A Systematic Review of the Development and Implementability of Complex Interventions After Hospitalization for Survivors of Intensive Care

Abstract

Background

Survivors of the ICU can experience physical, mental, and cognitive impairments, limiting activities and societal participation. Limited evidence supports the effectiveness of complex interventions after hospitalization, raising questions regarding how these interventions are developed and evaluated. Recommendations from implementation science and complex intervention research may provide further insight.

Research Question

What methods have informed the development and evaluation of complex interventions after hospitalization for survivors of the ICU. How have implementability (acceptability, fidelity, and feasibility) and efficacy been considered in the development and evaluation of these interventions?

Study Design and Methods

Studies were included if they developed or evaluated, or both, a complex, structured intervention after hospitalization aimed at improving recovery outcomes for survivors of the ICU. MEDLINE, Embase, PsycINFO, CINAHL, and PEDro were searched through June 4, 2024. Extracted data included intervention development processes; intervention description; and if and how acceptability or satisfaction, fidelity, feasibility, and efficacy were evaluated. Synthesis methods included deductive analysis and scoring using the Template for Intervention Description and Reporting (TIDieR) and the National Institutes of Health’s Treatment Fidelity Framework. Quality appraisal was completed using the applicable Johanna Briggs Institute (JBI) guidelines.

Results

Seventy-one publications were included involving 62 unique patient cohorts. Twelve studies (19%) used intervention development frameworks, whereas 24 studies (39%) engaged stakeholders in development processes. The median TIDieR score was 16 (interquartile range [IQR], 14-20) of 24. Twenty-two studies (35%) evaluated patient acceptability, of which 2 studies also evaluated clinician acceptability. Median treatment fidelity score was 6 (IQR, 6-9) of 21 with training, delivery, receipt, and enactment domains described poorly. The median consent rate was 48% (IQR, 34%-68%). Thirteen of the 22 studies (59%) designed to test efficacy achieved their sample size. Eight studies (13%) evaluated cost and 20 studies (34% of studies delivering interventions) reported safety. The median JBI score was 61% (IQR, 50%-70%).

Interpretation

Few studies reported applying theory-informed methods or engaging stakeholders in intervention development. Treatment fidelity focused on delivery with little description of receipt or enactment. Future efforts may consider applying implementation science theory and complex intervention approaches.

Clinical Trial Registration

International Prospective Register of Systematic Reviews; No.: CRD42023444648; URL: https://www.crd.york.ac.uk/prospero/