Optimal Protocol Selection for Cardiopulmonary Exercise Testing in Severe COPD

Roberto P. Benzo, Sriram Paramesh, Sanjay A. Patel, William A. Slivka and Frank C. Sciurba Chest 2007;132;1500-1505; Prepublished online October 9, 2007; DOI 10.1378/chest.07-0732

The online version of this article, along with updated information and services can be found online on the World Wide Web at: http://chestjournal.org/cgi/content/abstract/132/5/1500

CHEST is the official journal of the American College of Chest Physicians. It has been published monthly since 1935. Copyright 2007 by the American College of Chest Physicians, 3300 Dundee Road, Northbrook IL 60062. All rights reserved. No part of this article or PDF may be reproduced or distributed without the prior written permission of the copyright holder (http://www.chestjournal.org/misc/reprints.shtml). ISSN: 0012-3692.

Downloaded from chestjournal.org on March 5, 2008 Copyright 2007 by American College of Chest Physicians

Original Research
COPD

Optimal Protocol Selection for Cardiopulmonary Exercise Testing in Severe COPD*

Roberto P. Benzo, MD, MSc, FCCP; Sriram Paramesh, MD; Sanjay A. Patel, MD, MPH, FCCP; William A. Slivka, RPFT; and Frank C. Sciurba, MD, FCCP

Background: The current recommendations of 8 to 12 min for the optimal targeted duration of symptom-limited maximal cardiopulmonary exercise testing (CPET) to attain maximal oxygen consumption are based on results from healthy individuals and may not be applicable to patients with severe COPD. We aimed to determine the optimal duration for a CPET to attain the peak O2peak) in a group of patients with severe COPD using different carefully oxygen consumption (V conducted workload protocols. Methods: We studied 11 subjects with severe COPD (mean FEV1, 32% predicted; 95% confidence interval [CI], 27 to 38% predicted). They completed four incremental, symptom-limited exercise tests on a cycle ergometer using four protocols (4, 8, and 16 W/min continuous ramp protocols, and 8 W/min step protocol) using a randomized double-blind design. Results: The mean duration of these 44 tests was 6.3 min (95% CI, 5.0 to 9.0 min). The duration of the exercise tests differed significantly for the protocols used, as follows: 16-W ramp protocol, 4.0 min (95% CI, 3.0 to 5.1 min); 8-W ramp protocol, 6.6 min (95% CI, 5.0 to 9.0 min); 8-W step protocol, 6.0 min (95% CI, 4.0 to 8.0 min); and 4-W ramp protocol, 8.7 min (95% CI, 4.4 to 13.0 min; p < 0.001). The maximal workload significantly increased as the ramp slope increased from 4 to 8 to 16 W/min (maximal workload, 35.6 vs 50.7 vs 64.3 W, respectively; p < 0.001). Maximal O2 peak, were not different among the four minute ventilation, heart rate, Borg ratings, and V protocols. No differences were found between the ramp and step protocols. Conclusions: In patients with severe COPD (Global Initiative for Chronic Obstructive Lung Disease stages IIIIV), a targeted duration of 5 to 9 min for a CPET appears to be more appropriate than the 8 to 12 min proposed in the current guidelines. Maximal workload, in O2peak, is highly dependent on the ramp incrementation rate. contrast to V (CHEST 2007; 132:1500 1505)
Key words: COPD; emphysema; exercise test; health outcomes Abbreviations: CI confidence interval; CPET cardiopulmonary exercise test; MVV maximal voluntary ventila co2 carbon dioxide output; V e minute ventilation; V o2 oxygen consumption; V o2peak peak oxygen tion; V consumption

clinical use of maximal cardiopulmonary exT he ercise testing (CPET) in patients with COPD is extensive. The data generated from CPET can be used in the following applications: (1) to tailor exercise prescriptions in pulmonary rehabilitation (percentage of maximal workload)1; (2) to evaluate the responses to an intervention2; (3) to aid in the construction of prediction rules for defining risk3; (4) for preoperative selection of potentially high-risk patients for cancer
1500

resection4 8 and lung volume reduction surgery2; and (5) heart transplantation.9 Although widely used, some aspects of CPET standard protocol methodology have received little attention in the literature, and in fact may need refinement. One aspect of particular importance regards the optimal test duration needed to achieve the o2peak). The highest peak oxygen consumption (V American Thoracic Society/American College of Chest Physicians cardiopulmonary exercise guidelines10 recOriginal Research

Downloaded from chestjournal.org on March 5, 2008 Copyright 2007 by American College of Chest Physicians

ommend 8 to 12 min as the desired duration of an exercise test in order to be efficient and provide diagnostic information. This recommendation, however, is based on data generated from a group of five healthy individuals11 and may not be applicable to patients with severe COPD, who often have factors such as deconditioning,12 peripheral muscle dysfunction,13,14 or ventilatory limitation15 affecting exercise endurance. We performed a randomized double-blind study in patients with severe COPD to determine the optimal CPET protocol duration resulting in the highest o2peak by varying workload protocols. We primarily V o2peak in patients with severe hypothesize that V COPD will be achieved in 8 min when a range of durations for exercise testing is effected through variation in the workload ramp rate protocol, in contrast to current exercise duration target guidelines. We further hypothesize that such variations in workload step or ramp protocol will affect maximal workload but not o2peak measurements. V Materials and Methods
Eleven stable subjects (5 women, 6 men) with severe COPD (Global Initiative for Chronic Obstructive Lung Disease stage IIIIV) were recruited for the study based on history, and findings of physical examination, chest radiographs, and pulmonary function testing. No subjects were receiving long-term oxygen therapy. Spirometry and single-breath diffusing capacity of the lung for carbon monoxide (corrected for hemoglobin) were max 6200; SensorMedics; Yorba performed in all subjects (V Linda, CA) following the guidelines of the American Thoracic Society and using standard reference equations.16,17 CPET Each subject underwent four incremental, symptom-limited exercise testing regimes using an electronically braked cycle ergometer (Corival 400; Lode B.V.; Groningen, the Netherlands) using four protocols (continuous ramp rates of 4, 8, 16 W/min or step rate of 8 W/min) in a randomized fashion. Exercise mea*From the Division of Pulmonary & Critical Care Medicine (Drs. Benzo, Patel, and Sciurba, and Mr. Slivka), University of Pittsburgh, Pittsburgh, PA; and Georgia Lung Associates (Dr. Paramesh), Austell, GA. Dr. Benzo is supported by grant No. 5k23CA106544 from the National Institutes of Health. Dr. Sciurba is supported by grant No. 1P50HL084948-0 from the National Institutes of Health. The authors have reported to the ACCP that no significant conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article. Manuscript received March 22, 2007; revision accepted July 11, 2007. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml). Correspondence to: Roberto P. Benzo, MD, MSc, FCCP, Division of Pulmonary and Critical Care Medicine, University of Pittsburgh, Suite 1211 Kaufmann Building, 3471 Fifth Ave, Pittsburgh, PA 15213; e-mail: benzorp@upmc.edu DOI: 10.1378/chest.07-0732
www.chestjournal.org

surements were preceded by 3 min of rest (seated on the ergometer) and 2 min of unloaded pedaling. Dyspnea and leg fatigue were measured using the modified Borg scale at rest, at 2-min intervals during exercise, and at maximal exercise. Tests ended when the cadence dropped to 40 revolutions per minute and did not return despite encouragement, when a patient requested termination of the test, or when the technician terminated the test for safety reasons. The standard encouragement used for these test were as follows: Nice job; Keep it up; and You are doing fine. Subjects completed the protocols on 2 separate days with at least 2 h of rest between tests. Subjects and the investigator were blinded to the protocol used. Standardized submaximal comparisons of oxygen consumption o2), carbon dioxide output (V co2), minute ventilation (V e), (V respiratory rate, heart rate, Borg leg fatigue, and Borg dyspnea were calculated as follows. For each subject, the lowest maximal wattage level achieved among the protocols was used as a reference wattage. Wattage was then standardized for each subject as a percentage of this reference wattage. The highest ramp rate (16 W/min) was compared to the lowest ramp rate (4 W/min). Also the 8-W ramp protocol was compared to the 8-W step protocol. e, tidal volume, heart o2peak, carbon dioxide production, V V rate, respiratory rate, and external workload were measured using a computerized breath-by-breath metabolic system (model CPET-D; Medical Graphics Corp; St. Paul, MN) and o2peak percent predicted was averaged at 15-s intervals. The V calculated using the equation of Jones et al.18 Oxygen saturation by pulse oximetry (model 504-USP; Criticare Systems Inc; Waukesha, WI) and three-lead ECG (Sirecust 732; Siemens Medical Systems, Inc; Danvers, MA) were also continuously recorded. The study was approved by the University of Pittsburgh Institutional Review Board. Informed consent was obtained from all subjects. Statistical Analysis Data are summarized as the mean and SD. We also calculated the 95% confidence intervals (CIs) for the main outcome variable (CPET duration). Comparisons of maximal and submaximal exercise parameters among the ramp protocols were made using repeated measures analysis of variance. Post hoc comparisons to identify differences among groups were made using the Tukey test. All data were analyzed using a commercially available software package (SPSS for Windows, version 13; SPSS Institute; Chicago, IL). A p value of 0.05 was considered to be statistically significant.

Results Subject Characteristics Baseline characteristics of the study group are shown in Table 1. Duration of the Exercise Tests The mean duration for all 44 tests was 6 min (95% CI, 5 to 9 min). The test durations also differed significantly according to the protocol used (p 0.001): 16-W ramp protocol, 4.0 min (95% CI, 3.0 to 5.1 min); 8-W ramp protocol, 6.6
CHEST / 132 / 5 / NOVEMBER, 2007

1501

Downloaded from chestjournal.org on March 5, 2008 Copyright 2007 by American College of Chest Physicians

Table 1Baseline Characteristics*

Subject 1 2 3 4 5 6 7 8 9 10 11 Sex M M F M M M F F F M F Age, yr 67 58 64 52 73 73 58 70 67 72 58 Height, cm 178 173 157 178 173 175 175 163 163 188 168 BMI 25.7 21.5 31.3 17.5 22.9 24.8 21.6 22.8 22.6 27.8 29.6 FEV1, % Predicted 43 26 40 37 22 31 18 43 35 33 32 Dlco, % Predicted 49 61 41 44 61 58 23 24 46 39 41 MVV, % Predicted 31 24 38 29 18 30 22 37 35 33 32

*BMI body mass index; Dlco diffusion capacity of the lung for carbon monoxide; M male; F female.

min (95% CI, 5.0 to 9.0 min); 8-W step protocol, 6.0 min (95% CI, 4.0 to 8.0 min); and 4-W ramp protocol, 8.7 min (95% CI, 4.4 to 13.0 min). Thirty-three of the 44 tests (75%) lasted 8 min. Patients undergoing tests that lasted 8 to 12 min (recommended duration) did not have a higher o2peak compared to those in studies that lasted V 8 min (difference not significant) in all participants (Fig 1). Notably, 7 of 11 subjects (64%) were unable to sustain exercise beyond 8 min despite the lowest possible increment technically available using our exercise system (4 W/min). Maximal Tolerated Workload (Peak Watts) The maximal tolerated workload increased as the ramp rate increased with similar variability (4-W protocol, 35.6 W/min; 8-W protocol, 50.7 W/min;

16-W protocol, 64.3 W/min) [Table 2, Fig 2]. In contrast, there were no differences in the mean ( SD) maximum tolerated workload between the 8-W ramp protocol and the 8-W step protocol (50.7 23.1 vs 50.1 24.4 W/min; difference was not significant). Other Outcomes Measurements co2, VE, respiratory rate, heart rate, o2 (Fig 3) V V Ve/maximal voluntary ventilation (MVV) ratio, and Borg ratings of dyspnea and leg fatigue at maximal exercise were not different among subjects undergoing the four protocols (difference is not significant) [Table 2]. Comparing the 16-W ramp protocol to the o2, 4-W ramp protocol at submaximal values, the V Vco2, Ve, respiratory rate, heart rate, and Borg scores (dyspnea and fatigue) were shifted to the right

Duration of CPET and V O2peak (%pred) 120 100 80 VO2peak 60 40 20 0 0 4 8 12 16 CPET Dur ation (minutes ) 20 24

p at ient 1 p at ient 2 p at ient 3 p at ient 4 p at ient 5 p at ient 6 p at ient 7 p at ient 8 p at ient 9 p at ient 10 p at ient 11

o2peak percent predicted and the duration of the maximal exercise Figure 1. Relationship between V test (incremental time only); 75% of the tests lasted 8 min. Patient 3 achieved 100% of the predicted o2 due to the underestimation of predicted values in individuals with extreme anthropometric V characteristics (Table 1).
1502
Original Research

Downloaded from chestjournal.org on March 5, 2008 Copyright 2007 by American College of Chest Physicians

Table 2Peak Exercise Measures for Four Different Exercise Protocols

Measure Maximal load o2 V mL/min mL/kg/min Vco2, mL/min e, L/min V Respiratory rate, breaths/min e /MVV ratio V Heart rate, beats/min Exercise time, min:s 4 W/min Ramp Speed 35.6 26.1 866.6 251.1 12.3 2.8 792.3 268.3 32.7 9.8 33.4 5.4 1.01 0.22 123.3 14.9 8:46 6:23 8 W/min Ramp Speed 50.7 23.1 885.8 233.3 12.4 2.4 810.8 236.6 32.7 9.3 32.4 6.5 0.99 0.17 122.0 10.4 6:40 3:00 8 W/min Step Speed 50.1 24.4 885.4 251.0 12.3 2.6 823.6 258.0 33.1 10.2 31.6 6.0 0.99 0.15 122.7 11.7 5:52 3:02 16 W/min Ramp Speed 64.3 25.5 909.2 198.5 12.5 2.3 840.6 234.8 33.7 6.2 33.4 9.5 1.00 0.14 120.0 10.9 4:04 1:35

Values are given as the mean SD. Maximal load maximal exercise capacity measured on cycle ergometer. p 0.05 (by repeated measures analysis of variance): 4-W vs 8-W ramp protocol, p 0.01; 4-W vs 16-W ramp protocol, p 0.01; 8-W vs 16-W ramp protocol, p 0.01; 8-W ramp protocol vs 8-W step protocol, difference was not significant.

and were significantly lower (p 0.05) for any given workload (Fig 3). There were no such differences found in any parameter between the ramp and step protocols. Discussion o2peak, in a group of patients with In this study, V severe COPD, was achieved in a shorter time frame than that proposed in the current exercise guidelines.10 The mean duration of CPET in our study based on 44 carefully executed tests was 6.3 min (95% CI, 5.0 to 9.0

min) [Fig 1]. This is shorter than the recommended duration of 8 to 12 min for CPET10 that was based on the only available data from a group of five healthy subjects.11 We found no significant differences in the o2peak achieved by the four protocols (Fig 3, TaV ble 2) nor were there any significant differences in o2peak between tests that lasted 8 min compared V to shorter tests. In fact, even at the minimal ramp rate of 4 W, in 7 of these 11 subjects the exercise duration was 8 min. This is the first study that has addressed CPET duration in patients with severe COPD. In addition, we found that protocols involving

Peak Watts attained per protocol

120

4 watts 8 watts
100

16 watts

80

Watts

60

40

20

0 1 2 3 4 5 6 7 8 9 10 11

Subjects
Figure 2. Maximal tolerated workload (in watts) attained in subjects using each protocol.
www.chestjournal.org CHEST / 132 / 5 / NOVEMBER, 2007

1503

Downloaded from chestjournal.org on March 5, 2008 Copyright 2007 by American College of Chest Physicians

1300

140

1000

VO2 ml/min

120

Heart Rate

700

100

400

80

50

60 70 80 90 % highest W

100
50

50

60 70 80 90 % highest W

100

50

40

Minute Ventilation

Respiratory Rate
40

30

30

20

20

10

50

60 70 80 90 % highest W

100

10

50

60 70 80 90 % highest W

100

Figure 3. Effect of ramp rate on submaximal exercise parame o 2, V e, heart rate, and respiratory ters at iso-watt workloads. V rate are lower during higher work-rate increments (dotted line 16 W/min) compared to lower work-rate increments (solid line 4 W/min) at every iso-workload.

different ramping increments will produce clinically important differences in peak watts while not differ o2peak. Specifically, higher incrementation ing in V protocols achieve higher workloads. While external workload (ie, maximal watts) is a useful measurement as a predictor of mortality in COPD19 and an important discriminator of the response to lung volume reduction surgery,2 our results suggest that an identical methodology must be followed when o2peak is a workload results are compared and that V more reliable parameter with which to quantify exercise tolerance. The analysis of the maximal response indicates for e, heart rate, and symptom scores were o2, V V independent of the exercise protocol. Because the e, and intramuscular adjustment in cardiac output, V o2, O2 uptake is not instantaneous, the responses in V Ve, and heart rate during exercise lag behind the increase in the work rate. This phenomenon has been well reported in healthy subjects and COPD subjects.20 The magnitude of the difference between the observed values and the theoretical values for these parameters is directly proportional to the increment in work rate and time constant of the response.21 o2, V e, The latter explains that for a given workload, V heart rate, and symptoms scores become lower as the incremental rate increases (Fig 3). Our findings confirm previous reports22,23 but extend that work by improving gender distribution and by demonstrating that a greater time for un1504

loaded pedaling (2 min) does not abrogate the discrepancy in maximal watts. A further contribution of our work involves the comparison of the effects of a ramp vs a step protocol on exercise performance. It is conceivable that the application of intermittent abrupt step increases may be perceived by the patient more than a continuous ramp incrementation and thus impact exercise performance. The 8-W ramp and step protocols were comparable with respect to both maximal and submaximal exercise parameters. Two studies in nonCOPD populations also found no difference in the maximal work rate achieved between ramp and step protocols.24,25 This study has limitations. First, COPD is a very heterogeneous disease, and our modest sample size precludes a determination of variation in our findings between disease subtypes. For example, our population represents predominantly ventilatory e/MVV ratio, 0.8) who limited individuals (mean V may behave differently than peripheral muscle fatigue-limited subjects. Further, we make no attempt to define the impact of protocol variations using other important modalities such as treadmill testing, as our hypotheses surrounded variations in cycle ergometry methodology. Cycle ergometry is the standard modality used in most pulmonary exercise laboratories due to its ability to provide true quantification of workload incrementation and the ability to linearly increase workload. Further studies to understand the implications of methodological variations in treadmill testing are indeed warranted. Based on our findings in patients with severe, nondesaturating COPD (Global Initiative for Chronic Obstructive Lung Disease stage IIIIV), our main conclusion is that a targeted duration for CPET of between 5 o2peak. We and 9 min is appropriate to achieve V further confirm that higher incrementation protocols achieve higher maximal workloads while having no o2peak. We believe that these results may impact on V have important implications for the clinical application of CPET as it is commonly applied to patients with severe COPD. References
1 Troosters T, Casaburi R, Gosselink R, et al. Pulmonary rehabilitation in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2005; 172:19 38 2 Fishman A, Martinez F, Naunheim K, et al. A randomized trial comparing lung-volume-reduction surgery with medical therapy for severe emphysema. N Engl J Med 2003; 348: 2059 2073 3 Epstein SK, Faling LJ, Daly BD, et al. Predicting complications after pulmonary resection: preoperative exercise testing vs a multifactorial cardiopulmonary risk index. Chest 1993; 104:694 700
Original Research

Downloaded from chestjournal.org on March 5, 2008 Copyright 2007 by American College of Chest Physicians

4 Benzo R, Kelley GA, Recchi L, et al. Complications of lung resection and exercise capacity: a meta-analysis. Respir Med 2007; 101:1790 1797 5 Bolliger CT, Jordan P, Soler M, et al. Exercise capacity as a predictor of postoperative complications in lung resection candidates. Am J Respir Crit Care Med 1995; 151:14721480 6 Brutsche MH, Spiliopoulos A, Bolliger CT, et al. Exercise capacity and extent of resection as predictors of surgical risk in lung cancer. Eur Respir J 2000; 15:828 832 7 Richter Larsen K, Svendsen UG, Milman N, et al. Exercise testing in the preoperative evaluation of patients with bronchogenic carcinoma. Eur Respir J 1997; 10:1559 1565 8 Win T, Jackson A, Sharples L, et al. Cardiopulmonary exercise tests and lung cancer surgical outcome. Chest 2005; 127:1159 1165 9 Butler J, Khadim G, Paul KM, et al. Selection of patients for heart transplantation in the current era of heart failure therapy. J Am Coll Cardiol 2004; 43:787793 10 American Thoracic Society, American College of Chest Physicians. ATS/ACCP statement on cardiopulmonary exercise testing. Am J Respir Crit Care Med 2003; 167:211277 11 Buchfuhrer MJ, Hansen JE, Robinson TE, et al. Optimizing the exercise protocol for cardiopulmonary assessment. J Appl Physiol 1983; 55:1558 1564 12 Montes de Oca M, Celli BR. Peripheral muscles in COPD: deconditioning or myopathy? Arch Bronconeumol 2001; 37: 82 87 13 Spruit MA, Gosselink R, Troosters T, et al. Resistance versus endurance training in patients with COPD and peripheral muscle weakness. Eur Respir J 2002; 19:10721078 14 OShea SD, Taylor NF, Paratz J. Peripheral muscle strength training in COPD: a systematic review. Chest 2004; 126:903 914 15 Neder JA, Jones PW, Nery LE, et al. Determinants of the

16 17 18 19 20 21 22 23

24 25

exercise endurance capacity in patients with chronic obstructive pulmonary disease: the power-duration relationship. Am J Respir Crit Care Med 2000; 162:497504 Hankinson JL, Odencrantz JR, Fedan KB. Spirometric reference values from a sample of the general US population. Am J Respir Crit Care Med 1999; 159:179 187 Crapo RO, Morris AH. Standardized single breath normal values for carbon monoxide diffusing capacity. Am Rev Respir Dis 1981; 123:185189 Jones NL, Makrides L, Hitchcock C, et al. Normal standards for an incremental progressive cycle ergometer test. Am Rev Respir Dis 1985; 131:700 708 Martinez FJ, Foster G, Curtis JL, et al. Predictors of mortality in patients with emphysema and severe airflow obstruction. Am J Respir Crit Care Med 2006; 173:1326 1334 Nery LE, Wasserman K, Andrews JD, et al. Ventilatory and gas exchange kinetics during exercise in chronic airways obstruction. J Appl Physiol 1982; 53:1594 1602 Whipp BJ, Davis JA, Torres F, et al. A test to determine parameters of aerobic function during exercise. J Appl Physiol 1981; 50:217221 Debigare R, Maltais F, Mallet M, et al. Influence of work rate incremental rate on the exercise responses in patients with COPD. Med Sci Sports Exerc 2000; 32:13651368 Miyahara N, Eda R, Takeyama H, et al. Cardiorespiratory responses during cycle ergometer exercise with different ramp slope increments in patients with chronic obstructive pulmonary disease. Intern Med 2000; 39:1519 Bader DS, Maguire TE, Balady GJ. Comparison of ramp versus step protocols for exercise testing in patients or 60 years of age. Am J Cardiol 1999; 83:1114 Schmid JP, Gaillet R, Noveanu M, et al. Influence of the exercise protocol on peak VO2 in patients after heart transplantation. J Heart Lung Transplant 2005; 24:17511756

www.chestjournal.org

CHEST / 132 / 5 / NOVEMBER, 2007

1505

Downloaded from chestjournal.org on March 5, 2008 Copyright 2007 by American College of Chest Physicians

Optimal Protocol Selection for Cardiopulmonary Exercise Testing in Severe COPD Roberto P. Benzo, Sriram Paramesh, Sanjay A. Patel, William A. Slivka and Frank C. Sciurba Chest 2007;132;1500-1505; Prepublished online October 9, 2007; DOI 10.1378/chest.07-0732 This information is current as of March 5, 2008
Updated Information & Services References Updated information and services, including high-resolution figures, can be found at: http://chestjournal.org/cgi/content/full/132/5/1500 This article cites 25 articles, 17 of which you can access for free at: http://chestjournal.org/cgi/content/full/132/5/1500#BIBL Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://chestjournal.org/misc/reprints.shtml Information about ordering reprints can be found online: http://chestjournal.org/misc/reprints.shtml Receive free email alerts when new articles cite this article sign up in the box at the top right corner of the online article.

Permissions & Licensing

Reprints Email alerting service

Images in PowerPoint format Figures that appear in CHEST articles can be downloaded for teaching purposes in PowerPoint slide format. See any online article figure for directions.

Downloaded from chestjournal.org on March 5, 2008 Copyright 2007 by American College of Chest Physicians