Midazolam
| Evidence Level: L2 | Predicted Indications: 10 |
Table of Contents
Midazolam: From Anxiety to Insomnia
One-Sentence Summary
Midazolam is a short-acting benzodiazepine, originally used for anxiety and sedation. The TxGNN model predicts it may be effective for insomnia. The data compiled for this report include 2 clinical trials (neither tests midazolam for chronic insomnia) and 10 retrieved publications, of which about 5 are directly relevant. These include several 1980s–1990s studies of oral midazolam as a hypnotic.
Quick Overview
| Item | Content |
|---|---|
| Original Indication | Anxiety (no approved-indication text is recorded in the Singapore licence data) |
| Predicted New Indication | Insomnia |
| TxGNN Prediction Score | 99.74% |
| Evidence Level | L2 |
| Singapore Market Status | ✓ Marketed |
| Number of Registrations | 6 |
| Recommended Decision | Hold |
Why is This Prediction Reasonable?
Detailed mechanism-of-action data is not available in the compiled data. Midazolam is known to be a GABA-A receptor positive allosteric modulator. This produces sedation, anxiolysis and sleep induction, so the mechanism is consistent with a hypnotic effect.
Anxiety and insomnia often occur together, and both respond to enhanced GABAergic inhibition. Oral midazolam was studied as a sleep medication in the 1980s–1990s. These studies included dose-finding work and a 14-day comparison with flurazepam in chronic insomnia patients. This prediction is therefore closer to an established-use signal than to novel repurposing.
Its current relevance is limited by tolerance, dependence, next-day impairment, and the availability of newer hypnotics.
Clinical Trial Evidence
| Trial Number | Phase | Status | Enrollment | Key Findings |
|---|---|---|---|---|
| NCT02142595 | Phase 4 | Completed | 111 | Dexmedetomidine vs midazolam sedation with spinal anaesthesia in prostate resection, comparing postoperative sleep quality. This is a perioperative setting, so it is indirect evidence only. |
| NCT00886314 | N/A | Unknown | 80 | Clown doctors vs sedative premedication for preoperative anxiety in children. It does not address insomnia, so the match is likely spurious. |
Literature Evidence
| PMID | Year | Type | Journal | Key Findings |
|---|---|---|---|---|
| 2121802 | 1990 | RCT | J Clin Psychopharmacol | Randomised, double-blind multicentre study of sleep, performance and plasma levels in chronic insomniacs over 14 days of flurazepam vs midazolam. |
| 2229461 | 1990 | RCT | J Clin Psychopharmacol | Executive summary of the same multicentre flurazepam vs midazolam study. |
| 6138072 | 1983 | Clinical study | Br J Clin Pharmacol | In a double-blind study of 30 patients with insomnia secondary to neuromuscular disease, midazolam 15 mg was an effective hypnotic. It was better tolerated than Vesparax, which caused hangover. |
| 6120704 | 1981 | Dose-finding study | Arzneimittel-Forschung | Multicentre pilot in 75 inpatients with mild to moderate insomnia, testing oral midazolam 10–30 mg for efficacy and tolerance. |
| 2883820 | 1986 | Review | Acta Psychiatr Scand Suppl | Review of hypnotic use. Benzodiazepines with differing profiles are all effective, but dose and pharmacokinetics matter. |
| 17988972 | 2007 | Review | Orv Hetil | General review of insomnia. It is not specific to midazolam. |
| 36615100 | 2022 | Clinical protocol | J Clin Med | Lemborexant for insomnia and delirium prevention after endoscopy under sedation. Benzodiazepines are discussed only as a delirium risk. |
Singapore Market Information
There are 6 valid single-ingredient registrations. The five main ones are listed below. No approved-indication text is recorded for any of them, so that column is omitted.
| Authorization Number | Product Name | Dosage Form |
|---|---|---|
| SIN00476P | DORMICUM TABLET 15 mg | Film-coated tablet |
| SIN03800P | DORMICUM INJECTION 5 mg/5 ml | Injection |
| SIN00478P | DORMICUM INJECTION 15 mg/3 ml | Injection |
| SIN11039P | DOMI INJECTION 5 mg/ml | Injection |
| SIN12549P | MIDAZOLAM-HAMELN 1 MG/ML INJECTION | Injection |
The only oral product is the 15 mg tablet. This is the form relevant to insomnia, since the other products are injections.
Safety Considerations
- Disease associations from the knowledge graph (not label text): midazolam is linked to contraindication or caution signals in the following conditions. These are chronic obstructive pulmonary disease, cirrhosis of the liver, angle-closure glaucoma, and chronic kidney disease or end-stage renal failure.
- Drug interactions: 100 interacting drugs were identified. The main moderate-level ones are:
- CYP3A4 inhibitors and related agents: clarithromycin, miconazole, clotrimazole, cimetidine, aprepitant, omeprazole, ranitidine.
- CNS-active drugs: morphine, bupropion, dronabinol, nabilone, metoclopramide.
- Others: eluxadoline, teduglutide, glycerol phenylbutyrate.
Minor-level interactions include dexamethasone, betamethasone, and antacids (magnesium oxide, aluminium hydroxide, calcium carbonate).
- Class risks from the literature and rationale: tolerance, dependence, next-day impairment, and respiratory depression.
Label warnings and contraindications were not available. Please refer to the package insert for full safety information.
Conclusion and Next Steps
Decision: Hold
Rationale: Oral midazolam has historical evidence as a hypnotic, including a 14-day RCT in chronic insomniacs. However, that evidence is 35–45 years old. The two registered trials are not about insomnia treatment, and no approved sleep indication is recorded in Singapore. Dependence and next-day impairment risks, together with newer hypnotic options, argue against advancing without further review.
To proceed, the following is needed:
- Review of the HSA package insert for the oral tablet, to confirm the approved indications and the warnings and contraindications.
- Detailed mechanism of action data (MOA) from DrugBank.
- Appraisal of the full text of the 1990 flurazepam vs midazolam RCT, and comparison with current hypnotic standards of care.
- A risk-management plan covering dose and duration limits, dependence monitoring, and respiratory and renal/hepatic precautions.
Related note: among the other predictions, anxiety disorder has much stronger evidence (L1, with Phase 3 and Phase 4 trials). That evidence is mainly for acute procedural or preoperative anxiety, so it is closer to on-label use than to true repurposing. It may be a more productive direction for further evaluation.
Disclaimer
This content is for research purposes only and does not constitute medical advice. Clinical validation is required before any clinical application.