This issue of Critical Care launches the first review in the new Health Technology Assessment section. As outlined in the editorial 1, the format is a combination of information from the developer and a balanced independent review. These articles should be read in conjunction as they are designed to assess the technology from two different perspectives.

The technology under review is a continuous cardiac output monitor based on lithium dilution (LiDCO™plus, LiDCO Ltd, Cambridge, UK). The first section is based on a structured questionnaire derived from the SCCM Working Group of HTA 2. This provides answers from the manufacturer relating to the technology's background, usage and outcome data. The responses are presented unaltered for the reader to form their own opinion. Clearly there is a potential for product promotion, but the formalised structure and narrow scope of the questions are designed to minimise this. There then follows a review by Dr Rupert Pearce, who has experience with the technology but no competing interests.

We hope this combination of articles will provide some added value in an area of our specialty where the truth often lies buried. The questionnaire-and-review structure of the assessment is intended as a template for development of the HTA section, and will be maintained as a consistent format for future device reviews.

Kashif Ikram and John Barry

PAC = pulmonary artery catheter; PiCCO = pulse-induced contour cardiac output.

K Ikram is an employee of LiDCO Ltd and J Barry is a Director of LiDCO Plc.

Measurement of cardiac output and its role in clinical management remains a controversial topic. Although the pulmonary artery catheter (PAC) has not been shown to cause excess mortality 58, concerns remain about the morbidity associated with such an invasive technique.

Some practitioners do not accept a role for cardiac output measurement in clinical practice. Others believe fluid and/or inotropic therapies should be guided by flow measurements wherever possible. This debate is beyond the scope of the present review. Any data provided by a monitoring device should be interpreted with care and used in conjunction with other physiological and biochemical parameters. Clinical estimation of cardiac output, even by an experienced physician, is unreliable 59, whereas the use of flow monitoring has proved beneficial in various patient groups both with 6061 and without the use of targets for oxygen flux 62636465.

The LiDCO™plus™ system (LiDCO Ltd, Cambridge, UK) is one of several cardiac output measurement devices that are now commercially available. This review provides a critical analysis of the technological basis for the product and discusses its clinical applications. The aim is to equip the reader with an understanding of the strengths and limitations of the system (Table 1), thereby allowing safe and more effective use.

The LiDCO™plus system employs two technologies. Initial cardiac output measurement is performed by lithium indicator dilution. A bolus of lithium chloride is injected intravenously and then detected by a lithium-sensitive electrode attached to an arterial cannula. This measurement is then used to calibrate pulse contour analysis software, which provides continuous cardiac output data by analyzing the arterial pressure waveform. The technique is minimally invasive, requiring only arterial and venous cannulae. A peripheral venous cannula may be used although a central venous catheter is preferable. Patients in whom cardiac output monitoring is useful generally require invasive arterial and central venous pressure monitoring, and the use of this system does not usually require additional cannulation.

The underlying technology is similar to that employed by the PiCCO™ (Pulsion systems, Munich, Germany) system, which uses transpulmonary thermodilution to calibrate pulse contour analysis software. Both systems allow minimally invasive measurement of cardiac output in conscious and unconscious patients for as long as necessary. This allows wider clinical application than oesophageal Doppler, which is poorly tolerated in conscious patients, or the PAC, the duration of use of which is limited by infection risk.

Several studies have evaluated the lithium dilution technique of cardiac output measurement, most frequently in comparison with thermodilution using the PAC. Only three studies in humans have been published in peer-reviewed journals, two in cardiac surgical patients 6667 and one in critically ill paediatric patients 68. The accumulated evidence in animal and human studies does suggest a good correlation with thermodilution using the PAC. However, whether thermodilution can be regarded as a 'gold standard' of cardiac output measurement is doubtful. It is estimated that a 22% change in cardiac output is necessary before any difference is detected by this technique 69. It is reasonable to accept the body of evidence for the accuracy of the lithium dilution technique for the time being, but further validation in a general population of critically ill adults would be helpful.

The pharmacokinetics of intravenous lithium have been described 70. No additional side effects of the administration of lithium by this route have been reported. The recommended dose of lithium required for calibration may be used on 10 successive occasions in a 40 kg anephric patient without exceeding the therapeutic range for oral lithium therapy. The use of intravenous lithium chloride is not recommended in patients who weigh under 40 kg, those who are pregnant and those receiving oral lithium therapy.

The various features of the arterial pressure waveform are determined by the physiology of both the heart and the peripheral circulation. Any complex waveform may be analyzed by separation into a number of contributory waveforms or harmonics. PulseCO™ (LiDCO Ltd) calculates change in stroke volume by power analysis of the first harmonic of the arterial pressure waveform. This approach differs slightly from that of the PiCCO™ system; PulseCO™ analyses the arterial waveform throughout the cardiac cycle whereas PiCCO™ utilizes only the area under the systolic portion of the curve. There are no published reports directly comparing these two approaches of pulse contour analysis.

It is only possible to calculate changes in stroke volume rather than absolute values, hence the requirement for calibration by lithium dilution. Because lithium dilution measures cardiac output rather than stroke volume, significant change in heart rate during the calibration process will result in misleading data.

Recalibration is recommended every 8 hours. This technique of pulse contour analysis has been validated by comparison with lithium dilution and thermodilution techniques 71. Once again this raises the question of whether there is a reliable standard against which a new technology can be compared. This concern applies to all cardiac output measurement techniques. The system has been used successfully with arterial cannulae in various sites, although not all have been scientifically validated. The PiCCO™ system may only be used with a cannula placed in the femoral or axillary artery.

Changes in the damping coefficient of the arterial pressure transducing system may profoundly alter cardiac output measurements. While air bubbles or blood clots in the arterial cannula may be removed, kinking of the cannula may necessitate recalibration or even replacement of the arterial cannula followed by recalibration. As circulatory compliance changes in response to primary physiological changes or vasoactive drugs, the morphology of the arterial waveform alters. Although studies do not report measurement error as a result of this phenomenon, this must be an inherent risk in any form of pulse contour analysis. Regular scrutiny of the arterial pressure waveform for changes in morphology is necessary, and recalibration may be required. Because cardiac output is estimated every cardiac cycle, atrial fibrillation, and occasionally other arrhythmias, may result in irregular data output, limiting clinical usefulness. This is not problematic unless the pulse rate is particularly irregular. Adjustments to the system may improve data quality.

One indication for the use of flow monitoring is the prediction of fluid responsiveness. The LiDCO™plus system also calculates the pulse pressure, systolic pressure and stroke volume variations that occur through the respiratory cycle.

These dynamic markers of fluid responsiveness are more reliable than traditional techniques 72 and more practical to use than fluid challenges guided by stroke volume change. This feature combined with more traditional parameters permits more appropriate fluid management in the ventilated patient. However, variations in stroke volume or pulse pressure may not be as readily attributed to hypovolaemia in the spontaneously breathing patient or in the presence of an irregular cardiac rhythm. As a result, these parameters may not be reliable in a large proportion of critical care patients.

The equipment provides a valuable guide to fluid and inotropic therapy in high-risk patients in the intensive care unit, operating theatre and other critical care areas. Because of the minimally invasive nature of the technology, the device may be used more readily than the PAC.

Training in the use of the system is necessary, but with practice it is possible to perform an initial calibration within 10 min and subsequent recalibrations within 5 min. This is faster than pulmonary artery catheterization and comparable to the PiCCO™ system but slower than the oesophageal Doppler. Any error in the calibration process once the lithium bolus is injected will result in a delay of approximately 15 min while the background plasma lithium concentration subsides. The use of muscle relaxants may interfere with calibration (although not continuous measurement) for up to 45 min. Lithium chloride is safe in the doses used and the maximum dose is rarely a limiting factor. The equipment is generally reliable, although there have been manufacturing problems with the lithium sensors in the past.

There are as yet no published interventional studies utilizing the LiDCO™plus system but a randomized trial of postoperative goal-directed haemodynamic therapy is under way. The benefits of cardiac output measurement using various devices have been repeatedly demonstrated 606162636465. What is important in any new method of cardiac output measurement is its accuracy and reliability rather than validation in interventional studies.

It is not clear whether data provided by the LiDCO system are accurate during the use of the intra-aortic balloon pump. Use is also not recommended in the presence of aortic regurgitation; whether mild valve dysfunction has any clinically relevant effect on data accuracy is unclear. Further validation in these two areas may allow wider use of the technology.

The new generation of cardiac output measurement techniques include the LiDCO™plus system, oesophageal Doppler and PiCCO™, as well as other technologies. Each device provides a safe and reliable alternative to the PAC. The choice of monitor depends mainly on the clinical application. The advantages of the LiDCO™plus system are that it is minimally invasive and may be used in conscious and unconscious patients.

None declared.