Key moments, in order
The DSM's insomnia describes a symptom, not an entity. "Primary insomnia" just means we could not find the cause; it is defined by our ability to assess, not by shared characteristics, family history or a constellation of signs.
A night or two of insomnia after something stressful, a death, a divorce, a hospitalization, is a normal and inevitable part of life.
If you meet DSM criteria for a psychiatric disorder, you are excluded from the sleep medication studies. So the primary insomnia research and the sleep guidelines are nearly worthless for almost every patient in a psychiatric clinic.
The medications with regulatory approval are benzodiazepines, Z-drugs, DORAs, low-dose doxepin and ramelteon. Follow those guidelines and start Z-drugs first line, and you get a whole host of problems.
Tailor the approach to the cause: is it really psychological, or is it coming from something like the ruminations of a primary anxiety disorder?
When there is no more than a mild Axis I condition, insomnia is more likely psychological, and the appropriate intervention is CBT-I and sleep hygiene.
In anxiety disorders, patients cannot engage with the psychological and behavioral changes they need if they are not sleeping.
Sleep medications can be sorted by system: GABA, melatonin and orexin, cholinergic, histamine, plus agents with mixed histamine, serotonin and alpha effects. Gabapentin is a favorite that does not actually work through GABA.
Hydroxyzine, trazodone and gabapentin are the broadly applicable sleep medications for most to all disorders, with melatonin as a runner-up.
With the liquid, have the patient get a small 1 mL syringe: each mark is a milligram, so 3 mg is easy to draw up, and it is covered.
It is a TCA, but at lower doses it is probably the most selective antihistamine we have.
It is a regular antihistamine and overly sedating in the daytime for most people, so we use it almost only at night, and it is not the first sleep aid we reach for.
Start incredibly low, half of a 50 mg tablet, and take it one or two hours before the desired bedtime, because onset is slow. It is long lasting and good for sleep maintenance.
In our opinion it is just too anticholinergic for sleep. We do not like to use it unless we are looking for anticholinergic effect alongside an antipsychotic.
Even at low doses there is considerable appetite increase and weight gain, which is a reason to avoid it as a sleep medication.
Say: this works great for a lot of people, and for some it does not work at all and might even make sleep worse. We do not know which you are, so try it. A small share do have that paradoxical reaction.
Its half-life is probably about 20 hours, which is why it is good for sleep maintenance.
If someone is not very sensitive to it, it can go to 200 mg and beyond, and then it is an antidepressant too.
Low-dose trazodone is not a small antidepressant. At 50 mg it is not reuptaking serotonin, which is its antidepressant mechanism, so you have to reach higher doses for that.
Depressed patients get less deep sleep, more light sleep and a higher REM percentage.
Oddly, REM deprivation (waking people before they reach REM) can treat depression.
In depression, the goal is not just more sleep but a habit of sleeping at night and being active during the day, built into behavioral activation. Sleep problems do tend to improve once the SSRI dose is adequate, but one medication rarely fixes everything right away.
Disrupted sleep makes cycling more likely, so stabilize sleep alongside the mood stabilizer.
As an adjunct, it is a nice sleep and anxiety support while you truly treat with a mood stabilizer. In bipolar disorder we are much more likely to reach for it for sleep, though it is only ever adjunctive, for things like anxiety and insomnia.
Do not use paroxetine for sleep in bipolar disorder: it is a potent SSRI with a lot of withdrawal effects.
In bipolar disorder, be wary and careful above 150 mg.
Move all antipsychotic dosing to once a day at night. It covers the psychosis and the sleep, keeps patients more active and less sedated by day, and improves side effects you would not expect.
At a high enough dose you will not absorb enough with one meal, so you may have to split it twice a day.
Imagery rehearsal therapy works well for nightmares, and DBT has a good nightmare protocol.
Tell every patient: this medication will not solve your problems or take away what happened, but it will help. It supports the primary treatment, which is the psychotherapy and what you do in your life.
Recommend any named, specific trauma psychotherapy: CPT, trauma-focused CBT, even EMDR.
Personality disorders can disrupt sleep through maladaptive coping and emotional dysregulation.
Go off the clinical history rather than sending everyone for a sleep study. Refer when you hear something at least suspicious for apnea, such as multiple nighttime awakenings.
A basic screener such as sleephealthscreen.com covers insomnia disorders, circadian disorders, obstructive sleep apnea, restless legs, narcolepsy and parasomnias.
Do not give sleep medications to someone with untreated sleep apnea. You will be suppressing the respiratory drive they need.
Recognize when you are dealing with a circadian disorder rather than insomnia. There, go more for melatonin and a bright light regimen.
It has a double effect, shifting the sleep clock and sedating. To move the clock, give it earlier than you would think, several hours before the intended bedtime.
A simple start: 300 mg, go to 600 if it is not enough and call; if it is way too much, call and switch to 100s. Titrate to effect.
If you are on your phone, you cannot be asleep. People up on their screens may still ask for a sleep medication.
Handing someone a generic sheet of sleep hygiene rules does not work, but that does not mean the things on it do not work.
Get a good sense of what they actually do at night, then ask what they are willing to change by the next visit. Small steps: even one small thing, even if they try it once.
Move the phone charger out of the bedroom. People go to bed swiping, wake at 3 a.m., can't fall asleep and go back on the phone. That one change has cured insomnia and made drastic improvements in many other areas of patients' lives.
There is no sleep algorithm guiding psychiatrists, and the American sleep guidance recommends against all the medications we actually use for insomnia.
Each moment is a point made in the episode, linked to the second it was said. The topic links open everything the podcast has said on that topic.