Insomnia State of the Science: An Evolutionary, Evidence-Based Assessment

EDITORIAL Insomnia State of the Science: An Evolutionary, Evidence-Based Assessment Daniel J. Buysse, MD Sleep and Chronobiology Program, University of Pittsburgh, Pittsburgh, PA Disclosure Statement Dr. Buysse has served as a paid consultant for Actelion, Neurogen, Biosciences, Cephalon, Lilly, Merck, Pfizer, Respironics, Sanofi-Aventis, Sepracor, Servier, and Takeda; has received research support from Pfizer, SanofiAventis, Sepracor, and Takeda; and has received support from the International Meetings and Science supported by Pfizer. Address correspondence to: Daniel J. Buysse, MD, 3811 O’Hara Street, Room E-1127, Pittsburgh, PA 15213; Tel: (412) 246-6413; Fax: (412) 2465300; E-mail SLEEP, Vol. 28, No. 9, 2005 1045 Editorial—Buysse of insomnia. The Conference Statement, reprinted in this issue of SLEEP, marks a critical juncture in the history of insomnia research and clinical practice. The statement is a well-balanced and accurate assessment of the current state of knowledge regarding chronic insomnia and its management. More importantly, it identifies several key issues to be addressed in future research. To fully appreciate the potential impact of the Conference Statement, it is worth reviewing some of the major points of the last consensus statement, “Drugs and Insomnia: The Use of Medications to Promote Sleep,” published in 1983.8 Among the enduring legacies of that statement are the view of insomnia primarily as a symptom of other underlying disorders; advocacy for the role of sleep disorders centers in the evaluation of chronic insomnia; the distinction between transient, short-term and chronic insomnia subtypes; the recommendation for short-term and intermittent use of hypnotics; the recommendation for short-acting benzodiazepines in most situations; and the recommendation for treatment of underlying conditions as the first approach in chronic insomnia, with adjunctive treatment of insomnia only as a second step. The conclusions of the 1983 statement had wide-ranging effects on clinical practice, research and policy. For instance, the consensus statement supported the adage taught in seemingly every lecture on the topic, “Insomnia is a symptom, not a diagnosis.” But arguably the greatest effect related to how drugs were to be used in treatment of chronic insomnia. The NIH consensus statement, together with a set of other self-reinforcing factors including FDA class labeling, medical regulatory board practices, third party payer restrictions, and pharmaceutical industry reluctance, all emphasized the clinical practice of short-term, intermittent hypnotic administration for chronic insomnia. There was only one problem with this approach: It was not based on empirical data. Simply put, there were no studies actually examining the comparative efficacy and safety of different modes and durations of hypnotic administration for chronic insomnia. Some of the available studies at the time had included drugs no longer in common use today (eg, flurazepam), and others included doses of medications higher than currently recommended (eg, triazolam). Some of the 1983 recommendations may also have stemmed from fears regarding the serious adverse effects seen with barbiturates and the so-called “non-B non-B” (non-barbiturate, nonbenzodiazepine) drugs such as methyprylon and meprobamate. The major issue, however, was that the consensus development process in 1983 had a much lower criterion for evidence than the analagous process in 2005. The clinical reality is that short-term pharmacologic treatment with short-acting hypnotic drugs is not adequate to treat chronic insomnia in many cases. Clinicians and patients have borne the burden of therapeutic guilt when using hypnotics for longer-term treatment, even for patients in whom this strategy is apparently successful. Thus, the 1983 consensus statement had a generally conservative tone, but it was not rigor- INSOMNIA IS WIDELY RECOGNIZED TO BE THE MOST COMMON SLEEP PROBLEM. CHRONIC IMPAIRING INSOMNIA AFFECTING APPROXIMATELY 5-10% of the adult population.1 The consequences and morbidity of chronic insomnia can be substantial and can include increased health care utilization, impaired quality of life, increased risk of psychiatric disorders (including depression), increased risk of falls and hip fractures, and worse outcomes for comorbid disorders.2-6 New treatments and new applications of existing treatments are being developed among both psychological/behavioral and pharmacologic strategies. Unfortunately, despite the burgeoning interest in insomnia, its consequences and its treatment, there are no empirically-validated treatment algorithms or clinical pathways to guide clinicians. The absence of such guidelines may be one of the reasons that few patients have access to empirically-validated behavioral treatments, and that a non-approved and relatively littlestudied drug, trazodone, is among the most frequently-prescribed medications for insomnia.7 Given this knowledge vacuum, clinicians, researchers, and policy-makers look for guidance where they can find it. One of the most authoritative and trusted sources is the National Institutes of Health and its consensus development process, run by the Office of Medical Applications of Research (OMAR). Over the past 25 years, OMAR has conducted more than 100 consensus development conferences on a wide range of medical topics. As a result of petitions from investigators in the field, the National Center on Sleep Disorders Research (NCSDR), and organizations including the American Academy of Sleep Medicine, Sleep Research Society, and National Sleep Foundation, OMAR agreed to hold a State of the Science conference addressing chronic insomnia. OMAR conducts 2 types of conferences: Consensus Conferences in cases where a substantial body of high-grade evidence exists, and State of the Science conferences for topics where the quality and quantity of evidence is more limited. The insomnia conference, entitled “Manifestations and Management of Chronic Insomnia in Adults” was held on June 13-15, 2005. It included 1.5 days of expert presentations, evidence-based literature reviews, and input from the scientific and general public. The expert panel that crafted the final statement included scientists with outstanding credentials in related fields, but who have not published on the specific topic 1983 Statement, which was at once both more directive and less data-based. The 2005 Statement also offers a sober assessment of future research needs, and calls for “a substantial public and private research effort” to address those needs. Now is the time for us to get to work, so that we have substantial progress to show for the next conference. Hopefully that one will be a true Consensus Conference, and hopefully it will occur sooner than another 20 years. ously evidence-based according to today’s standards. Some of its conclusions limited clinical research and practice for the past 20+ years. Th (...truncated)