Today, nobody can be certain that any of the prescription sleeping pills is safe for long-term use or even that they help sleep with long-term use. Although the potential risk of intermittent use (a few times a month) seems to be less than the risk of nightly use, our CPSI and CPSII studies indicated that sleeping pills may be unsafe in any amount. It is important to recognize that both studies were done before some of the drugs popular today came on the market. In particular, zolpidem, zaleplon, eszopiclone and ramelteon (the most popular U.S. sleeping pills in early 2006) were not tested. Different experts may have different opinions about the likelihood of long-term safety, but the fact is that nobody knows for sure about particular drugs. The only way to assure public safety is to do more long-term randomizing trials.

It is not a unique idea that long-term studies should be done. The National Institutes of Health (NIH) are doing long-term studies of diet and exercise, hormone replacement and vitamins, cholesterol-lowering drugs and aspirin. Of the psychoactive drugs, long-term studies have been supported for anti-depressant and anti-psychotic drugs, mood stabilizers (such as lithium), tranquilizers, narcotic agonists and antagonists, and stimulants for treatment of hyperactive kids. Truthfully, hypnotics are the only class of psychotropic drugs for which long-term studies have not been supported by the NIH. In a 1997 search of the CRISP list of NIH-supported projects, I found about 140 clinical studies of other psychotropic drugs, but none on sleeping pills. This is neglectful, considering the proportion of total psychotropic drug sales which the sleeping pill market occupies.

Although I receive generous grant funding in other areas of sleep research, the federal agencies have never been willing to support any studies of sleeping pills which I have proposed. I do not take it personally, since I know other investigators have the same problem. There is a mind set against funding studies of sleeping pills and a lot of passing the buck among different NIH Institutes. You could count on one hand (and have fingers left) the total number of clinical trials of hypnotic drugs funded by NIH in its entire half-century history. Of the few funded, the purpose was usually to compare some other treatment with sleeping pills, not directly to test sleeping pills.

Lack of government curiosity about sleeping pill prescribing is exceptional. As mentioned above, government agencies denied that they had data on overall U.S. consumption of sleeping pills. They are certainly more careful in studying other addicting drugs.

You might expect that with the new evidence that new sleeping pills cause cancer, no company would want to do large long-term studies of their product which might demonstrate cancer risk. On the other hand, a company which really believed their product was safe should do such trials to prove it. If they are not doing the trials, you can assume that they themselves are not sure their product is safe.

Before I denounce the drug companies excessively, let me mention that the same heavy prescribing of hypnotics existed in Communist countries before the dissolution of the Soviet block, and there was a similar lack of studies behind the Iron Curtain. One should not ascribe the scientific neglect of sleeping pills entirely to the profit motive.

The cost of studies which would establish the long-term benefits/risks ratios of the most popular hypnotics would probably be $10-20 million for a period of several years. This would mean devoting less than 1% of sleeping pill costs for long-term research, in a sleeping-pill-giving industry which grosses over 2 billion dollars a year. The retail costs of the hypnotics drugs themselves may surpass two billion, and in addition, the fees of the prescribing doctors, the necessary laboratory tests, etc. amplify the total cost. Compare this with other areas of drug research, where the research costs may reach 25% of sales, or aircraft, computer and defense industries, where the R&D costs may reach 50% of total costs.

The Congress would have several choices for how to finance this research. It could require the manufacturers by law to perform the needed research. It could impose a user-fee on hypnotics (either at the retail level or the manufacturer level) to do the studies. A cost of a nickel a pill added to pills which cost $3 each would be a trivial cost to the consumer and well worth while in keeping consumers safe. To the extent that this small cost might discourage people from using sleeping pills, it would be doing them a favor. Such a user-fee might be regarded like the levies on other addicting substances such as cigarettes and alcohol. Congress could take the money out of the overall appropriations for health services research or for the National Institutes of Health, or it could conceivably appropriate new funds from general revenues. It is for the Congress to decide from where the money should come, but one way or another, the studies should be done.

Continued in Chapter 10