Interventions for improving critical care in low- and middle-income countries: a systematic review

Purpose: To systematically review the typology, impact, quality of evidence, barriers, and facilitators to implementation of Quality Improvement (QI) interventions for adult critical care in low- and middle-income countries (LMICs). Methods: MEDLINE, EMBASE, Cochrane Library and ClinicalTrials.gov were searched on 1st September 2022. The studies were included if they described the implementation of QI interventions for adult critical care in LMICs, available as full text, in English and published after 2000. The risks of bias were assessed using the ROB 2.0/ROBINS-I tools. Intervention strategies were categorised according to a Knowledge Translation framework. Interventions' effectiveness were synthesised by vote counting and assessed with a binomial test. Barriers and facilitators to implementation were narratively synthesised using the Consolidated Framework for Implementation Research. Results: 78 studies were included. Risk of bias was high. The most common intervention strategies were Education, Audit & Feedback (A&F) and Protocols/Guidelines/Bundles/Checklists (PGBC). Two multifaceted strategies improved both process and outcome measures: Education and A&F (p = 0.008); and PGBC with Education and A&F (p = 0.001, p < 0.001). Facilitators to implementation were stakeholder engagement, organisational readiness for implementation, and adaptability of interventions. Barriers were lack of resources and incompatibility with local systems. There was a lack of evidence from low-income countries. Conclusions: The evidence for QI in critical care in LMICs is sparse and at high risk of bias but suggests that multifaceted interventions are most effective. Co-designing interventions with and engaging stakeholders, communicating relative advantages, employing local champions and adapting to feedback can improve implementation. Hybrid study designs, process evaluations and adherence to reporting guidelines would improve the evidence base.