Central  South  Dakota  Medical News
The Clinical View
  by P.E. Hoffsten, M.D.
12 November 2003

TREATMENT STATISTICS – 2:  ADVERTISING MISREPRESENTATION

     Last week’s column dealt with direct-to-consumer advertising for prescription medications.  In that column a study from Finland using simvastatin (Zocor) to prevent a second heart attack in men who had already had their first heart attack was shown to be both effective and economically balanced.  The risk of disease in the untreated group not receiving simvastatin was substantially high with 12% of these men developing heart attacks over the 5 years of the study.  The treated group only had 8% of the men  developing a heart attack.  It was claimed that the risk of heart attack was cut by 33% in the group of men who were treated with the drug.  Last week’s column introduced the idea of “the number needed to treat”. This is a statistical term that describes how many men had to be treated before one heart attack was prevented.  In the case of Zocor in this study, one heart attack was prevented for every 25 men who took the drug.

     I have presented “the number needed to treat” type of information to groups before.  The groans of disappointment that roll through the audience are usually prominent.  People seem to have the idea that if a person takes a medication, it will protect that person from having the heart attack.  But this isn’t really true. As I pointed out in last week’s column, 88% of the men in the untreated group did not have heart attacks and 8% of the men in the treated group did have heart attacks. Thus, the concept of perfection in drug effect really doesn’t apply.  Be that as it may, it was pointed out that from a health policy standpoint, the use of this drug made sense in its overall saving to society. The study described above was related to so called “secondary prevention.”  The object of the study was to prevent a heart attack from occurring in a group of people who had already had one heart attack.  This provided a high profile group with substantial risk.

     But watch what happens when the advertisers start doing their job.  Another drug company started a study to show that their drug would prevent the first heart attack in a group of people who had high blood cholesterol but no heart disease known.  Note this is a much lower risk group of patients.  In 5 years there were 248 heart attacks in 3,293 men not receiving the drug. This is a 7½% incidence of heart attack in the untreated group.  Notice this is lower than the heart attack rate in the treated group of the Finnish men described above.  It was seen that in the treated group of men who received a drug called pravastatin (Pravachol) there were only 174 heart attacks in 3,302 men studied.  Thus, the rate of heart attack was reduced from 7 ½% in the untreated group to 5.2% in the treated group.  This amounts to a 33% decrease in the heart attack rate just like the first study noted above. But when one goes through the calculation of “the number needed to treat” it seems that more than 43 men have to be treated with the drug before one heart attack is stopped.  In other words, the cost of a medication for the prevention of a first heart attack was $172,608.00.

     The company advertising this medication didn’t mention numbers such as shown above.  Instead, they said that their drug decreased the rate of heart attack by 33%, the same way that Zocor did.  Indeed, that statistic is exactly correct, but the cost factor becomes unmanageable when one thinks of that type of expense to prevent one heart attack unless, of course, it is your heart attack that is prevented.

     A third study done in Texas looking at the affect of a drug called lovastatin (Mevacor) included people who had normal blood cholesterol and no history of a previous heart attack.  In other words, these people were felt to be completely normal at the start of the study.  It was shown that the rate of heart attack in the untreated group was 5 ½ heart attacks per hundred men as compared to 3 ½ heart attacks in the group of men who were treated with the drug Mevacor.  Now if one does the calculation on the number needed to treat, it is seen that one heart attack is prevented for every 50 men that take the drug.  The cost of this program becomes $192,00.00.  Yet if one looks at the percent decrease in the heart attack rate, it is almost a 38% decrease in the rate of heart attack in this group of men.  That is what the advertisement emphasized, not the exorbitant cost of trying to stop one heart attack in 50 people.

     Current Food and Drug Administration requirements mandate that a company do a study comparing their drug either to another established drug or to an inert pill with no drug affect.  If it cannot be shown that the drug company’s new medication is better than an inert pill, or better than a pill that is already on the market, the medication does not get approved.  Thus, there are virtually hundreds of studies such as described above that are published periodically.  The drug companies will emphasize the statistics that will make their medications seem attractive.  That is the American free enterprise system.

     Drugs such as the cholesterol medications present a perplexing problem.  People taking them don’t feel any better.  The drug doesn’t stop a pain.  The only benefit one derives from the cholesterol-lowing drugs is the hope of never having the heart attack or the stroke that comes with high blood cholesterol. Thus, the final question always becomes “what is an appropriate number needed to treat” in order to make a drug worth taking. This answer has to be an individual one.  It depends upon the individual person’s risk profile and what the individual person feels is an appropriate amount of “health insurance”.  Notice that the use of a drug such as a cholesterol lowering medication is really a form of term life insurance.

     Whether or not a medication is worth taking is considered carefully by the health care providers at your local clinics.  They think  about whether or not a medication is worth using as compared to the cost of that medication.  They consider the side effect profile, the amount of expense monitoring the drug, the likelihood of success, and “the number needed to treat”. Then a rational decision on the value of your medications can be made.  “The number needed to treat” provides a sobering statistic to consider.