Ignacio Martin-Loeches, Marc Leone, Craig M Coopersmith, Flavia R Machado, Carol Hodgson, Carolyn S Calfee, Marcus J Schultz
Across haemodynamic support, mechanical ventilation, renal replacement therapy, antimicrobial treatment, nutrition, glucose control, transfusion, and sedation, landmark RCTs describe more than therapeutic expansion. Early biological and haemodynamic trials exposed the limits of reductionist strategies in syndromic critical illness. Protocol-driven approaches improved timeliness and consistency, but later pragmatic trials showed that rigid targets and invasive algorithms often added little once high-quality usual care was established. More recent trials challenge the assumption that greater intervention intensity improves outcomes, emphasising timing, disease phase, baseline risk, heterogeneity, patient selection, and treatment-related harm. Modern ICM RCTs have clarified not only what works, but what can be reduced, delayed, avoided, or applied selectively. Future progress requires biologically informed, context-sensitive trials focussed on meaningful survival, recovery, and long-term function.
BACKGROUND: Randomised controlled trials (RCTs) have reshaped intensive care medicine (ICM), a discipline defined by acute, interacting organ failures, time-sensitive decisions, biological uncertainty, and heterogeneous recovery trajectories. Many interventions supported by strong physiological rationale failed to improve patient-centred outcomes, whereas durable advances often emerged from optimisation of supportive care, avoidance of iatrogenic harm, and reassessment of established practices.
METHODS: We conducted an interpretive historical review of landmark multicentre RCTs in adult ICM from the early 1990s onwards. PubMed searches, reference lists from major trials and reviews, and international guidelines were used to identify trials with conceptual influence on practice, guidelines, physiological reasoning, therapeutic strategy, research priorities, de-implementation, or outcome framing. Evidence was organised into overlapping paradigms rather than rigid chronological periods.
FINDINGS: Across haemodynamic support, mechanical ventilation, renal replacement therapy, antimicrobial treatment, nutrition, glucose control, transfusion, and sedation, landmark RCTs describe more than therapeutic expansion. Early biological and haemodynamic trials exposed the limits of reductionist strategies in syndromic critical illness. Protocol-driven approaches improved timeliness and consistency, but later pragmatic trials showed that rigid targets and invasive algorithms often added little once high-quality usual care was established. More recent trials challenge the assumption that greater intervention intensity improves outcomes, emphasising timing, disease phase, baseline risk, heterogeneity, patient selection, and treatment-related harm. Modern ICM RCTs have clarified not only what works, but what can be reduced, delayed, avoided, or applied selectively. Future progress requires biologically informed, context-sensitive trials focussed on meaningful survival, recovery, and long-term function.