Authors: Rob Hirst, Liz Farah, Andy Neill, Dave McCreary/ Codes: CC9, MuC1, MuP1, PalC4, PC3 / Published: 04/12/2023

  • Andy Neill
  • Dave McCreary
Clinical question
  • do long waits in the ED for a bed in the hospital lead to increased mortality?
  • Roussel, M. et al. Overnight Stay in the Emergency Department and Mortality in Older Patients. JAMA Intern. Med. 183, (2023).
  • Hospitals are overcrowded and above capacity. Most institutions choose to allow this hospital overcrowding to remain centrally located in the ED. Multiple papers over recent years are ringing the call that long waits for patients once a decision has been made to admit leads to adverse outcomes including mortality
  • If you’re listening to this podcast then most likely this is your lived daily experience in your ED. And while it seems a no brainer that long waits are a bad thing it has become incumbent on EM to prove what we already know - that long waits are adverse to outcomes.
  • this is a large research network in France. (best healthcare in the world remember…)
  • looked at a 48 hr period in 2022 and split their admitted >75 cohort into two groups.
    • first group were admitted to the inpatient bed before midnight
    • second group were admitted after 0800 the next morning.
  • primary outcome of in hospital mortality which is a fairly solid and reproducible outcome
  • pre specified looking at secondary outcomes like infection and falls as a potential cause of any increased mortality
  • 97 EDs and 1600 pts (which if split evenly between each ED would be 16 >75 year old boarders in a 48 hr period, or 8 in a 24 hr period. This seems fairly tame compared to what many of us see in our EDs)
  • the diagnoses were fairly typical of what we might see in ED. They have a 25% trauma which i get the impression were the low level falls and minor trauma needing an admission under the gen med take for frailty rather than the hip fracture or high speed RTC.
  • median age 86.
  • Physiology between the two groups appears the same.
  • roughly half of the cohort got to their in patient beds before midnight, the other half waited until 0800 the next morning.
  • 15% vs 11% mortality with the longer waiting group having the worse outcomes. This is a clinically large difference and FWIW met statistical significance as well.
  • this increase in mortality correlated nicely with things like increased nosocomial infections and falls as potential aetiologies of increased mortality.
  • this is yet another study that suggests outcomes are worse with long waits. Add this to your list of papers to present to clinical directors, hospital management and commissioners and politicians.
  • it is not an RCT but we can only hope that no one asks us to randomise patients to a night on a trolley simply to prove bad outcomes…
  • they excluded people who went to a ward bed between midnight and 0800. A little unclear why but i suspect it gives a greater separation between groups. And of course if you have a large number of patients moving to a ward bed after midnight you have serious questions to ask of your bed managers. What patients are being discharged at night? And if there were no late night discharges it implies that those beds were available all along and were being delayed by wards.
  • there may well be unmeasured confounders to explain the mortality but in my mind they would skew the opposite direction of the mortality increase found. One would imagine sicker patients are more likely to move to the ward quicker and hence people who spent less time in the ED would have the higher mortality. But they found the opposite.
  • they didn’t just find an increase in mortality but nicely demonstrated the reasons why and could correlate the increased mortality with increased nosocomial infections and falls. Again this rings incredibly true with my experience.
  • the paper also suggests that the most vulnerable of an already vulnerable cohort (what they describe as “limited autonomy” patients) with dementia were the worst affected by long waits.
  • locating hospital crowding in the ED is now standard practice. It has become unfortunately standard of care. It is accepted by hospital management as an inevitability. Chairs in corridors are considered “clinical spaces”. Normal infection control procedures and fire safety rules do not seem to apply in the ED for reasons i do not understand. I am aware of a recent shiny new ED that opened with a well placed sequence of plug sockets down the length of the corridors to ensure that our parents and grandparents will have somewhere to charge their phone while they wait for a bed.
  • This is pure anecdote but I’m in a rant now… I was a domestic assistant and cleaner in an ED for 4 years in northern ireland as a medical student and would remember bringing toast and porridge to maybe 40 trolley waits spread down the corridors of the ED. I then graduated and worked in the same ED and in my early training, in the mid noughties in peak NHS I saw things improve and we hit our 98% 4 hr targets consistently for a number of years and trolley waits were no longer a thing. You can imagine things have not been good like that for a long time. I tell the story to suggest that things do not have to be like this. In that one hospital over my training career things went from bad to good to bad again. There are solutions, they may be costly but i think it’s becoming pretty clear that our choices not to change things are costing lives.

Clinical Question

Is a short course of opioids helpful analgesia for the management of acute, non-specific low back and neck pain?

Title of Paper

Opioid analgesia for acute low back pain and neck pain (the OPAL trial): a randomised placebo-controlled trial

Journal and Year

Lancet. July 2023

Lead Author

Caitlin Jones


  • Opioids are (probably deservedly) getting a worse and worse reputation for the management of acute pain.
  • Prescribing of opioids got a bit out of hand, particularly in the US, but Australia and UK not great on that front either.
  • Docs are still commonly prescribing opioids for the short term management of acute lower back and neck pain.

Study Design

  • Triple-blinded, placebo-controlled randomised controlled trial

Patients Studied

  • Adults presenting to primary care or EDs in NSW
  • with ≤12 weeks low back pain, neck pain or both, without radiculopathy
  • At least moderate severity


  • Targin (oxycodone/naloxone) 5mg/2.5mg BD
    • Uptitrated to max 10mg BD
  • Treatment continued until adequate improvement: pain 0-1/10 for 3/7 consecutively or for max 6 weeks
  • Plus guideline-care:
    • Reassurance of positive prognosis
    • Advice to stay active, avoid bed rest
    • Other guideline-recommended treatments including non-opiod analgesics as needed


  • Placebo
  • Plus guideline-care


  • Primary: pain intensity (0-10 scale) at 6 weeks
    • Brief Pain Inventory Pain Severity subscale: asks patients to rate the intensity of their pain at its worst, on average and currently using a scale of 1-10
  • Powered to find minimal clinical difference of 1/10 point difference between groups (needed 173 per group)

Summary of Results

  • 347 participants: 174 opioid vs 173 placebo → 172 after one placebo was diagnosed with boney mets
    • Mean age 44.7 years
    • Median duration of pain 7 days
    • Only 3% recruited from ED - the rest were from primary care
  • Primary outcome: No significant difference in pain scores at 6 weeks between groups.
    • Mean pain score opioids 2.78 | placebo 2.25
    • Unadjusted mean difference 0.53 [0.00-1.07]
    • Adjustments: results consistent following adjustment for site of pain and days since pain onset.
  • By weeks 26 and 52 there was a small difference favouring placebo
  • 6 week Roland-Morris Questionnaire (self-rated assessment of physical disability caused by low back pain) favouring placebo.
  • Small difference favouring placebo in mental health score at 12 weeks
  • Risk of misuse significantly higher in opioid group at 52 weeks on the current opioid misuse measure (COMM) - 10% in placebo vs 20% in opioid group (p 0.049)
    • Assesses key risk factors such as signs and symptoms of intoxication, emotional volatility, addiction and problematic medication behaviour.
    • Suggests that even a short course of opioids can increase the risk of long term misuse

Authors Conclusion

Short-term, judicious use of an opioid analgesic plus guideline care did not confer any benefits for people with acute low back pain or neck pain when compared with placebo plus guideline care and had a small but significant harmful effect on risk of opioid misuse in the long-term. There is no evidence that opioid should be prescribed for people with acute non-specific low back pain or neck pain.

Clinical Bottom Line

This well performed study gives us some useful data showing not only the lack of utility of opioids, but that they may lead to increased pain at 6-12 months, along with a potential for opioid misuse even after a short course.

While most of these patients weren’t actually recruited from the ED, and duration of pain was on the upper side of duration for acute pain we’d see in our practice, they are still probably close enough to the cohort we see to say that we should be avoiding opioid outpatient prescription in patients with uncomplicated low back and neck pain in the ED.

That’s not to say you mightn’t use single doses, while in the ED, during the acute spasm in the early stages of back pain. That just wasn’t this study population.

I don’t consider a 6 week course of opioids, as patient’s got in this study, to be a short course. If my patients get anything (and not for back pain), its about 10-12 tablets - that’s 2-3 days worth.