A Short-Term Sleep Prescription Is Not a Short-Term Treatment Plan

This is a fictional composite. It is not a description of an actual patient.

A patient being treated for depression reports that sleep has become the hardest part of the week. The psychiatrist and patient agree to try a sleep medication for a limited period. The note says “short term.” The prescription is sent. At the next contact, the conversation begins with whether another supply is needed.

Nothing in that sequence establishes that the initial decision was wrong. It does expose an unanswered question: what was supposed to happen between starting the medication and deciding whether to continue it?

“Short term” describes an intention. A treatment plan makes that intention reviewable.

In a September 3 JAMA Network Open study, Cloughesy and colleagues examined claims from 129,287 adults starting zolpidem, eszopiclone, or zaleplon. Inclusion required a year without Z-drug fills and at least eight years of subsequent enrollment in commercial or Medicare Advantage coverage.

In the first year after the index-fill year, adjusted proportions filling a Z-drug were 15.6% after an initial 1-14-day supply and 38.1% after 25-34 days. The adjusted difference was 22.4 percentage points, with a 95% confidence interval of 21.8-23.0. The association persisted, but narrowed, at seven years.

These were prescription fills, not observed ingestion or proof of uninterrupted treatment. Initial supply was not randomized. More severe or persistent insomnia could explain both longer initial prescriptions and later fills. Requiring prolonged enrollment also selected the cohort. The findings do not establish that shortening a prescription would prevent long-term use, identify an optimal duration, or diagnose dependence. [1]

The useful response is a prescribing question: when we select a quantity, have we also decided how the treatment will be evaluated?

For the psychiatrist, a sleep complaint belongs inside the current clinical assessment. The practical task is to establish what the patient wants changed and what else needs attention, rather than allowing “insomnia” on the medication list to stand in for the entire formulation.

The 2025 VA/DoD chronic-insomnia guideline directs clinicians to review sleep history and relevant contraindications before short-term pharmacotherapy, including other sleep disorders, daytime sleepiness, respiratory or cognitive impairment, substance-use history, and medication interactions. It also calls for discussion of a deprescribing plan when medication is appropriate. [2]

In the fictional encounter, the clinician could make the intended benefit concrete: less time awake at the beginning of the night, fewer prolonged awakenings, or better function the following day. Those examples are prompts for a conversation, not interchangeable outcomes. A patient who sleeps longer but cannot function safely the next morning has information the review needs to capture.

An instruction to contact the office “if needed” leaves much to interpretation. A scheduled reassessment gives both parties a place to bring what happened.

For chronic insomnia, the VA/DoD guideline recommends cognitive behavioral therapy for insomnia, or CBT-I, and suggests it over medication as initial treatment. It suggests against sleep-hygiene education as the sole treatment. Behavioral treatment may need adaptation or delay in particular clinical circumstances, including bipolar disorder or acute mental-health symptoms. [2]

The AASM's 2026 guideline adds a useful distinction. It conditionally favors CBT-I plus medication over medication alone, but conditionally advises against preferring that combination over CBT-I alone. Both recommendations carry low certainty. The evidence concerns starting the treatments together; it does not settle every sequence of adding, continuing, or withdrawing treatment. [3]

That leaves room for individualized prescribing. It also makes “we discussed sleep hygiene” an inadequate description of whether CBT-I was offered or started.

For the fictional patient, the next step might be a referral, a discussion of a suitable delivery format, or an explicit account of why the proposed treatment cannot yet begin. The note should make the situation recognizable at the next visit. “Referral placed,” “appointment scheduled,” and “treatment started” describe different events.

The purpose is not to turn access difficulties into a test of patient cooperation. It is to avoid writing a plan that assumes a service has begun when nobody knows whether it has.

FDA warnings for these Z-drugs include rare complex sleep behaviors that can cause serious injury or death. Such events can occur after the first dose or with continued use, and at lower doses. FDA advises stopping the medicine and contacting a health professional immediately if a complex sleep behavior occurs. It advises clinicians not to prescribe these agents after a previous such episode with any of them. FDA also warns about alcohol, combining sleep medicines, and next-morning impairment. [4,5]

This information belongs at initiation, with a clear way to obtain help. A short supply is not a substitute for counseling.

At reassessment, “it helped” deserves elaboration. What improved? What became harder? What did the patient actually take? The decision can then address an observed response instead of merely the arrival of a refill request.

An article about the first prescription should not become an instruction to abruptly stop an established one.

NICE NG215 recommends an agreed, documented management plan, regular reviews, and clarity about who holds prescribing responsibility. Its withdrawal guidance generally favors a slow, individualized, stepwise reduction for Z-drugs, with flexibility when symptoms become difficult. Abrupt cessation is reserved for exceptional medical circumstances. The FDA's instruction to stop after complex sleep behaviors is a specific safety exception requiring prompt clinical contact. [4,6]

The first-prescription problem and the established-treatment problem therefore require different conversations. In one, the clinician and patient are defining a trial. In the other, they are reassessing a treatment history. A chart phrase from months ago cannot do either job for them.

The following four questions are an editorial way to organize the next encounter, not a validated checklist:

The prescription can carry a limited quantity. The clinical plan has to carry the patient to the next decision.

1. Cloughesy JN, Thunell J, Liu X, et al. Initial Z-Drug Prescription Duration and Long-Term Use. JAMA Network Open. September 3, 2026. doi:10.1001/jamanetworkopen.2026.32006. Open full text; observational claims study.

2. VA/DoD. Management of Chronic Insomnia Disorder and Obstructive Sleep Apnea. 2025. Public guideline; relevant recommendations and clinical considerations reviewed.

3. Buysse DJ, Arnedt JT, Buenaver L, et al. Combination treatment for chronic insomnia disorder in adults: an AASM clinical practice guideline. J Clin Sleep Med. 2026;22:56. Open full text; conditional guidance.

4. FDA. FDA adds Boxed Warning for risk of serious injuries caused by sleepwalking with certain prescription insomnia medicines. Drug Safety Communication. April 30, 2019.

5. FDA. Taking Z-drugs for Insomnia? Know the Risks. Public patient-safety guidance.

6. NICE. Medicines associated with dependence or withdrawal symptoms: safe prescribing and withdrawal management for adults. NG215. Public recommendations; UK guidance informing general clinical principles.

Educational Disclaimer: The Psychiatric Record provides general educational information for psychiatric and mental-health professionals. Content does not constitute medical, legal, regulatory, compliance, billing, or other professional advice; does not establish a standard of care; and is not a substitute for independent professional judgment. Appropriate assessment and treatment depend on the patient, setting, available interventions, evidence, and current authoritative guidance. This article does not recommend starting, continuing, stopping, or changing any medication or treatment for any individual, and it does not set a universal prescribing duration or withdrawal schedule. The patient described is a fictional composite; no individual is depicted.