• the huge group of untreated patients,
• the chronicity of complaints.
• numerous methods recommended for treatment, and
• the somewhat unknown benefit/risk ratio of several drugs and pharmacological treatment strategies available to improve sleep in a long term.

Data obtained in Europe underline the importance to make great efforts to increase the proportion of recognition of sleep problems among general practitioners (GPs) and patients with insomnia. A representative survey in Germany (state-wide, random-route, birthday-key) showed that only 49% of subjects suffering from difficulties in initiating or maintaining sleep for more than two years ever consulted a doctor because of their sleep disturbances (Figure 1). That also may GPs are unaware of the existence of sleep problems in their patients is indicated by the results of an ongoing study in Germany (Göttingen and Freiburg, personal communication). Here, 24 GPs were evaluated regarding their knowledge about sleep disorders. 21% of 2490 patients older than 14 years suffered from a full-blown insomnia according to DSM-III-R. In only 46% of the cases the physicians were aware of their patients' severe insomnia. In many cases, insufficiently nonpharmacologic treatment strategies were recommended (relaxation training to 6% of patients). In 49% of the patients psychologic or psychotherapeutic techniques were never applied and inadequate behaviour was not corrected by the doctors (33% of the patients would watch TV and 12% would eat in their bed to get better sleep). Application of specific training and education programs to the physicians clearly improved their diagnostic and therapeutic quality. For this purpose, the GPs underwent a 6-hour training course on sleep disorders. This one-day training increased the diagnostic sensitivity for insomniac patients by 15% (severe insomnia), and by 11% (moderate insomnia) in the group of trained GPs compared to a group of untrained GPs (Figure 2).

In terms of the adequacy of therapeutic intervention, there is agreement between sleep experts that the treatment should aim at causes of the disease. Furthermore, treatment concepts need to consider that every hypnotic treatment of an insomniac patient should be combined with at least some basic elements of nonpharmacologic treatment strategies. Although the use of hypnotics is widely accepted, it is still debated whether physicians should reduce prescribing hypnotics or whether their patients would benefit from a more generous prescription of sleep inducing drugs. In Germany, only 13% of insomniacs in general practice take hypnotics three times or more per week, only 4% of all insomniacs in the whole population use hypnotics daily and more than a half (57%) of all these insomniacs have never taken any sleep inducing drug (Figure 3). However, it should be considered that as many as 45% of the patients taking their sleep medication daily, still continue to have sleeping problems frequently or always.

A common concern with hypnotic agents is the potential for drug abuse, misuse and addiction. In spite of the reduced risk of abuse and dependence associated with the use of newer hypnotics, sleep expert groups recommend that patients should restrict their daily hypnotic intake to the shortest possible period and not exceed a treatment period beyond 4 to 8 weeks. The corresponding recommendations are based on the view that good clinical practice discourages chronic use on the lack of data establishing the safety (e.g., dependency liability) and efficacy with chronic use. This is a pitfall for patients with insomnia which in 75% of all cases (representative survey, Germany) continuously or recurrently suffer from insomnia for a period of at least 2 years. 82% of patients in general practice (Göttingen and Freiburg, Germany) suffer from insomnia according to the diagnostic criteria of DSM-III-R for more than 1 year and 29% for more than 10 years. 51% of these insomniacs continued suffering from sleep problems in spite of having tried at least one way of treatment. Obviously, these patients need not only more efficacious but also longer term treatment.

It should be noted that the discussion about daily long term treatment with hypnotics must differentiate between the drugs used. There is evidence from an epidemiological study on 550 physicians in Germany that newer hypnotics (i.e., imidazopyridine, cyclopyrrolone) are prescribed for a significantly shorter average duration than benzodiazepine hypnotics. During a period of 5 years 29,302 patients were treated with hypnotics, 24,571 of the patients were drug naive before treatment. 5.5% to 10% of the patients treated with 3 different benzodiazepine hypnotics received their drug continuously for more than 90 days after the first prescription. In contrast only 1.6% of the patients taking the imidazopyridine zolpidem were long term users (Figure 4).

Some sleep experts underline that intermittent use of hypnotics is advisable for long term treatment, as this may benefit patients and prevent the development of dependence. The following concepts for the intermittent long term application of hypnotics have been proposed and are currently being tested:

• intermittent therapy according to patients' requirements with controlled prescription, restricting hypnotics to 10 tablets per 3 weeks,
• daily standard intermittent therapy with 2 randomized chosen drug-free nights per week,
• weekly standard intermittent therapy with 2 to 4 weeks of daily hypnotic intake followed by a 2- to 4-week drug-free interval, on the basis of continuing sleep improvement after treatment periods of 4 weeks,
• controlled prospective intermittent therapy according to patients' requirements combining the advantages of intermittent use of hypnotics (maximum of 3 to 4 daily dosages per week) with a prospective approach to medication intake in accordance with the patients own decision, and
• combination therapy combining low doses of sedating antidepressants given every night with intermittent use of benzodiazepine receptor agonists.

A preliminary study on controlled prospective intermittent therapy was performed in a small number of 12 patients with primary insomnia. Subjective sleep quality improved by 43% from baseline during the first three weeks and reached 68% improvement after 6 months of treatment (Figure 5).

In conclusion, a multidimensional approach for research and education is urgently needed to improve the diagnostic and therapeutic quality of sleep medicine.