Article Text

Download PDFPDF

SAT0700 Comparative risk of respiratory depression in patients treated with opioids for non-malignant pain
  1. M Jani1,
  2. K Kopec-Harding1,
  3. M Lunt1,
  4. WG Dixon1,2,3
  1. 1Arthritis Research UK Centre for Epidemiology, University of Manchester, Manchester
  2. 2Rheumatology department, Salford Royal NHS Foundation Trust, Salford
  3. 3Health eResearch Centre, University of Manchester, Manchester, United Kingdom


Background Opioid use for non-cancer pain has increased considerably over the last 30 years. The U.S. Food and Drug Administration announced several boxed warnings in 2016 to highlight serious opioid-related risks (1) in an effort to reduce fatal overdoses, 80% of which are unintentional (2). The most serious opioid related adverse event is respiratory depression (RD), which can be potentially fatal. There are few data on the incidence of RD in opioids users for non-malignant pain and no comparative data between drugs.

Objectives To assess the comparative risk of RD in new users of opioids for non-malignant pain using routinely collected electronic patient records (EPR).

Methods Opioid users from Salford hospital EPR were identified between 2014–2016. Patients with prior malignancy were excluded on the basis of ICD-10 codes and health issues. Those with prior history of opioid use were excluded using keyword searches within the medicines reconciliation document. The primary analysis was on drug + 1 day lag window, unless patient switched to another opioid. Exposure was categorised as opioid monotherapy by drug or combination of opioids (so that RD events could be assigned to a single exposure category). Electronic National Early Warning Scores, used to regularly monitor physiological parameters during inpatient visits, were used to classify RD. An RD event was defined as any one of the following: respiratory rate (RR) ≤8/min, RR ≤10/min and oxygen saturations<94%, RR ≤10/min and altered consciousness, or dispensed naloxone use. Crude rates per 1000 person years of follow up were calculated and Cox proportional hazards models were used to examine the comparative risk of dispensed opioids and RD. Patients contributed follow up time for a particular drug from dispensed drug start date until day after discontinuation, first RD event, death or end of last hospital admission (by 31/05/16).

Results 8007 opioid users were included in the study, 3,976 female (50%) and a mean age (SD) of 53 (21) years. There were 255 RD events observed on treatment, 87 with severe respiratory depression (RR <8/min), 114 requiring naloxone and 3 respiratory arrests. Patients on fentanyl, morphine, oxycodone and combination treatment had the highest crude rates of RD (table). In the age and gender adjusted Cox-model, using codeine as the referent, patients on fentanyl, morphine, oxycodone and combination opioids had the highest risk of RD [(table), adjusted HR (95% CI) for combination opioids: 2.22 (1.56, 3.16)]. Compared to morphine, combination opioids had an adjusted HR of 1.52 (1.09, 2.13).

Conclusions Fentanyl, oxycodone, morphine monotherapy have a significantly higher risk of RD than codeine, but these are not significantly different from each other. Combination opioids confer the highest risk of RD, compared to both codeine and morphine. The strengths of this study include capture of real time physiological parameters to define RD and dispensed medication use (rather than prescribed) to define exposure.


  1. FDA Drug Safety Communication: March 2016.

  2. CDC. Wide-ranging OnLine Data for Epidemiologic Research (WONDER); 2014. http://wonder.


Disclosure of Interest None declared

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.