Well it sounds right

More about the stark contrast between what we are sure is true, and what the facts indicate.

Medical misinterpretation

A big area for misinformation is medical studies. Some educators are calling for changes to Medical School curriculums to insure that future doctors can interpret statistics and studies. If they depend on the drug rep for their education after Medical School, as many do, they don’t understand enough to explain to their patients. They often don’t understand the study results well enough to make informed decisions that will help their patients.

Surviving prostate cancer

New York Mayor Giuliani announced a while back that he was a five-year survivor of prostate cancer. He explained how fortunate he was to live in New York where 82% of diagnosed prostate cancer patients survive five years. He contrasted that to living in Great Britain where 44% of such cases survived five years.

What he didn't understand, or at least didn't mention, is that comparable cases lived about as long in both jurisdictions. In New York, vigorous screening identified a lot of low-grade cancers that would never become dangerous. These added to the 5-year survival numbers in New York but weren't identified in Britain.

And, high-grade lesions were identified earlier in their progress. There was a longer time between diagnosing the cancer and the patient dying. The course of the disease was the same, but measuring from an earlier point in the disease’s progress also added to the 5-year survival number.

The patients with each type of tumor were given similar treatments and survived for five years at about the same rate. In New York, prostate cancers were identified earlier, not leading to cures but leading to longer times between being identified and causing death.

Is it safer to be diagnosed with prostate cancer in New York than in Great Britain? Nope.

Breast cancer: figures don't lie but liars distort figures 

Sometimes misleading information is deliberately published, such as when organizations seek funds.

The Susan G. Komen Foundation advertised (and may still, in some areas) that “The 5-year survival for breast cancer when caught early is 98%. When it’s not? 23%.

According to the British Medical Journal:

Screening mammography does not guarantee that a woman will survive breast cancer. The best evidence indicates that it decreases the chance that a 50 year old woman will die from breast cancer in the next 10 years roughly from 0.53% to 0.46%--a difference of 0.07 percentage points. Because breast cancer treatments are much more effective now than when trials of screening were done, some experts question whether screening mammography has any benefit.

And further:

The five year survival for early and late stage cancers tells you nothing about the benefit of screening. Because of biases caused by lead time (the time from diagnosis by screening to when a tumour can be felt) and overdiagnosis, the five year survival can improve regardless of whether cancer mortality is increased, decreased, or unchanged by the screening.

The article concluded “. . . The Komen advertisement campaign failed to provide the facts. Worse, it undermined decision making by misusing statistics to generate false hope about the benefit of mammography screening. That kind of behavior is not very charitable. . .”

Doctors should know how to read statistics

A respected medical columnist says, “Editors should enforce transparent reporting of evidence, for the benefit of their readers and of healthcare in general.”

If it's bad, don't measure it

Another example of misuse was a study on a new statin drug. (Statins cause LDL cholesterol lab values to decrease by blocking some liver functions.) The researcher noted that cardiac events were down, as is shown by many statin studies. However, he also noted that the overall death rate was higher for the study participants due to a higher rate of liver cancer.

That drug was withdrawn, and we can guess that the researcher had trouble finding funding from the pharmacy companies after that. However, subsequent studies of statins simply stopped measuring or reporting liver cancer or overall death rates.

The risk to the consumer wasn't changed, only the reporting of risks was changed.

Feinstein's gun death stats

This subject intersects the study of gun deaths quoted by a California Senator. Essentially, gun deaths are higher where there are more guns, so if we reduce the number of guns we’ll decrease violence. (Also see the previous blog.)

Studies show the contrary, however. The key here is the leap from gun deaths to violence. Fewer gun deaths are not related to less violence. Instead, more guns licensed for concealed carry markedly reduces violent crime.

In the study that Senator Feinstein quotes, removing the suicide deaths leaves the gun death rates about the same. (Suicide by car or medication isn't examined.) Considering only gang-violence numbers and numbers killed by guns of police officers reverses the rates: more gun deaths in low ownership areas. And this is reasonable to believe, since the most regulated areas have dense populations and high crime. Prime examples are Chicago, Washington D.C., Detroit, and most of Los Angeles.  

Leaping from fact to fantasy

But the leap from fact to fantasy serves Feinstein’s purpose, and it sells well in a press conference.

Leaps from fact to fantasy ran rampant during the last election. They’re still common in some columns and blogs in Costa Mesa.

Think about whether it makes sense

Take the time to think about what is said. It just may be that someone is using statistics as propaganda. That is, using statistics to convince you by emotion instead of logic.

Just because we wish there were a simple answer doesn't mean that there is one. Or that the writer or speaker knows what it is.

What’s in it for him? What does she want you to believe?