This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Siu, L. Articles by Monagle, J. Search for Related Content PubMed PubMed Citation Articles by Siu, L. Articles by Monagle, J.

---

TECHNOLOGY, COMPUTING AND SIMULATION

Does Patient Position Influence Doppler Signal Quality from the USCOM Ultrasonic Cardiac Output Monitor?

From the Department of Anaesthesia, Monash Medical Centre, Clayton, Victoria, Australia.

Address correspondence and reprint requests to Dr. LWL Siu, Department of Anaesthesia, Monash Medical Centre, 246 Clayton Rd, Clayton, Victoria, 3168. Address e-mail to lysiu{at}yahoo.com.

Abstract

BACKGROUND: The USCOM1A continuous wave cardiac output monitor (USCOM Pty Ltd., Sydney, NSW, Australia) is a novel Doppler-based device used to measure cardiac output noninvasively. The proper alignment of the transducer, and hence the ultrasound beam to the aortic or pulmonary outflow tracts, is essential to acquire accurate measurements and often much time is spent on transducer and/or patient positioning. In this prospective, observational, crossover study, we investigated the effect of patient positioning on the acquisition of cardiac output measurement with USCOM1A.

METHODS: We measured cardiac output using USCOM1A in 30 healthy adult volunteers, each in five different positions: sitting, supine, Trendelenburg (20 degrees), left lateral tilt (20 degrees), and right lateral tilt (20 degrees) and compared the time required to obtain acceptable measurements. We also compared the quality of the Doppler signal obtained in these positions using a scoring system designed for this study.

RESULTS: There was a higher rate of failed measurement, the mean time to obtain the first acceptable measurement was prolonged and the optimal measurement obtained within a 5-min period was of a lower quality in the sitting position compared with the other four positions.

CONCLUSIONS: Our results suggested the sitting position is the least suitable and least reliable position in which to perform cardiac output measurements using USCOM1A compared with the supine, Trendelenburg (20 degrees), left lateral tilt (20 degrees), and right lateral tilt (20 degrees) positions.

The USCOM1A (Ultrasonic Cardiac Output Monitor Model 1A) was developed by an Australian company, USCOM Ltd. (Sydney, Australia), as a continuous wave Doppler-based device used to measure cardiac output noninvasively. Although Doppler echocardiography has been in clinical use to measure cardiac output since the early 1980s,1,2 the USCOM1A is a novel device, which allows determination of cardiac output independent of two-dimensional ultrasonic measurement of outflow tract cross-sectional areas. Nidorf et al. demonstrated that cardiac linear dimensions vary uniformly with height in subjects aged 6 days to 76 yr.3 Based on this, the USCOM1A uses proprietary height-referenced algorithms to predict aortic and pulmonary outflow tract areas.

The USCOM1A device weighs 4.8 kg and measures 31 cm high x 35 cm wide x 18 cm deep (approximate size of standard cardiac monitors); it has a 12.1 inches touch screen and menu system interface and a hand-held ultrasound transducer (2.2 MHz for adults or 3.3 MHz for pediatrics). By placing the transducer in the suprasternal or left parasternal position and hence aligning the ultrasound beam with the aortic or pulmonary outflow tract, respectively, velocity of the blood ejected can be measured over time shown as waveforms of velocity–time integrals on screen. These waveforms are referred to as Doppler flow profiles. The predicted outflow tract cross-sectional area and measured heart rate are used with the Doppler flow profiles to calculate cardiac output. The noninvasive nature of this device, along with ease of use and minimal training requirement, has sparked recent interest in its potential use in emergency departments, operating rooms, and intensive care units. Concordance between the cardiac output measurements determined by the USCOM device and those determined by the thermodilution technique using a pulmonary artery catheter has been evaluated in previous studies in postoperative cardiac surgical patients4–6 and intensive care patients.7,8 Like all ultrasonic devices, results obtained are operator dependent.9–11 The proper alignment of the ultrasound beam to the aortic or pulmonary outflow tracts is important for acquiring accurate results and often much time is spent on probe and/or patient positioning to obtain optimal and reliable results. There is, however, no formal recommendation by the developers with regards to optimal patient positioning to facilitate cardiac output measurements, and hence the impact of positioning on measurement is unclear.

This study aimed to determine the effect of patient positioning on the acquisition of cardiac output measurement with USCOM1A. In particular, the quality of the signal and the time required to obtain an "acceptable" signal were evaluated.

METHODS

In this prospective, observational, crossover study, 30 healthy adult volunteers were recruited for cardiac output measurements using USCOM1A after ethics committee approval. Written consent was obtained from all recruited volunteers. The exclusion criteria were previous cardiac surgery and known cardiac disease.

Continuous cardiac output measurements, using the aortic approach, were performed for 5 min in each of five positions: sitting, supine, Trendelenburg (20 degrees), left lateral tilt (20 degrees), and right lateral tilt (20 degrees). Measurements were performed on an operating table to facilitate positioning and the order of positions in which the measurements were performed was randomized according to a computer-generated schedule. Randomization was used to minimize bias from progressive ease of measurements on the same subject due to practice.

Each of the two researchers with comparable training and experience in usage of USCOM1A performed measurements on 15 subjects. Both researchers had received two 90-min training sessions with the developers' representative and had no other experience with the device, before undertaking the study.

With each position, the objective was to attain a Doppler flow profile of optimal quality during the 5-min period. Each 5-min period commenced immediately upon placement of the transducer on the suprasternal notch. The quality of Doppler flow profiles was judged based on 10 features listed in Table 1. These features were selected by the researchers based on the Freemantle criteria (Table 2) and characteristics described by the manufacturer.12 The Freemantle criteria for Doppler flow profile quality assessment is a six-point scoring system developed by Dey and Sprivulis as a way of assessing the quality of the Doppler flow profiles based on Australasian College of Emergency Medicine published guidelines for ultrasound training and technical information obtained from the manufacturer.13

All measurements were video recorded with a portable digital video recorder mounted directly in front of the USCOM1A device. Recorded measurements were then replayed and analyzed by both researchers. Both researchers separately analyzed the measurements according to the set protocol described below.

Analysis and Scoring of Video-Recorded Doppler Flow Profiles
Video recording of all measurements was reviewed and analyzed. For each measurement epoch of 5 min duration (150 epochs in total), the data were collected as follows:

1. The video recording was paused when the first "acceptable " Doppler flow profile was obtained and the time elapsed recorded. A Profile Quality Score was then assigned based on 10 profile features listed in Table 1. A score of more than 5 denoted an "acceptable" profile. Each feature was assigned 0 points if absent or 1 point if present. A feature was judged present if it was present on at least three consecutive Doppler flow profiles. Two further discretionary points judged on the "overall impression" of the profile quality (determined by the researchers) were either added or deducted from the feature-based score to constitute a final Profile Quality Score. A minimum score of 0 and maximum score of 10 could be assigned. A Freemantle score (out of 6) was also assigned for comparison.
   
2. For all subsequent Doppler flow profiles of improved quality (ones that were as good or better than the first), the video recording was paused and the formal scoring process repeated. The time elapsed from the beginning of measurement was also recorded.
   
3. Either or both researchers could initiate video pause and profile analysis provided the profile in question will be assigned a higher Profile Quality Score compared with the previous profile analyzed. The subjective nature of the profile analysis meant that both researchers differed in their Profile Quality Scores.
   
4. Where no acceptable profile was obtained, the video was stopped at 60, 180, 300 s, and a Profile Quality Score was assigned to the best Doppler flow profile achieved during those respective time periods.
   

Table 3 shows an example of data collection and scoring for one 5-min measurement period. In this 5-min period, the video recording was paused three times. The first pause occurred at 59 s after commencement of measurement when the first "acceptable" profile was attained. This was then assigned a Profile Quality Score of 6 points (7 points from feature-based score minus one discretionary "overall impression" point) according to our scoring criteria. The best quality profile during this 5-min period was attained at 177 s scoring 8 points. All profiles after 177 s until the end of the 300-s (5 min) measurement period did not score above 8 points. Freemantle Scores were also assigned for comparison.

Statistics
SPPS (version 14) was used in all statistical analyses and the Student's t-test, ² and one-way ANOVA tests were used as appropriate when comparing differences among positions depending on the nature of the data.

Bonferroni correction was applied in all one-way ANOVA tests. P value of <0.05 denoted statistical significance.

RESULTS

One hundred and fifty measurement epochs of 5 min duration each were obtained, recorded on video, and analyzed. The demographics of subjects with regards to age, height, weight, and body mass index are listed in Table 4.

Failed measurement was defined as the inability to obtain an acceptable Doppler flow profile within 5 min. This occurred in 1–3 subjects (3%–10%) for each position for each researcher in all positions, except the sitting position. Acceptable profiles could not be obtained in 13 subjects (43%) in the sitting position (Table 5), a significantly higher number than in the other four positions (P < 0.01). This was identified by both researchers. There was no statistically significant difference in the failure rate between both researchers in all positions.

Both researchers identified similar subjects in which measurements failed. In the Trendelenburg, left lateral, and sitting positions, the same subjects were identified by both researchers to have had failed measurements. In the supine and left lateral positions, one subject (subject 8) was identified to have had failed measurements by one researcher whereas the other researcher judged the same measurements to be acceptable (Table 6).

Our results also showed Profile Quality Scores correlated well with Freemantle Scores (Pearson correlation of 0.84) across 824 pairs of scores.

There was no statistically significant difference in the times required to obtain the first "acceptable" (Profile Quality Score >5) Doppler flow profile among all positions, except in the sitting position. In the sitting position, the mean time was longer at 66 s, and was consistent between researchers (Table 7).

Both the Profile Quality Scores and Fremantle Scores for the first acceptable profiles were similar among all positions (mean 7.0 (±2.0) in all positions, P = NS). This was also consistent between researchers (P = NS).

There was no statistically significant difference demonstrated among the five positions in the times required to obtain the optimal profile (with best possible Profile Quality Score) within the 5-min period (Table 8). However, in the sitting position, the optimal profile obtained was of a lower quality based on both Profile Quality Score and Freemantle Scores (Tables 9 and 10).

The times required to obtain both acceptable and optimal profiles within 5 min and the quality of these profiles were similar across the four positions: supine, Trendelenburg, left, and right lateral tilt. This suggested that patient repositioning among these positions to facilitate measurements would be unhelpful and hence unnecessary.

The sitting position, however, was the least ideal position in which to obtain a useful signal with longer measurement times, lower quality optimal profiles, and significantly higher measurement failure rate. These findings may be explained by the cranial and posterior displacement of the heart in the supine positions,14 which may facilitate the necessary alignment between the transducer (placed in the suprasternal notch) and the ascending aorta/aortic valve. This finding may have more clinical significance in awake patients with acute cardiac failure and pulmonary edema in which sitting is the position of choice. However, in intubated and ventilated patients with unstable hemodynamics, this is perhaps of lesser relevance as sitting would unlikely be the patient position. As such, the semirecumbent position (i.e., 30 degrees), which was not included in this study, may be worth future investigation. The semirecumbent position is a more clinically feasible position to perform measurements in critically unwell patients than the sitting position. When concurrent increased intracranial pressure or respiratory compromise is present, it may be the preferred position.

In this study, healthy subjects with no known cardiac disease or previous cardiac surgery were recruited. Given the researchers' limited experience with USCOM1A and the paucity of published evidence in methods of profile assessment, a homogenous group was needed to focus our investigation on the effect of positioning on Doppler flow profile acquisition and to minimize potential factors, which may complicate a relatively subjective method of profile quality assessment. For example, the relatively higher maximum velocity in aortic stenosis may inadvertently improve the Profile Quality Score independent of effects from positioning. Surgical adhesions postcardiac surgery may affect or limit intrathoracic movement of the heart; this may mask potential effects of positioning in the general population without adhesions. As a result of the exclusions, our subject group may not be reflective of the entire patient population, which may benefit from continuous noninvasive cardiac output measurements in the clinical setting. The inclusion of these patient groups in larger future studies would be appropriate to confirm the generalized applicability of our results.

Five-minute measurement was chosen, because the researchers deemed USCOM1A use clinically impractical if more than 5 min was required for each acceptable measurement. This has particular relevance in anesthesia where hemodynamic monitoring should provide information with minimal delay and distraction. In the emergency department or intensive care unit setting, a longer time period of time may be acceptable.

The subjective nature of profile analysis is a problem with a study of this nature. More importantly, in practice, the accuracy of the cardiac output measurement depends upon the "quality" of the Doppler flow profile obtained. There is currently no "gold standard" or objective measure of an optimal Doppler flow profile obtained from this device.

Another potential confounder is the physiological effect of positioning to the actual cardiac output. The Trendelenburg position increases venous return and cardiac output.15–18 An increase in cardiac output may result in an increase in the maximum velocity recorded (across the aortic outflow tract). Maximum velocity is part of our scoring criteria and would affect the Profile Quality Score obtained. For example, if the maximum velocity is increased to over 0.8 m/s (Table 1) because of an increase in cardiac output, one extra point would be gained as a result of increased cardiac output rather than actual improved quality of the Doppler flow profile. However, we believe that the impact on our results was minimized by our scoring system, which consisted of 10 variables and two discretionary "overall impression" points.

In this study, we investigated the variability in signal acquisition related to patient positioning and did not address the relationship between good signal acquisition on USCOM1A and accurate cardiac output measurement as measured by gold standard techniques, e.g., thermodilution using a pulmonary artery catheter. This has been demonstrated in several previous studies.4–8 Of note, however, these studies did not specify the criteria used to determine the adequacy of Doppler flow profile quality; it was assumed that they followed the manufacturer's guidelines to attain adequate profile quality, which was also the basis of our scoring criteria. Furthermore, we cannot conclude from this study what threshold for the signal quality will consistently provide a reliable cardiac output measurement.

Our results did not identify an ideal patient position to facilitate cardiac output measurements using USCOM1A. The times required to obtain both acceptable and optimal profiles within 5 min are similar for four positions studied: supine, Trendelenburg, left, and right lateral tilt. However, our results suggested that the sitting position is the least suitable and least reliable position in which to perform measurements.

ACKNOWLEDGMENTS

The authors thank Ms Bev Jacobson (Marketing Executive, USCOM Pty. Ltd., Sydney, NSW, 2000, Australia) for providing technical support and training.

Footnotes

Accepted for publication February 22, 2008.

REFERENCES

1. Phillips RA, West MJ, Burstow DJ. Trancutaneous continuous wave Doppler monitoring is feasible producing reliable and reproducible signals. J Am Coll Cardiol 2002;39:283
2. Kusumoto F. Measurement of aortic blood flow by Doppler echocardiography: temporal, technician, and reader variability in normal subjects and the application of generalizability theory in clinical research. J Am Soc Echocardiogr 1995;8(5 Pt1):647–53[Medline]
3. Nidorf SM, Picard MH, Triulzi MO, Thomas JD, Newell J, King ME, Weyman AE. New perspectives in the assessment of cardiac chamber dimensions during development and adulthood. J Am Coll Cardiol 1992;19:983–8[Abstract]
4. Tan HL, Pinder M, Parsons R, Roberts B, Ven Heerden PV. Clinical evaluation of USCOM ultrasonic cardiac output monitor in cardiac surgical patients in intensive care unit. Br J Anaesth 2005;94:287–91[Abstract/Free Full Text]
5. Knobloch K, Lichtenberg A, Winterhalter M, Rossner D, Pichlmaier M, Phillips R. Non-invasive cardiac output determination by two-dimensional independent Doppler during and after cardiac surgery. Ann Thorac Surg 2005;80:1479–83[Abstract/Free Full Text]
6. Chand R, Mehta Y, Trehan N. Cardiac output estimation with a new Doppler device after off-pump coronary artery bypass surgery. J Cardiothorac Vasc Anesth 2006;20:315–19[Web of Science][Medline]
7. O'Driscoll G, Wright JJ, Wright IW, Green DJ, Phillips RA. Cardiac output measurement with USCOM Ultrasonic Cardiac Output Monitor. Abstract Presented at the 2nd Asian Pacific Congress of Heart Failure (APCHF), Singapore, January 12, 2005
8. Haas LEM, Tjan DHT, van Wees J, van Zanten ARH. Validation of the USCOM1A monitor in the hemodynamic unstable intensive care patients. Abstract Presented at the Annual Intensive Care Society Congress, the Netherlands, February 2006
9. Ehler D, Carney DK, Dempsey AL, Rigling R, Kraft C, Witt SA, Kimball TR, Sisk EJ, Geisser EA, Gresser CD, Waggoner A. Guidelines for Cardiac Sonographer Education: Recommendations of the American Society of Echocardiography Sonographer Training and Education Committee. J Am Soc Echocardiogr 2001;14:77–84[Web of Science][Medline]
10. Quinones MA, Otto CM, Stoddard M, Waggoner A, Zoghbi WA. Recommendations for Quantification of Doppler Echocardiography: A Report from the Doppler Quantification Task Force of the Nomenclature and Standards Committee of the American Society of Echocardiography. J Am Soc Echocardiogr 2002;15:167–84[Web of Science][Medline]
11. Tiit K. Non-invasive cardiac output determination: state of the art (Review article). Curr Opin Anaesthesiol 1999;12:9–13[Medline]
12. USCOM Ltd—User Reference Guide. 2005;6
13. Dey I, Sprivulis P. Emergency physicians can reliably assess emergency department patient cardiac output using USCOM continuous wave Doppler cardiac output monitor. Emerg Med Australas 2005;17:193–99[Medline]
14. O'Rahilly R, Muller F, Carpenter S, Swenson R. Basic human anatomy—a regional study of human structure. Chapter 23: The pericardium and the heart. Philadelphia: WB Saunders, 1983
15. Johnson S, Henderson SO. Myth: the Trendelenburg position improves circulation in cases of shock. Can J Emerg Med 2004;6:48–9
16. Terai C, Anada H, Matsushima S, Kawakami M, Okada Y. Effects of Trendelenburg versus passive leg raising: autotransfusion in humans. Intensive Care Med 1996;22:613–4[Web of Science][Medline]
17. Sibbald WJ, Paterson NA, Holliday RL, Baskerville J. The Trendelenburg position: haemodynamic effects in hypotensive and normotensive patients. Crit Care Med 1979;7:218–24[Web of Science][Medline]
18. Reich DL, Konstadt SN, Hubbard M, Thys DM. Do Trendelenburg passive leg raising improve cardiac performance? Anesth Analg 1988;67:S184

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Siu, L. Articles by Monagle, J. Search for Related Content PubMed PubMed Citation Articles by Siu, L. Articles by Monagle, J.