Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Chapter 6 CONCLUS IONS The study committee orig inally intended to examine the use of hypnotics in medical practice' and compare the barbiturate hypnotics with newer pharmacologic agents. As the inquiry proceeded, it became clear that this approach required first addressing fundamental issues of the physiology of sleep and the meaning of subj ective complaints of sleep disturbance. Thus, this study has explored medical issues -- scientific knowledge and research relating to sleep, clinical practice, and phys ic fan educe t tori; and it has exam ined soc ie tal i s sue s incl ud ing public health risks associated with the availability of hypnotic medica- tion and ways of enhancing consumer information. The observations contained in this chapter are the committee' s best judgment as to steps that might be taken to improve the physician's therapeutic response to the complaint of unsatisfactory sleep, and to enhance the public' s understanding of the hazards associated with the casual, unnecessary, or excessive use of hypnotic medication. A. Clinical Practice, Patient Information, and Public Health To relate the scientific study of sleep, insomnia, and hypnotic medication to contemporary health care, the committee offers the follow- ing suggestions for prudent clinical practice, patient education and preventive medicine, and improvement of public health. Diagnostic Appraisal A thorough medical and psychosocial appraisal of the insomniac patient should precede any decision to prescribe drugs for relief of the sleep complaint. Recent developments in sleep disorders research permit a more detailed diagnostic classification of insomnia complaints than was previously possible. Diagnostic possibilities include a wide range of syndromes from the more common medical or psychiatric illnesses ~ such as arthritis or depression) which lead to sleeplessness, to the much less common disorders of sleep-induced ventilatory impair- ment ~ sleep apnea) causing numerous life-saving arousals each night. Psychosocial aspects of daytime living often play a role in the nocturnal problems of patients in various age groups. 137
Choosing the most appropriate and least hazardous approach to providing relief depends upon interpreting the best possible diagnostic information about the particular patient. For example, in the three specific syndromes cited above (arthritis, depression, sleep apnea), prescribing a hypnotic drug would be inappropriate as the primary response to the patients' complaints of insomnia. For some patients under certain forms of psychosocial stress, hypnotics may be useful as temporary adjuncts to other measures, but their long-term clinical benefit in chronic use is not established. Prudent Prescribing As a class of drugs, hypnotics should have only a limited place in contemporary medical practice; it is difficult to justify much of the current prescribing of sleeping medication. As a standard of prudent ambulatory medical care, the committee favors the prescription of only very limited numbers of sleeping pills for use for a few nights at a time, to aid in specific situations such as travel or transient psycho- social stress. Hypnotic drugs should be avoided for patients with respiratory impairment, mental depression, or substance abuse problems such as alcoholism. Hypnotic drugs should be selected carefully and prescribed cautiously, if at all, for patients who are or may become pregnant, who operate machinery, who have kidney or liver disease, or who are old. Data are lacking with which to judge the efficacy and safety of long-term nightly use of hypnotics; currently, the consensus of sleep disorders specialists seems to be that hypnotic drugs should not be the primary form of treatment for most patients with persistent insomnia. The committee urges physicians to closely monitor patients receiving these medications, and generally refrain from providing auto- matic or "as needed" refills. Long-term users of sleeping pills should be re-appraised frequently for changes in health status, alcohol consumption and drug use in order to prevent untoward effects. Patient Information Patients should receive clear directions and warnings about the use of hypnotic drugs. Important examples are as follows: -- Patients receiving the benzodiazepine, flurazepam (Dalmane~ R)), should be warned about the lingering presence of an active drug in their bloodstream all day long and be told to avoid driving if they are even slightly drowsy; if they do drive (or operate machinery) they -138-
should abstain from alcoholic beverages during the entire period -- day and night -- of treatment with this drug. Patients receiving flurazepam should also be informed that the drug has a cumulative action due to its active metabolite, so that nightly (or alternate nightly) use may lead to side effects such as drowsiness or ataxia appearing after the initial few days of drug treatment; these should be reported to the physician promptly. -- Patients receiving barbiturates (or most other nonbenzodiazepine hypnotics) should be advised of possible hangover effects during the first few mornings after treatment has begun. Patients should be told to inform all health care personnel in the future if they continue to take the drugs, so as to forestall complications should the need arise for them to receive anticoagulant medication. -- Women of childbearing age who receive prescriptions for either benzodiazepine or barbiturate hypnotics should be told they may use them in future episodes of insomnia only if they are certain that they are not pregnant at that time, or if instructed to do so by a physi- cian . -- With any hypnotic, patients should be advised of the potential for the development of habituation in the form of nightly reliance on sleeping pills. In addition to oral instructions from health care personnel, many patients could be assisted by informational brochures, available in the medical office, which would provide guidance on proper use of the medica- tion (and, perhaps, other sleep-related information of a non-pharmaco- logical nature). Suicide Prevention The reduction in the ready availability of barbiturates since 1970 because of changes in prescribing patterns has reduced the number of deaths, particularly suicides, from barbiturate overdose. Only a relatively small decline can be detected, however, in the overall drug suicide rate. The use of multiple drugs, including alcohol, is the increasingly predominant mode in suicides and accidental deaths due to drugs. It would thus appear that efforts to control the avail- ability of drugs should not be based solely upon the toxicity of a drug used by itself in overdose, but should take into account the dangers of combining different kinds of medication with each other and with alcohol. -139-
Implusive pill-taking in suicide attempts conceivably could be discouraged by (a) larger pills; (b) individually wrapped pills; or (c) inclusion of an emetic compound in small amounts in each pill so that vomiting would occur in case of overdose. The last proposal may not be clinically feasible, requires further pharmacological research, and may have medico-legal drawbacks; but the first two could be explored and tested for those drugs which are perennially attractive to suicidal individuals (analgesics, hypnotics, tranquilizers and anti-depressants). Drawbacks to be considered include greater difficulty for those with dexterity problems (e.g., the elderly) and increased costs passed on to all consumers. Clearly, an alert health care team can be the first line of attack on a suicide problem. A great many suicidal drug-ingesting episodes have been preceded by visits to physicians; often it is recently pre- scribed medication that is used self-destructively. Toxicology In studies of drug deaths and traffic accident victims, modern toxicological analyses for the commonly prescribed benzodiazepines and their metabolites should be performed, in addition to traditional tests for alcohol, opiates and barbiturates. Nationally, there is a need for adequate toxicological analyses to provide more reliable data with which to assess the role played by psychotropic drugs in suicide and various kinds of accidental death or injury. Epidemiology Current national mortality and morbidity statistics are inadequate for determining the extent and source of drug problems in the United States. The currently used coding manual, International Classification of Diseases, Adapted (Ninth Revision) does not allow sufficient dis- aggregation of data on deaths by ingestion of various types of drugs. What is needed is a code system which would allow the coding of indi- vidual drugs so that their patterns of use may be accurately determined. Such a system could be an expansion of the international code. B. Benzodiazepines vs. Barbiturates as Hypnotic Drugs Benzodiazepines are now used as hypnotics or bedtime sedatives far more often than barbiturates or other types of drugs in this group. In fact, one benzodiazepine, flurazepam, is prescribed more often than all other hypnotics combined. In Europe, nitrazepam -140-
is similarly popular. Other benzodiazepines used primarily as anxiolytics or daytime sedatives are also often taken at bedtime for sleep. This represents a significant change from the state of affairs a decade ago, when barbiturate hypnotics and glutethimide were overwhelmingly predomi- nant; this in part reflects the medical profession's response to some well-known advantages of the benzodiazepines -- especially their relative safety when taken alone in attempted suicidal overdose and the absence of interference with actions of certain other drugs, especially anti- coagulants. From anecdotal and clinical reports, the committee noted that physicians have found patient satisfaction with benzodiazepines generally to be about equal to that obtained from barbiturate hypnotics; and physicians have preferred to continue prescribing benzodiazepines because of their reputation for safety. There is a lack of standardized criteria for defining insomnia and a lack of clear documentation of efficacy in alleviating the patients' actual complaints. These problems are compounded by the diverse methods by which efficacy studies are performed and data are presented. Little can be reliably said about the comparative benefits of various hypnotics. Until such knowledge is available, the committee concludes that relative risks of individual hypnotics for particular patients should be the dominant factor in choosing which to prescribe. Insufficien;t recognition has been given to undesirable or possibly hazardous features of the benzodiazepines. Some potentially hazardous attributes of the longer-acting benzodiazepines which are not found qualitatively in barbiturate hypnotic use are: · the accumulation of long-lived active metabolites when used on consecutive or alternate nights, especially with flurazepam and nitraze- pam. o an increasing likelihood of adverse drug reactions with advanc- ing age of the patient. · a greater likelihood of adverse drug reactions in patients with diminished kidney function. Benzodiazepines also share the following potentially hazardous attri- butes with barbiturate hypnotics: ~ In combined overdose (either suicidal or accidental) with other depressant drugs, including alcohol, benzodiazepines act additively to increase the risk of death. The level of each of the substances may be distinctly sub-lethal, but the end result may still be fatal. (On a dose-for-dose basis, this appears to be less of a hazard than it is with barbiturates.) -141-
· During the day after nocturnal drug use, long-acting benzodiaze- pine hypnotics -- flurazepam and nitrazepam -- produce deficits on tests of visual-motor coordination which may be related to automobile driving skills. (Following seven nights' drug use, the impairment caused by these drugs is distinctly greater than that resulting from such use of a barbiturate hypnotic.) o When alcohol is ingested the day after a benzodiazepine hypnotic has been used, driving-related skills are very much impaired. (This potential problem is greater with flurazepam and nitrazepam than with with barbiturate hypnotics, especially following seven nights of drug use.) o Some benzodiazepines (diazepam but not, as yet, flurazepam) have been used in serious drug dependence of the barbiturate type. The extent of addiction to benzodiazepines is difficult to ascertain, but it is a source of growing concern in the medical community. (The potential for fatality from established barbiturate addiction is greater than that resulting from addiction to benzodiazepines.) 0 Nightly reliance on drugs for sleep is equally likely to arise with either benzodiazepines or barbiturates. e The benzodiazepines are respiratory depressants (quantitatively less potent than barbiturates on a dose-for-dose basis). O The teratogenic effects of either class of drug have not been fully explored, but -- if confirmed by future studies -- are likely to be of very low frequency. On the the basis of limited evidence, the benzodiazepi-nes would seem to increase the risk of a baby being born with cleft lip. This finding warrants further research, as does the very recent finding of increased risk of brain tumor in children whose mothers took barbiturates during pregnancy. · Like the barbiturates, most benzodiazepines are tolerated poorly by patients with liver disease. The committee believes that benzodiazepines are not innocuous drugs. They pose hazards that have been insufficiently appreciated, while the narrow margin of overdose safety of barbiturate hypnotics appears to be well-known to the medical profession. As for efficacy, the committee concludes that both types of drugs are probably equally effective in short-term use. In long-term use, the side effects and risks of benzodiazepine hypnotics are likely to increase, and tolerance is likely to develop to sleep-promoting properties of barbiturates. A number of older patients have become reliant on barbiturates for nightly use; it is likely that a number of younger patients will become reliant upon nonbarbiturate hypnotics. The safety and efficacy of nightly use of any type of hypnotic for prolonged periods is not established at this time. The committee believes that, until long-term safety and efficacy of regular -142-
hypnotic use is established, physicians should rarely, if ever, prescribe hypnotic drugs for periods beyond two to four weeks for patients who have not yet become reliant upon regular use of hypnotics. In the case of patients who are already reliant, the committee recommends frequent re- appraisal of diagnosis, risks and benefits of hypnotic use, and main- tenance of vigilance for development of either toxic side effects or risk factors which would make continuation of the drug hazardous (e.g., preg- nancy, renal disease, alcoholism, depression). In light of its review of safety, efficacy and public health problems associated with various hypnotic drugs, the committee finds no justification for more restrictive scheduling of barbiturate hypnotics under the Controlled Substances Act; this removal of barbi- turates from general availability in ambulatory medical care would be unwarranted at the present time. The "ideal hypnotic" has not been found. Drugs that might more safely relieve various types of sleep complaints should be investigated.* For example, the reality of chronic drug use emphasizes the need to find and evaluate drugs that do not accumulate in the body with repeated doses and do not result in adverse daytime effects; and the reality of overdose deaths dramatizes the need for drugs with the widest possible margin of safety, even when combined with other drugs and alcohol. As long as flurazepam continues to be the most widely prescribed hypnotic in North America, a suggestion regarding its use is in order: o The usual dose should be 15 ma. Studies of flurazepam generally describe only modest differences between a 15 mg and 30 mg dose in terms of the effect on objective (EEG) sleep measures or subjective satis- faction. Yet there are strong dose-related correlations when it comes to toxic side effects. The drug manufacturer (with FDA approval) continues to propose the 30 mg dose as the customary one for adults, except for "elderly" or "debilitated'' individuals, for whom 15 mg is suggested. The recommendations of clinical investigators that the 15 mg dose be used more frequently 1/-3/ have not been heeded. Occasional trials of doses less than 15 mg in some patients is currently impossible because the manufacturer does not make a capsule contain- ing less than 15 ma, and flurazepam is not marketed in scored tablets. C. Professional Education The findings of current and future research into sleep disorders and medication for their treatment must be translated for possible clinical use by practicing physicians. The committee finds that . *See Section E., below, on research needs -143-
although hypnotics are prescribed widely, most physicians receive little training in their use. Medical school curricula should include training in the diagnosis and management of sleep disorders. (In 1977, 42 out of 91 responding medical schools indicated that they did not offer even one lecture on sleep problems or sleep-related physiology within their required curriculum 4/~. The committee believes that major professional education efforts should be developed in the diagnosis and therapy of insomnia, and that these efforts should not be underwritten primarily by the pharma- ceutical industry. Programs on insomnia should become part of the continuing medical education that has become mandatory in many jurisdictions and in several specialty societies. The incentive to become proficient in this area would be increased if licensing and specialty board examinations would give sleep disorders a prominence commensurate with their occurrence in clinical practice and if research in this area were encouraged by appropriate federal support. However, before such professional education campaigns are launched, it would seem appropriate to convene a "consensus development exer- cise" -- in 1979, to be re-convened every two or three years there- after -- on the therapeutic and clinical research issues posed in this study. Modeled on recent NIH conferences on therapeutics (such as radical vs. simple mastectomy, surgical treatment of extreme obesity, treatable dementia in the elderly), such a meeting would bring together experts in several relevant disciplines to explore areas of consensus on diagnosis and management of various kinds of insomnia and the appropriate use of various therapeutic approaches, including drugs. Opportunity for input from consumers should also be provided. The consensual viewpoint could then be published and disseminated through specialty societies and continuing medical education courses. The need for such periodic "therapeutic reviews" is suggested by the rapid pace of developments in the sleep research field. A vigorous program of continuing medical education based on current consensual therapeutic reviews and sponsored by specialty societies, medical associations and the federal government would help diminish casual or inappropriate prescribing of hypnotic drugs. In addition, the committee recognizes that computer data files of ambulatory prescribing patterns exist in several states, including those in which PSROs are monitoring Medicaid practice, or law enforcement or medical disciplinary boards are checking on potentially unprofessional or illegal prescribing patterns.* With proper safeguards of patient privacy, each prescribing physician in those states could be sent a print-out showing his or her prescribing patterns compared with state norms or with standards generated by peers or professional organizations. The committee believes that this should be explored and would be a method of self- education to which many Physicians are likely to respond. *Described further in a Technical Supplement to this report. -144-
D. Advertising and Labeling "Detail men," sales representatives of pharmaceutical companies, have been found to constitute the most influential single source of information for physicians during the period when a drug is first being prescribed. 5/ Advertising in medical journals and material in the Physician's Desk Reference (PDR) have been shown to be frequently relied upon by physicians in matters such as dosage and adverse reactions for drugs. 5/ It is essential that the information conveyed by these promo- tional materials be accurate and complete. The PDR, distributed annually without charge to most practicing physicians, is essentially a collection of professional package inserts containing the "complete labeling" (also called "complete prescribing information") which has met with FDA approval. Drug manufacturers list their products by paying the publishing company a fee. The information on each drug usually includes data on chemistry, animal pharmacology and toxicity, indications, dosages, contraindications, precautions, side effects, and special warnings, if any. This approved prescribing information is a product of negotiation between the manufacturer and the FDA. On some occasions, the FDA insists that a specific warning be added to the professional information for a given product or group of drugs. Such is the case, for example, with all benzodiaze- pines, which carry the warning against usage during the first tri- mester of pregnancy because of an increased risk of teratogenicity. Recently, the FDA has required the drug manufacturers to submit for its approval a clinical (human) pharmacology section to be added to the prescribing information contained in the professional package insert and the PDR, although it need not be in the advertising. This requirement has resulted in disclosure of the existence of flurazepam's long-lived metabolite in the new labeling, which first reached physicians in a supplement to the PDR in late 1978 -- five years after publication of the research article that first described flurazepam's pharmaco- kinetics in man. Even in the new labeling, however, the half-life of the metabolite is stated without indicating the implications of its accumulation. Instead of stating that blood levels on the eighth morning after a week of consecutive nightly use are likely to be four to six times that found on the first morning, with concomitant potential for adverse daytime effects, the new prescribing information places the clinical implications of the drug's "pharmacokinetic profile" in a wholly favorable light.* *This pharmacokinetic profile may be responsible for the clinical observation that flurazepam is increasingly effective on the second or third night of consecutive use and that one or two nights after the drug is discontinued both sleep latency and total wake time may still be decreased." 7/ -145-
The "warning" that physicians are encouraged to give their fluraze- pam patients about driving or operating machinery is deficient in that it specifies that the patient be told to avoid such activities "shortly after ingesting the drug" (presumably on the way to bed). It fails to mention that the drug can produce adverse effects on coordination the next day or can interact adversely with alcohol the next day as well. In advertisements aimed at health professionals, manufacturers of prescription drugs are not required to include the "complete prescribing information," but must offer a "summary" which includes the special warn- ings such as avoiding use during pregnancy. The advertising claims for a drug are supposed to be consistent with the indications and other infor- mation originally approved by the FDA when the drug was first allowed on the market. Unlike the approved labeling, advertising copy is not submitted to FDA for pre-clearance. If, after the fact, an advertisment is found to be misleading, the FDA may step in and order corrective action. Only three hypnotic drugs -- none of them a barbiturate -- are currently advertised to the medical profession in the United States. Examination of current advertisements raises doubts about whether they present balanced, clinically relevant information to aid the con- scientious physician in prescribing. Advertisements for flurazepam for example, demonstrate the limitations of advertising as an information source. A typical artistically attractive, three-page full color spread will emphasize the laboratory procedures showing the drug's "effectiveness" lasting up to 28 nights in insomniac patients. An implication of these ad- vertisements is that insomnia is a unitary disorder, appropriately treatable for 28 nights at a time with flurazepam. Nowhere do these advertisements reveal that the claim of effectiveness for 28 nights is based on studies of only ten patients and that hundreds of individ- uals with sleep complaints had to be screened to select these severe insomniacs for research purposes. The emphasis on the sleep laboratory technology distracts attention from the absence of data on subjective relief and daytime measures of drug-induced improvement or impairment. Finally, the "summary" of prescribing information in current advertise- ments does not include even the limited information about flurazepam's long-acting metabolite which now appears in the PDR. The flow of information about drugs provided to the practicing physician is dominated by printed advertising in medical journals, manufacturers' sales representatives and the Physician's Desk Reference. In the case of hypnotic drugs the committee finds much of this informa- tion is apt to be incomplete and of questionable relevance to physicians. At present, responsibility is left to the professional journals whose vigilance in holding advertisers to high standards of accuracy and balance is unknown. Except in rare instances, the Food and Drug -146-
Administration is not legally authorized to screen advertising prior to its appearance, but it could be more vigilant and timely in insisting on "complete labeling" (prescribing information) which is maximally useful in format and content to the physician. Consideration should be given to requiring that such complete information should appear in advertisements, rather than being limited to the PDR and professional package inserts. Pharmaceutical sales representatives should be required to leave copies of this information with the physician whenever a product is discussed. E. Research Needs Insomnia and Appropriate Therapeutic Responses Clinical Investigations As in many other fields, critical links between basic science research and clinical practice are forged in the research arena of clinical investigation. No research effort consonant with the prevalence of the complaint of insomnia has been undertaken. Because insomnia complaints are in great part subjective complaints but also lend themselves to objective measurement, intensive clinical investigation of patients offers promise for greater clarifica- tion of these problems in the future. Clinical investigations are needed on two levels: analyses of fundamental issues regarding human sleep, and studies of series of patients. Development of normative data is required for respiratory, cardiac, and endocrine function during sleep; it has become clear that the physiology of these functions differs greatly in the sleeping and waking states. Further work on experimentally induced sleep disturbance could shed light on the sequelae of interrupted or short sleep for later comparison with the daytime results of insomniacs' sleep patterns (with and without therapeutic intervention). Clinical studies should be conducted (1) of patients with sleep complaints, but whose findings on objective sleep measures are well within "normal" limits; (2) of individuals of various ages whose sleep is not a source of distress, in order to extend the normative data base; (3) of individuals who only use sleeping medications intermit- tently (several times a year for a few days); (4) of individuals who are nightly users of sleeping medication, who appear to be satisfied, and whose insomnia complaints are "remedied"' by such nightly use. Their daytime functioning as well as nighttime sleep measurements need to be studied; (5) of the actual mechanisms whereby hypnotic drugs promote sleep. In the latter, there is a need for fundamental neurochemical research as well as sophisticated psychological investigations of satisfied users of hypnotic medications. -147-
Epidemiological research is needed to distinguish the potential . ~ from the actual hazards to drivers, pedestrians, and industrial workers of residual daytime effects from nocturnal hypnotic medication. All such studies must include the psychiatric condition of the subjects as a variable. Studies are also needed to discern what hazards, if any, accompany a poor night's sleep that has not been treated with drugs. Further studies should be undertaken of the industrial accident rate by shift workers, both on and off hypnotic medication. Further studies on teratogenicity and carcinogencity of hypnotic drugs are essential. Longitudinal studies are required to develop knowledge about the duration of various kinds of insomnia complaints and the crucial determinants of the duration in such cases. In one large follow-up survey, a significant relationship between stated sleep time and mor- tality was found. 7/ The less a person slept (by report) the more likely that person was to be dead in six years. The same was true for prolonged (by report) sleep and for those who reported taking sleeping pills frequently. These relationships need to be explored in depth. Health services research could help to determine how primary health . . care practitioners respond to patients with behavioral problems such as insomnia. Primary care physicians do not customarily receive payment for much of their time spent in office counseling on problems like insomnia. Demonstrations and studies should be made of greater utili- zation of paramedical personnel for this purpose (e.g., nurses, social workers, physician's assistants) as well as third party payment for physicians' time spent talking with patients. Treatment research With the present inadequate data base, the committee is unable to endorse the efficacy of specific therapies, drugs or classes of drugs in the management of insomnia complaints. The committee recommends that new, more sophisticated multidisci- plinary research be initiated to test the efficacy of various pharmacological, psychotherapeutic, behavioral, and psychosocial treatment approaches to various types of insomnia. Patients with an insomnia complaint tend to differ widely from each other in their objective and subjective sleep abnormalities, age, sex, primary diagnosis, psychopathology and daytime mood and psychomotor functioning. Prior to selection for study, the patients need to be carefully evaluated from a multidisciplinary point of view, with special attention to objective and subjective sleep characteristics, including appropriate tests to diagnose specific sleep disorders such as sleep apnea and nocturnal myoclonus; the patients also need to be evaluated from the point of view of daytime functioning, mood, and psychopathology. -148-
The effect of any particular treatment must be evaluated on many dimen- sions, including objective and subjective sleep characteristics, daytime measures of mood, psychopathology, functioning, and cognition and memory, with a view to establishing both the potential benefits and risks of the treatment. The duration of effectiveness must also be established. Someday, it may be possible to match a pharmacological (or nonpharmaco- logical) treatment to an individual's chief sleep complaint, once adequate diagnostic rigor has been applied and there has been more clinical research comparing diagnoses and outcomes. With regard to research on pharmacologic treatment methods, every effort must be made to standardize and upgrade hypnotic drug efficacy studies. There must be clear definitions of the specific problems and of the goals of treatment. Above all these goals must be related to the actual problems. Many sleep laboratory studies have omitted significant descriptions of the subjective complaints of the individuals whose responses to hypnotic drugs were being measured. In fact, in a good number of studies the subjective effects upon the individual have gone unreported in the midst of a flurry of EEG tracings. Most surprising, many sleep laboratory studies have presented only meager objective data. They have not specified when the drug was taken, how long the patient spent in bed (and whether this was at the discretion of the patient or the experimenter), total sleep time, sleep efficiency, early morning wake time, individual differences in treatment response, or a host of other potentially significant measures of effectiveness. More- over, many of the measures that were reported have remained undefined. With regard to daytime effects, it is necessary to assess mood, mental alertness, and visual-motor coordination. New and sophisticated tests have been developed from aviation and traffic safety sources which should be applied in modeled evaluations. Special attention to the risks associated with the accumulation of long-acting metabolites is of major concern with some drugs; these could impair driving skill, potentiate the effects of alcohol and be involved in late-developing adverse reactions. Clinical experimentation with low doses of flurazepam and other agents with long-lasting metabolites should be pursued. Shorter-acting benzodiazepines - oxazepam (Serax(R)) and lorazepam (Ativan(R)) which are generally marketed for use as daytime tranquilizers, should be tested in the United States specifically for their hypnotic utility and their liabilities, such as drug abuse potential. Other promising drugs which have been incompletely evaluated for efficacy and safety in various types of patients include methyprylon (Noludar(R)), chloral hydrate, diphenhydramine (Benadryl(R)), L-tryptophan, and antidepressants like amitryptiline (Elavil(R)) and doxepin (Sinequan (R)~. -149-
Nonpharmacological treatments for different types of insomnia also must be further investigated. Various forms and combinations of behavioral techniques have shown some promise for the individual who suffers from uncomplicated chronic insomnia, though these have not been documented in controlled studies. The utility of psychotherapy for such individuals, even in the absence of a specific mental disorder, could be tested by controlled clinical research. As with all such studies, it would be best if these were collaborative studies in which representatives from different disciplines engaged in mutual evaluation and consultation. Much more research is necessary on the sleep difficulties of elderly patients, with special attention to nonpharmacological, psychosocial therapeutic approaches. The elderly are likely to remain relatively refractory, in many cases, to pharmacological efforts. Taking any medication for sleep may add new hazards to their already complicated drug regimens. Low doses of carefully selected and monitored medications should be evaluated, but special efforts should be made to develop and evaluate psychosocial and behavioral intervention. Finally, for the individual who awakens with a sense of unsatisfactory sleep and may even claim that he or she did not sleep (though objective findings show otherwise), there is a need for a multifaceted attempt to understand the problem. Indeed, here is an area where an intersection of clinical psychiatry and psychoanalysis (which originated the 20th century investigation of sleep and dreams) and the big-behavioral disciplines of neurophysiology and sleep medicine may prove fruitful. The need for Independent Research and Evaluation of Hypnotics Although the problem of insomnia is widespread and the inappropriate use of hypnotics can have serious and even life-threatening consequences, the Federal government has supported few studies of hypnotic efficacy and safety. Nearly all of the investigations into prescription drugs and insomnia conducted in the United States have been sponsored by drug manufacturers. In the United Sates, there is no independent clinical research, supported either by the government or by philanthropic founda- tions, with which to seek answers to legitimate public health questions about hypnotic drugs. Although various agencies of the Public Health Service (especially the National Institute of Mental Health) have supported studies of sleep physiology and sleep disturbance associated with such specific mental disorders as depression or schizophrenia, the amount of clinical investigation directed toward treatment methods for insomnia has been negligible. A few studies of the efficacy of L-tryptophan, a dietary supplement not marketed as a drug, have been supported by the National Institute on Drug Abuse. There are three studies currently funded by the National Institute on Mental Health of behavioral -150-
techniques in the therapy of insomnia. The National Institute on Aging has begun funding investigations of the changes of sleep with age, but these are not as yet treatment-oriented. Some studies have been funded by the National Institutes of Health specifically on sleep apnea and nocturnal myoclonus, and the NIH has provided support for the multi- faceted studies of the Boston Collaborative Drug Surveillance Program, including several clinical studies of hypnotic drugs in hospital patients. There has been a modest amount of intramural clinical study within the Veterans Administration hospital system. These appear to be the entire inventory of Federally-funded clinically-oriented investigations of sleep problems associated with insomnia over the past decade. In contrast to the notable efforts of the National Institute of Mental Health and the Veterans Administration to collaboratively examine the efficacy and safety of antipsychotic drugs and antidepressant medication, hypnotic prescription drugs have been relatively neglected. Furthermore, relatively few investigators have presented research proposals to the Public Health Service to study drugs and insomnia. Approximately 150 studies of hypnotic drug efficacy* were reviewed in the course of preparing this report; all but a handful were sponsored by pharmaceutical companies. The results of most of these are extremely difficult to interpret. There has been a failure to set high standards of interpretability, replicability, and general validity in the published studies. In the design of the original experiments as well as in the report of the outcomes, the manufacturer has a strong influence on which aspects are to be emphasized. Reports of manufacturer-sponsored research are generally presented to the Food and Drug Administration by the manufacturer, not directly by the investigators themselves. There is no adequate, independent peer review of protocols or healthy competi- tion for support. Rigorous peer review depends upon the availability of a reasonably large group of individuals knowledgeable about the discipline. Until recently, there has been little to encourage the career development of sleep disorders specialists and sleep pharmacologists. It is noteworthy that nearly all the investigations of residual adverse effects of hypnotics on daytime psychomotor performance tests, have taken place overseas where support was provided by the respective foreign governments. The one American preliminary study was partially supported by the National Aeronautics and Space Administration (NASA), not by a drug manufacturer. Federal support of research in this area would increase the opportunity to obtain objective data concerning these and other adverse effects of these drugs. The committee believes that independent clinical pharmacology should be strengthened in the United States through the creation of support and networks of peer review, particularly for studies of hypnotic drugs. - *Tabulated in the Technical Supplement to this report. -151-
Within the present authority of the Alcohol, Drug Abuse and Mental Health Administration (ADAMHA) as well as the National Institutes of Health (NIH) there should be increased research into the nature of various types of sleep disturbance and appropriate measures for their relief -- including non-pharmacological methods. In funding this research, these federal agencies must take care not to repeat errors of past research. The committee recognizes that the drug regulatory system in this country relies heavily upon data produced by companies that have a com- mercial interest in the way the data are interpreted. Ways must be found to collect data and sponsor research which will supplement the information put forth by pharmaceutical companies in New Drug Applications and in post-marketing reports to the FDA. One approach recommended by the committee is to greatly expand the FDA's access to and use of qualified outside consultants. Establishment of a new standing committee on hypnotic drugs should be considered. If such a committee were formed its members should include independent sleep laboratory researchers, clinical pharmacologists, basic pharmacologists, drug addiction experts, consumer representatives, and specialists in appropriate branches of in- ternal medicine, primary care, psychiatry, and the behavioral sciences. F. Initiation and Coordination of Federal Efforts The committee recommends a search within the Public Health Service to identify an ongoing mechanism or mechanisms to identify and fund quality research and collect and disseminate information on the practical questions that have been raised in this study. As a first step, the committee suggests that an organizational entity be foraged within the Public Health Service to review the current scientific, educational, and administrative issues associated with sleep disorders and their treatment. Working with specialty societies, medical associations and academic health centers, this authority could begin the process of improving research and education and encouraging high standards of medical practice and clinically relevant research. It could-coordinate the consensus development and professional and public education efforts in the field of sleep disorders, as well as identify promising research for grant funding or intramural programs of the Public Health Service. The committee recognizes that many of the problems associated with treatment of insomnia and the medical use of hypnotic drugs resemble those encountered in the clinical management of other health care pro- blems -- which also require evaluation and professional education by those without a commercial stake in the results. The establishment of this central authority and coordination effort might eventually lead the way to a more organized focus on therapeutics and health care evaluation, ultimately resulting in better health care for those who suffer from a variety of widespread conditions which heretofore have been given low priority on the health science agenda. -152-
REFERENCES 1 2 3 4 5 6 Salkind, M.R. and Silverstone, T., A clinical and psychometric evaluation of flurazepam. British Journal of Clinical Pharmacology 2:223-226, 1975. Kales, A., Bixler, E.O., Scharf, M. and Kales, J.D., Sleep labora- tory studies of flurazepam: A model for evaluating hypnotic drugs. Clinical Pharmacology and Therapeutics 19:576-583, 1976. Dement, W.C., Carskadon, M.A., Mitler, M.M., Phillips, R.L. and Zarcone, V.P., Prolonged use of flurazepam: A sleep laboratory study, Behavorial Medicine 5:25-31, 1978. Association of Sleep Disorders Centers, Survey on Teaching About Sleep Disorders and Proper Use of Sleeping Pills in American Medical Schools. Unpublished manuscript available from W.C. Dement, Stanford University, Palo Alto, California. Miller, R.R., Prescribing habits of physicians: A review of studies on prescribing of drugs. Drug Intelligence and Clinical Pharmacology 8:81-91, 1974. Physicians' Desk Reference 33rd edition, p. 1432. (Oradel, New Jersey: Medical Economics Company) 1979. Kripke, D.F., Simons, R.M., Garfinkel, L. and Hammond, E.C., Sleep duration, insomnia and sleeping pill use. Psychiatry 36:103-116, 1979. -153- Archives of General
