RCGP SCA Algorithm — UK Primary Care
Look for the dangerous and treatable causes first: mood and suicide risk, mania, OSA and driving, night sweats with systemic features, nocturnal chest pain or breathlessness, nocturnal seizures, substance withdrawal.
Insomnia is a common symptom of depression, mania, OSA, pain and withdrawal states. Treating the sleep with a hypnotic while missing the cause can be dangerous — sedatives worsen untreated OSA, mask withdrawal and add overdose risk in someone who is suicidal.
The sleep diary is the single most useful tool: it shows the pattern, separates time in bed from time asleep, and provides the numbers for sleep restriction in CBT-I.
Causes change over time, so the differential is re-run at every review — especially before any further prescription. Restless legs, OSA and circadian disorders each have their own treatment, and none responds well to a hypnotic.
Scores support but do not replace clinical judgement: an Epworth score alone does not decide OSA referral, and the ISI tracks response rather than making the diagnosis.
The three situations need different plans. Acute insomnia usually settles once the trigger passes; chronic insomnia is maintained by behaviours and beliefs that CBT-I targets; long-term hypnotic users need a planned, patient-paced reduction rather than abrupt cessation.
Most insomnia is managed in primary care. Referral is for the dangerous causes, for specific sleep disorders that need a study or specialist treatment decisions, and for insomnia that persists despite genuine CBT-I.
CBT-I addresses the maintaining factors and its benefit can last after treatment ends; hypnotics mainly give short-term symptom relief and do not treat them — which is why CBT-I comes first for chronic insomnia.
At every hypnotic prescription: the stop date agreed now, no automatic repeat, the next-day impairment warning. New prescribing: avoid combining with alcohol, opioids or other sedatives wherever possible. Existing co-prescribing (for example a hypnotic with an opioid) needs prompt medication review, risk assessment, a deprescribing plan and clear safety-netting — warn about respiratory depression and overdose, particularly with OSA, COPD, frailty or alcohol use. Take extra care in older adults (falls), respiratory disease and suspected OSA. Check doses in the SPC and your local formulary.
What to say: "Sleeping tablets can help for a short spell, but they tend to stop working if taken every night and can be hard to stop. The sleep programme is harder for the first week or two, but it's the treatment that keeps working after it ends."
NICE TA77 limits Z-drugs to severe insomnia, at the lowest dose, for short periods, and found no compelling difference between them. NICE TA922 recommends daridorexant for long-term insomnia only when CBT-I has been tried and not worked or is unavailable or unsuitable, with assessment within 3 months. The repeat prescription is where long-term use starts — no automatic repeats.
Next steps: see Benzodiazepines and z-drugs · Snoring and sleep apnoea · Restless legs syndrome · Depression · Anxiety disorders · Menopause and HRT · Driving and the DVLA
Gradual, symptom-led tapering with psychological support is more successful than abrupt withdrawal and avoids withdrawal seizures. The pace belongs to the patient; the direction is agreed together.
Sleep hygiene alone is rarely enough for chronic insomnia, but it supports CBT-I. The special groups are where sleep restriction and hypnotics carry the most risk.
Persistent insomnia most often reflects a missed cause, CBT-I that was never really delivered, or a maintaining factor such as caffeine, alcohol or medicine timing — not the need for another hypnotic.
Fixed review points stop short courses becoming long-term repeats, and each review is a chance to re-check mood, OSA and driving.