Protocol for natural experiment examining glucocorticoids versus no glucocorticoids on opioid use in hip and knee arthroplasty

IntroductionHigh-dose glucocorticoids are increasingly used in patients undergoing primary total hip arthroplasty (THA), unicompartmental knee arthroplasty (UKA), and total knee arthroplasty (TKA). However, the evidence supporting the use of high-dose glucocorticoids to manage acute postoperative pain is based mainly on randomised controlled trials, which may have limited external clinical validity due to strict exclusion criteria. In Denmark, the treatment was implemented stepwise, comprising a natural experiment. ObjectiveTo examine the real-world effect of a single dose of high-dose glucocorticoids on postoperative opioid consumption and adverse events in patients undergoing THA, UKA, and TKA. MethodsThis protocol describes a natural experiment study that will be reported as a target trial emulation, i.e. by attempting to mimic a randomised clinical trial. Inclusion and exclusion criteriaAll adults (>18 years) undergoing primary, elective THA, UKA, and TKA in the Capital Region and Region Zealand in Denmark are eligible. At each centre, patients who underwent surgery within one year before and one year after the implementation of high-dose glucocorticoids are screened for eligibility. InterventionThe intervention is the intended use of a single high dose of glucocorticoids, administered after induction of anaesthesia, as part of the standard treatment. ComparatorThe comparator is no administration of glucocorticoids. Patients in the control arm may have received a lower dose of glucocorticoids for postsurgical nausea and vomiting Assignment of interventionFor each hospital, patients who underwent surgery before the implementation of glucocorticoids serve as controls, while patients operated on after the implementation constitute the treatment arm. OutcomesThe primary outcome is opioid consumption, expressed in morphine equivalent doses (MEQs) in the first 24 hours after surgery. Secondary outcomes are two adverse effect-outcomes: incidence of opioid-related adverse events and serious adverse events, and three efficacy-outcomes: length of postoperative observation area-stay, length of hospital stay, and worst pain intensity-score measured. Data collectionThis study uses data from the TRIPLE-A project (www.triplea.dk), comprising validated electronic health record data. Statistical analysesThe analyses will be based on a pre-defined mixed effects model with hospital as a random effect and adjusted for important presurgical (and thus pre-intervention) variables. A per protocol population will be analysed as a sensitivity analysis. For the primary outcome, a difference of 5 MEQs between treatment arms is considered clinically important. Knowledge disseminationThe results will be shared at conferences and made publicly available. Registrationhttps://doi.org/10.1101/2025.11.11.25339982