Comparison of Non-Invasive Methods of Assessing Fluid Responsiveness in ED and ICU Patients

Comparison of Non-Invasive Methods of Assessing Fluid Responsiveness in ED and ICU Patients

Accurate assessment of fluid responsiveness (FRes) is central to guiding fluid management in septic and critically ill patients. As evidence accumulates that both inadequate and excessive fluid resuscitation are associated with increased morbidity and mortality, it is simultaneously becoming increasingly clear that current widely used methods to predict FRes are of questionable accuracy. The optimal technique to predict FRes would be a non-invasive point-of-care test with not only a high degree of accuracy, but also one which requires minimal training to perform correctly and may be easily performed repeatedly for serial evaluation of FRes during the ongoing management of the critically ill patient.

To date, three major ultrasonographic modalities have emerged as viable candidates for the bedside assessment of FRes: 1) measurement of dynamic changes in inferior vena caval diameter (IVC-CI), 2) measurement of dynamic changes in peripheral arterial waveform derived variables (PA Doppler), and 3) echocardiographic measurement of dynamic changes in left ventricular outflow tract waveform derived variables (LVOT Doppler). In this study, the investigators will perform the first direct comparison of techniques representing all three of the above modalities in the prediction of FRes against a non-invasive bioreactance cardiac output monitor (the Cheetah NICOM™), which has been extensively validated against gold-standard invasive methods of cardiac output measurement. The investigators will compare the accuracy of these modalities in both spontaneously breathing and mechanically ventilated patients using passive leg raise testing (PLR) as a surrogate for volume challenge. In addition, the investigators will also elicit information from the treating physician(s) on their clinical assessment of FRes.

The investigators will recruit adult patients in the ED and ICU with sepsis, who have received an initial bolus of 20-30 cc/kg of IV fluid, and can tolerate the PLR and US procedures. Prior to US and NICOM measurement, investigators will ask the attending physician managing the patient regarding their assessment of the their fluid status. Then, separate investigators will perform US and NICOM measurements independently to ensure blinding. NICOM data, which has been well validated in septic patients, will be shared with the attending physician. They will then be asked if this new data will change management.

This is a diagnostic accuracy study comparing selected ultrasound techniques to a well validated standard for predicting fluid responsiveness in septic patients (NICOM). The primary outcomes will be reported in sensitivity, specificity, positive predictive value, negative predictive value, likelihood ratios, and receiver operating characteristic curves.

The timeframe for measurement is within 72 hours of the subjects' admission. Determination of the outcome (i.e., whether a subject is fluid responsive) is based upon the data generated by the NICOM device. The NICOM generates data within 5 minutes of assessment. The ultrasound modalities to be studied will generate data within 5 minutes of assessment as well. The US data will be compared to the NICOM data to generate the statistics listed in this protocol. There is no further planned follow-up for the subjects.

1. The investigators will compare the ability of dynamic changes in inferior vena caval diameter (IVC-CI), dynamic changes in peripheral arterial waveform derived variables (PA Doppler), and dynamic changes in left ventricular outflow tract waveform derived variables (LVOT Doppler) to detect fluid responsiveness (FRes) relative to FRes as assessed by bioreactance cardiac output monitoring before and after passive leg raise (PLR).

   Hypothesis: PA Doppler and LVOT Doppler will have superior sensitivity and specificity to predict FRes in comparison to IVC-CI.

2. The investigators will evaluate data acquisition duration and data acquisition failure rate of IVC-CI, PA Doppler, and LVOT Doppler. Hypothesis: PA Doppler will have shorter data acquisition durations in comparison to LVOT velocity variation and IVC-CI.

   Hypothesis: PA Doppler will have lower data acquisition failure rates in comparison to LVOT velocity variation and IVC-CI.

3. The investigators will compare the ability of IVC-CI, PA Doppler, and LVOT Doppler to predict FRes in comparison to FRes as predicted by clinical judgment.

   Hypothesis: PA Doppler and LVOT Doppler will have superior sensitivity and specificity to predict FRes in comparison to clinical judgment.

4. The investigators will evaluate the ability of dynamic changes in PA corrected flow times (cFT) to detect FRes relative to FRes as assessed by bioreactance cardiac output monitoring before and after PLR. Hypothesis: cFT will have greatest sensitivity and specificity to predict FRes at a Δflow time threshold of ≥15% after PLR.

Hypothesis: At a Δflow time threshold of ≥15%, cFT will have superior sensitivity and specificity to detect FRes in comparison to IVC-CI.

Inclusion Criteria:

• Subjects presenting to the ED will be screened in accordance with the Olive View Medical Center Sepsis Waiver, and subjects meeting two of the following four criteria will be eligible for inclusion in the study: Temperature >38/100.4 or <36/96.8, heart rate >90, respiratory rate >20 or PaCO2<32, and WBC >12K or <4K or > 10% bands (neutrophils). The clinical team will be contacted regarding patients identified based on this initial screening, and if the patient is determined to have a suspected infection and a fluid bolus is planned, the patient will be enrolled in the study if they meet the following criteria:
• Adult between the age of 18 to 90 years old;
• Not pregnant;
• Not using non-invasive positive pressure ventilation (CPAP or BiPAP);
• No history of heart, lung, and/or liver transplant;
• Not determined by the clinical team to be acutely unstable requiring immediate intervention at the time of data collection, and;
• No contraindications to a passive leg raise test (known or suspected cranio-cerebral or C-spine injury, venous thrombosis, orthopedic injury requiring immobilization or traction, or other contraindications as per clinical team).

Exclusion Criteria:

• Pregnant;
• Using non-invasive positive pressure ventilation (CPAP or BiPAP);
• Presence of or suspected elevated intracranial pressure;
• Presence of central venous catheters or other medical devices that prevent the use of bedside ultrasound;
• History of heart, lung, and/or liver transplant;
• History of critical aortic stenosis or severe aortic insufficiency;
• Determined by the clinical team to be acutely unstable and requiring immediate intervention at the time of study enrollment or data collection or;
• Has a contraindication to passive leg raise maneuver, e.g. known cervical spinal injury, deep venous thrombosis, orthopedic injury requiring immobilization or traction.