The inherent bias in studies reporting outcomes using total opioid burden metrics: a simulation study

Authors
Affiliations

Dylan Whitaker

Institute of Health Informatics, University College London, London, United Kingdom

Department of Critical Care, University College London Hospitals NHS Foundation Trust, London, United Kingdom

Matthew Wilson

Department of Anaesthesia and Perioperative Medicine, University College London Hospitals NHS Foundation Trust, London, United Kingdom

Division of Surgery and Interventional Science, University College London, London, United Kingdom

Steve Harris

Institute of Health Informatics, University College London, London, United Kingdom

Department of Critical Care, University College London Hospitals NHS Foundation Trust, London, United Kingdom

Cover Letter

Professor Hugh Hemmings,
Weill Cornell Medical College,
New York,
NY,
USA

Dear Professor Hemmings,

We thank you for considering our Research Communication manuscript titled “The inherent bias in studies reporting outcomes using total opioid burden metrics: a simulation study” to the British Journal of Anaesthesia.

Our paper is a methodological critique of total opioid burden metrics (such as oral morphine equivalent) that finds an important measurement bias. The bias is present in any study comparing groups with unmatched mean opioid doses opioid or where there is an error in the conversion values used with regards to the biological ground truth. The measurement bias can skew or even invert the observed treatment association depending on the specific circumstances of the study. We conclude that any study using total opioid burden metrics should asses the degree of bias in the context of their minimum clinically important difference to ensure the results can be interpreted accurately. Whilst this applies predominantly to observational research, there are key implications for randomised controlled trials with regards to analgesia protocols.

Despite multiple attempts to unify the methodology with conversion tables and calculator tools, the inherent bias continues to be overlooked. Given over 4,000 published papers have used total opioid burden metrics in the last decade, and it remains an important quantitative outcome measure for perioperative pain research, the potential implications of this bias are significant 1.

We hope this paper will lead to a critical review of how total opioid burden metrics are used in research and improve the accuracy in the metrics already widely used.

We can confirm no parts of this work have been presented before and has not been under consideration at any other journal.

Many thanks, The authors

  1. From a PUBMED search of “(Milligram morphine equivalent) OR (Oral morphine equivalent)” filtered to the last 10 years on 16th September 2026↩︎