10. Glycemic Control in Critical Illness

Quick Recap

Cross-cutting supportive-care protocol — part of the daily ICU bundle. This topic has one of critical care's most instructive historical arcs: a single-center trial's dramatic positive findings were widely adopted as practice, then substantially reversed by larger, more rigorous multicenter evidence — a cautionary tale about single-center trial generalizability worth understanding, not just the final number to target.

1. Definition

Stress hyperglycemia: elevation of blood glucose in critically ill patients (including those without pre-existing diabetes) driven by counter-regulatory hormone release (cortisol, catecholamines, glucagon), cytokine-mediated insulin resistance, and increased hepatic gluconeogenesis — a physiological response to acute illness, not necessarily pathological in itself, but associated with worse outcomes when severe or poorly controlled.

Intensive glycemic control: historically defined as a target blood glucose of 80–110 mg/dL (4.4–6.1 mmol/L), the "tight control" target popularized by the early Leuven trials and subsequently found to be harmful in larger trials (Section 11/22).

Conventional/moderate glycemic control: current standard-of-care target, generally 140–180 mg/dL (7.8–10.0 mmol/L), avoiding both severe hyperglycemia and the hypoglycemia risk associated with tighter targets.

Glycemic variability: the degree of fluctuation in blood glucose over time, increasingly recognized as a potentially important outcome-associated metric independent of mean glucose level or a single target range — an emerging, not yet fully established, area of interest (Section 22).

2. Pathophysiology

Critical illness produces a state of stress-induced insulin resistance superimposed on counter-regulatory hormone excess, resulting in hyperglycemia even in patients without underlying diabetes. This is mechanistically similar to, but distinct from, chronic type 2 diabetes — it is an acute, illness-driven phenomenon that typically resolves as the underlying critical illness resolves.

Why intensive control seemed rational but proved harmful: hyperglycemia is associated with adverse outcomes (infection risk, polyneuropathy, mortality) in observational data, making tight normalization an intuitively appealing target. However, achieving very tight glucose targets (80–110 mg/dL) in the unpredictable, rapidly fluctuating metabolic environment of critical illness carries a substantial and difficult-to-avoid hypoglycemia risk, and severe hypoglycemia itself is independently associated with harm (arrhythmia, seizure, neurological injury, death) — the net effect of pursuing a tight target can be negative even if moderate hyperglycemia is genuinely harmful, because the iatrogenic risk of overshooting into hypoglycemia while chasing a narrow target outweighs the benefit of glucose normalization itself.

Diabetes status may modulate the relationship: emerging evidence suggests the harm associated with a given degree of hyperglycemia may differ between patients without diabetes (or with well-controlled diabetes) versus those with poorly-controlled chronic hyperglycemia at baseline — a patient whose tissues are chronically adapted to higher glucose levels may tolerate moderate ICU hyperglycemia differently than a previously normoglycemic patient, though this individualization is not yet incorporated into standard, universally validated targets.

3. Immediate Stabilization (ABCDE) — Glycemic Management as a Bundle Element

Not an acute stabilization scenario in itself, though severe hypo- or hyperglycemia can be acutely destabilizing:

Checklist:

4. Focused History

• Pre-existing diabetes status and prior glycemic control (HbA1c if available)
• Current nutritional intake/route (affects insulin requirement and hypoglycemia risk with feed interruption — cross-reference Nutrition Support in Critical Illness protocol)
• Concurrent corticosteroid therapy (a major, common driver of hyperglycemia in ICU patients)
• Renal function (affects insulin clearance and hypoglycemia risk)
• Hepatic function (affects gluconeogenic capacity and hypoglycemia risk, particularly in liver failure)

5. Comprehensive System-wise Examination

Not a primary examination-driven protocol; relevant findings are indirect — signs of hypoglycemia (altered mental status, diaphoresis, tachycardia — though these may be masked by sedation or critical illness itself, making point-of-care glucose monitoring more important than clinical suspicion alone in this population) and signs of severe hyperglycemia-related complications (though acute DKA/HHS are addressed in their own dedicated protocols, Endocrine & Metabolic System).

6. Syndrome Identification — Reframed as Target-Setting Classification

• Standard ICU patient, no specific modifying factors: target 140–180 mg/dL per current standard of care
• Concurrent corticosteroid therapy: anticipate higher insulin requirements; do not under-treat hyperglycemia on the assumption it will resolve once steroids are weaned if the ICU stay is prolonged
• Pre-existing poorly-controlled diabetes: consider whether a somewhat higher target within the accepted range may be reasonable given baseline tissue adaptation, though this remains an individualized judgment rather than a validated separate protocol
• Hypoglycemia risk factors present (renal/hepatic failure, feed interruption risk, sepsis): heightened monitoring frequency and lower threshold for insulin infusion adjustment

7. Differential Diagnosis (of Unexplained Glycemic Instability)

Must-not-miss:

• Occult sepsis/new infection (worsening glycemic control can be an early, nonspecific sign of a new infectious process)
• Adrenal insufficiency (can present with unexplained hypoglycemia or glycemic lability)
• Unintended feed interruption in a patient on a fixed insulin infusion rate (common, preventable cause of hypoglycemia)

Common:

• Corticosteroid dose changes
• Changes in nutritional delivery rate/route
• Changes in renal function affecting insulin clearance

Iatrogenic:

• Insulin infusion rate not adjusted proactively with anticipated feed interruption (e.g., for a procedure)
• Inadequate monitoring frequency during dose titration

8. Severity/Risk Assessment

Severe hypoglycemia: blood glucose ≤40 mg/dL (2.2 mmol/L) — the threshold used in NICE-SUGAR; associated with a markedly increased risk when intensive glucose targets are pursued (NICE-SUGAR: 6.8% vs. 0.5% in the intensive vs. conventional arms, a striking absolute difference).

Emerging risk indices (observational evidence only, not yet incorporated into validated targets): time-in-target-range, glycemic variability (coefficient of variation), and the stress hyperglycemia ratio (a measure comparing acute glucose elevation to estimated chronic baseline glycemic status) are all under active investigation as potentially more nuanced outcome predictors than a single mean glucose value or target range — worth being aware of as an evolving area, not yet a basis for changing standard target practice.

9. Investigations

Immediate bedside: point-of-care glucose monitoring, frequency dictated by clinical stability and insulin infusion status (typically hourly during active titration, less frequent once stable)

Routine labs: HbA1c on admission if diabetes status is unclear or baseline control is relevant to individualizing the target discussion

Repeat frequency: glucose monitoring should be more frequent, not less, around any change likely to affect glycemic stability — nutrition rate changes, corticosteroid dose changes, hemodynamic instability, initiation/cessation of RRT

10. Point-of-Care Ultrasound — Not Applicable

11. Evidence-Based Management

The Historical Arc — Why 140–180 mg/dL, Not 80–110 mg/dL, Is Current Standard of Care

• The Leuven trials (Van den Berghe et al., early 2000s, single-center, surgical then medical ICU populations): intensive insulin therapy targeting 80–110 mg/dL showed a striking mortality benefit in the surgical ICU trial (median estimated mortality reduction ~32%), along with reduced bloodstream infections, acute renal failure requiring dialysis, and critical illness polyneuropathy — these results drove widespread adoption of tight glycemic control as standard practice through the mid-2000s
• Subsequent multicenter trials failed to replicate the Leuven findings: VISEP and Glucontrol were both terminated prematurely (VISEP due to increased hypoglycemia in the intensive arm; Glucontrol due to a high rate of unintended protocol violations) — neither could confirm the dramatic benefit seen in the single-center Leuven work
• NICE-SUGAR (2009, n=6,104, 42 hospitals across Australia, New Zealand, and Canada — by far the largest and most rigorous trial in this space): intensive control (81–108 mg/dL) vs. conventional control (≤180 mg/dL) — intensive control significantly INCREASED 90-day mortality (OR 1.14, 95% CI 1.02–1.28; absolute risk increase 2.6%, NNH 38), with severe hypoglycemia occurring in 6.8% of the intensive-control group vs. 0.5% of the conventional-control group (OR 14.7, 95% CI 9.0–25.9) — a dramatic, clinically important difference in hypoglycemia risk
• Practical synthesis and the broader lesson: this arc — promising single-center results, followed by failed replication attempts, followed by a large, rigorous multicenter trial demonstrating harm — is one of critical care's clearest illustrations of why single-center trial findings, however dramatic, require multicenter confirmation before becoming standard of care. The Leuven trials were later understood to likely reflect single-center-specific factors (patient population, glucose measurement methodology, insulin protocol specifics, non-blinding) that did not generalize

Current Standard of Care

• Target blood glucose 140–180 mg/dL (7.8–10.0 mmol/L) — this is the current, broadly endorsed standard across major critical care societies, directly informed by NICE-SUGAR
• Avoid targets below 140 mg/dL — current guidance explicitly cautions against tighter targets given the accumulated harm evidence, not merely "tighter control is unproven," but specifically "harms are likely to increase" below this threshold
• Insulin infusion protocols (rather than sliding-scale subcutaneous dosing alone) are preferred for consistent, titratable control in unstable or NPO ICU patients, transitioning to subcutaneous regimens once the patient is stable and tolerating regular nutrition

Emerging/Unresolved Areas

• Glycemic variability, time-in-range, and stress hyperglycemia ratio: increasingly studied as potentially more informative than a single target range, well-described in observational studies, but not yet incorporated into or validated within the original landmark RCTs — these remain research-phase concepts rather than a basis for changing current practice
• Diabetes-status-specific individualization: worse outcomes from hyperglycemia have been observed specifically in patients without diabetes and those with previously well-controlled diabetes, compared to patients with poorly-controlled chronic hyperglycemia at baseline — suggesting a single universal target may not be optimal for all patients, though a validated, diabetes-status-stratified target protocol does not yet exist for routine practice
• Continuous glucose monitoring (CGM) in the ICU: an active, expanding area of investigation — small single-center RCTs suggest CGM may improve time-in-target-range and reduce glycemic variability compared to conventional point-of-care monitoring, but evidence remains preliminary (small trials, single-center, 2022–2024 vintage) and CGM has not yet displaced point-of-care monitoring as ICU standard of care; accuracy in the hypoglycemic range remains a specific technical limitation worth noting (mean absolute relative difference ranging 6.6–14.8% across devices, with reduced accuracy specifically in the low-glucose range where accuracy matters most for safety)

12. Organ Support

Interacts directly with Nutrition Support in Critical Illness (feeding rate/interruption affects insulin requirement) and with renal replacement therapy (insulin clearance).

13. Disease-Specific Therapy

• Insulin infusion: titrated per institutional protocol to maintain blood glucose 140–180 mg/dL; avoid protocols or ad hoc titration targeting <140 mg/dL
• Transition to subcutaneous insulin: once hemodynamically stable and tolerating consistent nutrition, transition from infusion to a scheduled subcutaneous regimen (basal-bolus preferred over sliding-scale-only) with appropriate overlap during transition to avoid a glycemic gap

14. Consultation Matrix

15. Monitoring Framework

• Clinical: point-of-care glucose monitoring frequency matched to stability and titration status
• Escalation triggers: any severe hypoglycemic event (≤40 mg/dL) — treat as a reportable safety event warranting root-cause review, not simply corrected and forgotten
• De-escalation criteria: stable glucose within target range on a steady insulin rate → can extend monitoring interval per protocol

16. ICU Bundle Checklist (Daily)

17. Complications

Early:

• Severe hypoglycemia (arrhythmia, seizure, neurological injury, death) — the dominant, most consequential complication of overly aggressive glycemic targeting
• Uncontrolled severe hyperglycemia contributing to infection risk, osmotic effects, and (in extreme cases) DKA/HHS (cross-reference relevant Endocrine & Metabolic System protocols)

Late:

• Critical illness polyneuropathy/myopathy (historically associated with poor glycemic control in early single-center data, though the relative contribution of glycemic control specifically vs. other critical-illness factors to this complication remains debated in light of the broader trial evidence reversal)

Prevention: adherence to the 140–180 mg/dL target, insulin infusion protocols with adequate monitoring frequency, proactive rate adjustment around anticipated feed interruptions

Rescue: prompt dextrose administration and insulin infusion adjustment for hypoglycemia; escalating insulin dosing with search for an unaddressed driver (steroids, occult infection) for refractory hyperglycemia

18. Escalation & De-escalation

Escalation: persistent hyperglycemia above target despite reasonable insulin titration → reassess for unaddressed driver (new infection, steroid dosing, inadequate infusion protocol adherence) before simply increasing insulin dose further.

De-escalation: transition from insulin infusion to subcutaneous regimen once stable; reduce monitoring frequency once glucose is consistently within target on a steady regimen.

19. ICU Discharge Criteria (Glycemic-Relevant Context)

Cross-reference ICU Discharge Criteria & Step-Down protocol. Glycemic-specific consideration: transition to a ward-appropriate subcutaneous regimen completed (not transferring a patient still on IV insulin infusion without an explicit plan), and any new or adjusted diabetes management plan communicated clearly to the receiving team, including whether the current insulin requirement reflects a likely-transient stress response or an anticipated ongoing need.

20. Documentation & Medicolegal Checklist

• Glucose target documented explicitly (140–180 mg/dL) at initiation of any glycemic management protocol
• Insulin infusion rate changes and rationale documented
• Any severe hypoglycemic event documented with root-cause assessment
• Transition from infusion to subcutaneous regimen documented with rationale and overlap plan
• Diabetes status and baseline control (if known) documented as context for the management plan

21. Key Guidelines

• Current major critical care society guidance broadly converges on a 140–180 mg/dL target, directly informed by NICE-SUGAR, representing a durable practice standard since 2009 rather than an area of recent controversy at the target-range level (though see Section 22/23 for genuinely unresolved adjacent questions)

22. Landmark Trials

23. Controversies

• Why did Leuven and NICE-SUGAR disagree so starkly? This remains a topic of ongoing methodological discussion — proposed explanations include single-center vs. multicenter generalizability, differences in glucose measurement methodology and insulin protocol specifics, non-blinding effects, patient population differences (Leuven's surgical ICU population skewed toward post-cardiac-surgery patients), and possibly chance/play of a single-center result. No single explanation is fully settled, and this remains a genuinely instructive case study in evidence hierarchy rather than a fully resolved methodological question.
• Diabetes-status individualization: the observation that hyperglycemia harm may differ by baseline diabetes control status is real and biologically plausible, but has not yet been translated into a validated, diabetes-status-stratified target protocol for routine practice — clinicians should be aware of this nuance without over-applying it as if it were established practice guidance.
• Glycemic variability and stress hyperglycemia ratio: genuinely promising observational signals, but these metrics were not part of the original landmark trials' design and have not themselves been prospectively tested as targets in a large RCT — worth monitoring as an evolving evidence area rather than treating as current practice-changing evidence.
• Continuous glucose monitoring in the ICU: technologically appealing and showing early positive signals in small trials, but not yet validated at the scale or rigor needed to displace point-of-care monitoring, and with a specific, safety-relevant accuracy limitation in the hypoglycemic range that is particularly important to resolve before widespread ICU adoption, given how central hypoglycemia avoidance is to the entire evidence base underlying current practice.

24. References

1. NICE-SUGAR Study Investigators. Intensive versus conventional glucose control in critically ill patients. N Engl J Med. 2009;360(13):1283-1297.
2. Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in critically ill patients. N Engl J Med. 2001;345(19):1359-1367.
3. Van den Berghe G, Wilmer A, Hermans G, et al. Intensive insulin therapy in the medical ICU. N Engl J Med. 2006;354(5):449-461.
4. Brunkhorst FM, Engel C, Bloos F, et al; VISEP Investigators. Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med. 2008;358(2):125-139.
5. Preiser JC, Devos P, Ruiz-Santana S, et al. A prospective randomised multi-centre controlled trial on tight glucose control by intensive insulin therapy in adult intensive care units: the Glucontrol study. Intensive Care Med. 2009;35(10):1738-1748.
6. Glycemic control in the critically ill: less is more. Cleve Clin J Med. 2022;89(4):191-199.
7. Fong KM, Au SY, Ng GWY. Glycemic control in critically ill patients with or without diabetes. BMC Anesthesiol. 2022;22:227.
8. Continuous glucose monitoring versus conventional glucose monitoring in the ICU: a randomized controlled trial. 2024.
9. Continuous glucose monitoring for inpatient diabetes management: an update on current evidence and practice. 2023.
10. The Washington Manual of Critical Care, 4th ed. 2025 — glycemic control chapter.
11. ICU Protocols: A Step-wise Approach, 2nd ed. — hyperglycemia and hypoglycemia in the ICU chapter.