- Email: [email protected]
Resting single-breath diffusing capacity as a screening test for exercise-induced hypoxemia
Resting Single-Breath Diffusing Capacity as a Screening Test for Exercise-Induced Hypoxemia
MARK A. KELLEY, M.D. REYNOLD A. PANETTIERI, Jr., M.D. ANN V. KRUPINSKI, B.S., R.C.P.T. Philadelphia,
Pennsylvania
From the Cardiovascular-Pulmonary Division, Department of Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania. This work was supported in part by Trainee National Institutes of Health Grant 5-T35-HD07217-OlAl. Requests for reprints should be addressed to Dr. Mark A. Kelley, 100 Centrex, Hospital of the University of Pennsylvania, 3400 Spruce Street, Philadelphia, Pennsylvania 19104. Manuscript accepted April 4, 1985. A continuing medical education quiz on this article (one hour of Category 1 credit) appears on page A 18 1 of this issue.
Recent reports in selected patients have suggested that a reduced resting single-breath carbon monoxide diffusing capacity may be associated with a fall in arterial oxygen saturation durlng exercise. To determine if the diffusing capacity could serve as a screening test for changes in oxygen saturation in an unselected population, results of exercise studies were examined in 106 patients consecutively referred to an exercise laboratory. Nearly half of the patients underwent exercise testing to evaluate interstitial disease whereas the remainder were referred for unexplained dyspnea or for disabilii evaluations. Arterial desaturation was seen within all patient subgroups and was closely associated with reduced diffusing capacity. For detecting changes of 4 percent or more in oxygen saturation, a diising capacity of less than 50 percent of predicted gave the best combination of sensitivity (69 percent) and specificity (93 percent), whereas a diiusing capacity of 60 percent or less of predicted was 100 percent sensitive and 64 percent specific. For detecting lesser degrees of desaturation, sensitivities were slightly reduced but specificities were preserved. Thus, a diffusing capacity of less than 50 percent of predicted was associated with substantial arterial desaturation during exercise, whereas patients with a diffusing capacity of more than 60 percent of predicted had no desaturation during exercise. These results suggest that the resting diffusing capacity can serve as a screening test for exercise-induced hypoxemia in an unselected population. Exercise testing is often used to detect exercise hypoxemia particularly in the evaluation of patients with interstitial lung disease, disability assessment, or unexplained dyspnea [ 1,2]. Recently, the resting single-breath carbon monoxide diffusing capacity has been related to gas exchange during exercise. In selected patients with interstitial lung disease [3] and chronic obstructive lung disease [4], reductions in diffusing capacity were associated with arterial desaturation during exercise. Since these studies were performed in special patient subgroups, we wondered if these observations were applicable to an unselected group of patients. Specifically, we wished to document how well the resting diffusing capacity could serve as a screening test for exercise-induced hypoxemia. To address these questions, we studied all patients referred to our exercise facility and analyzed the relationships between clinical diagnosis, pulmonary function, resting diffusing capacity, and exerciseinduced hypoxemia.
May
1986
The American
Journal
of Medicine
Volume
60
607
DIFFUSING
CAPACITY
TABLE I
AS SCREENING
Clinical Diagnoses Hypoxemia
TEST
FOR HYPOXEMIA-KELLEY
ET AL
and Exercise
Exercise Protocol. Before exercise, the exercise protocol was explained to each patient and 12-lead etectrocardiography was performed. The patient breathed room air through a mouthpiece connected to a Hans-Rudolph valve. Expired ventilation was measured continuously by using a heated pneumotachograph (Fleisch no. 3) connected to a differential pressure transducer (Validyne MP451). The heart rate was continuously measured by a cardiotachometer. Expired gases were sampled continuously from an 11.5 liter mixing chamber. Percentages of carbon dioxide and oxygen were measured on a Beckman Medical Gas Analyzer LB-2 and Oxygen OM-11, respectively. Data were recorded on FM tape and later analyzed with a PDP-12 computer (Digital Corporation). All exercise studies were performed on a programmable treadmill (Quinton Instruments). Each patient underwent exercise testing according to a modified Naughton protocol so that oxygen uptake rose every minute by approximately 3.5 ml/kg per minute [9]. Patients terminated the study when they could no longer continue to exercise. Predicted values for maximal oxygen uptake during exercise were obtained from age-related regression equations [IO]. Measurement of Arterial Blood Gases and Oxygen Saturation. Ail patients had arterial oxygen saturations determined at rest and during the last minute of exercise. in 29 studies, arterial blood was sampled directly from an indwelling catheter in the brachiai or radial artery; in 53 studies, arterial oxygen saturation was measured with an ear oximeter (Hewlett-Packard model 47201A); and in 24 studies, both techniques were utilized. Arterial blood specimens were kept on ice and analyzed within 20 minutes after sampling. Measurements of arterial oxygen tension, arterial carbon dioxide tension, and pH were determined in dupiicate on two different instruments (Corning 135 Automatic pH Blood Gas System and a Radiometer MB53 Microsystern) and were temperature-corrected. Duplicate arterial blood gas results agreed to within 3 mm Hg. From these measurements, arterial oxygen saturation was then caiculated [ll]. We found no significant differences between ear oximetry and arterial measurements of oxygen saturation except when values were below 60 percent, thus confirming previous observations [ 121. Therefore, if arterial oxygen saturation was below 60 percent by ear oximetry at rest or during exercise, arterial oxygen saturation was calculated from oxygen tension measured in arterial blood. Diffusing Capacity as a Screening Test. Previous reports have noted exercise hypoxemia with diffusing capacities of less than 50 percent to 60 percent of predicted [3,4]. Therefore, we chose to examine the diffusing capacity at two reference points: less than 50 percent of predicted and 60 percent or less of predicted. We also examined three levels of change in oxygen saturation during exercise: 2 percent or more, 3 percent or more, and 4 percent or more. With these reference points, we used standard methods to calculate sensitivity, specificity, and predictive values for diffusing capacity [ 131. Data Analysis. All pulmonary function test results are expressed as percent of predicted. Ail arterial oxygen saturations are expressed as percent of hemoglobin saturated.
Numberof Patients Numberof Patients Studied wilh Desaturalion’ No known disease Chronic obstructive pulmonary disease Sarcoidosis idiopathic pulmonary fibrosis Collagen vascular disease Miscellaneous interstitial diseaset Coal workers’ pneumoconiosis Asbestosis -Miscellaneous occupational diseases Total
27 (25)+ 11 (10)
5 (14) 4(ll)
32 (30)
8 (22)
6 (6)
6 (17)
4 (4) 4 (4)
3 (6) 3 (6)
16 (15)
4(11)
3 (3) 3 (3)
2 (6)
106
1 (3) 36
* Defined as change in oxygen saturation of 2 percent or more. t Numbers in parentheses are percent of total number of patients in the column. t Three eosinophilic granulomas, one drug-induced fibrosis. 5 One each of berylliosis, smoke inhalation, chest trauma.
PATIENTS
AND
METHODS
Patients. This study included ail patients referred to our exercise laboratory over a four-year period. The laboratory is located in a 660-bed, urban, university teaching hospital. Ail exercise tests were ordered for clinical purposes by patients’ physicians who were unaware of the nature of this study. No attempt was made to recruit any subgroup of patients. Pulmonary Function Tests. Before exercise, each patient underwent resting pulmonary function tests. With a 13.5 liter water seal spirometer (W.C. Collins), measurements were made of forced vital capacity and forced expiratoty volume in one second. Lung volumes were determined by body plethysmography. The diffusing capacity was determined by the single-breath method [5] and considered valid only if the inspiratory volume was at least 90 percent of forced vital capacity. No patient had anemia or qualitative hemoglobin abnormality. From published sources, predicted values were obtained for lung volumes [6], spirometric data [7], and diffusing capacity [6]. Each patient was categorized by resting pulmonary function test criteria. Patients with normal results of pulmonary function tests had a forced vital capacity and total lung capacity of 60 percent or more of predicted with a onesecond forced expiratory volume/forced vital capacity ratio of 70 percent or greater. Restrictive pulmonary function test data were defined as one-second forced expiratory volume/forced vital capacity ratios of 70 percent or greater with a forced vital capacity or total lung capacity of 66 percent to 79 percent of predicted (mild restriction) or with a forced vital capacity or total lung capacity of less than 65 percent of predicted (moderate to severe restriction). Patients with obstructive disease had one-second forced expiratory volume/forced vital capacity ratios of less than 70 percent and no evidence of restrictive disease.
808
May
1988
The American
Journal
of Medicine
Volume
80
DIFFUSING
TABLE Ii
AS SCREENING
TEST FOR HYPOXEMIA-KELLEY
ET AL
Pulmonary Function Test Results and Maximal Exercise Performance
Pulmonary Function Category Normal (n = 55) Obstructive (n = 11) Mild restriction (n = 23) Moderate to severe restriction (I? = 17) All data expressed
CAPACITY
Forced Vital Capacity
One-Second Forced Expiratory Volume/Forced Vital Capacity
Diffusing Capacity
Total Lung Capacity
Maxlmal Oxygen Uptake
98 f
3
83 f
1
100 f
2
74 f
3
63 f 3
90 f
6
47 f
4
131 f
8
52 f
6
53 f
5
75 f
1
84 f
1
75 f
3
56 f
6
45 f
4
53 f
3
80 f
3
60 f
3
36 f
6
36 f
4
volume/forced
vital capacity
as percent
of predicted
except
one-second
forced
expiratory
ratio.
was associated with a fall in oxygen saturation with exercise (Figure 1). More severe desaturation occurred in patients with markedly reduced diffusing capacities. Of 28 patients with a diffusing capacity of 50 percent or less of predicted, 23 had desaturation of 4 percent or more. in contrast, the vast majority of patients with a diffusing capacity of 50 percent or more of predicted had desaturation of 1 percent or less during exercise (Figure 1). Desaturation of at least 2 percent occurred in ail pulmonary function test subgroups: 11 of 55 patients with normal pulmonary function test results; 21 of 40 with restrictive changes; and four of 11 with obstructive changes (Figure 1).
The difference between resting arterial oxygen saturation and that during exercise is termed a change in oxygen saturation. A positive change in oxygen saturation denotes desaturation with exercise whereas a negative value indicates improvement in oxygen saturation with exercise. All results are reported as means f standard errors. RESULTS Patients. A total of 152 patients underwent exercise studies in our laboratory over a four-year period. Of these, 106 patients had complete data available for analysis. The 57 men and 49 women had a mean age of 45 f 2 years (range 17 to 84). Thirty-eight patients were referred for unexplained dyspnea; 46 for interstitial lung disease; and 22 for disability assessment. in Table I, clinical diagnoses on referral are presented with the numbers of patients in each category who demonstrated changes in oxygen saturation of at least 2 percent. These data show that a variety of clinical problems were referred to our laboratory. Of note is that 25 percent of the patients had no known lung disease at the time of the exercise study. In these patients, the exercise test was used to detect any clinically occult cardiopulmonary disorder. Of the 36 patients who had desaturation, no single diagnosis dominated the group. However, 28 of these 36 patients could be classified as having idiopathic or occupationally related interstitial lung disease. Pulmonary Function and Maximal Exercise Performance. The patients were characterized by their pulmonary function test data ITable ii): 55 were normal; 11 had obstructive changes; 23 showed mild restriction; and 17 had moderate to severe restriction. The patients with more abnormal pulmonary function test results had larger reductions in diffusing capacity and maximal oxygen uptake. Diising Capacity, Resting Arterial Oxygen Saturation, and Arterial Hypoxemia with Exercise. Of 106 patients, 36 had changes in oxygen saturation of 2 percent or more whereas 26 had changes in oxygen saturation of 4 percent or more. A reduced resting diffusing capacity
PFT
.
0 0
CATEGORIES
NORMAL MILD RESTRlCTlON
.
MODERATE RESTRICTION X OBSTRUCTION
.
RESTING
DC0
OR SEVERE
(% PREDICTED)
Figure 1. Resting single-breath diffusing capacity for carbon monoxide (DCO) compared with the fall in arterial oxygen (02) saturation during exercise. PFT = pulmonary function test.
May
1986
The American
Journal
of Medicine
Volume
80
809
DIFFUSING
CAPACITY
AS SCREENING
TEST
FOR
HYPOXEMIA-KELLEY
ET AL
Diiing Capacity as a Screening Test. Results of sensitivity, specificity, and predictive value calculations are presented in Table Iii. For detecting arterial desaturation of at least 4 percent, a diffusing capacity of 60 percent or less of predicted had 100 percent sensitivity and 64 percent specificity. Diffusing capacity of less than 50 percent of predicted had a sensitivity of 89 percent and a specificity of 93 percent. For lesser degrees of desaturation, diffusing capacity was less sensitive but specificity remained unchanged. For all degrees of desaturation, the diffusing capacity of 60 percent or less had somewhat superior sensitivity but markedly inferior specificity compared with the diffusing capacity of less than 50 percent. This trend was also seen in predictive values where a diffusing capacity of less than 50 percent gave the best combination of positive and negative predictive values.
5-
A
*
h
j-
A A
SAT 2 4%
A Or
0 A 02 SAT 2-3% . A02 SAT<2% La 1
)O
20
I
I
1
40
I
60
80
100
I
RESTING
DC0
120
I
140
(% PREDICTED)
COMMENTB
Figure 2. Resting single-breath diffusing capacity for carbon monoxide (DCO) compared with resting arterial oxygen (0,) saturation. AO,&AT = change in arterial oxygen sateration during exercise.
TABLE iii
Our results demonstrate a close association between resting diffusing capacity and exercise-induced hypoxemia in patients referred to a general pulmonary exercise laboratory. Our patients had a variety of clinical diagnoses, indications for exercise testing, and pulmonary function results. The usefulness of the diffusing capacity as a screening test for exercise hypoxemia depends on the desaturation threshold desired by the clinician. In monitoring oxygen saturation during exercise, measurement errors of at least 1 percent could be encountered. Therefore, the minimal change in oxygen saturation of clinical interest would likely be 2 percent. For detecting this degree of desaturation, the diffusing capacity was reasonably sensitive and specific. For desaturation of at least 4 percent, a diffusing capacity of 60 percent or less identified all patients who had desaturation whereas a diffusing capacity of less than 50 percent gave a useful combination of high sensitivity (89 percent) and specificity (93 percent). The predictive value of any diagnostic test depends not only on the test characteristics but also on the prevalence of the disease within the studied population [ 131. In our patients, the “disease” of diagnostic interest was arterial desaturation. Desaturation of 2 percent or more was seen in 34 percent of our study group whereas desaturation of 4 percent or more occurred in 25 percent of our patients. In populations with higher prevalences of arterial desaturation, the positive predictive values of the diffusing capacity would be even higher than in our study, thus reducing the number of false-positive results. In populations in whom desaturation is unusual, the diffusing capacity, like any other screening test, would have a reduced positive predictive value, and thus a large number of false-positive results. However, regardless of the prevalence of desaturation, the diffusing capacity would retain its high negative predictive value because of its high sensitivity. Thus, the
Diffusing Capacity (DCO) as a Screening Test for Exercise Hypoxemia Detected Decrease in Arterial Oxygen Saturation during Exercise L2 Percent 23 Percent 14 Percent
Sensitivity DC0 560% DC0 <50% Specificity DC0 560% DC0 <50 % Positive Predictive DC0 160% DC0 <50 % Negative Predictive DC0 560% DC0 <50%
81% 67%
93% 82%
100% 89%
63 % 92%
62% 92%
64% 93%
53% 86%
47% 82%
47% 82%
86 % 82%
96% 92%
100% 92%
Value
Value
Desaturation was more closely associated with reduced diffusing capacity than with reduced resting arterial oxygen saturation (Figure 2). Although 29 of 36 patients with desaturation had a diffusing capacity of 60 percent or less of pred/cted, 18 of these 36 patients had apparently normal resting arterial blood gas values with resting arterial oxygen saturation of at leas? 95 percent. Particularly striking were nine patients with changes in oxygen saturation of more than 4 percent and resting arterial oxygen saturation of 95 percent or more. All of these patients had diffusing capacities of less than 60 percent of predicted (Figure 2).
810
May
1986
The
American
Journal
of Medicine
Volume
80
DIFFUSING
diffusing capacity with a minimal number of false-negative results has an excellent ability to exclude arterial desaturation in any study population. Our results are consistent with two previous reports. In patients with interstitial disease, Risk et al [3] found that exercise-induced hypoxemia was associated with a diffusing capacity of less than 70 percent of predicted and, in most cases, less than 50 percent of predicted. Owens et al [4] studied patients with severe, obstructive disease and found that patients with diffusing capacities of less than 55 percent of predicted had falls in oxygen saturation of at least 3 percent. Our study, based on an unselected patient population, revealed several applications of the resting diffusing capacity. First, we found that the diffusing capacity is more closely associated with exercise hypoxemia than the resting arterial oxygen saturation. Some patients had severe hypoxemia during exercise despite a resting arterial oxygen saturation of at least 95 percent, which corresponds to an arterial oxygen tension of about 75 mm Hg under normal resting conditions. This arterial oxygen tension represents only a slight impairment in gas exchange and is not suggestive of arterial desaturation on exercise. However, all patients with this level of resting arterial oxygen saturation and diffusing capacities of less than 60 percent of predicted had substantial desaturation on exercise. These results suggest that the diffusing capacity reduction was a more accurate clue to the presence of exercise hypoxemia than was the resting arterial oxygen saturation. Second, we found that an abnormal diffusing capacity may be the only resting physiologic parameter that suggests clinically important pulmonary disease. Over half of our patients had normal lung volumes and spirometric values and yet 20 percent of this group had desaturation, as predicted by reduced diffusing capacity. These patients were later found to have occult pulmonary disease, unmasked by the resting diffusing capacity and exercise hypoxemia. A third contribution of our study is that our observations were consistent within a variety of patient groups, whether categorized by physiologic testing or by clinical diagnosis. For all patients who had desaturation, no disease was dominant compared with its occurrence in the overall study group. Thus, our results are not skewed by any special subgroup. On the basis of these data, we believe that our patient population was sufficiently diverse to be
CAPACITY
AS SCREENING
TEST
FOR HYPOXEMIA-KELLEY
ET AL
representative of most patients undergoing exercise evaluations in a general exercise facility. The diffusing capacity has been recently criticized as an imprecise measurement with considerable interlaboratory variability [ 141. However, the remarkable similarity of our observations to those of three other exercise facilities [3,4] suggests that diffusing capacity results can be quite consistent among experienced laboratories. Therefore, recent efforts to standardize the diffusing capacities, if successful, should make this test a more useful clinical tool. Our study results deserve one major caveat. We defined exercise-induced hypoxemia as a fall in arterial oxygen saturation and not as a change in arterial oxygen tension. A disadvantage of measuring oxygen saturation is that small changes in arterial oxygen tension do not substantially alter oxygen saturation in the upper, flat segment of the oxyhemoglobin dissociation curve. Thus, the ear oximeter, which measures arterial saturation only, may not detect changes in arterial oxygen tension in the range of 80 to 100 mm Hg. Although arterial desaturation with exercise has obvious clinical significance, more subtle changes in arterial blood gas values may also be important, especially in evaluating and treating interstitial disease [ 151. Our data do not allow us to comment on whether the diffusing capacity would be a useful screening test for predicting those subtle changes. On the basis of our results, we believe that a diffusing capacity below 50 percent of predicted is strongly suggestive of exercise-induced hypoxemia. We recommend that in all such patients, gas exchange be m~SI.md during exercise, even if resting arterial blood gas values and spirometric data are normal. Patients with diffusing capacities of 50 percent to 60 percent of predicted have intermediate probability of exercise desaturation and may also benefit from diagnostic testing. A corollary of our study is that patients with diffusing capacities of more than 60 percent of predicted have very little likelihood of hypoxemia during exercise. Therefore, in such patients, exercise testing for the sole purpose of detecting arterial desaturation is likely to be unrewarding. ACKNOWLEDGMENT We wish to thank Thomas Nusbickel for technical assistance; Andrea Cornitcher for assistance in manuscript preparation; and Drs. Alfred P. Fishman and Allan I. Pack for their helpful review of the manuscript.
REFERENCES 1.
Crystal RG, Fulmer JD, Roberts WC, Morton ML, Line DR, Reynolds HY: Idiopathic pulmonary fibrosis: clinical, histologic, radiologic, physiologic, scintigraphic, cytologic, and biochemical aspects. Ann Intern Med 1976: 85: 769-788.
2.
3.
May
1986
American Thoracic Society: Evaluation of impairment/disability secondary to respiratory disease. Am Rev Respir Dis-1982; 126: 945-951. Risk C. Eoler GR, Gaensler EA: Exercise alveolar-arterial oxygen’ pressure difference in interstitial lung disease.
The American
Journal
of Medicine
Volume
80
811
DIFFUSING
4.
5.
6.
7.
8. 9.
812
CAPACITY
AS SCREENING
TEST
FOR HYPOXEMIA-KELLEY
ET AL
Chest 1984; 85: 69-74. Owens GR, Rogers RM, Pennock BE, Levin D: The diffusing capacity as a predictor of arterial oxygen desaturation during exercise in patients with chronic obstructive pulmonarydisease. N Engl J Med 1984; 31: 1218-1221. Ogilvie CM, Forster RE, Blakemore WS, Morton JW: A standard breath-holding technique for the clinical measurement of the diffusing capacity of the lung for carbon monoxide. J Glin Invest 1957; 36: 1-17. Naimark A, Cherniack RM, Protti D: Comprehensive respiratory information system for clinical investigatlon of respiratory disease. Am Rev Respir Dis 1971; 103: 229239. Knudson RJ, Slatin RC, Lebowitz MD, Burrows B: The maximal expiratory flow-curve. Am Rev Respir Dis 1976; 113: 587-600. Cotes JE: Lung function. Philadelphia: FA Davis, 1968; 375. Naughton JP, Haider R: Methods of exercise testing. In: Naughton JP, Helierstein HR, Mohler IC, eds. Exercise
May
1988
The American
Journal
of Medicine
Volume
10.
11. 12.
13.
14.
15.
80
testing and exercise training in coronary heart disease. New York: Academic Press, 1973; 79-91. Jones NL, Campbell EJ, Edwards RHT, Robertson DG: Clinical exercise testing. Philadelphia: WE Saunders, 1975; 202. Severinghaus JW: Blood gas calculator. J Appl Physiol 1966; 21: 1108-1116. Poppius H, Viljanen AA: A new ear oximeter for assessment of exercise induced arterial desaturation in patients with pulmonary disease. Stand J Respir Dis 1979; 58: 279-283. Griner PF, Mayewski RJ, Mushlin Al, Greenland P: Selection and interpretation of diagnostic tests and procedures. Ann Intern Med 1981; 121: 553-600. Clausen J, Crapo R, Gardner R: Interlaboratory comparisons of pulmonary function testing (abstr). Am Rev Respir Dis 1984; 129: A37. Kelley MA, Daniele RP: Exercise testing in interstitial lung disease. Clin Chest Med 1984; 5: 145-156.
MARK A. KELLEY, M.D. REYNOLD A. PANETTIERI, Jr., M.D. ANN V. KRUPINSKI, B.S., R.C.P.T. Philadelphia,
Pennsylvania
From the Cardiovascular-Pulmonary Division, Department of Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania. This work was supported in part by Trainee National Institutes of Health Grant 5-T35-HD07217-OlAl. Requests for reprints should be addressed to Dr. Mark A. Kelley, 100 Centrex, Hospital of the University of Pennsylvania, 3400 Spruce Street, Philadelphia, Pennsylvania 19104. Manuscript accepted April 4, 1985. A continuing medical education quiz on this article (one hour of Category 1 credit) appears on page A 18 1 of this issue.
Recent reports in selected patients have suggested that a reduced resting single-breath carbon monoxide diffusing capacity may be associated with a fall in arterial oxygen saturation durlng exercise. To determine if the diffusing capacity could serve as a screening test for changes in oxygen saturation in an unselected population, results of exercise studies were examined in 106 patients consecutively referred to an exercise laboratory. Nearly half of the patients underwent exercise testing to evaluate interstitial disease whereas the remainder were referred for unexplained dyspnea or for disabilii evaluations. Arterial desaturation was seen within all patient subgroups and was closely associated with reduced diffusing capacity. For detecting changes of 4 percent or more in oxygen saturation, a diising capacity of less than 50 percent of predicted gave the best combination of sensitivity (69 percent) and specificity (93 percent), whereas a diiusing capacity of 60 percent or less of predicted was 100 percent sensitive and 64 percent specific. For detecting lesser degrees of desaturation, sensitivities were slightly reduced but specificities were preserved. Thus, a diffusing capacity of less than 50 percent of predicted was associated with substantial arterial desaturation during exercise, whereas patients with a diffusing capacity of more than 60 percent of predicted had no desaturation during exercise. These results suggest that the resting diffusing capacity can serve as a screening test for exercise-induced hypoxemia in an unselected population. Exercise testing is often used to detect exercise hypoxemia particularly in the evaluation of patients with interstitial lung disease, disability assessment, or unexplained dyspnea [ 1,2]. Recently, the resting single-breath carbon monoxide diffusing capacity has been related to gas exchange during exercise. In selected patients with interstitial lung disease [3] and chronic obstructive lung disease [4], reductions in diffusing capacity were associated with arterial desaturation during exercise. Since these studies were performed in special patient subgroups, we wondered if these observations were applicable to an unselected group of patients. Specifically, we wished to document how well the resting diffusing capacity could serve as a screening test for exercise-induced hypoxemia. To address these questions, we studied all patients referred to our exercise facility and analyzed the relationships between clinical diagnosis, pulmonary function, resting diffusing capacity, and exerciseinduced hypoxemia.
May
1986
The American
Journal
of Medicine
Volume
60
607
DIFFUSING
CAPACITY
TABLE I
AS SCREENING
Clinical Diagnoses Hypoxemia
TEST
FOR HYPOXEMIA-KELLEY
ET AL
and Exercise
Exercise Protocol. Before exercise, the exercise protocol was explained to each patient and 12-lead etectrocardiography was performed. The patient breathed room air through a mouthpiece connected to a Hans-Rudolph valve. Expired ventilation was measured continuously by using a heated pneumotachograph (Fleisch no. 3) connected to a differential pressure transducer (Validyne MP451). The heart rate was continuously measured by a cardiotachometer. Expired gases were sampled continuously from an 11.5 liter mixing chamber. Percentages of carbon dioxide and oxygen were measured on a Beckman Medical Gas Analyzer LB-2 and Oxygen OM-11, respectively. Data were recorded on FM tape and later analyzed with a PDP-12 computer (Digital Corporation). All exercise studies were performed on a programmable treadmill (Quinton Instruments). Each patient underwent exercise testing according to a modified Naughton protocol so that oxygen uptake rose every minute by approximately 3.5 ml/kg per minute [9]. Patients terminated the study when they could no longer continue to exercise. Predicted values for maximal oxygen uptake during exercise were obtained from age-related regression equations [IO]. Measurement of Arterial Blood Gases and Oxygen Saturation. Ail patients had arterial oxygen saturations determined at rest and during the last minute of exercise. in 29 studies, arterial blood was sampled directly from an indwelling catheter in the brachiai or radial artery; in 53 studies, arterial oxygen saturation was measured with an ear oximeter (Hewlett-Packard model 47201A); and in 24 studies, both techniques were utilized. Arterial blood specimens were kept on ice and analyzed within 20 minutes after sampling. Measurements of arterial oxygen tension, arterial carbon dioxide tension, and pH were determined in dupiicate on two different instruments (Corning 135 Automatic pH Blood Gas System and a Radiometer MB53 Microsystern) and were temperature-corrected. Duplicate arterial blood gas results agreed to within 3 mm Hg. From these measurements, arterial oxygen saturation was then caiculated [ll]. We found no significant differences between ear oximetry and arterial measurements of oxygen saturation except when values were below 60 percent, thus confirming previous observations [ 121. Therefore, if arterial oxygen saturation was below 60 percent by ear oximetry at rest or during exercise, arterial oxygen saturation was calculated from oxygen tension measured in arterial blood. Diffusing Capacity as a Screening Test. Previous reports have noted exercise hypoxemia with diffusing capacities of less than 50 percent to 60 percent of predicted [3,4]. Therefore, we chose to examine the diffusing capacity at two reference points: less than 50 percent of predicted and 60 percent or less of predicted. We also examined three levels of change in oxygen saturation during exercise: 2 percent or more, 3 percent or more, and 4 percent or more. With these reference points, we used standard methods to calculate sensitivity, specificity, and predictive values for diffusing capacity [ 131. Data Analysis. All pulmonary function test results are expressed as percent of predicted. Ail arterial oxygen saturations are expressed as percent of hemoglobin saturated.
Numberof Patients Numberof Patients Studied wilh Desaturalion’ No known disease Chronic obstructive pulmonary disease Sarcoidosis idiopathic pulmonary fibrosis Collagen vascular disease Miscellaneous interstitial diseaset Coal workers’ pneumoconiosis Asbestosis -Miscellaneous occupational diseases Total
27 (25)+ 11 (10)
5 (14) 4(ll)
32 (30)
8 (22)
6 (6)
6 (17)
4 (4) 4 (4)
3 (6) 3 (6)
16 (15)
4(11)
3 (3) 3 (3)
2 (6)
106
1 (3) 36
* Defined as change in oxygen saturation of 2 percent or more. t Numbers in parentheses are percent of total number of patients in the column. t Three eosinophilic granulomas, one drug-induced fibrosis. 5 One each of berylliosis, smoke inhalation, chest trauma.
PATIENTS
AND
METHODS
Patients. This study included ail patients referred to our exercise laboratory over a four-year period. The laboratory is located in a 660-bed, urban, university teaching hospital. Ail exercise tests were ordered for clinical purposes by patients’ physicians who were unaware of the nature of this study. No attempt was made to recruit any subgroup of patients. Pulmonary Function Tests. Before exercise, each patient underwent resting pulmonary function tests. With a 13.5 liter water seal spirometer (W.C. Collins), measurements were made of forced vital capacity and forced expiratoty volume in one second. Lung volumes were determined by body plethysmography. The diffusing capacity was determined by the single-breath method [5] and considered valid only if the inspiratory volume was at least 90 percent of forced vital capacity. No patient had anemia or qualitative hemoglobin abnormality. From published sources, predicted values were obtained for lung volumes [6], spirometric data [7], and diffusing capacity [6]. Each patient was categorized by resting pulmonary function test criteria. Patients with normal results of pulmonary function tests had a forced vital capacity and total lung capacity of 60 percent or more of predicted with a onesecond forced expiratory volume/forced vital capacity ratio of 70 percent or greater. Restrictive pulmonary function test data were defined as one-second forced expiratory volume/forced vital capacity ratios of 70 percent or greater with a forced vital capacity or total lung capacity of 66 percent to 79 percent of predicted (mild restriction) or with a forced vital capacity or total lung capacity of less than 65 percent of predicted (moderate to severe restriction). Patients with obstructive disease had one-second forced expiratory volume/forced vital capacity ratios of less than 70 percent and no evidence of restrictive disease.
808
May
1988
The American
Journal
of Medicine
Volume
80
DIFFUSING
TABLE Ii
AS SCREENING
TEST FOR HYPOXEMIA-KELLEY
ET AL
Pulmonary Function Test Results and Maximal Exercise Performance
Pulmonary Function Category Normal (n = 55) Obstructive (n = 11) Mild restriction (n = 23) Moderate to severe restriction (I? = 17) All data expressed
CAPACITY
Forced Vital Capacity
One-Second Forced Expiratory Volume/Forced Vital Capacity
Diffusing Capacity
Total Lung Capacity
Maxlmal Oxygen Uptake
98 f
3
83 f
1
100 f
2
74 f
3
63 f 3
90 f
6
47 f
4
131 f
8
52 f
6
53 f
5
75 f
1
84 f
1
75 f
3
56 f
6
45 f
4
53 f
3
80 f
3
60 f
3
36 f
6
36 f
4
volume/forced
vital capacity
as percent
of predicted
except
one-second
forced
expiratory
ratio.
was associated with a fall in oxygen saturation with exercise (Figure 1). More severe desaturation occurred in patients with markedly reduced diffusing capacities. Of 28 patients with a diffusing capacity of 50 percent or less of predicted, 23 had desaturation of 4 percent or more. in contrast, the vast majority of patients with a diffusing capacity of 50 percent or more of predicted had desaturation of 1 percent or less during exercise (Figure 1). Desaturation of at least 2 percent occurred in ail pulmonary function test subgroups: 11 of 55 patients with normal pulmonary function test results; 21 of 40 with restrictive changes; and four of 11 with obstructive changes (Figure 1).
The difference between resting arterial oxygen saturation and that during exercise is termed a change in oxygen saturation. A positive change in oxygen saturation denotes desaturation with exercise whereas a negative value indicates improvement in oxygen saturation with exercise. All results are reported as means f standard errors. RESULTS Patients. A total of 152 patients underwent exercise studies in our laboratory over a four-year period. Of these, 106 patients had complete data available for analysis. The 57 men and 49 women had a mean age of 45 f 2 years (range 17 to 84). Thirty-eight patients were referred for unexplained dyspnea; 46 for interstitial lung disease; and 22 for disability assessment. in Table I, clinical diagnoses on referral are presented with the numbers of patients in each category who demonstrated changes in oxygen saturation of at least 2 percent. These data show that a variety of clinical problems were referred to our laboratory. Of note is that 25 percent of the patients had no known lung disease at the time of the exercise study. In these patients, the exercise test was used to detect any clinically occult cardiopulmonary disorder. Of the 36 patients who had desaturation, no single diagnosis dominated the group. However, 28 of these 36 patients could be classified as having idiopathic or occupationally related interstitial lung disease. Pulmonary Function and Maximal Exercise Performance. The patients were characterized by their pulmonary function test data ITable ii): 55 were normal; 11 had obstructive changes; 23 showed mild restriction; and 17 had moderate to severe restriction. The patients with more abnormal pulmonary function test results had larger reductions in diffusing capacity and maximal oxygen uptake. Diising Capacity, Resting Arterial Oxygen Saturation, and Arterial Hypoxemia with Exercise. Of 106 patients, 36 had changes in oxygen saturation of 2 percent or more whereas 26 had changes in oxygen saturation of 4 percent or more. A reduced resting diffusing capacity
PFT
.
0 0
CATEGORIES
NORMAL MILD RESTRlCTlON
.
MODERATE RESTRICTION X OBSTRUCTION
.
RESTING
DC0
OR SEVERE
(% PREDICTED)
Figure 1. Resting single-breath diffusing capacity for carbon monoxide (DCO) compared with the fall in arterial oxygen (02) saturation during exercise. PFT = pulmonary function test.
May
1986
The American
Journal
of Medicine
Volume
80
809
DIFFUSING
CAPACITY
AS SCREENING
TEST
FOR
HYPOXEMIA-KELLEY
ET AL
Diiing Capacity as a Screening Test. Results of sensitivity, specificity, and predictive value calculations are presented in Table Iii. For detecting arterial desaturation of at least 4 percent, a diffusing capacity of 60 percent or less of predicted had 100 percent sensitivity and 64 percent specificity. Diffusing capacity of less than 50 percent of predicted had a sensitivity of 89 percent and a specificity of 93 percent. For lesser degrees of desaturation, diffusing capacity was less sensitive but specificity remained unchanged. For all degrees of desaturation, the diffusing capacity of 60 percent or less had somewhat superior sensitivity but markedly inferior specificity compared with the diffusing capacity of less than 50 percent. This trend was also seen in predictive values where a diffusing capacity of less than 50 percent gave the best combination of positive and negative predictive values.
5-
A
*
h
j-
A A
SAT 2 4%
A Or
0 A 02 SAT 2-3% . A02 SAT<2% La 1
)O
20
I
I
1
40
I
60
80
100
I
RESTING
DC0
120
I
140
(% PREDICTED)
COMMENTB
Figure 2. Resting single-breath diffusing capacity for carbon monoxide (DCO) compared with resting arterial oxygen (0,) saturation. AO,&AT = change in arterial oxygen sateration during exercise.
TABLE iii
Our results demonstrate a close association between resting diffusing capacity and exercise-induced hypoxemia in patients referred to a general pulmonary exercise laboratory. Our patients had a variety of clinical diagnoses, indications for exercise testing, and pulmonary function results. The usefulness of the diffusing capacity as a screening test for exercise hypoxemia depends on the desaturation threshold desired by the clinician. In monitoring oxygen saturation during exercise, measurement errors of at least 1 percent could be encountered. Therefore, the minimal change in oxygen saturation of clinical interest would likely be 2 percent. For detecting this degree of desaturation, the diffusing capacity was reasonably sensitive and specific. For desaturation of at least 4 percent, a diffusing capacity of 60 percent or less identified all patients who had desaturation whereas a diffusing capacity of less than 50 percent gave a useful combination of high sensitivity (89 percent) and specificity (93 percent). The predictive value of any diagnostic test depends not only on the test characteristics but also on the prevalence of the disease within the studied population [ 131. In our patients, the “disease” of diagnostic interest was arterial desaturation. Desaturation of 2 percent or more was seen in 34 percent of our study group whereas desaturation of 4 percent or more occurred in 25 percent of our patients. In populations with higher prevalences of arterial desaturation, the positive predictive values of the diffusing capacity would be even higher than in our study, thus reducing the number of false-positive results. In populations in whom desaturation is unusual, the diffusing capacity, like any other screening test, would have a reduced positive predictive value, and thus a large number of false-positive results. However, regardless of the prevalence of desaturation, the diffusing capacity would retain its high negative predictive value because of its high sensitivity. Thus, the
Diffusing Capacity (DCO) as a Screening Test for Exercise Hypoxemia Detected Decrease in Arterial Oxygen Saturation during Exercise L2 Percent 23 Percent 14 Percent
Sensitivity DC0 560% DC0 <50% Specificity DC0 560% DC0 <50 % Positive Predictive DC0 160% DC0 <50 % Negative Predictive DC0 560% DC0 <50%
81% 67%
93% 82%
100% 89%
63 % 92%
62% 92%
64% 93%
53% 86%
47% 82%
47% 82%
86 % 82%
96% 92%
100% 92%
Value
Value
Desaturation was more closely associated with reduced diffusing capacity than with reduced resting arterial oxygen saturation (Figure 2). Although 29 of 36 patients with desaturation had a diffusing capacity of 60 percent or less of pred/cted, 18 of these 36 patients had apparently normal resting arterial blood gas values with resting arterial oxygen saturation of at leas? 95 percent. Particularly striking were nine patients with changes in oxygen saturation of more than 4 percent and resting arterial oxygen saturation of 95 percent or more. All of these patients had diffusing capacities of less than 60 percent of predicted (Figure 2).
810
May
1986
The
American
Journal
of Medicine
Volume
80
DIFFUSING
diffusing capacity with a minimal number of false-negative results has an excellent ability to exclude arterial desaturation in any study population. Our results are consistent with two previous reports. In patients with interstitial disease, Risk et al [3] found that exercise-induced hypoxemia was associated with a diffusing capacity of less than 70 percent of predicted and, in most cases, less than 50 percent of predicted. Owens et al [4] studied patients with severe, obstructive disease and found that patients with diffusing capacities of less than 55 percent of predicted had falls in oxygen saturation of at least 3 percent. Our study, based on an unselected patient population, revealed several applications of the resting diffusing capacity. First, we found that the diffusing capacity is more closely associated with exercise hypoxemia than the resting arterial oxygen saturation. Some patients had severe hypoxemia during exercise despite a resting arterial oxygen saturation of at least 95 percent, which corresponds to an arterial oxygen tension of about 75 mm Hg under normal resting conditions. This arterial oxygen tension represents only a slight impairment in gas exchange and is not suggestive of arterial desaturation on exercise. However, all patients with this level of resting arterial oxygen saturation and diffusing capacities of less than 60 percent of predicted had substantial desaturation on exercise. These results suggest that the diffusing capacity reduction was a more accurate clue to the presence of exercise hypoxemia than was the resting arterial oxygen saturation. Second, we found that an abnormal diffusing capacity may be the only resting physiologic parameter that suggests clinically important pulmonary disease. Over half of our patients had normal lung volumes and spirometric values and yet 20 percent of this group had desaturation, as predicted by reduced diffusing capacity. These patients were later found to have occult pulmonary disease, unmasked by the resting diffusing capacity and exercise hypoxemia. A third contribution of our study is that our observations were consistent within a variety of patient groups, whether categorized by physiologic testing or by clinical diagnosis. For all patients who had desaturation, no disease was dominant compared with its occurrence in the overall study group. Thus, our results are not skewed by any special subgroup. On the basis of these data, we believe that our patient population was sufficiently diverse to be
CAPACITY
AS SCREENING
TEST
FOR HYPOXEMIA-KELLEY
ET AL
representative of most patients undergoing exercise evaluations in a general exercise facility. The diffusing capacity has been recently criticized as an imprecise measurement with considerable interlaboratory variability [ 141. However, the remarkable similarity of our observations to those of three other exercise facilities [3,4] suggests that diffusing capacity results can be quite consistent among experienced laboratories. Therefore, recent efforts to standardize the diffusing capacities, if successful, should make this test a more useful clinical tool. Our study results deserve one major caveat. We defined exercise-induced hypoxemia as a fall in arterial oxygen saturation and not as a change in arterial oxygen tension. A disadvantage of measuring oxygen saturation is that small changes in arterial oxygen tension do not substantially alter oxygen saturation in the upper, flat segment of the oxyhemoglobin dissociation curve. Thus, the ear oximeter, which measures arterial saturation only, may not detect changes in arterial oxygen tension in the range of 80 to 100 mm Hg. Although arterial desaturation with exercise has obvious clinical significance, more subtle changes in arterial blood gas values may also be important, especially in evaluating and treating interstitial disease [ 151. Our data do not allow us to comment on whether the diffusing capacity would be a useful screening test for predicting those subtle changes. On the basis of our results, we believe that a diffusing capacity below 50 percent of predicted is strongly suggestive of exercise-induced hypoxemia. We recommend that in all such patients, gas exchange be m~SI.md during exercise, even if resting arterial blood gas values and spirometric data are normal. Patients with diffusing capacities of 50 percent to 60 percent of predicted have intermediate probability of exercise desaturation and may also benefit from diagnostic testing. A corollary of our study is that patients with diffusing capacities of more than 60 percent of predicted have very little likelihood of hypoxemia during exercise. Therefore, in such patients, exercise testing for the sole purpose of detecting arterial desaturation is likely to be unrewarding. ACKNOWLEDGMENT We wish to thank Thomas Nusbickel for technical assistance; Andrea Cornitcher for assistance in manuscript preparation; and Drs. Alfred P. Fishman and Allan I. Pack for their helpful review of the manuscript.
REFERENCES 1.
Crystal RG, Fulmer JD, Roberts WC, Morton ML, Line DR, Reynolds HY: Idiopathic pulmonary fibrosis: clinical, histologic, radiologic, physiologic, scintigraphic, cytologic, and biochemical aspects. Ann Intern Med 1976: 85: 769-788.
2.
3.
May
1986
American Thoracic Society: Evaluation of impairment/disability secondary to respiratory disease. Am Rev Respir Dis-1982; 126: 945-951. Risk C. Eoler GR, Gaensler EA: Exercise alveolar-arterial oxygen’ pressure difference in interstitial lung disease.
The American
Journal
of Medicine
Volume
80
811
DIFFUSING
4.
5.
6.
7.
8. 9.
812
CAPACITY
AS SCREENING
TEST
FOR HYPOXEMIA-KELLEY
ET AL
Chest 1984; 85: 69-74. Owens GR, Rogers RM, Pennock BE, Levin D: The diffusing capacity as a predictor of arterial oxygen desaturation during exercise in patients with chronic obstructive pulmonarydisease. N Engl J Med 1984; 31: 1218-1221. Ogilvie CM, Forster RE, Blakemore WS, Morton JW: A standard breath-holding technique for the clinical measurement of the diffusing capacity of the lung for carbon monoxide. J Glin Invest 1957; 36: 1-17. Naimark A, Cherniack RM, Protti D: Comprehensive respiratory information system for clinical investigatlon of respiratory disease. Am Rev Respir Dis 1971; 103: 229239. Knudson RJ, Slatin RC, Lebowitz MD, Burrows B: The maximal expiratory flow-curve. Am Rev Respir Dis 1976; 113: 587-600. Cotes JE: Lung function. Philadelphia: FA Davis, 1968; 375. Naughton JP, Haider R: Methods of exercise testing. In: Naughton JP, Helierstein HR, Mohler IC, eds. Exercise
May
1988
The American
Journal
of Medicine
Volume
10.
11. 12.
13.
14.
15.
80
testing and exercise training in coronary heart disease. New York: Academic Press, 1973; 79-91. Jones NL, Campbell EJ, Edwards RHT, Robertson DG: Clinical exercise testing. Philadelphia: WE Saunders, 1975; 202. Severinghaus JW: Blood gas calculator. J Appl Physiol 1966; 21: 1108-1116. Poppius H, Viljanen AA: A new ear oximeter for assessment of exercise induced arterial desaturation in patients with pulmonary disease. Stand J Respir Dis 1979; 58: 279-283. Griner PF, Mayewski RJ, Mushlin Al, Greenland P: Selection and interpretation of diagnostic tests and procedures. Ann Intern Med 1981; 121: 553-600. Clausen J, Crapo R, Gardner R: Interlaboratory comparisons of pulmonary function testing (abstr). Am Rev Respir Dis 1984; 129: A37. Kelley MA, Daniele RP: Exercise testing in interstitial lung disease. Clin Chest Med 1984; 5: 145-156.