Oct 20, 1973 - or withdrawn from the study on ethicalgrounds. Administration of Oxygen. .... air until the Pao, rose above the steady-state value predicted in each case. .... clinical trial of treatment will be partly determined by the proportion of ...
154 Cathcart, E. S., Ritchie, R. F., Cohen, A. S., and Brandt, K., American Journal of Medicine, 1972, 52, 93. 4Glenner, G. G., Ein, D., and Terry, W. D., American_Journal of Medicine, 1972, 52, 141. 5 Magnus-Levy, A., Zeitschrift fur klinische Medizin, 1931, 116, 501.
3
BRITISH MEDICAL JOURNAL
20 OCTOBER 1973
6 Jones, N. F., Hilton, P. J., Tighe, J. R., and Hobbs, J. R., Lancet, 1972, 2, 616. 7British Medical,Journal, 1973 3, 120. 8 Limas, C., Wright, J. R., Matsuzaki, M., and Calkins, E., American 7ournal of Medicine, 1973, 54, 166.
Occasional Survey Oxygen Therapy in Chronic Respiratory Failure R. HUME, P. J. ROONEY, D. L. McLELLAN
British Medical Journal, 1973, 4, 154-156
Summary In hypercapnic chronic respiratory failure red cell volume correlates with the arterial oxygen tension during the quiescent phase of the disease. During an acute exacerbation measurement of the red cell volume can give a reasonable index of the arterial oxygen tension existing in the previous quiescent state. This level of arterial oxygen tension appears to determine the tolerance of the individual to treatment with oxygen.
Introduction It is now well recognized that oxygen therapy can be dangerous in patients with chronic respiratory failure characterized by hypoxaemia and hypercapnial unless monitored closely by frequent estimation of the arterial blood gases and pH. In these patients the hypoxaemia provides the major stimulus to ventilation, and in acute exacerbations of the disease treatment with oxygen can, by abolishing the hypoxaemia, diminish the stimulus, resulting in hypoventilation and respiratory acidosis. Thus treatment with oxygen in these cases aims to raise the arterial oxygen tension (Pao,) as high as possible without inducing respiratory depression. Various arbitrary rules have been devised to help in this-for example, it has been recommended that the arterial carbon dioxide tension (Paco,) should not be allowed to exceed 90 mm Hg2 nor the pH allowed to fall below 7 25;3 at the same time the recommended safe low limit of Pao2 has been set as low as 30 mm Hg2 and as high as 50 mm Hg.3 To achieve these levels workers have varied in their recommendations regarding the optimal concentrations of oxygen in the inspired air for such patients. The current recommendation of 24%3 while only occasionally producing a dangerous degree of respiratory depression will not, in every case, provide the highest tolerance level of blood oxygen. To date there has been
no adequate explanation of this individual variation in the tolerance to oxygen in the acute stages of the disease. It has been postulated4 that the factor which determines a patient's tolerance to oxygen therapy during an acute exacerbation of chronic respiratory failure is the level of arterial oxygen to which his respiratory centre has become accustomed during the quiescent phase of the disease-that is, the "central oxygenstate" has been reset at a new low level. It has been claimed that this "steady-state" Pao, can be predicted during an acute phase of the disease by determining the red cell volume, which is directly proportional to the degree of chronic hypoxaemia.4 This study sets out to test whether during an acute exacerbation of the respiratory failure the steady-state Pao, can be predicted from measurement of the red cell volume and whether, by not exceeding this steady-state Pao, during treatment with oxygen, hypoventilation and hypercapnia will be averted. Patients and Methods
All patients studied were suffering from chronic hypercapnic respiratory failure.' Group 1. Thirty patients were studied during the quiescent phase of their disease (see chart). These included 15 patients previously studied by Hume4 and six patients from group 2 who were also studied while the disaese was quiescent and when a steady-state Pao, had again been achieved.
40
-
E 50-
0 60
-
X70
-
80
90
.o
0
E
L
Southern General Hospital, Glasgow G51 4TF R. HUME, M.D., F.R.C.P., Consultant Physician T. ROWAN, PH.D., A.R.I.C., Principal Biochemist Centre for Rheumatic Diseases, Glasgow G4 OEH P. J. ROONEY, M.B., M.R.C.P., Senior Medical Registrar Department of Neurology, London Hospital, London El 2AD D. L. McLELLAN, M.B., M.R.C.P., Senior Registrar
n=30 r=0-88 P< 0.001
30-
~~
~
%
0
S
,
.
80 100 120 140 160 R.C.V. (per cent. of normal)
180
2Q0
Arterial oxygen tensions and red cell volumes (R.C.V.) in 30 patients (group 1) during quiescent phase of their disease.
20 OCTOBER 1973
BRITISH MEDICAL JOURNAL
155
Group 2. Ten patients were studied during an acute exacerbation of their respiratory failure (see table II). Measurement was made of their red cell volume and their steady-state Pao, as predicted from the relation between red cell volume and Pao, established in group 1. Increasing concentrations of inspired oxygen were then administered, and two hours after each increment measurement was made of the arterial blood gases and pH until the measured Pao, exceeded the value predicted as steadystate for that patient. Patients with very acid pH were excluded or withdrawn from the study on ethical grounds. Administration of Oxygen.-The concentration of oxygen in the inspired air was varied by the use of a Venturi mask (British Oxygen Company), concentrations of approximately 24°o and 280' being delivered. In addition, in one patient in whom a higher Pao, was required an M.C. mask was used. This device normally delivers about 60", oxygen. Blood Gas Analysis.-Pao, was measured by the radiometer Po, electrode. Paco2 and pH were measured by the interpolation technique, using the micro-Astrup apparatus. Blood Volume Measurements.-The plasma volume was measured by the radio iodinated human serum albumin (131 I HSA) technique. The red cell volume was derived from the plasma volume after correcting the packed cell volume by a factor of 091.5 The predicted normal red cell volume was calculated by the method of Hume and Goldberg,6 which is based on the prediction of the total blood volume from height and weight.7 All measured red cell volumes were expressed as percentages of the predicted "normal." The packed cell volume was measured in the micro haematocrit (Hawksley). Haematological data are summarized in table I. TABLE i-Haeniatological Data in Group 2 Case No. 1 2 3
4 5 6 7 8 9 10
Measured Plasma Volume
(ml) 1,730 2,157 1,587 1,694 2,652 2,687 3,155
1,090 2,194 1,958
Packed Cell Volume (,,)
Measured
60 57 50 54 57 61 48 52 59 58
2,880 1,992 1,285 1,563 2,775 3,284 2,695 2,276 2,523 2,187
Red Cell Volume
(ml)
Predicted Steady-state
Predicted Red Cell Volume (ml)
,, Rise in Red Cell Volume
Po., (mm Hg)
1,957 1,204 1,257 1,410 1,696 1,917 1,859 1,794 1,683
44 65 8 10 58 71 45 28 50 52
58 48 72 70 52 46 56 63 54 54
1,404
TABLE iI-Blood Gas Data in Group 2 during Acute Exacerbation Case No.
Blood Gases at Various Oxygen Concentrations in Inspired Air
Steady-state Po2 Measured Predicted
Air f
58
1
2
52
48
3
70
72
L
4 5
56
Po. pH Pco., Po., pH Pco2 Po.,
PHO Pco., Po,
70
pH
52
Pco, P0.. pH
Pco2, 46
6
r
Po.
pH
Pco.
Po., 7
59
56
8
65
63
8 Po.6
g
l
54
9 10
pH Pco.,
55
.
54
Po ., and Pco, are cxpressed in
g
mm
pH
Pco2 Po., pH Pco., Po., pH
Pco2
32 7-34 79 43 7-36 68 66 7-37 55 58 7-45 57 42
7.38 61
24', 41 7-31 81 46 7-34 72
71 7-38 53
54 7.33 75
282" 62 7-25
115I
58 7-28 92 83 7-26 80 69 230 7-44 7-24 56 85 55 7 823 94
36
48
52
7-39 52
7-30
7-26 80
67
43 7.37 50 58 7-38 50 47
54
61
7-34 64 62 7-39
7.24 8.,5 67 7-29 74
7-35
7-27
48
55
7141 51
50 56 70
60Appox
Approx.
67 7-29 71
Hg, and pH is expressed in units.
Results Group 1.-Analysis of the chart shows that a linear relationship existed between the Pao2 and the red cell volume (expressed as a percentage of normal). The linear regression is given by y=112 0-0 3752x. The correlation coefficient (r) between Paoa and red cell volume was 0-8792 (P
