It is well documented that patients with schizophrenia demonstrate disturbances in sleep continuity and sleep architecture as well as disrupted circadian rhythmicity. These sleep disturbances are associated with reduced quality of life and social functioning. Insomnia might be secondary to the hyperarousal caused by positive psychotic symptoms. However, in recent years, it has been acknowledged that insomnia often persists beyond adequately dosed antipsychotic treatment. Residual insomnia in schizophrenia, defined as insomnia persisting beyond otherwise adequate antipsychotic treatment, covers a range of insomnia symptoms that could be secondary to other potentially treatable conditions such as medical or psychiatric comorbidity, substance use, use of medication that might interfere with sleep or a primary sleep disorder. A thorough evaluation of the insomnia diagnosis is therefore warranted. This includes a careful patient history and examination that addresses sleep and wake function, as well as common medical, psychiatric and medication-/substance-related conditions.
The complex problem of residual insomnia in schizophrenia is currently addressed by substituting or adding sedating drugs, but current treatment guidelines suffer from a lack of information on the best clinical practice. Substituting an otherwise efficient antipsychotic drug-therapy with a more sedating drug carries the risk of psychotic relapse during the switch, and adding another drug carries the disadvantages of polypharmacy including increased rate of side-effects, drug–drug interactions, patient non-compliance and medication errors.
To investigate the evidence base for the different treatment options in residual insomnia in schizophrenia, we performed a systematic literature review. We searched the major bibliographic databases for randomized clinical trials (search terms: sleep and schizophren*) supplemented by review of reference lists of relevant articles. We found only five randomized clinical trials investigating pharmacological treatment options for residual insomnia in schizophrenia. Two trials examined the efficacy of switching antipsychotic drug to, respectively, risperidone  or its active metabolite, paliperidone . The data suggest that risperidone might be a good choice for this particular indication, but methodological limitations and the lack of comparison with other second-generation antipsychotics weaken the suggestion. Three studies evaluating the efficacy of melatonin showed improvement in certain sleep characteristics (e.g. sleep duration and sleep efficiency) [3-5], but two of the studies [3, 4] came from the same group of researchers presumably using the same sample of patients. We found no published randomized clinical trials investigating non-pharmacological treatment of residual insomnia in schizophrenia.
Thus, our attempt to review this clinically important subject was limited by the few interventions tested experimentally and difficulties making a thorough comparison of the tested interventions due to methodological limitations. These included small sample size, selective reporting, lack of specification of primary outcome measure and a high level of variability regarding assessment tools and allowed comedication.
The benzodiazepines and benzodiazepine-like drugs (Z-drugs) are frequently prescribed as hypnotics, but they have not been investigated for residual insomnia in schizophrenia. These drugs are generally only recommended for short-term therapy due to the development of tolerance and dependence. Because residual insomnia in schizophrenia is a chronic condition, these drugs are generally not recommendable. An additional reason to avoid prescription of benzodiazepines is their deteriorating effect on cognitive function and their impairment (rather than correction) of sleep architecture. Particular care should be taken in patients' comorbid for a sleep-related breathing disorder or a comorbid history of alcohol or drug abuse.
Other possible treatment options mentioned in the literature, but not systematically investigated in schizophrenia, include add-on treatment with drugs with antihistamine properties including various antidepressants, antipsychotics and antihistamines. They are extensively prescribed (or sold over-the-counter for some antihistamines), but scarce evidence exists regarding efficacy, tolerability and safety when used off-label as hypnotics. The use of classic antihistamines (e.g. diphenhydramine) is limited by anticholinergic (including cognitive impairment) and cardiovascular adverse effects. Some evidence suggests the use of low-dose doxepin and trazodone (tricyclic antidepressants) as efficacious hypnotic drugs in primary insomnia and not associated with tolerance, weight gain or abuse potential. However, they have only been sparsely investigated in insomnia not associated with depression and are not available in all countries. The addition of a low-dose quetiapine as hypnotic agent has gained widespread acceptance without any evidence base of randomized controlled trials to support its use. Observational studies have found that even the low doses are associated with the development of adverse metabolic effects.
In conclusion, the evidence base is currently too scarce to extract any valid clinical recommendations regarding the treatment of residual insomnia in schizophrenia. Future studies should focus on rigorous evaluation of the switching and augmenting strategies currently applied in clinical practice and also on evaluation of psychological behavioural treatment methods. The latter is especially important in this vulnerable population of patients already chronically medicated.