If you wish to respond to a paper or other item already published in the BJA, please go to the abstract/full text version of that item and click on the link "E-Letters: Submit a response to the article".
Electronic Letters to:
|
|
E-letters: Electronic letters published:
-
![[Read E-letter]](/icons/misc/sectionDOWN.gif)
Acceptable limits of agreement need to be defined a priori
- Anna Lee, Lester A Critchley (22 March 2005)
|
|
|||
|
Anna Lee, Associate Professor The Chinese University of Hong Kong, Lester A Critchley
Send letter to journal:
|
We would like to comment on the methods and conclusions reached by Tan and colleagues in a recent paper title “Clinical evaluation of the USCOM ultrasonic cardiac output monitor in cardiac surgical patients in intensive care unit” [1]. The USCOM is a newly developed Continuous wave Doppler Ultrasound device that can measure cardiac output either via the precordium or sternal notch. Although, the use of Doppler ultrasound to measure cardiac output is nothing new, some of the features of the USCOM make it technically better than previous devices. Therefore, the USCOM could potentially solve the long standing problem in clinical practice of not having a reliable and easy to use method of measuring cardiac output. Thus, one can expect to see a number of publications over the next couple of years validating this device. Unfortunately, validation studies in the area of cardiac output monitoring have a long history of inadequate statistical analysis and conclusions, culminating in the introduction of Bland and Altman’s method in 1986 [2]. The original Bland and Altman method [2] that the authors have used is incorrect as: (1) the systematic variation in the difference in cardiac output measurements across the range of cardiac output has not been taken into consideration. Therefore, the assumption of uniform variance has not been met for the original Bland and Altman method; and (2) the analysis in Figure 4 is based on more than one measurement per patient. The standard deviation of the difference is likely to be underestimated as some of the effect of measurement error has been removed [3]. Therefore, the quoted mean bias and limits of agreement may be incorrect. A re-analysis of the data using the newer methods of agreement [3] taking into account of these issues would be helpful before recommending the USCOM monitor in the intensive care unit. However, far more important than their use of Bland and Altman is that the authors fail to define a priori what limits of agreement were acceptable for clinical practice. They conclude that the USCOM is accurate but this conclusion is based on no objective test against well defined criteria. All they provide are limits of agreement that range -1.43 to 1.78 L/min, which is meaningless. What are acceptable limits of agreement when comparisons are made with an imprecise reference method such as thermodilation is a question that many researchers in this field have found difficult. However, it is a question that needs to be addressed if future publications on validating the USCOM are going to be of true scientific worth. Editors should make sure that authors in the future address the issue. Based on a meta-analysis, we believe that the limits of agreement for the interchangeability of thermodilution with a new technique, such as the USCOM, should be approximately 14% to 28% (1.96 SD/mean) [4]. As the mean cardiac output from the USCOM was not provided in Tan et al.’s paper [1], we are uncertain if their results lie within the above range. Nevertheless, our data collected with a dog model indicate that the USCOM provides reliable cardiac output measurements over a wide range of cardiac output [5], providing further evidence of the potential role of this device in clinical monitoring. References 1. Tan HL, Pinder M, Parsons R, Roberts B, van 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 2. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307-10 3. Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res 1999; 8: 135-160 4. Critchley LA, Critchley JA. A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. J Clin Monit Comput 1999; 15: 85-91 5. Critchley LA, Peng ZY, Fok BS, Lee A, Phillips RA. Testing the reliability of a new ultrasonic cardiac output monitor, the USCOM, by using aortic flowprobes in anesthetized dogs. Anesth Analg 2005; 100: 748- 53 Conflict of Interest:None declared |
|||