The IMPROVE-GAP trial

Background/objectives: Community-acquired pneumonia (CAP) is a common illness associated with high levels of hospitalisation, morbidity and mortality. As it frequently manifests in the elderly and those with complex chronic disease, representative recruitment is a challenge for interventional studies in CAP. This trial sought to implement a multidisciplinary team approach to CAP management, designed to improve adherence to interventions with prior proven efficacy (early mobilisation, routine malnutrition screening, corticosteroids, and early switch to oral antibiotics), and evaluate their effectiveness in a routine clinical setting. Method: Within a pragmatic, embedded, stepped-wedge, randomised controlled design, all patients hospitalised under a General Internal Medical (GIM) unit meeting a standard case-definition for CAP were included. Eight GIM units at two Australian hospitals were randomised to either: i) usual clinical care, or ii) care supported by a dedicated “CAP Service”: a multidisciplinary team deploying algorithm-based implementation of four evidence-based interventions. Outcome measures included: i) compliance with evidence-based practice, ii) length of hospital stay (LOS), iii) mortality, and iv) readmission rates. Results: 415 and 401 individuals were randomised to usual care and the CAP Service respectively. Under the CAP Service, adherence to evidence-based protocols improved across all four interventions (early mobilisation: 19.3% (usual care) vs. 71.6% (CAP Service), malnutrition screening: 54.9% vs. 83.0%, corticosteroids: 1.9% vs. 53.4%, switch to oral antibiotics: 69.2% vs. 77.3%). However, no difference in outcome was observed for LOS (predicted mean 3.5 vs 3.4 days, geometric mean ratio [95% CI] 0.96 [0.79-1.17]), 90-day mortality (15.8% vs 15.7%, odds ratio 0.92 [0.49-1.75]) or 90-day readmission (27.6% vs 31.1%, 1.17 [0.70-1.97]). Conclusion(s): This study demonstrates an innovative, analytically robust approach to embedding health services research, where effectiveness can be evaluated in a routine setting. A multidisciplinary team was able to increase adherence to evidence-based practice, however, there was no associated improvement in study outcomes.