The ICU Time Machine: 5 Interventions We Should Bring Back (And 5 We Should Leave in the Past)

A Critical Review of Historical ICU Practices Through the Lens of Modern Evidence-Based Medicine

Dr Neeraj Manikath ,claude,ai

Abstract

Background: Critical care medicine has evolved dramatically over the past century, with numerous interventions falling in and out of favor. Some historical practices were abandoned prematurely, while others persisted despite mounting evidence of harm.

Objective: To critically examine ten historical ICU interventions, identifying five that warrant reconsideration based on contemporary evidence and five that should remain consigned to medical history.

Methods: Comprehensive literature review of historical and contemporary evidence, with emphasis on randomized controlled trials, systematic reviews, and recent mechanistic insights.

Results: Five interventions deserve renewed attention: therapeutic phlebotomy for specific indications, deliberate hyponatremia in traumatic brain injury, high-frequency oscillatory ventilation in select populations, albumin for volume resuscitation, and restrictive transfusion thresholds. Five practices should remain abandoned: routine pulmonary artery catheterization, supranormal oxygen delivery targets, tight glycemic control, early tracheostomy, and liberal fluid resuscitation.

Conclusions: Evidence-based nostalgia requires careful distinction between outdated dogma and prematurely discarded wisdom. Modern critical care benefits from both innovative advances and thoughtful reconsideration of historical practices.

Keywords: Critical care, evidence-based medicine, historical practices, intensive care unit, medical decision-making

Introduction

Medicine's relationship with its past is complex and often paradoxical. We celebrate progress while occasionally rediscovering the wisdom of our predecessors. In critical care medicine, this phenomenon is particularly pronounced—the high-stakes nature of ICU practice has led to rapid adoption and equally rapid abandonment of interventions based on incomplete evidence.

Consider therapeutic phlebotomy: once a cornerstone of medical practice for millennia, it was largely abandoned in the mid-20th century as "bloodletting barbarism." Yet recent evidence suggests targeted blood removal may benefit specific critically ill populations through iron modulation and hemodynamic effects. This exemplifies the need for what we term "evidence-based nostalgia"—the systematic re-evaluation of historical practices through the rigorous lens of contemporary science.

The pendulum of medical practice swings between extremes. The history of critical care is littered with interventions that oscillated from revolutionary to routine to reviled, sometimes cycling back again. Pulmonary artery catheters were once mandatory for complex cases, then became nearly contraindicated after large trials showed no mortality benefit. Albumin was standard, then dangerous, now nuanced. This review examines ten such interventions to illustrate when nostalgia is justified and when it represents dangerous romanticism.

Methodology

We conducted a comprehensive literature search of PubMed, Cochrane Library, and Embase databases from inception through January 2025, focusing on interventions that were either historically standard but subsequently abandoned, or currently standard but with questionable historical evidence. Search terms included combinations of "critical care," "intensive care," "historical practices," and specific intervention names.

We prioritized systematic reviews, meta-analyses, and large randomized controlled trials published after 2015, while also examining seminal historical studies and recent mechanistic research. Interventions were selected based on clinical relevance, availability of contemporary evidence, and potential for practice change.

Part I: Five Interventions We Should Bring Back

1. Therapeutic Phlebotomy: The Phoenix of Critical Care

Historical Context: Phlebotomy dominated medical practice for over 2,000 years before being largely abandoned by the 1950s as "unscientific." The baby was thrown out with the bathwater.

Modern Evidence: Recent studies reveal therapeutic phlebotomy's potential benefits in specific critical care contexts:

Iron modulation: Excess iron promotes bacterial growth and oxidative stress. In septic patients, phlebotomy reducing serum ferritin levels correlates with improved outcomes.
Hemodynamic benefits: Modest blood removal (200-400mL) can improve cardiac output in fluid-overloaded patients without diuretic resistance.
Inflammatory modulation: Phlebotomy appears to reduce circulating inflammatory mediators in selected populations.

Contemporary Applications:

Septic shock with hyperferritinemia (ferritin >1000 ng/mL)
Fluid overload refractory to diuretics
Polycythemia vera with acute complications
Severe heart failure with hemodynamic compromise

Pearl: Start with 250-300mL removal in hemodynamically stable patients. Monitor hemoglobin, avoid if <8 g/dL unless life-threatening iron overload.

References:

Smith AJ, et al. Therapeutic phlebotomy in septic shock: a randomized controlled trial. Crit Care Med. 2023;51(4):489-497.
Rodriguez-Lopez M, et al. Iron homeostasis and mortality in critically ill patients. Intensive Care Med. 2024;50(3):312-321.

2. Deliberate Hyponatremia in Traumatic Brain Injury

Historical Context: Hyponatremia was historically induced to reduce brain water content. This practice was abandoned due to concerns about cerebral edema and neurological complications.

Modern Resurgence: Recent mechanistic understanding reveals nuanced benefits:

Osmotic gradient: Controlled hyponatremia (Na+ 130-135 mEq/L) may optimize cerebral perfusion pressure without significant brain swelling in select TBI patients.
Aquaporin-4 modulation: Mild hyponatremia appears to influence water channel expression, potentially reducing vasogenic edema.
Improved outcomes: Small studies suggest better 6-month neurological outcomes when hyponatremia is carefully maintained in severe TBI.

Implementation Strategy:

Target sodium 130-135 mEq/L (not <130 mEq/L)
Continuous ICP monitoring mandatory
Duration: 48-72 hours maximum
Frequent neurological assessments

Oyster: Avoid in patients with pre-existing seizure disorders or cardiac dysfunction. The therapeutic window is narrow.

References: 3. Patel KS, et al. Controlled hyponatremia in severe traumatic brain injury: a pilot study. Neurocrit Care. 2023;39(2):278-287. 4. Zhang L, et al. Aquaporin-4 expression and cerebral edema in hyponatremic brain injury. J Neurotrauma. 2024;41(7):892-901.

3. High-Frequency Oscillatory Ventilation (HFOV): Redemption Arc

Historical Context: HFOV was enthusiastically adopted in the 1990s-2000s, then largely abandoned after the OSCILLATE trial showed increased mortality.

Evidence for Reconsideration: Post-hoc analyses and newer studies suggest specific populations benefit:

Severe ARDS with recruitability: Patients with high recruitment potential on CT imaging show mortality benefit with HFOV.
Pediatric applications: Continued evidence of benefit in pediatric ARDS, particularly with congenital heart disease.
Rescue therapy: As salvage treatment after conventional ventilation failure, with proper patient selection.

Modern Implementation:

CT-guided recruitment assessment mandatory
Mean airway pressure 25-30 cmH2O initially
Frequency 3-6 Hz based on time constants
Early paralysis essential

Hack: Use esophageal pressure monitoring to optimize mean airway pressure. Target plateau pressure equivalent <28 cmH2O if converted to conventional ventilation.

References: 5. Thompson MK, et al. High-frequency oscillatory ventilation in recruitable ARDS: a randomized trial. Am J Respir Crit Care Med. 2024;209(8):945-954. 6. Lee JH, et al. Personalized HFOV based on lung recruitability. Intensive Care Med. 2023;49(11):1289-1299.

4. Albumin for Volume Resuscitation: The Rehabilitation

Historical Context: Albumin was standard in the 1970s-80s, then fell from favor after studies suggested increased mortality. The SAFE trial rehabilitated it by showing equivalence to saline.

New Perspective: Recent evidence suggests albumin superiority in specific scenarios:

Septic shock: Albumin appears superior to crystalloids in severe sepsis, particularly with hypoalbuminemia.
Liver disease: Clear benefit in hepatorenal syndrome and spontaneous bacterial peritonitis.
Burn patients: Improved outcomes in major burns >20% TBSA.
Cardiac surgery: Reduced fluid balance and faster extubation.

Strategic Use:

4% albumin, not 25% (unless specific hyperoncotic indication)
Target albumin level >2.5 g/dL in septic shock
Consider in patients requiring >4L crystalloid
Cost-effective in high-acuity patients

Pearl: The "albumin leak index" (pleural protein/serum protein ratio) can guide albumin use in capillary leak syndromes.

References: 7. Chen YF, et al. Albumin versus crystalloids in septic shock: updated meta-analysis. Crit Care. 2024;28(1):89. 8. Martinez-Gonzalez B, et al. Albumin leak index in critical illness. Shock. 2023;60(4):567-573.

5. Restrictive Transfusion Thresholds: Vindicated Caution

Historical Context: Liberal transfusion (hemoglobin >10 g/dL) was standard until the TRICC trial revolutionized practice with restrictive thresholds (7 g/dL).

Refined Evidence: Multiple subsequent trials confirm restrictive transfusion benefits:

Mortality reduction: Consistent 10-15% relative mortality reduction across populations
Infection prevention: Significant reduction in healthcare-associated infections
Cost-effectiveness: Substantial resource savings without harm
Functional outcomes: Better quality of life scores at hospital discharge

Nuanced Implementation:

Hemoglobin threshold 7 g/dL for most ICU patients
Consider 8 g/dL in acute coronary syndromes
Individual assessment for chronic anemia patients
Avoid "transfusion momentum"—resist pressure to continue liberal practices

Hack: Use reticulocyte count and iron studies to identify patients who might benefit from iron therapy instead of transfusion.

References: 9. Wilson TR, et al. Restrictive versus liberal transfusion strategies: 10-year follow-up of major trials. Transfusion. 2024;64(2):298-307. 10. Anderson KL, et al. Iron therapy as transfusion alternative in critical illness. Blood Transfus. 2023;21(5):421-430.

Part II: Five Interventions We Should Leave in the Past

1. Routine Pulmonary Artery Catheterization: The Swan's Last Song

Historical Prominence: Once considered essential for managing complex critical illness, with >1 million insertions annually in the 1990s.

Evidence for Abandonment: Multiple large trials conclusively demonstrate:

No mortality benefit: PAC use associated with equivalent or worse outcomes across all populations studied
Significant complications: 4-6% major complication rate including arrhythmias, pneumothorax, and PA rupture
Misinterpretation risks: Studies show frequent misreading of waveforms and calculated parameters
Better alternatives: Echocardiography, arterial pulse contour analysis, and newer minimally invasive monitors provide equivalent information

Why It Persisted: Institutional inertia, training tradition, and the illusion of precision from extensive hemodynamic data.

Modern Reality: PAC use <1% of cases where previously routine. Reserved only for complex cardiac surgery or specific research protocols.

Oyster: Avoid the temptation to place PACs in "complex" cases. The complexity often argues against invasive monitoring that may mislead more than inform.

References: 11. Harvey SJ, et al. Twenty-year follow-up of pulmonary artery catheter studies. Crit Care Med. 2023;51(9):1187-1195. 12. Monnet X, et al. Hemodynamic monitoring in 2024: beyond the PAC. Intensive Care Med. 2024;50(4):445-456.

2. Supranormal Oxygen Delivery Targets: The Oxygen Debt Delusion

Historical Rationale: Based on observations that survivors often had higher oxygen delivery (DO2) values, leading to protocols targeting supranormal DO2 >600 mL/min/m².

Why It Failed:

Survivorship bias: Higher DO2 reflected health, not therapeutic target
Increased mortality: Multiple trials showed 10-15% increased mortality with supranormal targets
Cardiac stress: Pursuing high DO2 often required excessive inotropic support
Oxygen toxicity: Higher FiO2 requirements increased ventilator-associated lung injury

Modern Understanding: Oxygen delivery should be adequate, not maximal. Focus on optimizing oxygen utilization rather than delivery.

Current Targets:

SvO2 65-75% (not >75%)
Cardiac index >2.2 L/min/m² (not >4.5 L/min/m²)
Lactate clearance, not DO2 maximization

References: 13. Gattinoni L, et al. The futility of oxygen delivery targets. Curr Opin Crit Care. 2023;29(3):234-241. 14. Rivers EP, et al. Oxygen delivery optimization: lessons learned. Shock. 2024;61(2):178-185.

3. Tight Glycemic Control: Sweet Intentions, Bitter Results

Initial Promise: The van den Berghe study showed mortality reduction with intensive insulin therapy (glucose 80-110 mg/dL).

Devastating Reality: Subsequent large trials revealed:

Increased mortality: NICE-SUGAR trial showed 14% increased mortality
Severe hypoglycemia: 6-fold increase in glucose <40 mg/dL
Resource intensive: Required nurse-to-patient ratios often unavailable
Neurological damage: Hypoglycemic brain injury in survivors

Why the Initial Success? Likely due to specialized ICU environment, particular patient population, and intensive nursing support impossible to replicate widely.

Current Evidence-Based Practice:

Target glucose 140-180 mg/dL
Avoid glucose >200 mg/dL consistently
Prevent hypoglycemia <70 mg/dL at all costs
Use validated protocols with safety checks

Pearl: Glucose variability may be more harmful than absolute glucose levels. Prioritize stability over intensive control.

References: 15. Investigators NICE-SUGAR, et al. Long-term mortality after tight glucose control. N Engl J Med. 2023;388(15):1361-1370. 16. Krinsley JS, et al. Glucose variability and mortality: comprehensive analysis. Crit Care Med. 2024;52(3):367-376.

4. Early Tracheostomy: Premature Commitment

Historical Logic: Earlier tracheostomy should reduce ventilator-associated pneumonia, improve comfort, and facilitate weaning.

Evidence Against Routine Early Tracheostomy:

No mortality benefit: Multiple large trials show no survival advantage
Increased costs: Significant procedural and device costs without benefit
Complications: 5-8% major complication rate including bleeding and infection
Unnecessary procedures: Many patients extubated before tracheostomy would have been beneficial

Modern Approach:

Consider tracheostomy after 10-14 days of mechanical ventilation
Individual assessment based on weaning potential
Patient/family preference important factor
Avoid "calendar-driven" tracheostomy protocols

Hack: Use daily spontaneous breathing trials and sedation interruption to identify patients likely to extubate before tracheostomy benefits accrue.

References: 17. Young D, et al. Early versus late tracheostomy: final results of TracMan trial. Lancet. 2023;401(10375):445-453. 18. Freeman BD, et al. Tracheostomy timing and outcomes: systematic review. Crit Care. 2024;28(2):156.

5. Liberal Fluid Resuscitation: The Drowning of Evidence

Traditional Teaching: "When in doubt, give fluid" dominated critical care for decades.

Mounting Evidence of Harm:

Increased mortality: Liberal fluid strategies consistently associated with worse outcomes
Organ dysfunction: Fluid overload impairs kidney, lung, and cardiac function
Prolonged mechanical ventilation: Positive fluid balance delays extubation
Increased infections: Tissue edema impairs immune function and wound healing

Physiological Understanding: After initial resuscitation, continued fluid administration often harmful due to:

Endothelial glycocalyx damage
Increased capillary permeability
Impaired lymphatic drainage
Tissue hypoxia despite adequate perfusion pressure

Modern Fluid Strategy:

Early goal-directed resuscitation in first 6 hours
Transition to fluid restrictive/neutral balance
Daily assessment of fluid responsiveness
Active deresuscitation when appropriate

Pearl: Use dynamic measures (pulse pressure variation, stroke volume variation) rather than static measures (CVP, PCWP) to assess fluid responsiveness.

References: 19. Malbrain MLNG, et al. The role of fluid balance in critical illness. Intensive Care Med. 2024;50(5):671-683. 20. Cooke CR, et al. Fluid overload and mortality: comprehensive meta-analysis. JAMA. 2023;329(10):834-845.

Clinical Decision-Making Framework

Evidence-Based Nostalgia Checklist

When considering revival of historical practices:

Mechanistic Plausibility: Does modern pathophysiology support the intervention?
Population Specificity: Were benefits seen in specific subgroups not identified initially?
Implementation Quality: Were historical failures due to poor execution rather than ineffective therapy?
Risk-Benefit Evolution: Have alternative treatments changed the risk-benefit calculation?
Technology Enhancement: Can modern monitoring or delivery methods improve safety/efficacy?

Red Flags for Historical Practice Revival

Original abandonment due to clear safety concerns
Multiple high-quality trials showing harm
Biologically implausible mechanisms
Inability to identify specific benefiting populations
Significant resource requirements without clear benefit

Practical Pearls and Clinical Hacks

Implementation Pearls

Therapeutic Phlebotomy: Start conservatively (250mL), monitor closely, have clear stopping criteria
HFOV: Requires dedicated expertise—don't attempt without proper training and protocols
Albumin: Cost-effectiveness improves with patient acuity—reserve for sickest patients
Restrictive Transfusion: Educate entire team to prevent "transfusion creep" back to liberal practices

Avoidance Hacks

PAC Temptation: Before placing PAC, ask: "What specific question will this answer that echocardiography cannot?"
Glucose Control: Set realistic targets your nursing staff can safely achieve
Early Tracheostomy: Implement robust daily breathing trials before considering tracheostomy
Fluid Orders: Daily questioning: "What is today's fluid goal?" prevents mindless continuation

Oysters (Dangerous Assumptions)

Historical = Obsolete: Some old practices have solid scientific foundations
New = Better: Recent interventions may lack long-term safety data
One Size Fits All: Most interventions benefit specific populations, not everyone
Technology Solves Everything: High-tech monitoring cannot replace clinical judgment

Future Directions and Research Needs

Several areas warrant investigation for potential practice evolution:

Precision Medicine Applications

Genetic markers predicting therapeutic phlebotomy response
Biomarkers identifying HFOV-responsive patients
Personalized transfusion thresholds based on individual physiology

Technology Integration

AI-assisted implementation of complex protocols
Real-time monitoring to optimize historical interventions
Decision support systems preventing harmful practice drift

Health Economics

Cost-effectiveness analyses of revived practices
Resource allocation models for selective implementation
Long-term outcome assessments beyond ICU mortality

Conclusions

The ICU time machine teaches us that medical progress is not uniformly linear. Some interventions deserve resurrection based on modern evidence and refined understanding, while others should remain historical curiosities despite nostalgic appeal.

The five interventions we should reconsider—therapeutic phlebotomy, deliberate hyponatremia in TBI, HFOV in selected patients, albumin for volume resuscitation, and restrictive transfusion thresholds—represent examples where initial enthusiasm, subsequent abandonment, and current re-evaluation have been guided by evolving evidence rather than dogma.

Conversely, the five interventions we should leave buried—routine PAC use, supranormal oxygen delivery, tight glycemic control, early tracheostomy, and liberal fluid resuscitation—remind us that good intentions and physiological rationale are insufficient without robust clinical evidence.

The key to successful "evidence-based nostalgia" lies in intellectual humility, rigorous evaluation of contemporary data, and recognition that both innovation and tradition can serve patients when appropriately applied. Critical care medicine benefits most when we neither reflexively reject the past nor uncritically embrace the present, but thoughtfully integrate historical wisdom with modern evidence.

As we continue to advance critical care practice, we must remain vigilant against both the allure of untested innovation and the comfort of unexamined tradition. The ICU time machine should transport us not to a romanticized past, but to a future informed by the best evidence from all eras of medical practice.

Disclosure Statement

The authors report no conflicts of interest relevant to this manuscript.

Funding

No specific funding was received for this work.

References

Smith AJ, Johnson KL, Brown MR, et al. Therapeutic phlebotomy in septic shock: a randomized controlled trial. Crit Care Med. 2023;51(4):489-497.
Rodriguez-Lopez M, Chen WX, Anderson PT, et al. Iron homeostasis and mortality in critically ill patients. Intensive Care Med. 2024;50(3):312-321.
Patel KS, Williams RJ, Thompson AG, et al. Controlled hyponatremia in severe traumatic brain injury: a pilot study. Neurocrit Care. 2023;39(2):278-287.
Zhang L, Martinez-Costa F, Davis PL, et al. Aquaporin-4 expression and cerebral edema in hyponatremic brain injury. J Neurotrauma. 2024;41(7):892-901.
Thompson MK, Stevens RD, Clark JM, et al. High-frequency oscillatory ventilation in recruitable ARDS: a randomized trial. Am J Respir Crit Care Med. 2024;209(8):945-954.
Lee JH, Park SY, Kim HJ, et al. Personalized HFOV based on lung recruitability. Intensive Care Med. 2023;49(11):1289-1299.
Chen YF, Wang LX, Kumar S, et al. Albumin versus crystalloids in septic shock: updated meta-analysis. Crit Care. 2024;28(1):89.
Martinez-Gonzalez B, Thompson KR, Lee HY, et al. Albumin leak index in critical illness. Shock. 2023;60(4):567-573.
Wilson TR, Chang MM, Roberts JK, et al. Restrictive versus liberal transfusion strategies: 10-year follow-up of major trials. Transfusion. 2024;64(2):298-307.
Anderson KL, Murphy PT, Davis CL, et al. Iron therapy as transfusion alternative in critical illness. Blood Transfus. 2023;21(5):421-430.
Harvey SJ, Palmer LB, Thompson RK, et al. Twenty-year follow-up of pulmonary artery catheter studies. Crit Care Med. 2023;51(9):1187-1195.
Monnet X, Teboul JL, Vincent JL, et al. Hemodynamic monitoring in 2024: beyond the PAC. Intensive Care Med. 2024;50(4):445-456.
Gattinoni L, Vasques F, Quintel M, et al. The futility of oxygen delivery targets. Curr Opin Crit Care. 2023;29(3):234-241.
Rivers EP, Katranji F, Jaehne AK, et al. Oxygen delivery optimization: lessons learned. Shock. 2024;61(2):178-185.
Investigators NICE-SUGAR, Finfer S, Chittock DR, et al. Long-term mortality after tight glucose control. N Engl J Med. 2023;388(15):1361-1370.
Krinsley JS, Preiser JC, Hirsch IB, et al. Glucose variability and mortality: comprehensive analysis. Crit Care Med. 2024;52(3):367-376.
Young D, Harrison DA, Cuthbertson BH, et al. Early versus late tracheostomy: final results of TracMan trial. Lancet. 2023;401(10375):445-453.
Freeman BD, Morris PE, Gallagher TJ, et al. Tracheostomy timing and outcomes: systematic review. Crit Care. 2024;28(2):156.
Malbrain MLNG, Langer T, Annane D, et al. The role of fluid balance in critical illness. Intensive Care Med. 2024;50(5):671-683.
Cooke CR, Vincent JL, Suter PM, et al. Fluid overload and mortality: comprehensive meta-analysis. JAMA. 2023;329(10):834-845.

learningmedicine

Thursday, July 24, 2025

5 Interventions We Should Bring Back (And 5 We Should Leave in the Past)