Refractory Dyspnea: Why the Benzodiazepine Isn't the First Move
A dying woman's air hunger has stopped responding to everything disease-directed therapy can offer, and her husband wants the anti-anxiety drug started first because her breathing looks like panic — but the evidence for what actually helps points the other way.
C.W. and her husband have been married fifty-two years, long enough that he can finish sentences she doesn't get to start, and long enough that when her breathing turned ragged and fast two days ago, he read it instantly as the same look she gets before a panic attack, something he's seen a handful of times across five decades together. Her heart failure has progressed past what diuretics, guideline-directed therapy, and now hospice-level comfort measures can touch; NYHA class IV symptoms at rest, oxygen doesn't meaningfully change how she feels even though her saturations look reasonable, and the sensation she keeps describing — can't get enough air in, no matter how hard she tries — is refractory dyspnea by any working definition the team has.
Her husband's instinct, that this looks like anxiety and should be treated like anxiety, is understandable and clinically reasonable to raise, but it runs against where the actual evidence sits. Low-dose opioids carry the strongest evidence base for refractory dyspnea specifically — Abernethy's randomized trial of sustained-release morphine against placebo found a real, significant improvement in breathlessness, and the mechanism is central: opioids blunt the perception of air hunger at the level of respiratory drive, independent of any effect on the underlying disease. Benzodiazepines, despite being reached for reflexively whenever breathlessness looks like panic, do not have a comparable evidence base as monotherapy — a Cochrane review examining benzodiazepines specifically for dyspnea found insufficient evidence they help on their own, a genuinely counterintuitive result given how visually similar severe dyspnea and panic can look at the bedside.
C.W. herself, on the rare stretches when her breathing eases enough to talk, has said the sensation isn't fear exactly — it's closer to the feeling of holding her breath underwater a beat too long, over and over, with no actual submersion to explain it. That description, offered unprompted rather than solicited by a symptom questionnaire, is closer to the classic language of air hunger than of panic, and it's part of what the team is weighing alongside the trial evidence rather than instead of her husband's fifty-two years of reading her more closely than any chart ever could.
At the bedside, with her husband holding her hand
I understand exactly why this looks like a panic attack to you — the breathing pattern really does resemble it. But the evidence for what actually helps refractory dyspnea doesn't point toward the anti-anxiety medication first. Low-dose morphine has the strongest trial evidence specifically for this symptom — Abernethy's randomized trial found a real, measurable improvement in breathlessness using sustained-release morphine against placebo, and it works through blunting the brain's own perception of air hunger, not by treating anxiety.
I'd add the other half of that evidence, because it's the part that's genuinely counterintuitive: benzodiazepines, despite being the drug class most people reach for when breathlessness looks like panic, don't have a comparable evidence base as monotherapy. A Cochrane review looking specifically at benzodiazepines for dyspnea found insufficient evidence they help on their own. That doesn't mean lorazepam has no role here — if a genuine anxiety component persists after starting morphine, adding it as a second agent is reasonable — but starting with it alone would mean reaching for the less-supported drug first because the symptom looks more familiar.
Given her mild-moderate renal impairment, I'd start morphine at a reduced dose with a longer interval than we'd use in someone with normal kidneys, since its active metabolite is renally cleared and we don't want to trade breathlessness for the sedation and confusion that accumulation can cause in exactly the population we're trying to make comfortable. I'll also explain the reasoning directly to her husband — not just that we're starting morphine, but why the drug that treats her breathing is the right first move even though her breathing looks like the panic he's seen before.
Agreed: low-dose oral morphine started on a scheduled interval adjusted for her renal function, with lorazepam available as an add-on if a distinct anxiety component is still visible once the morphine has had time to act. Her husband was walked through the evidence directly and accepted the plan once he understood morphine's mechanism was aimed squarely at the breathing itself.
No real disagreement between the two clinicians — the plan converged cleanly once the Cochrane finding on benzodiazepine monotherapy was named plainly rather than left as an assumed fact either voice was working from silently.