There is no influence on the thermodilution measurements as long as the patient’s temperature is stable. Cooled injectate should be used.
Temperature fluctuations from the baseline are compensated for by the device. However, a thermodilution measurement is not recommended when a stable baseline cannot be detected, as shown by a change > 0.05°C /min.
- Vasoconstrictors / Inotropes / Volume Therapy
All parameters are correctly calculated. Where there are significant changes in the catecholamine requirements or volume therapy, recalibration of the pulse contour analysis is recommended. Also changes in Central Venous Pressure (CVP) should be updated regularly.
- Intra-aortic Balloon Pump (IABP)
The thermodilution measurement with the PiCCO is not influenced by the IABP, but the Pulse Contour Analysis is unable to provide valid continuous output. The PiCCO-Technology can be used to measure cardiac output (CO/CI) preload volume (GEDI), lung water (ELWI), and contractility (CFI, GEF) with every thermodilution measurement.
- Continuous Renal Replacement Therapy (CRRT)
Dufour et al. (ICM 2012), Sakka et al. (Anesth Analg 2007), Heise et al. (Min Anesth 2012) and Pathil et al. (EJA 2012) investigated the influence of CRRT on PiCCO results. All concluded that CRRT does not have any significant influences on PiCCO thermodilution values.
To have accurate results in CRRT it is important that:
- PiCCO thermodilution measurements should be avoided directly after the CRRT is switched on or off
- A stable blood temperature baseline needs to be reached before performing PiCCO measurements
- The CRRT catheter out- and inflow should not lie in the PiCCO indicator passage track
- PiCCO during extra corporal membrane oxygenation (ECMO)
Clinical experience shows that PiCCO thermodilution does not give reliable results during ECMO. The reasons for this are the relatively high flow rates and the positioning of the ECMO catheters. Therefore it is recommended that PiCCO thermodilution measurements are performed only before or after ECMO treatment. During running ECMO the PiCCO pulse contour analysis should be correct as long as no significant changes in the vascular compliance are present and provided that the initial calibration from thermodilution was correct.
- Valve Insufficiency
Valve insufficiency may cause regurgitation of the thermodilution injectate and prolong the transit time of the indicator, or interfere with the thermodilution curve. However, where a thermodilution curve is possible, the calculation of the cardiac output should be correct. In mitral valve insufficiency the accuracy of the PiCCO cardiac output measurement has been confirmed (Staier et al., EJA 2012).
- Aortic Stenosis
In aortic stenosis arterial thermodilution accurately reflects cardiac output. The arterial pressure curve may have reduced systolic and elevated diastolic pressures. However the area under the arterial curve still reflects stroke volume. Recalibration of the pulse contour (with thermodilution) substantially improves reliability in severe aortic stenosis (Petzold et al., ICM 2013).
- Intra-cardiac Shunts
Due to the marked alteration in the thermodilution curve, no valid values are able to be calculated. In less severe shunts, measurements may be possible. However the occurrence of a premature hump in the thermodilution-curve can lead to the detection of a previously unknown R-L-shunt.
- Aortic aneurysms
Theoretically GEDI will be increased by the volume of the aortic aneurism, as the indicator will also have to pass through this (from CV injection to arterial detection). In case of abdominal aneurysms, this can be avoided by placing the PiCCO catheter in the axillary artery.
- Cardiac Arrhythmia
The thermodilution parameters are still measured correctly. The pulse contour analysis is correct in mild to moderate arrhythmias (normal rate atrial flutter/fibrillation, bigeminal, trigeminal or occasional extra systoles). In severe cardiac rhythm disturbances (tacharrhythmia, supraventricular tachycardia), pulse contour analysis may be inaccurate. Evaluate the quality of the pressure curve by checking for the white lines (PulsioFlex horizontal/ PiCCO2 vertical) below the pressure curve whether the algorithm was able to detect the systolic portion. It is also recommended to recalibrate with 3-5 thermodilution measurements.
- Partial Lung Resection
Lung resection procedures (lobectomy, bilobectomy, pneumectomy) theoretically reduce the Pulmonary Blood Volume (PBV) and may lead to inaccurate calculation of the Extravascular Lung Water (EVLW). To evaluate this theoretical assumption a double indicator dilution technique is required to determine PBV before and after lung resection. Published clinical data on this topic is very limited. Two studies with 35 patients (Schroder et al., Internet J Thorac Cardiovasc Surg 2005; Naidu et al., Int Cardio Vasc Thorac Surg 2009) show that:
- the amount of extracted lung tissue and pulmonary blood volume do not correlate
- clear correction factors for PBV calculation cannot be determined
- an initial effect on PBV is generally physiologically compensated for 1-2 days post-operatively
Thus, it is not recommended to correct the measured values for PBV and EVLW with fixed calculation factors. Clinical evidence is not available and such corrections may lead to unexpected and unpredictable errors in the calculation of EVLW in patients after lung resection.
- Pulmonary perfusion disturbances (e.g. pulmonary embolism)
In order to receive accurate thermodilution measurements and thereby accurate parameters, the injectate must pass from the point of injection through the central line, the right heart, across the lungs, the left heart and pass the tip of the PiCCO catheter in the chosen artery without any interruptions. In case of a lung blockage (e.g. caused by a large pulmonary embolism) measurements may be erroneous or misleading, particularly with the extravascular lung water as the whole lung volume is not measured correctly by the thermodilution. Nevertheless, in this case the cardiac output and the global end-diastolic volume would be measured correctly.
- Pleural effusion
Pleural fluid is not included in the ELWI measurement. The capillary surface of the lung parenchyma that is in contact with the pleural fluid is very small in comparison to the pulmonary capillary network. The relation is the size of a tennis court to the size of the surface of two hands. Thus, the temperature loss to the pleural fluid is negligible.
- On pump (with extracorporeal circulation – heart-lung-machine):
The initial PiCCO calibration should be done after anaesthesia but before opening the chest. When performing pulse contour calibration by thermodilution measurements the patient should be haemodynamically stable without significant changes in body temperature. Another thermodilution measurement can be conducted immediately before going on to cardiopulmonary bypass (CPB). During extracorporeal circulation the PiCCO is not able to give valid results due to the lack of an existing arterial pressure curve. Therefore thermodilution measurements are of no use during extracorporeal circulation. As soon as the heart is pumping again the PiCCO will display the cardiac output derived from the pulse contour analysis. Recalibration of the pulse contour should be performed as soon as possible. This comes along with an update on the volume status (GEDI), which is normally of interest immediately after bypass and after closing of the chest.
- Off pump:
Initial calibration is done after induction of anaesthesia. During the whole procedure, continuous cardiac output can be followed on a beat to beat basis. Recalibration during the procedure is only possible when the patient is in a widely stable condition. During the procedure the index of Left Ventricular Contractility (dPmx) gives additional contractility information and can serve as an early warning indicator for ischemic events.
Stroke Volume Variation (SVV) and Pulse Pressure Variation (PPV) serve as indicators of volume responsiveness, even under open chest conditions. However, patients must be ventilated with a tidal volume >8ml/kg, and have a sufficient sinus rhythm.
The paediatric PiCCO catheter PV2013L07 has an outer diameter of 3F (= 1mm) and a usable length of 7 cm. It is intended for use in the femoral artery of paediatric patients. The decision in which kind of patient (age, weight) this catheter is used should be made by the treating physician. Recommendations on the body weight can be derived from publications e.g. Cecchetti et al., Min Anest 2013, where a 3F catheter was used with a body weight less than 10kg and 4F catheters for paediatrics with at least 10kg body weight. In other publications (e.g. Lemson et al., Crital Care 2010; Szekely et al., Ped Card 2010; Anton et al., An Ped 2009; Egan et al., Intensive Care Med 2005; Cecchetti et al., Min Anest 2003) the youngest patients were 2 months with a body weight of 3 kg. However the appropriateness of the catheter in the individual patient hast to be checked each time again. A review of the PiCCO in paediatrics was published by Proulx et al., (Pediatr Crit Care Med 2011).
Main indications for PiCCO in paediatric patients are:
- Head trauma (Cecchetti et al., Min Anest 2013)
- Severe burn injury (Branski et al., Critical Care 2011; Kraft et al., J Surg Res 2012)
- Cardiac surgery (Keller et al., J Clin Mon Comp 2011; Szekely et al., Ped Card 2010; Cherqaoui et al., Ped Anesth 2006; Mahajan et al., Anesth Analg 2003)
- Acute respiratory failure (Lubrano et al., Intensive Care Med 2010)
- Liver transplantation (Torgay et al., Transpl Proc 2005)
- General intensive care (Cecchetti et al., Crit Care Med 2008, Cecchetti et al., Min Anest 2003)
Please be advised that normal range areas are slightly different to those of adult patients. It has been shown that GEDI tends to be lower and ELWI tends to be higher the younger and lighter the patient is (Lemson et al., Pediatrics 2011).
No, the respiratory cycle does not influence PiCCO measurements, as the thermodilution curve is approximately 20 seconds long, including approximately 3 respiratory cycles. This means that compared to the pulmonary artery catheter, the PiCCO thermodilution cardiac output has a much lower coefficient of variation.
In terms of accuracy the two methods are comparable. However, the PiCCO method has a much lower coefficient of variation. In other words, the PiCCO method is less user dependent and gives more stable measurements. When compared to the gold standard (Fick method) the PiCCO shows an excellent correlation. PiCCO pulse contour cardiac output shows a high correlation and low bias to the PiCCO arterial thermodilution cardiac output.
The Pulse Contour Cardiac Output deviates from the transpulmonary thermodilution Cardiac Output. What could be possible reasons?
The thermodilution CO will only change when a new set of thermodilution measurements are performed. The pulse contour CO, based on the patient’s specific aortic compliance is updated beat by beat based on the systolic part of the arterial curve.
In case of haemodynamically unstable patients, differences between the pulse contour and thermodilution cardiac outputs may occur. In such cases frequent recalibration (via thermodilution) is recommended.
Other causes include errors in the detection of the arterial wave form and therefore errors in the wave form analysis and extreme arrhythmias or frequent extra systoles.
Cardiac preload volume is not equivalent to Global End-Diastolic Volume Index (GEDI), so how can this clearly indicate cardiac preload?
Strictly defined, cardiac preload is the myocardial fibre stretch at the end of ventricular diastole. A parameter that accurately reflects preload in clinical practice is not yet available. However, studies have demonstrated that GEDI (or ITBI) is a reproducible and sensitive parameter in good approximation of preload (e.g. Umgelter et al., BMC Gastro 2008; Sander et al., Critical Care 2007; Michard et al., Chest 2003; Della Rocca et al., Anesth Analg 2002).
In contrast, it has been repeatedly shown that the central venous pressure, pulmonary artery occlusion pressure, and Right Ventricular End-Diastolic Volume Index do not reflect cardiac preload.
As with all products on the market that provide these parameters, to correctly interpret SVV or PPV, the patient must be
- on fully controlled positive pressure ventilation with a tidal volume ≥ 8ml/kg (no spontaneous breathing or assisted breaths) and
- have a sufficient sinus rhythm with no artefacts.
Under these conditions the relative difference between maximum and minimum stroke volume or pulse pressure over a continuous time span of 30 seconds will indicate how volume responsive a patient is. A SVV or PPV above 13% indicate that a patient will react to administered volume, whereas a SVV or PPV below 10% usually represents sufficient volume status. Values between 10 and 13% require further interpretation as they may depend on the individual patient’s clinical picture. When using lower tidal volumes (<8ml/kg) SSV and PPV have been shown to become less accurate and are not recommended. Under open chest conditions, SVV and PPV are dependent on cardiac filling, thus cardiac preload (GEDI) can be used for volume management.
Cardiac Function Index is calculated as the Cardiac Output divided by the Global End-diastolic Volume. Global Ejection Fraction is calculated as Stroke Volume multiplied by four and divided by Global End-diastolic Volume. As Cardiac Index is calculated by multiplying Stroke Volume with Heart Rate the difference between the two parameters is that CFI includes heart rate in its calculation. Thus, interpretation of GEF for cardiac contractility might be advantageous, as low Stroke Volume (due to low contractility) can be compensated by a high heart rate.
The Extravascular Lung Water Index (ELWI) in ml/kg is calculated from the absolute volume of Extravascular Lung Water in mls divided by the body weight in kg. This causes an underestimation (in obese patients) or an overestimation (in underweight patients) of ELWI, if the actual body weight is used for indexing. As the lung size does not change with the body weight, the correct way of ELWI calculation is to use and idealized body weight, or predicted body weight (PBW). PBW is calculated from the height of the patient but also including other specifics such as age and gender.
The PiCCO devices have been using PBW for the calculation of ELWI since 2007, making it accurate even in severely obese patients. In the PiCCO-Technology integrations in patient monitors (Philips, Draeger, Mindray, GE) only the latest software versions provide the ELWI calculation related to PBW. In older software versions it is recommended to enter an estimated ideal body weight which is calculated by the simplified formula: height in cm minus 100 minus 10% for males, or minus 15% in females.
In addition to PBW the Predicted Body Surface Area (PBSA) is used to calculate Global End-diastolic Volume Index (GEDI in ml/m2). For the calculation of Cardiac Index (CI in ml/min/m2) standard Body Surface Area (BSA) is used which is calculated from the real body weight.
What is the effect of injecting the temperature indicator for the PiCCO into the femoral vein instead of the superior vena cava / right atrium?
If both the central venous catheter and PiCCO arterial catheter are placed on the same side (e.g. right femoral groin) a double hump indicator curve with resulting measurement errors may occur because of a disturbed temperature signal from the "cross talk" between the venous and arterial blood streams. In other words the cold temperature can pass across to the arterial thermistor without going through the cardiopulmonary system but directly from vein to artery. This is more common in paediatric patients.
Even when using long femoral venous catheters, there is a small effect from cross talk caused by thermal migration from the catheter into the vessel.
Cross talk can be avoided if the PiCCO arterial catheter is either placed in the opposite femoral artery or in the brachial/axillary artery.
If cross talk is avoided, thermodilution measurement is possible, but the PiCCO readings for GEDI will be slightly higher than the actual volumes. This is caused by the additional volume from the point of injection to the point of detection, because the catheter for indicator injection is not placed directly before or in the right atrium. The value for ELWI will be correct, as it is calculated from the difference of two overestimated volumes.
From PiCCO2 software version V3.1 onwards the PiCCO will ask for confirmation of where both the central venous and arterial catheters are placed to ensure accurate calculation of GEDI.
Can I perform a thermodilution measurement with a peripherally inserted central catheter (PICC) rather than a centrally inserted venous catheter?
In order to get an adequate thermodilution curve for accurate parameter calculation, the injectate bolus must be injected in under 7 seconds, and remain cool enough for the PiCCO catheter to detect a difference between the patient’s blood and the bolus. It may be that these conditions cannot be fulfilled and thus this approach is not generally recommended.
The injection volume is dependent on the patient’s body weight. If the patient has an increased amount of extravascular lung water (i.e. ELWI is more than 10 ml/kg body weight), the injection volume must be increased.
In clinical practice, most people use a standard volume of 15mls of cold saline solution in adults. This setting works in most patients and avoids confusion and mistakes. However, the injectate volume has to be adapted for patients with very low or very high body weight.
In the PiCCO measurement screen a recommendation of the most appropriate injectate volume is given.
It is recommend that three consecutive measurements, with less than 20% (+/-) variation compared to the mean value are performed within a 5 minute time frame. If the patient has elevated ELWI the first measurement may not be accurate, and more or cooler injectate may be required (e.g. if you are using room temperature injectate initially and ELWI reading is > 10ml /kg).
Yes, the PiCCO-Technology monitors and analyses the shape of the thermodilution curve for plausibility. On the display screen a status line will give an error code or error message if an abnormal curve is detected. In addition the thermodilution curve is also displayed on the screen and can be examined by the user.
What are the time intervals or under which circumstances is it recommended that new thermodilution measurements are performed, so that results from the pulse contour analysis are updated and more accurate according to the patient status?
In general the PiCCO should be calibrated every 8 hours by thermodilution; however individual patient needs vary greatly. For example, if your patient is in shock you may have to determine GEDI and ELWI hourly and confirm/recalibrate Pulse Contour Cardiac Output. Once the patient is stabilised you may be able to decrease the frequency of measurements to once every 2 hours and then, if the patient remains stable decrease to every 4-6 hours. Another guide may be to perform a thermodilution measurement if the continuous cardiac output has trended consistently in the same direction for 15 minutes or if there are large and/or sudden changes in the patient’s clinical status requiring changes to their vasopressors or inotropes or fluid requirements.
Why are the Cardiac Output values from the pulmonary artery and femoral artery thermodilution measurements different?
The results for simultaneous cardiac output measurements by thermodilution in the pulmonary (COpa from a pulmonary artery catheter) and femoral artery (COa from a PiCCO) should be identical, because the area under the thermodilution curve determines Cardiac Output, which should be identical in both sites, even when the shape of the curve is different. However, it is well documented by several publications from the 1980s (e.g. Lewis et al, Ann NY Acad Science 1982, Boeck et al, J Crit Care1989), that results may differ slightly. It was shown that COa is systematically 8-10% higher than COpa, although the correlation between both methods is excellent.
The difference was originally explained as an ‘indicator loss’ which today still remains unconfirmed. Other studies show that this effect is probably caused by physiology. Harris et al (Anesthesiology 1985) clearly showed that the heart rhythm significantly slowed down (around 10%) when 10ml of iced solution was injected central venously. This bradycardia resulted in a decrease in Cardiac Output. Because of the short distance between the central venous injection and the pulmonary arterial detection this directly afffected COpa measurements. The PiCCO arterial thermodilution signal is not usually affected due to the larger distance between injection and detection point.
An additional effect on the pulmonary artery catheter measurement is created by the ventilation cycle in controlled mechanically ventilated patients. COpa can give significantly different results in the inspiration versus expiration phase. Thus it is recommended to perform the injection for COpa measurements in the end-expiration phase and to perform at least three consecutive measurements. Basically this is not necessary with PiCCO CO as each measurement gives the average value over at least one complete ventilation cycle. However, when one single COpa measurement is compared to a COa measurement the results can differ significantly.
GEDI is the total volume of the heart divided by the time it takes for the indicator to traverse the cardiopulmonary circulation. Thus, it is not a value displaying the volume of the heart during a single cardiac cycle. Indicator injected into a compartment will be diluted into the largest available volume of that compartment. When applying this to the heart, all 4 compartments must be considered: the end-diastolic volume of the right atrium, right ventricle, left atrium, and left ventricle.
The indicator dilution curves from each compartment are added both individually and in time. GEDI is thus a sum of several cardiac cycles as opposed to a discrete value from one cardiac cycle. Furthermore a small volume must be added. These include volumes from the aorta and a small amount just prior to the right atrium.
Additionally, the volume passing through the atria during diastolic filling of the ventricles may be greater than the geometry of the atria. This phenomenon is due to the fact that there are no inflow valves in the atria and thus once the atrial-ventricular valve is open the pressure gradient between inflow pressure and current ventricular pressure dictates the amount of volume passing through the atria. As a result, the end-diastolic volume of the atria may appear larger than predicted by the atrial geometry.
First of all, Sakka et al (Intensive Care Med 2000) investigated the ability of measuring extravascular lung water (EVLW) with thermodilution. In 57 intensive care patients they used the so-called double indicator dilution technique (using thermodilution with indocyanine green, ICG) allowing them to measure intrathoracic blood volume (ITBV) and extravascular lung water (EVLW). All results lead to a needed calculation factor of 1.25 in order to calculate ITBV from the thermodilution derived global end-diastolic volume (GEDV). In another 209 intensive care patients this calculation factor was used to calculate ITBV and EVLW and when comparing those calculated values to the simultaneously real measured values (with double indicator dilution technique based on ICG), very close correlations were found.
This approach of measuring EVLW with the arterial thermodilution by PiCCO was later validated in several studies.
- Katzenelson et al (Crit Care Med 2004) compared in an experimental study in 15 dogs the PiCCO lung water to the direct assessment of lung water by gravimetric technique and found an excellent correlation.
- Kuzkov et al (Crit Care Med 2007) compared in an experimental study in 30 sheep in a complex setup the PiCCO lung water to the direct assessment of lung water by gravimetric technique and found acceptable accuracy.
- Tagami et al (Critical Care 2010) compared in a human study (30 brain dead subjects) the PiCCO lung water to the post mortem lung weight and found a very close correlation.
- Kirov et al (Critical Care 2004) compared in an experimental study in 18 sheep the PiCCO lung water to the direct assessment of lung water by gravimetric technique and found a close correlation.
It is recommended that saline solution is used. The use of glucose for example, may cause the small piston inside the injectate temperature sensor to stick to the housing, thus impeding movement during injection. Also, it is imperative that lipids are not administered through the injectate sensor housing.
Different BSA calculation formulas have been used since the PiCCOplus software V7.1 and included in PiCCO2 and PulsioFlex as well as in the latest software versions of monitors of Philips, Draeger, Mindray and GE.
BSA Body Surface Area (m2)
- Patients with BW < 15kg
BSA = (W 0.5378 x H 0.3964) x 0.024265 (Haycock et al., J Pediatrics 1978)
- Patients with BW ≥ 15kg
BSA = (W 0.425 x H 0.725) x 0.007184 (Du Bois & Du Bois, Arch Int Med 1916)
PBW Predicted body weight (kg)
PBW (Kg) = 50 + 0.91 (height (cm) – 152.4)
PBW (Kg) = 45.5 + 0.91 (height (cm) – 152.4)
PBW (Kg) = 39 + 0.91 (height (cm) – 152.4)
Paediatric > 152.4cm
PBW (Kg) = 42.2 + 0.91 (height (cm) – 152.4)
Paediatric > 152.4cm
PBW (Kg) = ((height (cm))2 x 1.65/1000
PBSA Predicted body surface area (m2)
Calculated with PBW instead of actual body weight (BW)
The concentration of the indicator is distributed over time because of the volume in the system, i.e. there is a given time for particle of the indicator to travel between the point of injection and point of detection. This time is called the transit time and each particle has its own transit time. The MTt is the mean value of all these transit times. The MTt multiplied with the CO gives the whole thermal volume (Intrathoracic Thermal Volume, ITTV) that the indicator has to go through.
The down-slope time is detected by plotting the thermodilution curve with the temperature change (indicator concentration) on a logarithmic scale (ln) and time change on a linear scale (lin). When you plot the thermodilution curve as a linear-ln graph, the indicator decay approximates a linear function. Two points, the starting point located at 85% of the maximum temperature response and an end point defined as 45% of the maximum temperature response, are identified. The time difference is determined and labelled as the down-slope time (DSt). DSt multiplied with CO gives the pulmonary thermal volume (PTV) which is the largest single volume in the series of “mixing chambers” of the cardiopulmonary system. PTV consists of Pulmonary Blood Volume (PBV) and Extravascular Lung Water (EVLW).