Bentzer P, Griesdale DE, Boyd J, MacLean K, Sirounis D, Ayas NT. Will This Hemodynamically Unstable Patient Respond to a Bolus of Intravenous Fluids?. JAMA. 2016;316(12):
Why I chose this article
We ask this question for every critically ill patient who remains hypotensive after what we believe is adequate fluid resuscitation. Knowing techniques to reassess these patients and understanding the strengths and limitations of each method is essential to our practice, and this article from JAMA’s The Rational Clinical Examination series provides us with a good overview.
Crystalloid infusion is the backbone of resuscitation with the goal to maximize cardiac output and tissue perfusion/oxygenation. Too little fluid and end organ perfusion is compromised. Too much fluid and there is an increased risk of ARDS, edema, morbidity, and mortality. The goal is finding that sweet spot and there are several tools that can be used.
- Physical Exam
- Mucus Membranes
- capillary refill
- Static measurement
- Central Venous Pressure (CVP) – representation of right ventricular volume, if low (<8 mmHg) then fluid bolus should improve preload and therefore cardiac output.
- Dynamic measurements
- Pulse Pressure Variation (PPV) – Inspiration and expiration causes changes in venous return and therefore cardiac output. These changes can be measured as differences in systolic and diastolic pressure at these time points. Accurate measurement requires that the patient be intubated without spontaneous respiratory effort and is dependent on the patient’s tidal volume as the tidal volume effects thoracic and venous return. Thresholds for Low tidal volume (<7 mL/kg) – 8% (above means fluid responsive); High tidal volume (>7 mL/kg) – 11%.
- Stroke Volume Variation (SVV) – similar to PPV but based on variation of stroke volume; from my research requires a special catheter or calculator. Also requires mechanical ventilation and higher tidal volumes (>8 ml/kg) – 13%
- Inferior Vena Cava Variation – If intubated, positive pressure breath would increase IVC size 15% during inspiration. Spontaneous breathing would decrease of IVC size 40% with inspiration.
- Passive Leg Raise – 45 degree leg raise of supine patient will imitate a fluid bolus due to increased blood return to heart – watch for 2 mmHg increase in CVP to know that volume improved. Increase in cardiac output (>11%) or increase in pulse pressure variation ( >10%) to know that patient is responsive.
What tests are best to determine if a patient will respond to a fluid challenge?
- Studies included adult ED/ICU patients with hemodynamic instability and objective findings of improvement of cardiac output.
- Results could be placed on 2×2 table to calculate sensitivity, specificity, and likelihood ratios (LR+, LR-).
- Excluded studies with small number patients (<20) and poor evidence.
The table above is all the money. Findings are consistent with my previous reading/clinical experience. Passive Leg Raise measurements are the best by far with LR+ of 11. Physical exam findings are poor with LR+ and LR- CI both crossing 1, and both in the wrong direction 0.93 and 1.2 respectively. IVC US had a good LR+ but had a wide confidence interval that approached 1. CVP does not have high LR+ or LR-.
Remember, likelihood ratios depend on pre-test probability to determine post-test probability. For PLR with CO, if results are positive (LR+ = 11) then the likelihood of benefit from fluid bolus moves from 50% -> 92%, and if negative (LR- = 0.13) then 50% -> 11%.
Application to our practice
The last four paragraphs of the discussion are worth a read because it lays out the bottom line for critically ill patients. Frequent reassessments and integration of multiple data points are essential for these patients. The critically ill patient is complex with many factors effecting the decision to bolus fluids; if they are not intubated, are you concerned that administration of extra fluid will lead to intubation? Do they have other co-morbid conditions limiting the liberal use of fluids? Are they receiving vasopressors or inotropes that may effect Frank-Starling curve and effect of fluid bolus? Use the data and your pre-test opinion to guide treatment.
In my future practice, when I have these type of patients I plan to do a better job of baseline assessment. What does their IVC look like originally, what is their CVP, and physical exam. After a fluid bolus how do these change and then go from there. Unfortunately, none of these techniques seems easy and most require equipment or training; so just saying “do a passive leg raise because it is the best” is a bit more challenging than it sounds.