Thorax: first published as 10.1136/thx.28.5.584 on 1 September 1973. Downloaded from Thorax (1973), 28, 584. Errors in the measurement of vital capacity A comparison of three methods in normal subjects and in patients with pulmonary emphysema D. C. S. HUTCHISON, C. E. BARTER', and N. A. MARTELLI2 Chest Unit, King's College Hospital, London SE5 Hutchison, D. C. S., Barter, C. E., and Martelli, N. A. (1973). Thorax, 28, 584-587. Errors in the measurement of vital capacity: a comparison of three methods in normal subjects and in patients with pulmonary emphysema. Three methods of measuring the vital capacity have been compared in six normal subjects and in six with pulmonary emphysema, according to a randomized design. The methods were (a) the inspiratory vital capacity (IVC), (b) the expiratory vital capacity (EVC), and (c) the forced vital capacity (FVC). In normal subjects, there was a small but significant difference between the methods. The residual standard deviation derived from analysis of variance was 94 ml (coefficient of variation 1.7 %). A slight but significant rise in vital capacity with repeated effort was observed. In emphysematous subjects, there was no significant difference between the IVC and EVC methods. The FVC gave values which were, on average, approximately 0.5 litre less copyright. than those obtained by the other methods. The standard deviation in all three methods was substantially greater than for the normal subjects. The FVC is not a suitable method for the measurement of vital capacity in patients with pulmonary emphysema. The EVC is satisfactory, provided it is used with caution, but in practice the IVC is the preferred method. http://thorax.bmj.com/ Hutchinson (1846) defined the vital capacity as the object of the study was to compare the three volume of gas displaced from the lungs by the methods in normal subjects and in patients with 'greatest voluntary expiration following the deepest pulmonary emphysema. A randomized design was inspiration', and although this is probably the used, provision being made for the detection of method most commonly used, the procedure is not time related effects, such as fatigue or training. infrequently carried out in the reverse order. There seems no obvious reason why the two methods SUBJECTS AND METHODS should yield differing results, provided that the SUBJECTS The six normal subjects were healthy male manoeuvres are performed without undue speed. members of the hospital staff, their ages ranging from on September 30, 2021 by guest. Protected It has become common practice, however, to make 25 to 35 years. All were apparently free from pul- the measurement at the same time as the forced monary disease, with forced expiratory volume (FEVy) expiratory volume, though this procedure may and vital capacity within the expected normal range. underestimate the true vital capacity in subjects Normal values were obtained from Cotes (1968). with severe airflow obstruction (Gilson and Hugh- The six emphysematous subjects (El to E6) all Jones, suffered from severe dyspnoea on exertion; in all six, 1949). attenuation of the peripheral pulmonary vessels In the present study, three variations on the (Laws and Heard, 1962) was observed on the chest measurement of vital capacity have been assessed: radiograph, and the FEV1 and the single-breath trans- 1. the expiratory vital capacity, here abbreviated fer factor for carbon monoxide (TF) were consider- to EVC; 2. the inspiratory vital capacity or IVC; ably less than the expected values (Table I). In patients and 3. the forced vital capacity or FVC. The El, E3, E4, and E6, emphysema was associated with at-antitrypsin deficiency; these correspond to patients 1 Supported by a grant from the Medical Research Council. Present 4, 5, 7, and 6 respectively of a previous report address: Repatriation General Hospital, Heidelberg West, 3077 Vic.toria, Australia (Hutchison et al., 1971). The nature and purpose of 2 Supported by a grant from the University of Buenos Aires. the procedure were explained to each patient. Present address: Centro de Rehabilitaci6n Respiratoria 'Maria Ferrer', E. Finochietto 849, Buenos Aires, Argentina LABORATORY METHODS The spirometer conformed to 584 Thorax: first published as 10.1136/thx.28.5.584 on 1 September 1973. Downloaded from Errors in the measurement of vital capacity 585 TABLE I only and so performed each type of test three times. MEASUREMENTS OF PULMONARY FUNCTION IN The normal subjects each completed three sessions, PATIENTS WITH EMPHYSEMA separated by one to two weeks, and so performed each of test nine times in all. FEV, VC" TF type Patient Age (5%' (5% (0% predicted) predicted) predicted) RESULTS El 46 25 106 29 E2 57 24 75 25 NORMAL SUBJECTS The mean values for vital E3 50 21 101 45 E4 39 14 65 12 capacity obtained by the three methods were very E5 45 16 85 48 similar (Table U), and the standard deviations E6 52 21 97 21 (within subjects) were small. Analysis of variance 5VC=vital capacity: mean of all values for EVC and IVC showed that there was a small but significant Predicted values from Cotes (1968) difference between the methods (P<005), and that there was a significant linear regression of volume the design of Bernstein, D'Silva, and Mendel (1952). on time (P<O0O1), the values being greater towards It was calibrated by displacement of air in steps of the end of each session. There was no consistent 1 litre from an accurately machined syringe and its rank order with respect to method, though the response was linear over the range used. The record IVC was not the greatest in any subject. was inscribed in ink on squared paper fixed to a rotating drum, adjusted so that the initial horizontal TABLE II and vertical reference lines were parallel to the grid- VITAL CAPACITY (LITRES BTPS) MEASURED BY THREE METHODS IN NORMAL AND EMPHYSEMATOUS SUBJECTS lines on the paper. The paper was 2 speed cm/sec No. of during the performance of the FVC, and 02 cm/sec Sub- IVC EVC FVC for the IVC and EVC. The deflections were measured jects Mean SD Mean SD Mean SD to the nearest millimetre and converted to litres BTPS. Normal 6 5 59 0-081 5-61 0116 563 0,087 was a Emphysema 6 3 97 0-189 3-91 0-217 3-48 0 340 One millimetre equivalent to volume of 44 ml. copyright. The tests were performed with the seated. subject IVC, EVC, and FVC For definition see text. The mouthpiece was inserted during tidal breathing BTPS =body temperature and ambient pressure, saturated and the manoeuvre was closely observed to ensure SD=within subjects standard deviation that there was no leakage round the mouthpiece or EMPHYSEMATOUS SUBJECTS The means and (within noseclip. During each test the subjects were encouraged subjects) standard deviations for the three methods to give a maximal response. are shown in Table II. http://thorax.bmj.com/ The EVC was measured by a slow expiration after The standard deviations for the deepest possible inspiration. The IVC was all three methods were significantly greater than in measured by a slow inspiration after the deepest pos- the normal subjects; this effect was apparently sible expiration. The FVC was measured by expira- greater for the FVC, though the range of values tion with maximal force after the deepest possible was particularly large in one patient (E5). inspiration. The time interval between each test was Analysis of variance showed that there was a two minutes in the normal subjects and three minutes highly significant difference between the methods in the emphysematous subjects. (P<OO1); there was no significant regression of FEV1 was measured from the same spirogram as volume on time. There was a significant patient- the FVC, the zero point for time and volume being method interaction (P<005) showing that the on September 30, 2021 by guest. Protected obtained by extrapolation of the steepest part of the differences between the methods were downward slope of the spirogram to a horizontal line not consistent through the maximum inspiratory plateau (American from patient to patient, an effect largely due to College of Chest Physicians, 1963). The transfer factor patient E3. The results in individual patients were for carbon monoxide was measured by the single- therefore analysed in greater detail. breath method of Ogilvie, Forster, Blakemore, and The values for the three methods in each patient Morton (1957), except that the breath-holding time are shown diagrammatically in the Figure. The was calculated as suggested by Jones and Meade rank order was more consistent than in the normal (1961) and the alveolar volume was obtained from subjects, the IVC being the greatest (or equal the simultaneous dilution of helium. greatest) and the FVC the smallest, in five subjects. The rank order was reversed in subject E3, though PLAN OF STUDY The study was carried out in in this 'sessions', each session consisting of nine 'tests', each particular case there was no significant test being one performance of either the EVC, the difference between the three methods. IVC or the FVC. The three types of test were each In view of the apparent similarity between the performed three times during one session, the order IVC and the EVC, these two methods were com- being determined on a Latin square basis. pared by the t test in each emphysematous indivi- The emphysematous patients completed one session dual, no significant difference at the 5 % level being 2U Thorax: first published as 10.1136/thx.28.5.584 on 1 September 1973. Downloaded from 586 D. C. S. Hutchison, C. E. Barter, and N. A. Martelli regression equations for the prediction of normal values.