Implementing the COMFORT Guidelines for Postpartum Pain Management

A new national clinical practice guideline (CPG) for pain management after childbirth aims to mitigate peripartum opioid-related risks without compromising or exacerbating existing inequities in pain management in the United States. Standard dissemination approaches are often insufficient to change clinical practice-more active implementation efforts are generally required. Replicating Effective Programs (REP) is a theory-driven implementation intervention that is publicly available and highly scalable, but REP alone may be insufficient for effectively embedding the CPG across all maternity sites. For sites needing more support, REP can be augmented with facilitation (e.g., individualized consultation with site champions to overcome local barriers to CPG adoption, "Enhanced-REP" \[E-REP\]). Because E-REP is more expensive and difficult to scale than REP, it is essential to identify those settings where REP alone is effective versus those where REP may need augmentation, but this has not been evaluated in maternity contexts. Our objective is to determine the effect of a new postpartum pain management CPG, as implemented by REP and E-REP, on postpartum opioid prescribing (primary outcome: rate and amount of opioid prescribed within three days of childbirth), overall, by hospital, and among key subgroups. This is a non-responder randomized trial within the Obstetrics Initiative (OBI), a perinatal collaborative quality initiative funded by Blue Cross Blue Shield of Michigan that includes 68 member hospitals serving more than 120,000 postpartum people over an approximately 15-month study time period. Hospitals not initially responding to REP (defined by performance below the top 15th percentile of all OBI hospitals for a) inpatient order for opioid-sparing postpartum pain management, \[e.g., scheduled acetaminophen and ibuprofen\], or b) amount of opioid prescribed at discharge, or c) provision of non-medication pain management interventions) will be allocated, via bloc