?
- Do you smoke?
- Do you have difficulty bathing?
- Do you have difficulty managing money because of health or memory problems?
- Do you have difficulty walking several blocks?
4. Add 1 point each for each of the following questions you answer YES to:
- Do you have diabetes or high blood sugar?
- Do you have difficulty pushing or pulling large objects, such as a heavy chair?
- Are you thin or normal weight (BMI < 25)?
Add up the total points (minimum score is 0 and maximum score is 26). Your risk of dying (expressed as a percent) in the next 10 years ranges from 2.8% with 0 points to 95% with 14 or more points (look up the percent associated with each point score on the table to the right).
Before you get too concerned about the results you or others may have obtained with this quick activity, let me clarify the methodology. Predictive indices are based on data garnered from large cohorts. Consequently, while they are effective at estimating average risk, they can never fully take into account individual variation. Researchers also did not calculate the effect of fluctuations in scores (i.e., quitting smoking, weight gains) on the mortality index so the index is static to health changes. Moreover, in older adults, low BMI (being thin) may be associated with illness, hence why it is actually a 1 point score in the index. However, for younger adults, lower BMI does not indicate illness and thus this index is not as useful for young, healthy adults. And finally, researchers caution that the utility of this index is for estimating whether older, sick adults may benefit from particular interventions that carry risk (certain cancer treatments and tests that have a delayed benefit and high immediate risk, for example). It is not meant to predict the date of one’s demise! On the other hand, if it can spur you to think critically about the health of you or your loved ones then perhaps the usefulness lies in behavior change and planning rather than absolute predictive ability.