Measuring RE‑AIM: a systematic review of metrics across intervention types and settings

Background The RE-AIM framework (Reach, Effectiveness, Adoption, Implementation, Maintenance) is widely used to evaluate how evidence-based practices are implemented and adapted within healthcare organisations. How its domains are measured in inpatient medical and surgical care is less well described. We asked which metrics have been used to measure each RE-AIM domain in inpatient care, and whether metric choice differs by intervention type or setting, so as to identify metrics transferable to spine and neurosurgical practice. Curating this evidence base can help inform future implementation strategies. Methods RE-AIM publications were identified via the curated list on re-aim.org, which links to a PubMed search of publications citing or applying RE-AIM since its 1999 description, run on 20 September 2025 and returning 1,943 records. Eligible studies explicitly applied RE-AIM to the design or evaluation of an intervention delivered to admitted patients, or to the staff caring for them, in an inpatient or emergency setting, in English; editorials, protocols without results, and reports citing but not applying RE-AIM were excluded. Two reviewers screened independently in Rayyan and extracted data into a proforma. Each RE-AIM domain was coded against a pre-specified definition, unit of analysis and decision rules, and findings synthesised narratively. Results Fifty-five studies were included, most from high-income countries (51/55, 92.7%), particularly the United States (33/55, 60.0%). Emergency medicine (20.0%) and internal medicine (18.2%) predominated; three studies (5.5%) were set in surgical wards or populations, seven (12.7%) in mixed medical-surgical settings, and none in spine surgery or neurosurgery. Evaluation metrics frequently relied on EMR/EHR logs (50.9%). RE-AIM domains were variably reported: Reach (100%), Effectiveness (94.5%), Adoption (98.2%), Implementation (100%), Maintenance (80.0%). Reach was most often evidenced by the proportion of eligible patients exposed (26/55, 47.3%). Barriers were often underreported, and only 58.2% achieved evaluative coverage of all five domains. Conclusion RE-AIM has been applied across diverse inpatient contexts, but reporting depth remains uneven and surgical settings are markedly underrepresented. Implementation and adoption indicators, particularly protocol adherence, workflow integration and record-derived uptake, are the most consistently quantified and the most portable between institutions, and are the starting point for a shared metric set in spine and neurosurgical care.