Thursday, July 9, 2015

Number Needed to Treat (NNT): A chasm between patients and clinicians

If you take some medications, you may believe that your drugs can maintain your health, of course. Your doctor says, "You have high blood cholesterol level. This Somestatin can reduce your cholesterol. If you quit taking it, you shall be attacked by a stroke. Remember it." Ok, this drug saves your life. Thanks, Doc.

Is it really true?

It is quite simple in the assumption that, all persons who take the drug will survive, and all who do not take will die. But the situation is more complicated. Some people are so unfortunate that they encounter a serious stroke in spite of daily medication. Ironically, there are also some people still healthy without any care of their cholesterol level. It is a matter of chance.

So, you should consider how much Somestatin is effective in the prevention of tragic events.

As an assumption, Somestatin reduce the risk of stroke by 50% for these 20 years. Sound fantastic? In this case, do you choose to take medication continuously?

Actually, more information is needed to make a decision. The information about how is the risk of stroke if you do not take Somestatin is crucial.

Considering your age, the current level of cholesterol, family history, habits, and so on, 5% of the risk is admitted in these 20 years, for an example. So, If you take Somestatin, the risk is decreased to 2.5%. Can you accept daily medication for the reduction by only 2.5% of the risk of stroke?

This discussion resembles the matter of health examination, as I wrote previously.

My past entry: Sensitivity and specificity of examination about H. Pylori

In the region of epidemiology, Number Needed to Treat (NNT) is adopted as a standard to evaluate the effectiveness of treatment. It means how many patients are needed to be treated for a successful prevention of an undesired result. If NNT=10, one patient is benefited from the treatment when ten patients take this treatment. Thus, other nine patients take treatment in vain.

In the example above, 95% of the patients will not encounter a stroke regardless of medication. And 2.5% will be betrayed by Somestatin. It is worthy for only 2.5% of the patients to take it. It means 25 persons in 1000 are saved by Somestatin.  Therefore, NNT is 40 (1000/25).

Surprisingly, only one person of forty can be benefited. Are you confident to be the one who will be attacked by stroke without medication and protected by the treatment? If you do not believe, it is rational to reject Somestatin.

In the real world, NNT of major drugs is various. Statins have an NNT of 60, according to This situation is more miserable than the assumption I presented.

Men's Journal: Drugs: Effective for the Few, Prescribed to the Many

You may think what a waste of money to prescribe statins is. However, I do not agree, notwithstanding the discussion above. Clinical practitioners desire to save the patients' life. We take any means to get the purpose. And there is no way to rescue the patient with 100% certainty. Instead, many patients lose their life despite any intervention. All we can do is to improve the possibility to protect them from probable mortality, regardless of the magnitude of efficacy. Therefore, we recommend taking statins to patients with high cholesterol level. We are happy to see one patient saved by our contribution, even if there are fifty-nine patients taking a meaningless drug behind him, because we cannot identify the one to be prescribed.

The standpoint of the patient is apparently different. The choice is yours. You may be annoyed to take daily medication. Or, you would regret when encountering the stroke if you rejected treatment. Concerning your health may be a proof of the love to your family.

In modern medicine, physicians look statistical view. But statistics is statistics. How interpret the result is your responsibility.

No comments:

Post a Comment