Opioid consumption and non-opioid multimodal analgesic treatment in pain management trials after hip and knee arthroplasties: A meta-epidemiological study

Background: The leading principle in peri-operative pain management is multimodal analgesia, which reduces opioid requirements and associated adverse effects. Pragmatic pain trials should optimally test interventions in addition to multimodal non-opioid analgesics and interventions to ensure clinical relevance and baseline levels of opioid consumption that reflect clinical settings. We aimed to investigate opioid consumption and use of non-opioid analgesics administered adjunct to interventions in post-operative pain trials after total hip and knee arthroplasty. Methods: A systematic literature search was conducted 7 January 2020 in The Cochrane Library's CENTRAL, PubMed, and EMBASE. Trials investigating analgesic interventions for post-operative pain in adults undergoing total hip or knee arthroplasty were included. The primary outcome was the aggregated median 0–24 h post-operative opioid consumption. Further, we assessed the use of paracetamol, non-steroidal anti-inflammatory drugs, gabapentinoids, high-dose glucocorticoids, local infiltration analgesia and nerve blocks administered as co-interventions equally to all participants. We assessed trends over time for all outcomes. Results: Of 14,200 records, 570 trials were included. Median 0–24 h opioid consumption was 21 and 22 mg iv morphine equivalents in hip and knee arthroplasty trials, respectively. Meta-regression showed no overall linear correlation between opioid consumption and publication year. The use of multimodal non-opioid analgesia increased over time, though only 48% of trials published from 2010 to 2020 administered two or more non-opioid analgesics. Applying more non-opioid analgesics was associated with lower opioid consumption in intervention groups. Conclusion: Post-operative 0–24 h morphine consumption was median 21–22 mg. The demonstrated differences in non-opioid multimodal analgesic regimens between research and clinical settings, can potentially diminish the demonstrated opioid-sparing effects of trial interventions when such are implemented in a clinical context.