Health Statistics -- View Them with Caution

Chrome 2001
.
Aetna Intelihealth InteliHealth Aetna Intelihealth Aetna Intelihealth
 
.
. .
Harvard Medical School

   Advertisement
Carepass Ad Carepass Ad .
Chrome 2001
Chrome 2001
.

Health Statistics -- View Them with Caution


Beth Israel Deaconess Medical Center

Surely, you've heard statistics like these before: 

            "A crime takes place on the streets of our city every 8 minutes."             

            "Every 10 minutes a new name appears on the organ transplant waiting list." 

            "One American dies every 12 minutes from lack of health insurance." 

            "If you're a woman, you have a 1 in 8 lifetime risk of breast cancer."  

            "If you're a man, you have a 1 in 6 chance of developing prostate cancer during your lifetime."

Did you ever wonder where these numbers come from or whether they're true? In fact, such statistics are usually used for dramatic effect. They're intended to make a point, which may be important:  Cancer is common and deadly, so get screened. Or, take care of your liver and be an organ donor. 

Statistics are often used to "fudge" the truth. The numbers can be misleading. And they're often not entirely (or at least literally) true.

Statistics are commonly used to make things seem worse than they are. Why? It makes the numbers easier to understand. And it gets our attention. The evening news on TV, the online news or the daily newspapers tend to go for the dramatic (though sometimes sacrificing accuracy).

Statistics have limitations. Here are four reasons to view them with a critical eye.

1. The numbers are based on averages.

No one is standing around with a stopwatch and counting events every few minutes. (You knew that, right?)  

The statistic above for transplant cases is based on simple math. There are 525,600 minutes in a year. You divide that number by an estimate of the total number of people added to the transplant list in a year — 47,782. (According to one government website, the number is about 49,000). 

What about a subject that varies by season, such as childhood drownings in the United States? Using the description of a death every so many minutes is particularly misleading. Whatever the estimate, it's likely way too high for winter months and way too low for summer months when most drownings occur. 

It's never made sense to me (other than the dramatic effect) to describe a number of cases or deaths as a new case every few minutes when they are not evenly spread over time. In my view, the numbers speak for themselves — and they are usually dramatic enough without this embellishment.

2. When a range of numbers is involved, the more dramatic figures are quoted.

The 1999 Institute of Medicine report estimated that "at least 44,000 people, and perhaps as many as 98,000 people, die in hospitals each year as a result of medical errors that could have been prevented." These statistics are estimates from two major studies, the report noted.  

I've seen this translated into "1 patient dies every 5 1/2 minutes from a medical error in the U.S." It's based on the 98,000 estimate. Reports rarely use the lower figure of 44,000. We don't want to lose sight of the fact that too many people are harmed or killed by medical errors. But, routinely using the higher estimates of a range may not be accurate.

3. Risk estimates may include insignificant cases.

Not all cases of a particular condition pose an equal risk. For example, some breast cancers in elderly women are so small or well contained that they are not considered a health risk.  

There are similar examples for prostate cancer. When detected, these tumors are monitored, but it's not clear that knowing about them is even helpful. Yet, these tumors tend to be included in statistics of lifetime risk. One could argue they should not be.

4. The numbers may not apply to you.

When it comes to health and disease, risk factors matter. If you learned that a mugging occurs in the United States every 14 minutes (a number I just made up), you may be afraid to go out. Yet, if you live in an area with a low crime rate, that statistic may not apply to you. In fact, it may overestimate your risk, while underestimating the risk in other neighborhoods. 

The same goes for certain cancers. The lifetime risk of lung cancer in the United States is about 1 in 13. But if you separate risk by smoking status (a process called "stratification"), you'd see that these lifetime risk numbers wildly overestimate a non-smoker's risk. In fact, 1 study found that the lifetime risk of lung cancer for person who has never smoked was 1 in 333, while much higher in heavy smokers (1 in 4 for men and 1 in 5 1/2 for women). If you've never smoked, the 1 in 13 lifetime risk doesn't seem to apply.

It's important to know the population the researchers studied. The more similar you are to the study group the more a statistic will apply to you. But lifetime risk statistics conceal this. So does the cases-by-minutes statistic.

The Bottom Line

I think it's fine to "repackage" statistics to make your point in a way that more people can relate to. But it's also important to understand where these numbers come from. And their potential limitations. Most people do not have a 1 in 13 chance of having lung cancer. Smokers have a much higher risk; nonsmokers have a much lower risk. 

The next time you hear about a health story in the news that quotes lifetime risk or an "every few minutes" quote, keep a critical eye open. And don't panic: These numbers usually make bad things seem even worse than they really are. 

Back to top

 

Robert H. Shmerling, M.D., is associate physician at Beth Israel Deaconess Medical Center and associate professor at Harvard Medical School. He has been a practicing rheumatologist for over 20 years at Beth Israel Deaconess Medical Center. He is an active teacher in the Internal Medicine Residency Program, serving as the Robinson Firm Chief. He is also a teacher in the Rheumatology Fellowship Program.

Last updated June 11, 2014


    Print Printer-friendly format    
   
.
.  
This website is certified by Health On the Net Foundation. Click to verify.
.