Opioids: Morphine, Hydromorphone, Fentanyl
CA-1 · draft
CA-1 Bootcamp day 10. Authored from cited abstracts; every claim carries a PMID.
Not reviewed. This deck cannot be opened, downloaded or sent
until you approve it. Read every point and its source first — they were drafted
from search results, not from clinical judgement.
Before induction
Fentanyl is 50 to 100 times as potent as morphine, and the short duration you see after a single intravenous dose is redistribution-limited rather than clearance-limited, in the same way thiopentone is — which is why repeated or large doses stop being short-acting and can leave a patient with delayed recovery and prolonged respiratory depression.
The half-life on the reference card is not the number that predicts when your patient will wake up: reported terminal half-lives for fentanyl range from about 1.5 to 6 hours in healthy volunteers and up to 15 hours in geriatric patients, and comparing half-lives between opioids is not a rational way to choose one — effect-site concentration and context-sensitive half-time are.
These drugs differ at the effect site and not just at the syringe: pharmacokinetic-pharmacodynamic studies give fentanyl and its derivatives a blood-to-effect-site equilibration half-life of only 0.2 to 9 minutes, while morphine's analgesic time course is observed with a prolonged delay behind its plasma concentration — so a second dose of morphine given because the first one 'is not working yet' is a dose you will meet again later.
Morphine's danger in renal impairment is not morphine but morphine-6-glucuronide, an active renally excreted metabolite: three patients with renal failure showed classical signs of morphine intoxication with prolonged respiratory depression while having no measurable morphine in plasma at all, and concentration-effect modelling in volunteers put morphine-6-glucuronide at four to eight times morphine's potency for miosis and for suppression of salivation.
Hydromorphone's principal metabolite, hydromorphone-3-glucuronide, carries no analgesic activity and is neuroexcitatory — allodynia, myoclonus and seizures on intracerebroventricular administration — and in anuric haemodialysis patients it accumulated more than twelvefold between treatments while hydromorphone itself barely accumulated at all, though dialysis removed the metabolite effectively.
Equianalgesic tables are estimates rather than conversions: the most consistently supported ratio is about 5 to 1 for oral morphine to oral hydromorphone, no large well-designed randomised trial underpins the tables in common use, and calculating a dose from one of them alone in a patient whose clinical status is changing is an oversimplification of pain management with real potential for harm.
Questions in the room
You gave fentanyl 45 minutes ago and she is comfortable. Is that dose gone, or is it just somewhere else in her?
What does a creatinine of 2.4 change about this choice — and which molecule are you actually worried about?
You want to switch her to hydromorphone. What ratio are you using, and where did that number come from?
If she becomes obtunded at 3 a.m., which drug did it, and why then rather than now?
Oral boards stem
A 71-year-old woman with a creatinine of 2.4 mg/dL has just had a hip hemiarthroplasty under general anaesthesia. She is comfortable in recovery on the fentanyl you gave intraoperatively, but she will need something that lasts for the ward. The examiner asks which of morphine, hydromorphone and fentanyl you would choose, how you would convert from what she has already had, and what specifically you are worried about at three o'clock tomorrow morning.
Sources
Mather, Clin Pharmacokinet 1983
PMID 6226471
Scholz et al., Clin Pharmacokinet 1996
PMID 8896944
Ing Lorenzini et al., Basic Clin Pharmacol Toxicol 2012
PMID 21995512
Osborne et al., Br Med J (Clin Res Ed) 1986
PMID 3087512
Westerling et al., Ther Drug Monit 1995
PMID 7624926
Smith, Clin Exp Pharmacol Physiol 2000
PMID 10874511
Davison et al., J Opioid Manag 2008
PMID 19192761
Mercadante et al., Palliat Med 2011
PMID 21708857
Treillet et al., J Pain Res 2018
PMID 30464578
Patanwala et al., Ann Pharmacother 2007
PMID 17299011
Approving says you read this deck, looked at its figures, and checked the board questions — including that every wrong option is actually wrong. Nothing else has checked those.