by echocardiography - NCBI

2 downloads 0 Views 1MB Size Report
Mar 3, 1980 - Serial measurements of left ventricular dimensions by echo septum could be clearly ..... American Society of Echocardiography. Our findings ...
Br HeartJ 1980; 44: 284-9

Serial measurements of left ventricular dimensions by echocardiography Assessment of week-to-week, inter- and intraobserver variability in normal subjects and patients with valvular heart disease G

A LADIPO,* F G DUNN, T H PRINGLE, B BASTIAN, T D V LAWRIE From the University Department of Medical Cardiology, Royal Infirmary, Glasgow 0

suMMARY The week-to-week, inter- and intraobserver variation in left ventricular echocardiographic measurements has been studied in 10 normal male volunteers and in five patients with stable valvular disease. A two-way analysis of variance showed no statistically significant variation either from week to week or between observers. Furthermore the within observer variation was minimal. Calculation of the coefficient of variation allowed confidence limits to be applied to each of the six ventricular measurements, thus providing ranges

of variation in follow-up studies using M-mode echocardiography.

The development of M-mode echocardiography has allowed much useful information to be obtained about left ventricular structure and function both in health and disease. One application of echocardiography in this regard is in sequential studies to assess the progression or regression of a disease process or the effect of pharmacological intervention. It is therefore important to establish the degree of variation among echocardiographers in their analysis of data. Recently, clear guidelines were suggested for echocardiographic measurements,' and, using these guidelines, we have looked at the variations in left ventricular dimensions from week to week for four weeks in 10 normal volunteers and five patients with stable valvular heart disease. We have also assessed the inter- and intraobserver error of three physicians in measuring these echocardiograms.

Subjects and methods Fourteen normal volunteers and six patients with stable valvular heart disease (on no medication and not showing evidence of cardiac decompensation) Previously Inter-University Council Senior Academic Present address: Department of Medicine, Faculty of Sciences, University of IFE, ILE-IFE, Nigeria. Received for publication 3 March 1980

*

Fellow. Health

284

were selected for study. Of these, four were excluded from the normal group and one from the group with valvular heart disease at baseline because of technically inadequate tracings. The 10 normal volunteers were all healthy men aged 20 to 35 years (mean 26&8) with no history of cardiac disease. Clinical details of the group with valvular heart disease are described in Table 1. The echocardiograms were performed with a Smith Kline Ekoline 20A ultrasonoscope and attached strip chart recorder at a paper speed of 50mm/s in all subjects after 15 minutes' rest. A lead II electrocardiogram was simultaneously recorded. The echocardiographic technique was as described by Henry et al.2 In brief, after a T-scan, a good septal-posterior wall transverse dimension was defined and gain settings were adjusted so that endocardium, epicardium, and both sides of the Table 1 Details offive subjects with stable cardiac disease Case no. Age (y) 11 12 13 14 15

46 56

65 59 50

Sex

Clinical diagnosis

F M M F M

Mitral valve prolapse Aortic stenosis Aortic stenosis and regurgitation Aortic regurgitation Aortic stenosis

Serial measurements of left ventricular dimensions by echo

285

septum could be clearly seen at the level of the Table 2 Guidelines for echocardiographic left chordae tendineae of the mitral valve apparatus. In ventricular measurements this position, all tracings were performed by one Left ventricular diastolic internal diameter (LVIDd) of us (GL). At the initial study, the position of the At level of chordae, measure vertical distance from most anterior septum at onset of first transducer was carefully noted with reference to the edge of endocardial echoes to left interspace and distance from the sternum. The deflection of QRS complex of ECG echocardiograms were performed at weekly intervals Left ventricular systolic internal diameter (LVIDs) vertical distance from most anterior edge of endocardial for four weeks using the same posture and trans- Measure echoes to left septum at nadir of septal motion (if septal motion is ducer position in each subject as were used in the normal), if not, at peak of posterior wall motion first recording. The tracings for measurement were Posterior left ventricular wall thickness (PLVWT) chosen and photocopied by one of us (GL) and At onset of first deflection of QRS complex of ECG, measure distance between most anterior edge of endocardial echoes read independently and in a completely blind man- vertical ner by the other three of us (called observers A, B, to epicardium at level of chordae and C). To minimise observer bias in measurement, Posterior left ventricular wall amplitude (PLVWA) Draw horizontal line between most anterior points of left only five cardiac cycles were included in each ventricular endocardium in systole, then measure maximal vertical tracing. Each observer was provided with a set of distance from this line to endocardium of LVPW at point just guidelines for left ventricular measurements (as before free wall begins to move anteriorly in systole shown in Table 2) based on the recommendations Interventricular septal thickness (IVST) Measure vertical distance from right ventricular side of IVS to its of Sahn et al.' In all, 60 technically acceptable echocardiograms left ventricular side at onset of first deflection of QRS of ECG (Fig.) were performed and each observer measured Interventricular septal amplitude (IVSA) Draw horizontal line between most posterior points of left six indices in each of 70 echocardiograms (including ventricular side of septum during systole, then measure maximal 10 duplicated tracings). vertical distance from this line to IVS just before septum begins to move posteriorly in systole.

7 iI

tt1

ft§t S

W;

. *. .:

*

t'0-SCM

LVI

LV