Patient controlled analgesia in the emergency department

BMJ 2015; 350 doi: https://doi.org/10.1136/bmj.h3240 (Published 21 June 2015) Cite this as: BMJ 2015;350:h3240
  1. Fiona Lecky, clinical professor of emergency medicine
  1. 1EMRiS Group, Health Services Research Section, School of Health and Related Research, University of Sheffield, Sheffield S1 4DA, UK
  1. f.e.lecky{at}sheffield.ac.uk

Reduces pain and increases autonomy

Acute severe pain that follows injury or onset of illness alerts patients to the likelihood of tissue damage (nociception) and drives much emergency department attendance.1 The prompt assessment and alleviation of pain is a quality benchmark for emergency departments internationally, and titrated intravenous opiates are the initial analgesic of choice for severe pain.2 3 Recent improvements in initial pain assessment and analgesic provision have occurred in UK emergency departments, but processes for referral to an inpatient team and time based standards for ward admission can interfere with effective pain management after the first dose of intravenous opiate. Recent surveys of patients and studies of care pathways suggest that current care is suboptimal in this regard.4

Two open label, multicentre, randomised trials of pain solutions in the emergency setting (PASTIES) by Smith and colleagues (doi:10.1136/bmj.h2988; doi:10.1136/bmj.h3147) provide new evidence to help us tackle this deficiency.5 6 The studies each enrolled a group of patients who commonly require parenteral analgesia throughout the first day of an emergency admission—adults with acute non-traumatic abdominal pain and adults with acute traumatic injury. Both studies included only patients who required initial treatment with boluses of morphine, titrated and given by a nurse in the emergency department. Participants in the intervention groups were given a patient controlled analgesia (PCA) device consisting of a volumetric pump safely delivering a fixed dose when a button is pressed, with a subsequent lockout period. Patients in the control group received standard prescriptions for intermittent opiates.

One novel feature of these studies was the measurement of patients’ own pain scores hourly for up to 12 hours. This allowed the authors to calculate the “area under the curve” for pain scores, a measure of total pain experienced in the emergency department and into the ward admission.

Interestingly, the studies reported different results. Adults with non-traumatic abdominal pain experienced significantly less pain overall when managed with PCA compared with usual treatment5; this observation was robust to sensitivity analyses. Participants in the PCA group spent 15% less time in moderate or severe pain throughout the 12 hour study period, received significantly more morphine, and had a 2.6-fold increase in the (adjusted) odds of being very or completely satisfied with pain management. In the study of patients with acute traumatic injuries, pain scores were lower and satisfaction scores were higher in those managed with PCA, but the differences were not statistically significant.6

The inconsistent results could be due to differing pain mechanisms in the two populations (visceral in abdominal pain, somatic in acute traumatic injury), a type 2 error in the study of acute traumatic injury (whereby the study failed to find an existing difference between treatments owing to methodological factors such as low power), or type 1 error in the abdominal pain study (whereby the study found a difference where none existed). As the confidence limits in the study of acute traumatic injury include a clinically relevant difference between the treatments, type 2 error seems more likely. The authors also acknowledge the strong possibility of a Hawthorne effect improving pain management in the control groups beyond the experience of patients in the “real world.”

However, somatic pain from limb fractures differs from visceral pain in that management in the emergency department can include reduction of fractures and dislocations, plaster splintage, and regional nerve blocks, which may markedly reduce subsequent analgesic requirements. In both studies, the relative increase in satisfaction scores with PCA was greater than the relative reduction in pain scores, suggesting that patients attach more value to autonomy in pain management than to the magnitude of pain experienced.

Together, these studies suggest that starting of PCA in emergency departments is likely to be beneficial for patients who have needed a bolus of intravenous opiate analgesia for initial pain management, particularly when severe pain recurs during their stay in the emergency department. Patients with a similar clinical profile to those in the PASTIES studies are most likely to benefit; however, extrapolation to other groups is not unreasonable, particularly when a second bolus of intravenous opiate is being considered. The counter arguments include concern about serious side effects of opiates, although in practice these are rare. Among the 400 patients in these studies, just one had a serious adverse event—excessive drowsiness from which the patient recovered fully.

Both PASTIES studies appropriately excluded patients with a less favourable risk-benefit balance in relation to intravenous opiates. Pain management guidance from the UK’s Royal College of Emergency Medicine also urges caution in the use of intravenous opiates in older adults,7 but the PASTIES investigators safely included patients up to 75 years old. It is important to acknowledge that benefit from PCA occurs in the context of multimodal analgesia that enables opiate sparing.8 The overall goal of emergency admission must be to treat the underlying illness or injury and reduce the need for parenteral analgesia.

We know that PCA devices are safe and effective in the postoperative setting, where they have been used extensively since the 1990s.9 Cost effectiveness—although not described in PASTIES—is unlikely to be a major barrier for emergency departments, as the small investment in reusable equipment and set-up time is more than likely to be recuperated by reductions in the nursing time spent administering additional bolus opiates. The acronym for the PCA device should perhaps be reattributed to “patient centred analgesia,” as these devices clearly deliver an autonomy that is highly valued by many patients.


Cite this as: BMJ 2015;350:h3240



View Abstract

Sign in

Log in through your institution