Few interventions in critical care medicine are administered as frequently, and scrutinized as inadequately, as intravenous fluid therapy. Almost every patient who passes through an intensive care unit receives fluids at some point in their trajectory, yet for decades the decision of how much, what type, and for how long was guided more by tradition and reflex than by physiological reasoning. Fluids have long occupied an ambiguous space in critical care: simultaneously life-saving and potentially harmful, indispensable in shock yet implicated in organ dysfunction when given indiscriminately. The growing recognition that fluid administration carries a dose-dependent risk profile, much like any pharmacological agent, has catalyzed a paradigm shift away from liberal, protocol-driven resuscitation toward individualized, physiology-guided, and increasingly technology-assisted fluid management. This editorial examines that evolution, situates it within the landmark trial evidence that has reshaped practice, and argues that guidelines inform but do not replace bedside judgement: the central task of modern intensive care is not simply to resuscitate, but to titrate.

The history of fluid resuscitation is, in many respects, a history of overcorrection. Early twentieth-century shock physiology emphasized the restoration of circulating volume as the primary therapeutic goal, and this principle was reinforced by battlefield and trauma experience through the mid-century, which demonstrated unequivocally that exsanguinating patients died without volume replacement. This foundational truth, however, was gradually generalized beyond its original context. By the late twentieth century, “more fluid is better” had become an unspoken axiom of critical care, embedded in early goal-directed therapy protocols that mandated fixed volumes of crystalloid within rigid time windows.

The Surviving Sepsis Campaign’s early iterations, while instrumental in standardizing sepsis recognition and reducing time to treatment, inadvertently entrenched a culture of aggressive, often unmonitored fluid boluses,(1) and a positive fluid balance accrued in this era is best understood as a physiological signal requiring reassessment rather than a therapeutic endpoint in itself.(2) It took nearly two decades of accumulating observational and trial data on fluid overload, capillary leak, and organ congestion before the pendulum began to swing back toward restraint.

Conceiving of fluid as a drug compels the clinician to define an indication before infusion, just as one would before prescribing an antibiotic or vasopressor. The decision to withhold fluid can be as therapeutically important as the decision to administer it. The dose must be individualized rather than weight-based or protocolized, the duration of therapy must be actively reassessed rather than open-ended, and de-escalation, including active fluid removal, must be planned as deliberately as initiation. Fluid therapy should be individualized rather than protocolized, a principle that this framework operationalizes at the bedside. This framework dovetails with the ROSE model of fluid management, which delineates four temporally distinct phases of critical illness: Resuscitation, in which fluid boluses address life-threatening hypoperfusion; Optimization, in which fluid administration is titrated against dynamic markers of responsiveness to fine-tune oxygen delivery; Stabilization, in which fluid balance is maintained near neutral while organ support continues; and Evacuation, in which active de-resuscitation, often pharmacological diuresis or renal replacement therapy, is employed to correct the cumulative positive fluid balance accrued during earlier phases.(6) The ROSE model reframes fluid therapy as a continuum requiring a different therapeutic posture at each stage, rather than a single undifferentiated intervention applied uniformly throughout the ICU stay.

The choice of fluid type has itself undergone substantial reappraisal. Normal saline, long the default crystalloid by virtue of historical inertia rather than physiological superiority, has been increasingly displaced by balanced crystalloid solutions whose electrolyte composition more closely approximates plasma, thereby avoiding the hyperchloremic metabolic acidosis and renal vasoconstriction associated with large-volume saline administration. The SMART trial demonstrated that balanced crystalloids reduced the composite outcome of death, new renal replacement therapy, or persistent renal dysfunction among critically ill patients compared with saline,(7) while the SALT-ED trial extended a similar, albeit more modest, signal to non-critically ill emergency department patients.(8) The subsequent PLUS study, a large multicenter trial comparing Plasma-Lyte (148) with saline in ICU patients, did not demonstrate a significant difference in ninety-day mortality,

The CLASSIC trial, which randomized septic shock patients to restrictive versus standard fluid strategies after initial resuscitation, found no significant difference in mortality but reinforced the safety of a more conservative approach beyond the initial resuscitation phase,(16,17) adding to a now substantial body of evidence that “less may be at least as good, and possibly better” once the resuscitation phase has concluded. De-resuscitation should be contemplated once hemodynamic stability has been achieved and ongoing fluid accumulation no longer provides physiological benefit, though diuresis or fluid removal should be guided by ongoing hemodynamic tolerance rather than by a fixed fluid balance target.

These risks have driven the adoption of dynamic, rather than static, indices of fluid responsiveness at the bedside. Dynamic indices should generally be preferred over static preload markers when evaluating fluid responsiveness. Passive leg raising, which transiently auto-transfuses approximately three hundred milliliters of venous blood toward the central circulation, functions as a reversible “fluid challenge” without the risk of fluid accumulation, and changes in cardiac output or stroke volume following this maneuver predict true fluid responsiveness with considerably greater accuracy than central venous pressure ever achieved;(3,18) passive leg raising provides this assessment without committing the patient to a non-retractable volume load, a property that distinguishes it from a conventional fluid bolus. In mechanically ventilated patients without arrhythmia, pulse pressure variation and stroke volume variation exploit the cyclical changes in venous return induced by positive-pressure ventilation, with variation thresholds reliably distinguishing responders from non-responders, though their utility diminishes in spontaneously breathing patients, those with low tidal volumes, or right ventricular dysfunction.(4) Point-of-care and transesophageal echocardiography have become indispensable adjuncts, allowing direct visualization of ventricular filling, contractility, inferior vena cava distensibility, and the velocity-time integral response to fluid or passive leg raising, while simultaneously excluding occult cardiac dysfunction that would otherwise be missed by hemodynamic numbers alone.(15) A fluid challenge, whichever modality is used to assess it, should always be paired with a predefined safety limit and a clear physiological endpoint,(19)

The cumulative weight of three decades of trial evidence and physiological research converges on a single, unifying message: fluids are drugs, and like all drugs they possess indications, contraindications, dose-response relationships, and adverse effect profiles that demand the same rigor in prescribing as any vasopressor, antibiotic, or sedative. The question facing the modern intensivist is no longer how much fluid a patient requires in absolute terms, but whether this patient, at this moment, in this physiological state, will benefit from this volume of this fluid, and whether a clear plan exists for de-escalation once the therapeutic window has closed. Optimal outcomes in critical illness will not be achieved by maximizing or minimizing fluid volume as a matter of dogma, but by consistently asking, and rigorously answering, whether the right fluid is being given to the right patient, at the right time, and in the right amount. Guidelines, landmark trials, and emerging technologies all serve this single end: they refine the question, but it is the clinician, at the bedside, who must still answer it.

References