Table 1 Potential factors contributing to the lack of “bench-to-bedside” translational success of preclinical sepsis research
Domain | Preclinical model | Human condition |
|---|---|---|
Construct validity | ||
Patient population | Historically male, however more female animals in recent years Healthy Young/juvenile Limited environmental exposure Genetically homogeneous | Female and male Medical comorbidities Old and young Environmental stressors Genetically diverse |
Site of infection | Non-bacterial surrogates (e.g. endotoxin) Polymicrobial abdominal/enteric Gram-negative Pneumonia (rare) Fungi/protozoa (rare) Virus (very rare) | Soft tissue Gram-positive Abdominal Gram-negative, including biliary Pneumonia (common) Virus (common) Fungi/protozoa |
Intercurrent therapy | None Antibiotics (monotherapy) Fluids Anesthesia/analgesia Experimental therapy | Antibiotics (poly-therapy) Fluids Blood products Vasopressors/Inotropes Sedation/analgesia Baseline medication regimen Adjunct therapies (e.g. steroids, heparin) |
Outcomes | Non-mortality surrogate Short term Organ failure (often single) Molecular biomarkers (common) Organ histology | Mortality Short and long term ICU/hospital length of stay Validated multi-organ failure score Molecular biomarkers (rare) Organ histology (very rare) |
Research methodology | ||
Biostatistics | Lack of sample size calculation | Study powered to detect difference in pre-specified outcome |
Reduce bias | Randomization rare Lack of blinding | Randomization Double-blinded |
Standardization | Single centre Variations in practice | Multicentre Shared protocol |
Reporting | Inconsistent Incomplete Difficult to synthesize | Required Comprehensive Conducive to systematic review |