đŸ©žđŸ’§Silent Infusion: Fluid Stewardship in the Maintenance Phase đŸ’§đŸ©ž

By Dr Prashant Nasa

 

Dr Prashant Nasa

MD, FNB, EDIC | Royal Wolverhampton NHS Trust, UK

Editor – Rational Use of IV Fluids | 1.3M+ reads
Global voice on Fluid Stewardship, ARDS & Sepsis
Excellence in Research Awardee, UAE 2023–24

Clinical Vignette

A 68-year-old woman with diabetic nephropathy, baseline creatinine 1.4 mg/dL, is admitted to the ICU with septic shock from perforated diverticulitis.

She was resuscitated appropriately (4.5L in the first 12 hours), started on noradrenaline and received early antibiotics and source control via laparotomy.

Now, on ICU Day 4:

  • MAP is stable on low-dose norad (0.03 mcg/kg/min)
  • FiO₂ 40% via HFNC
  • Urine output is 45–60 mL/h
  • Creatinine has plateaued at 2.1 mg/dL
  • CRP down-trending, vasopressors almost off
  • Abdomen soft, no signs of ACS
  • Weight has increased by 6 kg since from admission admission
  • Serum albumin: 2.2 g/dL
  • CVP: 11 mmHg, bedside echo IVC diameter ~2.4 cm with minimal respiratory variability
  • She’s on maintenance fluid: 60 mL/hr of D5 +0.45% saline + potassium chloride, as per pre-set ICU order set
  • Enteral feeds ongoing at 60 mL/hr

She is awake, but now intermittently confused and drowsy.

"*" indicates required fields

1. A 68-year-old man with septic shock secondary to pneumonia is receiving norepinephrine 0.08 ”g/kg/min. His serum creatinine has increased from 1.0 to 1.1 mg/dL. Urine output is 0.35 mL/kg/h over the previous 6 hours. Urinary [TIMP-2] × [IGFBP7] is 1.2. What is the most appropriate next step?
2. An ICU introduces an AI-based AKI prediction model. A patient with septic shock receives an alert predicting an 85% probability of Stage 2–3 AKI within 24 hours. Which statement best reflects current best practice?
3. Which factor most limits the generalizability of a single AKI prediction model across all ICU populations?
4. An intensivist is shown an AI-generated AKI risk score without any explanation of how the prediction was derived. Which feature would most improve clinical adoption?
5. Which intervention is LEAST likely to be included in a nephroprotective bundle for a septic patient identified as high risk for AKI?
6. A hospital implements a real-time AKI alert system. Initially, clinicians respond to alerts, but after several months most notifications are ignored. Which intervention is most likely to improve effectiveness?
7. A patient’s AKI risk score changes from 20% to 75% over 8 hours despite stable creatinine. Which factor most strongly supports recalculating risk dynamically?
8. A future ICU system combines genomics, proteomics, urinary biomarkers, and continuous physiological monitoring. What is the principal advantage of such an approach?
Name*
This field is for validation purposes and should be left unchanged.

Key Questions at the Bedside

(Click / Tap on Questions to Reveal Content)

Short answer: No.
This patient is no longer in the resuscitation phase, is hemodynamically stable, has functional enteral intake, and shows signs of fluid accumulation (6 kg weight gain = likely >5% from baseline).

Additionally:

  • CVP is 11 mmHg and IVC is plethoric → venous congestion
  • Serum albumin is low → capillary leak may persist, but ongoing iatrogenic input risks worsening interstitial edema
  • Urine output is adequate (>0.5 mL/kg/h), suggesting perfusion is preserved

Takeaway: Maintenance fluids in this context may be perpetuating fluid accumulation.

 (Fluid accumulation syndrome is defined by >=5% weight gain from baseline)

Let’s decode the “cocktail” she’s receiving:

IV Fluid: D5 0.45% NaCl + KCl @ 60 mL/hr

  • D5 = 5% dextrose → 50g/L glucose
  • 45% NaCl = 77 mEq Naâș and Cl⁻ per liter
  • Let’s say she gets 20 mEq KCl per liter (typical for maintenance)

→ Osmolarity (approximate):

  • Dextrose = ~278 mOsm/L (but becomes free water once glucose is metabolized)
  • 45% NaCl = ~154 mOsm/L
  • KCl = 20 mEq = 40 mOsm

→ Total: ~472 mOsm/L, though physiologically the net tonicity is hypotonic after glucose metabolism.y

Feeds @ 60 mL/hr

  • ~1440 mL/day or ~432 mOsm/day

The hypotonic fluids can cause mild but asymptomatic hyponatremia (provided they are justified). Hence, electrolytes need to be checked daily.

Ref: Van Regenmortel N, et al. 154 compared to 54 mmol per liter of sodium in intravenous maintenance fluid therapy for adult patients undergoing major thoracic surgery (TOPMAST): a single-center randomized controlled double-blind trial. Intensive Care Med. 2019 Oct;45(10):1422-1432

PS: For the above calculation do not forget the “fluid creep”

Nutritional feeds already contain water, electrolytes, glucose, and other substrates. Continuing maintenance IV fluids on top of feeds is often redundant and harmful.

Fluid from feeds:

  • 60 mL/hr = 1440 mL/day → enough for basal needs in most adults

So who’s in charge now? The feeds.

(With enteral feeds already running, are we duplicating volume and electrolytes?

Should maintenance fluids be stopped when feeds begin—or tailored?)

Fluid creep = hidden daily fluid load from ‘non-resuscitation’ sources.
Drug diluents, flushes etc.
Breakdown:
● IV “maintenance” = 60 mL/hr = 1440 mL/day
● Enteral feed = 60 mL/hr = 1440 mL/day
● Drug diluents, flushes (typical estimate) = 250–500 mL/day-mainly sodium chloride
This exceeds the daily physiological fluid need (~2–2.5 L), especially since she’s likely anuric or oliguric from AKI.

Yes. But first de-escalate

Stop IV fluids and check fluid creep

The key signs:

  • Positive fluid balance (+6 kg)
  • Stable hemodynamics on minimal vasopressors
  • No ongoing shock
  • Congestion markers: High CVP, plethoric IVC
  • Clinical signs: New confusion/drowsiness → possibly cerebral edema or dilutional hyponatremia
  • What other bedside tools (e.g., VExUS, bioimpedance, passive leg raise, cumulative input/output) help guide this decision?) Absence of preload responsiveness in this patient is a good initial trigger for de-resuscitation.

De-resuscitation can include:

  • Loop diuretics (if not oliguric/anuric)- Preferred initial strategy
  • Renal Replacement therapy- Consider if there is ongoing CRRT, AKI, unable to achieve targets with diuretics, or diuretic resistance.

The typical physiologic daily need (in stable adults):

  • Water: 25–30 mL/kg/day (~1.8–2.1 L)
  • Naâș/Kâș/Cl⁻: 1-1.5 mEq/kg/day
  • Glucose: 1g/kg/day minimum to avoid ketosis

But this includes all sources—feeds, drugs, oral intake, flushes.

So when enteral feeds are adequate:

  • No additional maintenance IV fluids are needed.
  • Electrolyte deficits can be corrected separately, not bundled into maintenance.

Fluid stewardship should mirror antimicrobial stewardship—targeted, reassessed, and goal-directed.

Key reframes:

  1. Stop “auto-pilot” prescriptions—default maintenance is not benign.
  1. Count all fluid inputs (creep, meds, feeds, flushes).
  1. Use fluid as a drug—right patient, right dose, right duration, timely de-escalation
  1. Link fluid decisions to organ function and context (e.g. AKI, hypoalbuminemia, gut integrity).

Leave Your Comment