Original Paper Respiration 1992;59:317-321
Departments of Medicine and Community and Family Medicine. University of California, San Diego. Calif.. USA
Effect of Posture on Arterial Oxygenation in Patients with Chronic Obstructive Pulmonary Disease
Key Words
Abstract
Chronic obstructive pulmonary disease Arterial blood gases Arterial oxygenation Posture Pulmonary rehabilitation
We studied 117 patients with chronic obstructive pulmonary disease (COPD) to evaluate (1) the frequency and magnitude of postural changes in resting ar terial oxygenation and (2) the relationship of these changes to other measures of pulmonary function and exercise arterial blood gases. Compared to the su pine measurement, room air P a0 2 measured while standing increased more than 3 mm Hg in 28 patients (group I ), did not change (±3 mm Hg) in 57 pa tients (group 2), and decreased more than 3 mm Hg in 32 patients (group 3) (range = 31 mm Hg increase to 20 mm Hg decrease). Patients in group I had significantly less severe disease than patients in the other two groups. There were no significant pulmonary function differences between groups 2 and 3. Supine P a0 2 was similar for all groups, suggesting that standing P a 0 2 ac counted for the postural change in P a 0 2. Because of unpredictable postural changes in P a0 2 in patients with COPD, we believe that body position should be noted for arterial blood gas measurements and should be kept constant for valid comparison of serial measurements. These findings may also be impor tant for other diffuse lung diseases.
Introduction
Measurement of arterial oxygenation is important in the evaluation and treatment of patients with chronic ob structive pulmonary disease (COPD). Typically, arterial blood is drawn with the patient in the most convenient body position (c.g.. supine or semi-recumbent in a hospital bed, seated in a laboratory, seated or standing for exercise testing). In normal subjects, there are small differences in arterial oxygenation between the supine and upright pos tures, which may depend upon the relationship between functional residual capacity (FRC) and closing volume [1, 2]. In patients with COPD, however, with nonhomogene
Research was supported bv the National Institutes of Health Grant No. KOI HI.34732 from the National Heart. Lung and Blood Institute and Grant No. RR00827 from the Division of Research Resources for the Clinical Research Center.
Received: October 15. 1991 Accepted after revision: August 3. 1992
ous ventilation and perfusion relationships, postural ef fects on P a0 2 are less predictable [3-5], Previous studies have reported minimal postural changes in P a 0 2in groups of patients with COPD, whereas in an individual patient P a02may increase or decrease significantly with change in posture [5-8], After observing significant postural changes in P a0 2 in many patients with COPD. we performed this analysis to describe systemically: (1) the frequency and magnitude of postural changes in resting arterial oxygenation and (2) the relationship of these changes to other measures of resting pulmonary function and exercise arterial blood gases.
Andrew !.. Ries, MD Division of Pulmonary and Critical Care Medicine
UCSD Medical Center 8377 225 Dickinson Street San Diego. CA 921«3-8377 (USA)
Downloaded by: Nagoya University 133.6.82.173 - 1/15/2019 12:49:09 AM
Andrew L. Ries Robert M. Kaplan Jae Chang
Fig. 1. PaO: measured supine and standing in II7 patients with COPD. Solid line is line of identity. Dashed lines indicate group placement [standing-supine PaO,: > 3 mm Hg (group 1): ± 3 mm Hg (group 2); < -3 mm Hg (group 3)].
Goldman and Becklake |!2| for lung volumes, Miller et al. [13] for DL(0, and Black and Hyatt |I4| for maximal respiratory' pressures. On a separate day, each patient completed an incremental, symp tom-limited exercise test to the maximal tolerable level on a tread mill. Prior to this test, a radial arterial catheter was inserted percutancously. A resting arterial blood sample was then drawn with the pa tient supine and on room air. The arterial blood was collected in a heparinized 3-ml syringe, immediately placed in ice. and analyzed in duplicate prior to the exercise test. Arterial oxygen saturation (SaO,) was monitored continuously by ear oximetry (Ohmeda. Boulder, Colo.. USA) and recorded simultaneously with each arterial blood sample. Patients then stood on the treadmill and breathed through a lowresistance breathing valve (Hans Rudolph. Kansas City. Mo.. USA) connected to a mixing chamber for expired gas measurements. Each patient rested until expired gas measurements were stable; at which time a second resting arterial blood sample was drawn. The patient then completed an incremental exercise protocol to a symptom-lim ited maximum. The test was terminated prior to symptom-limit for ear oximetry SAO, less than 80%. ST-T wave depression or serious arrhythmias on the ECO. or an excessively elevated blood pressure (systolic pressure greater than 250 mm Hg or decrease from baseline greater than 20 mm Hg). Arterial blood was sampled every 3 min dur ing exercise and at the maximal exercise level. Arterial blood samples were collected as previously described and analyzed in duplicate at the completion of the exercise test. All arterial blood samples were analyzed for pi I. PO-. and PaO, (1L813 and 513 blood gas analyzers. Instrumentation Laboratories. Lexington. Mass.. USA) as well as for SaO, and carboxyhemoglobin (COMb) using a four-wavelength spectrophotometer (IL282 COoximeter. Instrumentation Laboratories. Lexington, Mass.. USA). Group comparisons for physiological measurements were per formed by analysis of variance with post-hoc multiple comparison testing by Ncwman-Keuls procedures.
Methods Subjects
Pulmonary Function and Exercise Tests
Each patient performed pulmonary function tests including spi rometry. lung volumes and airway resistance (R.w ) by body plethys mography, single-breath diffusing capacity (D L ,,,). maximum volun tary ventilation, and maximum inspiratory and expiratory pressures. Spirometry, lung volumes, and RAVV tests were repeated after two puffs of inhaled mctaproterenol. Testing and quality-control proce dures followed standard and recommended methods |9. 1 0 ). Refer ence values used were those of Morris et al. [I I ], for spirometric data,
318
Ries/Kaplan/Chang
Results
For the purposes of this study, a change in PaO, from the supine to standing posture of greater than 3 mm Hg was considered to be significant. This threshold is greater than two standard deviations of repeat measurements of PC), in our laboratory. On this basis, we divided the 117 patients into three groups. Compared to the supine measurement, PaO, mea sured standing: increased in 28 patients [group 1: mean change = + 9.8 ± 6.2(SD) mm Hg]; did not change in 57 pa tients [group 2; mean change = -0.4 ±I.9(SD) mm HG|; and decreased in 32 patients [group 3; mean change = -9.2±4.1(SD) mm Hg|. Postural changes from supine to standing ranged from a 31 mm Hg increase to a 20 mm Hg decrease. Figure 1 presents the comparison of supine and stand ing resting PaO, measurements for all 117 patients. For each of the three patient groups, general character-
Effect of Posture on Arterial Oxygenation in Patients with Chronic Obstructive Pulmonary Disease
Downloaded by: Nagoya University 133.6.82.173 - 1/15/2019 12:49:09 AM
The study population comprised 119 patients with COPD entering a randomized clinical trial evaluating pulmonary rehabilitation. All patients met the following criteria: (I) diagnosis of COPD confirmed by pulmonary function evidence of obstructive lung disease and a compatible history, physical examination, and chest roentgenogram; (2) stable on an acceptable medical regimen; (3) no other significant lung disease, and (4) no unstable cardiac or other medical problem that would limit full participation in the rehabilitation program. For this analysis, data were available on 117 patients with resting arterial blood gas measurements on room air both supine and stand ing. Two of the 119 patients had a supine resting room air PaO, value less than 50 mm Hg and, subsequently, had the upright PaO, mea sured on supplemental oxygen.
Table 1. General characteristics anti pulmonary function test results for U7 COPD patients grouped according to postural change in PaO, (mean ±SD )
Age, years 1 Icight, cm Weight, kg VC, liters VC, % predicted FEV IU, liters FEV, ,/FVC, % FEF,,liters/s FEF,,_7S,;> % predicted RV/TLC. % TLC. liters TLC. % predicted DLa ) , ml/min/mm Hg DLtX), % predicted
Group 1 (n = 28)
Group 2 (n = 57)
Group 3 (n = 32)
62.8 ±7.5 168.3 ±8.4 72.5 ±14.9 3.56 ±0.95 95 ±19* 1.76 ±0.83 50± 13* 0.83 ±0.81* 30 ±29* 50 ±12* 7.09 ± 1.31 126 ± 17 16.7 ±8.5 67 ±30*
63.2 ±7.3 170.9 ±9.5 72.5 ±16.2 3.36 ±0.93 85 ±17 1.31 ±0.55 42 ±12 0.52 ±0.33 I9± 12 57 ±10 7.96 ±1.92 135 ±24 13.1 ±6.5 52 ±24*
63.2 ±6.9 171.8 ±7.5 78.8 ± 18.7 3.31 ±0.85 82 ±17 1.33 ±0.51 44 ± 12 0.49 ±0.27 18 ± 1 1 58 ± 9 7.89 ± 1.41 132 ± 18 14.0 ±6.3 56 ±23
P
> 0 .1 0 > 0 .1 0 > 0 .1 0 > 0 .1 0 0 .1 0
Exercise PaO,, mm Hg 3 min Maximum
74.9 ±16.1 76.1 ±21.8*
68.5 ± 13.7 65.9 ± 15.9
67.1 ±11.5 65.4 ±13.4
= 0.07