I would like to offer you a brief answer and follow it with a longer discussion of this important topic.

The short answer is: No, I would not prescribe Seroquel® (quetiapine) for insomnia unless the pt. also presented another indication for neuroleptic medication, i.e., delusions, auditory hallucinations, or even paranoia without frank delusions. It is true that antipsychotic medications are sometimes prescribed as off-label treatments for sleep, but there are important downsides. Chronic use of any antipsychotic drug is associated with weight gain, diabetes and other "metabolic syndrome" features, and also with extrapyramidal symptoms. Neuroleptic malignant syndrome is not necessarily expected to occur, but it can, especially in an older patient who has never previously taken antipsychotic medication or who has mild Parkinsonism. If your patient is in fact elderly (you didn't say so), you should be aware that black-box warnings were added last year for the use of atypical antipsychotics in elderly patients because of an increased risk of death, and also that the usefulness of antipsychotic drugs for behavioral symptoms of dementia has been seriously questioned in a recent collaborative study. Finally, my colleagues who have prescribed Seroquel for sleep induction have told me anecdotally that low doses are not effective for this purpose, which I find not surprising given the low potency (per milligram) of Seroquel compared to other atypical antipsychotics.

So, then, what are the other options? (This is the beginning of the longer discussion).

Non-Pharmacological Interventions
I usually begin by offering a patient several suggestions about how to improve sleep hygiene, depending upon which ones apply. In my practice, which is exclusively geriatric, I find that patients who complain of interrupted, unrestful or otherwise inadequate sleep often try to make up for what they lose at night by remaining in bed most of the morning and/or napping extensively in the afternoons. In extreme, but by no means rare, cases the result is nearly complete sleep cycle reversal (awake at night, asleep in the day); this occurs even more frequently in elders with dementia. In any event, late morning bed rest or afternoon napping are both to be discouraged, so that the original architecture of the individual's sleep-wake cycles can be given a chance to be reset. Getting morning sunshine and exercise, as is recommended for patients with seasonal affective disorder, is another part of the effort to restructure the sleep cycle. The judicious use of caffeine, e.g., for the morning after a poor night's sleep, or in the afternoon in lieu of napping, can be helpful.

I have found it important to assess how realistic the patient's expectations in this area actually are. Someone with a lifetime history of poor sleep is not likely to achieve uninterrupted 8-hour nights of sleep. Age, cognitive impairment, chronic pain, certain medications and illnesses, and urinary frequency are other confounds. I try to tell patients frankly what I think can be accomplished using available techniques and pharmacological agents.

Over the Counter Medications
This category includes over-the -counter sleep inducers such as antihistamines, some pain medications, L-tryptophan, valerian, melatonin, herbal supplements and alcohol.

For most there is no data either supporting or negating their use. However, caution should be used with herbal supplements, L-tryptophan and valerian, since their manufacture is largely unregulated, and there have been reports of toxicity. Antihistamines are associated with anticholinergic toxicity and delirium in vulnerable elderly patients. Alcohol can be helpful in reducing the time to sleep onset, but because of its short half-life it is not expected to sustain sleep; moreover, some individuals who have consumed alcohol before bedtime wake up abruptly with palpitations and anxiety after three or four hours of sleep.

Antidepressants
Trazodone is frequently prescribed in subtherapeutic (for depression) doses, such as 25-150 mg. QHS, for insomnia. It is especially popular among geriatricians and was considered "first line" for management of insomnia in elderly patients in a recent expert-consensus study. However, there is little evidence that it actually affects sleep latency, and it is itself associated with orthostatic hypotension and consequent risk of falls. Other antidepressant medications, when prescribed in therapeutic doses, are effective in addressing depression-related insomnia. Remeron ® (mirtazapine) has been observed to have particularly prominent sedating properties. However, all antidepressants have been shown to be effective across the entire range of depression symptoms, including insomnia. No antidepressant has been shown to be effective in reducing sleep latency in the absence of clinically significant depression.

Hypnotics
For most of the first half of the twentieth century, opioids and barbiturates were widely used to induce and sustain sleep. However, tolerance to dose, dependence, and lack of safety in overdose posed considerable problems. Then, in the 1960s, a group of gabanergeric enhancers, the benzodiazepines, were introduced. The most commonly prescribed of the early benzodiazepine hypnotics, Dalmane ® (flurazepam), was effective but had a long half-life (47-100 hours), leading to sedation or loss of alertness in the morning hours. Flurazepam was also associated with tolerance to dose and with dependence in all but occasional users.

The next generation of benzodiazepine hypnotics was developed in the 1970s and included, among others, Halcion® (triazolam) and Restoril® (temazepam). Triazolam and temazepam had shorter half-lives than flurazepam but were similarly associated with the development of tolerance to dose and dependence. Cases of amnesia and severe disinhibition were also reported for both drugs, especially triazolam.

The 1980s saw the development of a group of non-benzodiazepine hypnotics, collectively referred to as "selective benzodiazepine receptor agonists" because they work through the benzodiazepine receptor even though they are not benzodiazepines per se. This group now includes, in order of the date of development, Ambien® (zolpidem), Sonata® (zaleplon), Ambien-CR® (zolpidem-extended release), and Lunesta® (eszopiclone). In 2005, Rozerem® (remelteon) was approved by the FDA in a single 8-mg. dose. Remelteon is a melatonin receptor agonist which has been shown to mainly improve sleep induction; it differs from the hormonal melatonin available over the counter. The FDA indications for zolpidem-extended release, eszopiclone and remelteon are not limited to short-term use, because they do not appear to induce tolerance to dose. Together, remelteon and the selective benzodiazepine receptor agonists, except for zolpidem-extended release, have comparatively short half-lives.

However, regardless of half-life, any pharmacological agent which purports to "put the brain to sleep," let alone keep it that way, can be expected to produce confusional states in vulnerable individuals, most notably the elderly. Disinhibition and various "twilight" or somnambulistic behaviors have been reported with zolpidem and widely disseminated in the popular press. Eszopiclone, which has the longest half-life of the selective non-benzodizepine hypnotics, can produce dangerously inattentive states in some individuals; for example, an elderly patient for whom I prescribed it began to leave kettles boiling, toaster ovens burning, etc.

For these reasons, if I must prescribe medication for sleep in the elderly, I have been using short-acting benzodiazepines, such as Ativan® (lorazepam) or Xanax® (alprazolam). This is actually an off-label use, since these medications have very short half lives. However, I have found that if awakening after 3 or 4 hours occurs, it is not typically accompanied by palpitations or rebound anxiety, as with alcohol consumed before bedtime. Lorazepam and alprazolam are also extensively metabolized, i.e., have few active metabolites, another advantage for elderly patients with physiologically delayed drug clearance.

Conclusions
In general I prefer behavioral over pharmacological interventions for the treatment of chronic insomnia, but some individuals are either unwilling or unable to follow such recommendations. Assuming that the insomnia cannot be relieved with treatment of comorbid depression or pain syndromes, I might first try either low-dose Trazodone --notwithstanding the potential problems with trazodone I described earlier--or a short-acting non-hypnotic benzodiazepine such as lorazepam or alprazolam. If these approaches are ineffective, a risk-benefit analysis needs to be performed before prescribing a newer drug such as zolpidem, zaleplon or eszopiclone (my experience with remelteon is thus far very limited, and I do not consider zolpidem-extended release to be appropriate for elderly patients because of the prolonged half-life). If the patient is elderly, one must weigh the risks of chronic insomnia--itself associated with falls, daytime sedation, inattention, accidents, depression and increased perception of pain--against the potential for common and unusual side effects of hypnotic medications.

Reference
Glazer WM, Dodghramji PP, Erman MK. Supplement on treatment of insomnia. J Clinical Psychiatry. 2006;67: (suppl 13)

Schneider LS, Tariot PN, Dagerman KS, et al. Effectiveness of atypical antipsychotic drugs in patients with Alzheimer's disease. N Engl J Med 2006;255:1525-38

On the other hand, these considerations should be weighed against the emotional and physical consequences of asking the patient to "tough out" dose reductions for the promise of greater safety and fewer clinical sequelae later on. Stopping or sharply reducing benzodiazepine doses can cause considerable discomfort and complications, including the possibility of a medically dangerous withdrawal reaction if the tapering is done too abruptly. Factors influencing the decision include the age and level of functioning of the individual, as well as the length of time on the benzodiazepine. Your patient, a fairly well functioning and cognitively intact individual who has been taking the drug for a period of years might well remain on it, or at least on a somewhat lower dose (i.e., the current 3 mg. or slightly less). A very slow taper, over many months for example, might be the most appropriate course for some older patients. However, the "downregulation" of gabanergic neuroreceptors in the central nervous system that occurs over a period of years may mean that the individual may never feel entirely well off of the drug.

If I do decide to keep a patient on benzodiazepines chronically, I am careful not to increase the dose and to occasionally try small incremental decreases; the possibility of starting a slow taper can always be revisited. Maintaining an elderly patient chronically on benzodiazepines should done with an individual you know well and trust, not someone who deals with prescriptions unreliably or for other reasons is at risk to abuse the medication. Even then, I prefer to keep a "tight rein" on prescriptions.

Also, to avoid dealing with these problems later on, I try to avoid starting elderly patients on benzodiazepines whenever possible. Alternatives include the use of serotonin-enhancing antidepressants (SSRIs) or sedating antidepressants (tricyclics, trazodone, mirtazapine). SSRIs, however, appear to be more effective for obsessional than for generalized anxiety pictures. If reduction or stopping of benzodiazepines appears to be necessary in a frail or medically unstable patient, this intervention may need to be structured as a medical detoxification and carried out in an inpatient hospital setting. For most purposes in the elderly, short-acting, extensively metabolized benzodiazepines such as Lorazepam are preferred to older drugs with long half-lives such as diazepam, but detoxification is sometimes facilitated by substituting a benzodiazepine with a slightly longer duration of action, such as clonazepam, to achieve a "smoother" withdrawal involving less frequent peaks and troughs in serum drug levels.

Reference
Atkinson RM: Substance abuse, in Comprehensive Textbook of Geriatric Psychiatry. Edited by Sadavoy J, Jarkvik LF, Grossberg GT, Miller BS. American Association for Geriatric Psychiatry, WW Norton & Company, Ltd, New York, 2004. pp 723-761.