Midazolam and ritonavir: the one benzo you can never co-prescribe
Midazolam and ritonavir: the one benzo you can never co-prescribe
Your patient on a boosted HIV regimen is agitated, or headed for a quick procedure, and midazolam is the reflex reach. On ritonavir or cobicistat, that reflex is the trap.
- Oral midazolam + ritonavir (or cobicistat) is contraindicated. The boosted CYP3A4 block can raise oral midazolam more than tenfold — by some measures up to ~28-fold — risking deep sedation and respiratory depression.
- Triazolam sits on the same "never" list for the same reason. Midazolam is just the one you're more likely to reach for without thinking.
- Parenteral (IV) midazolam isn't absolutely off-limits — but only in a monitored/ICU setting, single dose, with the dose cut.
- Reach instead for a glucuronidated benzo: lorazepam, oxazepam, or temazepam. They skip CYP3A4, so the booster can't touch them.
- The one question that predicts all of it: is the regimen boosted (ritonavir/cobicistat) or is it an inducer (efavirenz)? Answer that first and every psych med falls into place.
Why is oral midazolam contraindicated with ritonavir?
Midazolam is cleared almost entirely by CYP3A4. Ritonavir and cobicistat are among the most potent CYP3A4 inhibitors in clinical use — that inhibition is exactly why they're added to HIV regimens as "boosters." Put the two together and midazolam's first-pass clearance collapses. Oral exposure doesn't climb by a little — studies put the AUC increase at roughly 14- to 28-fold — and the clinical result is prolonged, excessive sedation and respiratory depression.
This is not a "watch the levels" caution. It's a hard contraindication on the ritonavir label Avoid. There is no oral midazolam dose that reliably threads this needle on a boosted regimen.
Isn't midazolam the only one? What about triazolam?
Fair challenge — and the honest answer is that midazolam has company. Triazolam shares the identical mechanism and the identical contraindication with ritonavir. Both are potent-3A4-substrate benzodiazepines, and both can accumulate to dangerous levels when the enzyme is boosted shut.
Two others land in the caution zone rather than the never zone:
- Alprazolam — the interaction is biphasic and treacherous. Short-term ritonavir raises alprazolam exposure roughly 2.5-fold with markedly prolonged sedation; then chronic dosing induces CYP3A4 and pushes it back down (to about 12% below baseline by steady state). The danger is front-loaded in the first few days, which makes it unpredictable rather than safe. Caution
- Diazepam — raised by boosters (it leans on CYP3A4 and 2C19), so reduce the dose or, better, prefer a glucuronidated agent. Its long half-life and active metabolites make it a poor first choice here anyway. Caution
So the headline is really a shorthand: midazolam is the benzo you're most likely to reach for on instinct — for procedural sedation or acute agitation — which is precisely why it's the one that slips past the interaction check. Triazolam is barely used anymore; nobody reflexively pushes triazolam at 3 a.m.
What about IV midazolam for a procedure or in the ICU?
Here the picture softens. Parenteral midazolam is not an absolute contraindication with ritonavir — but it demands respect. Coadministration should happen only in an intensive-care or similarly monitored setting, where respiratory depression and prolonged sedation can be managed immediately. Use a single dose where possible, and reduce it. Because the IV route skips gut first-pass metabolism, the peak level is largely spared — but total exposure still rises sharply from reduced hepatic clearance (PBPK estimates put the AUC increase as high as roughly ninefold), so sedation runs long and accumulates with repeat dosing. Caution — monitored setting only
The distinction that matters at the point of care: oral midazolam is never; IV midazolam is "only with a monitor and a plan."
So which benzodiazepine is safe on a boosted HIV regimen?
The ones that never see CYP3A4. Lorazepam, oxazepam, and temazepam are cleared by Phase II glucuronidation, so a CYP3A4 booster has nothing to inhibit. All three carry the same low interaction potential — interaction references put them in the "use with any antiretroviral, no dose adjustment" column — with lorazepam the common workhorse simply because it's the most familiar and flexible. Preferred
| Benzodiazepine | On a ritonavir/cobicistat-boosted regimen | Verdict |
|---|---|---|
| Oral midazolam | Contraindicated — ~14–28× exposure increase; 3A4-dependent | Avoid |
| Triazolam | Contraindicated — same mechanism as midazolam | Avoid |
| IV midazolam | Peak spared, but AUC rises (~9×) — monitored setting, single reduced dose | Caution |
| Alprazolam | Biphasic — ~2.5× early, then induced back down; unpredictable | Caution |
| Diazepam | Raised (3A4/2C19) — dose-reduce or prefer a glucuronidated agent | Caution |
| Lorazepam | Glucuronidated — bypasses CYP3A4; use with any ARV, no adjustment | Preferred |
| Oxazepam | Glucuronidated — no meaningful 3A4 interaction | OK |
| Temazepam | Glucuronidated — no meaningful 3A4 interaction | OK |
Does this apply to cobicistat too — and what about efavirenz?
Yes to cobicistat. It's a pharmacokinetic booster with the same potent CYP3A4 inhibition as ritonavir; treat the benzo rules as identical whenever you see either agent on the regimen. That includes the fixed-dose combinations and any Paxlovid course, where the same oral-midazolam and triazolam contraindications apply.
Efavirenz is where the whole logic flips. It's a CYP3A4 inducer, not an inhibitor — so instead of amplifying 3A4-cleared drugs, it can grind their levels down toward subtherapeutic. Same enzyme, opposite direction, opposite prescribing problem. That single fork — booster or inducer — is the hinge the entire HIV framework turns on.
Psychopharmacology with HIV / AIDS
The full Rapid Decision Guide — the booster-versus-inducer framework, every drug class, and the psychiatric syndromes of HIV. Two chapters are free to read, no login required.
View the course →- The HIV Decision Framework: booster vs. inducerFree
- Antidepressants in HIV: what's safe to startFree
- Antipsychotics in HIVMembers
- Mood stabilizers & anticonvulsants in HIVMembers
- Benzodiazepines, anxiolytics & sedative-hypnotics (the full version of this post)Members
- Stimulants, wakefulness agents & apathy in HIVMembers
- Boosters & inhibitors: ritonavir & cobicistatMembers
- Plus chapters on HAND & cognition, delirium, and substance useMembers
Part of the Rapid Decision Guide series
Point-of-care prescribing guidance, one medical comorbidity at a time — cardiac, renal, hepatic, seizure, HIV/AIDS, and more on the way.
Browse the full series →The full sedative-hypnotic picture — Z-drugs, orexin antagonists, buspirone, and the clean non-benzo routes — lives in the members chapter. Join the Academy for $150/year — or see all plans.
This content is for the continuing education of licensed clinicians and is educational, not individual medical advice. Prescribing decisions should be adjudicated against current FDA labeling and primary interaction references (e.g., the Liverpool HIV drug-interaction resource and DHHS guidelines) for the specific patient and regimen in front of you.
Responses