HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:
Guest Access | Sign In via User Name/Password

This Article Abstract Full Text (PDF) Free Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Article Archive Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (18) Citing Articles via Google Scholar Google Scholar Articles by Rodger, M. A. Articles by Wells, P. S. Search for Related Content PubMed PubMed Citation Articles by Rodger, M. A. Articles by Wells, P. S.

Steady-State End-Tidal Alveolar Dead Space Fraction and D-Dimer^*

Bedside Tests To Exclude Pulmonary Embolism

^* From the Departments of Medicine (Drs. Rodger, Jones, Raymond, and Bredeson) and Radiology (Dr. Rasuli), University of Ottawa, Ottawa; the Department of Respiratory Therapy (Ms. Djunaedi), Ottawa Hospital - General Campus, Ottawa; and Clinical Epidemiology Unit (Dr. Wells), Ottawa Hospital Loeb Research Institute, Ottawa, Ontario, Canada.

Correspondence to: Marc A. Rodger, MD, Suite 7205, Ottawa Hospital - General Campus, 501 Smyth Rd, Ottawa, Ontario, K1H 8L6, Canada; e-mail: mrodger{at}ottawahospital.on.ca

	Abstract

TOP Abstract Introduction Materials and Methods Results Discussion References

 
Study objective: Less than 35% of patients suspected of having pulmonary embolism (PE) actually have PE. Safe bedside methods to exclude PE could save health-care resources and improve access to diagnostic testing for suspected PE. In patients with suspected PE, we sought to determine the sensitivity, specificity, and negative predictive value of (1) a steady-state end-tidal alveolar dead space fraction (AVDSf) of < 0.15, (2) a negative D-dimer result, and (3) the combination of a steady-state end-tidal AVDSf of < 0.15 and a negative D-dimer result.

Study design: Prospective cohort study.

Setting: Tertiary-care center in Ottawa, Ontario, Canada.

Patients: Consecutive inpatients, outpatients, and emergency department patients with suspected PE referred to the Departments of Nuclear Medicine or Radiology for investigation of suspected PE.

Interventions and measurements: All study patients had D-Dimer and alveolar dead space measurements prior to determining outcome (PE or no PE) with ventilation/perfusion scans and/or noninvasive leg vein imaging and/or pulmonary angiography.

Results: Two hundred forty-six eligible and consenting patients underwent diagnostic imaging that excluded PE in 163 patients, diagnosed PE in 49 patients, and was indeterminant in 34 patients. A negative D-dimer result excluded PE with a sensitivity of 83.0% (95% confidence interval [CI], 69.2 to 92.4%), a negative predictive value of 91.2% (95% CI, 83.4 to 96.1%), and a specificity of 57.6%. A steady-state end-tidal AVDSf of < 0.15 excluded PE with a sensitivity of 79.5% (95% CI, 63.5 to 90.7%), a negative predictive value of 90.7% (95% CI, 82.5 to 95.9%), and a specificity of 70.3%. The combination of a negative D-dimer result and a steady-state end-tidal AVDSf of < 0.15 excluded PE with a sensitivity of 97.8% (95% CI, 88.5 to 99.9%), a negative predictive value of 98.0% (95% CI, 89.4 to 99.9%), and a specificity of 38.0%.

Conclusion: This simple combination of bedside tests may safely rule out PE without further diagnostic testing in large numbers of patients with suspected PE.

Key Words: alveolar dead space • D-dimer • diagnosis • pulmonary embolism • reproducibility

	Introduction

TOP Abstract Introduction Materials and Methods Results Discussion References

 
Pulmonary embolism (PE), the third-leading cause of cardiovascular mortality in North America, has an estimated annual incidence of 23 cases per 100,000 population per year.¹ Untreated PE has a hospital mortality rate as high as 30%.² This mortality falls to 8% if PE is appropriately diagnosed and treated.²

The diagnosis of PE remains one of the most difficult problems confronting clinicians. PE is considered in the differential diagnosis of many clinical presentations, including chest pain, hemoptysis, or dyspnea, and in a wide variety of clinical settings, such as emergency departments, obstetrical units, surgical wards, and ICUs. However, < 35% of patients suspected of having PE are found to have PE.³ This leads to presumptive anticoagulation, hospital admission, and testing for a large number of patients without PE.

A bedside method that safely excludes PE would be desirable. For such a bedside method to be accepted into clinical practice, it must be near 100% sensitive (ie, very few false-negative findings). For a bedside method to be clinically useful, it must exclude a large proportion of the patients who do not have the disease (ie, it should have many true-negative findings).

Dead space ventilation represents ventilation of those parts of the lung not involved in gas exchange. Dead space ventilation has two components: anatomic dead space and alveolar dead space. Anatomic dead space ventilation represents ventilation of the airways. The airways conduct air to the alveoli and are not involved in gas exchange. Alveolar dead space represents ventilation of those alveoli that are not involved in gas exchange (ie, alveoli that are unperfused or poorly perfused, as occurs in PE). Kline et al⁴ demonstrated the high negative predictive value of alveolar dead space measurement combined with a negative D-dimer result in ambulatory patients with suspected PE.⁴ This combination of bedside tests has only been tested in an ambulatory patient population and has yet to be validated. We sought to prospectively determine, in consecutive ambulatory and hospitalized patients with suspected PE, the sensitivity, specificity, and negative predictive value of (1) a steady-state end-tidal alveolar dead space fraction (AVDSf) of < 0.15, (2) a negative D-dimer result, and (3) the combination of both negative test results.

	Materials and Methods

TOP Abstract Introduction Materials and Methods Results Discussion References

 
Patients
All inpatients and outpatients at the Ottawa Hospital - General Campus suspected of having PE and referred for a ventilation/perfusion (/) scan or pulmonary angiogram from January 1996 to August 1998 were approached for consent to participate in the study. Patients were excluded from the study if they (1) were < 18 years of age, (2) were unable to give informed consent, (3) had a contraindication to pulmonary angiography, (4) were receiving mechanical ventilation, or (5) were in the final stages of terminal illness. Participating patients signed informed consent approved by our hospital ethics review board.

Standardized Patient Assessment
The referring physician first assigned an index of clinical suspicion of PE (ie, pretest percentage likelihood of PE) based on clinical gestalt. The respiratory therapist (RT) performed arterial blood gas analysis and alveolar dead space analysis. Two methods of end-tidal alveolar dead space measurement were used over the course of the study. Both methods were taught in separate 2-h training sessions to 40 RTs in our institution. From January 1996 to February 1998, the sitting patient breathed through a one-way valve (Hans Rudolph, Inc.; Kansas City, MO). A sidestream port on the valve was connected to a capnograph (Datex CD-102–27-00; Datex; Helsinki, Finland). The capnograph provided continuous expired CO₂ vs time data. The capnograph was calibrated with room air (0.3% CO₂) and 4.0% CO₂ prior to each use. Once the patient was judged to be relaxed and breathing regularly by the RT, an arterial puncture was performed for arterial blood gas analysis. The blood gas was analyzed with an analyzer (Ciba-Corning 278; Ciba-Corning; San Diego, CA) that was calibrated twice daily with standard gases. The calibrations of the capnograph and the blood gas analyzer were cross-checked with a sample of expired gas. If a difference in the PCO₂ measurements with the capnograph and the blood gas analyzer was > 2 mm Hg, the patient’s dead space data were excluded from the study because small discrepancies in PCO₂ can result in large differences in the AVDSf (see below). The steady-state end-tidal AVDSf was calculated as follows:

where PETCO₂ = end-tidal carbon dioxide pressure concentration in last 10 breaths.

Because we were concerned with the importance of (1) patient stability at the time of arterial and end-tidal CO₂ sampling, (2) late-emptying alveoli (with higher PCO₂) being overrepresented with end-tidal PCO₂ measurement with time-based capnography, and (3) the accuracy of our end-tidal CO₂ monitor, we modified our methods and equipment. From February 1998 to August 1998, the sitting patient breathed through an airway adapter attached to a mouthpiece. The airway adapter had a mainstream CO₂ and volume sensor (CapnoSTAT and COSMO+; Novametrix Medical Systems; New Haven, CT). This device measures breath-by-breath CO₂ vs volume with an accuracy of ± 50 mL and CO₂ with an accuracy of 2 mm Hg. The calibration of the capnograph was verified prior to each use with a known gas (4% CO₂). Once the patient’s condition had stabilized (respiratory rate of plus or minus two breaths per minute over 2 min), the RT recorded this as the stable respiratory rate.

With the patient breathing at the stable respiratory rate, the end-tidal CO₂ was recorded if it was stable (± 1 mm Hg over 2 min). Arterial blood gas was obtained by a single arterial blood gas puncture only if the patient was breathing at the stable respiratory rate and had stable end-tidal PCO₂. Subsequently, the RT calculated the steady-state end-tidal AVDSf as follows:

We have shown that this technique has excellent reproducibility (unpublished data submitted for publication). The dead space measurements were not provided to the referring physician or to the interpreting nuclear medicine physician.

D-Dimer
A latex D-dimer (Accuclot; Sigma Diagnostics; St. Louis, MO) or a whole-blood agglutination D-dimer (SimpliRED; AGEN Biomedical Limited; Brisbane, Australia) was performed on venous blood within 24 h of the / scan or pulmonary angiogram.

Outcome Measure
The presence or absence of PE was determined independent of the standardized patient assessments. After the / scan, patients with a posttest probability of PE of < 5% were considered, for study purposes, not to have PE. This group was defined by (1) patients with normal or near-normal / scan results, (2) patients with a low index of pretest clinical suspicion who had low probability / scan results,³ and (3) patients with low-probability scan results with a negative leg vein ultrasound result at presentation.⁵ After / scan, patients with a posttest probability of 88% were considered for study purposes to have PE. This was defined as patients with a high or intermediate index of pretest clinical suspicion who had high-probability / scan results.³ We recommended that other patients proceed to pulmonary angiography, but this decision was left to the patient’s treating physician. Those patients with indeterminant scan results who did not undergo angiography were excluded from the analysis.

Analysis
Sensitivity, specificity, negative predictive were calculated with 95% exact binomial confidence limits.

	Results

TOP Abstract Introduction Materials and Methods Results Discussion References

 
We approached 293 patients, of whom 282 were eligible for participation in the study. Of these 282 patients, 246 consented. Of the 246 consenting patients, 49 patients (19.9%) had PE, 163 patients (66.3%) did not have PE, and 34 patients (13.8%) could not be classified with "gold standard" outcome measures (Table 1 ). Female patients were less likely to have PE. Patients with PE were significantly older (mean age, 58.9 years) than patients without PE (mean age, 50.6 years). More than 94% of our patients presented with chest pain or dyspnea.

	Discussion

TOP Abstract Introduction Materials and Methods Results Discussion References

The diagnosis and exclusion of PE remain problematic. The diagnostic "gold standard" is pulmonary angiography. Pulmonary angiography is an invasive and expensive procedure, with limited availability and potentially serious complications. Despite being the "gold standard," pulmonary angiograms are imperfect. A patient with a normal pulmonary angiogram can still expect a 2.2% (95% CI, 0.3 to 8.0%) venous thromboembolic event rate at 1-year follow-up.⁷ / scan results provide a definitive diagnosis in < 40% of cases.³ Furthermore, a normal/near-normal / scan result has a negative predictive value of only 96.4% (95% CI, 91.1 to 97.7%).³ Finally, more recent experience has shown that a patient with a normal / scan result can expect a 1.2% (95% CI, 0.3 to 3.0%) chance of having a venous thromboembolic event (PE or deep vein thrombosis) diagnosed over a 3-month follow-up period.⁵

Kline et al⁴ recently demonstrated the high negative predictive value of end-tidal alveolar dead space measurement combined with a negative D-dimer result in ambulatory patients with suspected PE. The derivation set of Kline et al⁴ only included ambulatory patients, thereby limiting generalizability.

Clinical decision rules often do not perform as well in validation studies as they do in derivation studies.⁸ This may be due to differences in surveillance strategies and definitions of outcome between the original studies and the validation studies.⁸ It is plausible that our bedside techniques to exclude PE will not perform as well in validation studies. This possibility is minimized by the fact that we utilized validated reference standards as outcome measures and that we developed the bedside techniques in consecutive inpatients and outpatients suspected of having PE. Furthermore, we have shown that with minimal training (2 h), a large number of RTs can be taught our technique with excellent reproducibility (unpublished data submitted for publication). We caution, however, that care must be taken to ensure that the equipment (blood gas analyzer and capnograph) is properly calibrated prior to each use, as small errors in PaCO₂ measurement may result in large differences in the steady-state end-tidal AVDSf. We also recommend using our new technique of alveolar dead space measurement, given that we have demonstrated its reproducibility in a "real world" setting (unpublished data submitted for publication).

The "excluded proportion" reflects the proportion of patients suspected of having PE who are safely excluded. The excluded proportion is a measure of the proportion of patients who can be excluded without further testing. Hence, the excluded proportion reflects the clinical and economic impact a bedside technique will have once validated and adopted into clinical practice. An excluded proportion of > 25% for our study represents significant numbers of patients who could be safely excluded at the bedside without further testing if our methods are validated and adopted.

There are limitations in our study. First, the CIs for sensitivity and negative predictive value are too wide to allow for widespread adoption of these bedside methods without further validation. Secondly, ideally, all patients would have had the reference standard (pulmonary angiography or a normal / scan result). Unfortunately, in clinical practice, it is not uncommon that clinicians do not pursue pulmonary angiography or other tests in patients with indeterminant / scan results.⁶ The reluctance of clinicians, even in the setting of a clinical trial (as we experienced), to use the "gold standard" (pulmonary angiography) necessitates that investigators must use imperfect outcome measures. Even if a group of investigators were able to convince clinicians to obtain pulmonary angiograms on all patients with suspected PE, as was done in the Prospective Investigation of Pulmonary Embolism Diagnosis study,³ it is likely that a highly selected study group would remain, as occurred in that study (1,493 patients consenting out of 3,016 eligible patients). This degree of selection almost certainly results in a biased study sample. In our study, we attempted to minimize selection bias and maximize the accuracy of diagnosis by utilizing the reference standard where possible (normal / scan result and pulmonary angiograms) and otherwise using validated combinations of / scan results and pretest probability or leg vein imaging. All imaging was interpreted by physicians unaware of patient presentation and outcomes, minimizing interpretation bias. Thirdly, not all patient data sets were complete, including 34 patients who were unclassified. The result of incomplete data sets may be that the study population is not representative of the total population of patients suspected of having PE. Reassurance that the unclassified and classified patients are comparable is given by the fact that these groups were not significantly different in gender or age. However, other important differences between these groups may exist. The fact that 14% of our population, in the context of a study, were unclassified provides further evidence that alternate diagnostic approaches, more acceptable to clinicians and patients, must be developed. Finally, it should be emphasized that this combination of tests is useful in excluding PE, not diagnosing PE. These tests individually and in combination have insufficient positive predictive value to diagnose PE.

If validated and adopted, the application of these bedside tests may safely exclude PE. Excluding PE at the bedside with these tests would eliminate the need for unnecessary presumptive treatment and further diagnostic testing in a large proportion of patients with suspected PE. This would save health-care resources and avoid inconveniencing and placing these patients at risk from presumptive therapy and further diagnostic testing.

	Acknowledgements

 
We are grateful to Dr. Alan Karovitch and the clinicians who completed data collection forms, including Dr. Jim Quinn and Dr. Mark Reardon. Many thanks to Dr. Francois Raymond and the staff of the Nuclear Medicine Department for their cooperation, Denise Blanchette and the RTs who provided valuable input to the alveolar dead space analysis, Jane Browing and Ted Tabort from Novametrix for valuable input and support with alveolar dead space measurement, and Michelle Willson for assistance in preparing this article.

	Footnotes

 
Abbreviations: AVDSf = alveolar dead space fraction; CI = confidence interval; PE = pulmonary embolism; PETCO₂ = end-tidal carbon dioxide pressure; RT = respiratory therapist; / = ventilation/perfusion

This study was conducted at the Ottawa Hospital - General Campus.

Financial support was provided by the Ottawa General Hospital

Research Fund and the Clinical Trials and Research Unit of the Division of Hematology, Ottawa Hospital- General Campus.

Dr. Rodger was the recipient of the Thrombosis Interest Group of Canada Research Fellowship. Dr. Wells was the recipient of a Research Scholarship from the Heart and Stroke Foundation of Canada.

	References

TOP Abstract Introduction Materials and Methods Results Discussion References

 

1. Anderson, FA, Wheeler, HB, Goldberg, RJ, et al (1991) A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. Arch Intern Med 151,933-938[Abstract]
2. Carson, JL, Kelley, MA, Duff, A, et al (1992) The clinical course of pulmonary embolism. N Engl J Med 326,1240-1245[Abstract]
3. . PIOPED Investigators. (1990) Value of the ventilation/perfusion scan in acute pulmonary embolism: results of the Prospective Investigation of Pulmonary Embolism Diagnosis study (PIOPED). JAMA 263,2753-2759[Abstract]
4. Kline, JA, Meek, S, Boudrow, D, et al (1997) Use of the alveolar dead space fraction (Vd/Vt) and plasma D-dimers to exclude acute pulmonary embolism in ambulatory patients. Acad Emerg Med 4,856-863[ISI][Medline]
5. Wells, PS, Ginsberg, JS, Anderson, DR, et al (1998) Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med 129,997-1005[Abstract/Free Full Text]
6. Schluger, N, Henschke, C, King, T, et al (1994) Diagnosis of pulmonary embolism at a large teaching hospital. J Thorac Imaging 9,180-184[Medline]
7. Hull, RB, Hirsh, J, Carter, C, et al (1993) Pulmonary angiography, ventilation lung scanning, and venography for clinically suspected pulmonary embolism with abnormal perfusion lung scan. Ann Intern Med 98,891-899
8. Charlson, ME, Ales, KL, Simon, R, et al (1987) Why predictive indexes perform less well in validation studies. Arch Intern Med 147,2155-2161[Abstract]

This article has been cited by other articles:

				I. Tillie-Leblond, C.-H. Marquette, T. Perez, A. Scherpereel, C. Zanetti, A.-B. Tonnel, and M. Remy-Jardin Pulmonary embolism in patients with unexplained exacerbation of chronic obstructive pulmonary disease: prevalence and risk factors. Ann Intern Med, March 21, 2006; 144(6): 390 - 396. [Abstract] [Full Text] [PDF]

				M. A. Rodger, C. N. Bredeson, G. Jones, P. Rasuli, F. Raymond, A. M. Clement, A. Karovitch, H. Brunette, D. Makropoulos, M. Reardon, et al. The Bedside Investigation of Pulmonary Embolism Diagnosis Study: A Double-blind Randomized Controlled Trial Comparing Combinations of 3 Bedside Tests vs Ventilation-Perfusion Scan for the Initial Investigation of Suspected Pulmonary Embolism Arch Intern Med, January 23, 2006; 166(2): 181 - 187. [Abstract] [Full Text] [PDF]

				K. Hogg, D. Dawson, T. Tabor, B. Tabor, and K. Mackway-Jones Respiratory Dead Space Measurement in the Investigation of Pulmonary Embolism in Outpatients With Pleuritic Chest Pain Chest, October 1, 2005; 128(4): 2195 - 2202. [Abstract] [Full Text] [PDF]

				P Leman, D Guthrie, R Simpson, and F Little Improving access to diagnostics: an evaluation of a satellite laboratory service in the emergency department Emerg. Med. J., July 1, 2004; 21(4): 452 - 456. [Abstract] [Full Text] [PDF]

				F. Verschuren, G. Liistro, R. Coffeng, F. Thys, J. Roeseler, F. Zech, and M. Reynaert Volumetric Capnography as a Screening Test for Pulmonary Embolism in the Emergency Department Chest, March 1, 2004; 125(3): 841 - 850. [Abstract] [Full Text] [PDF]

				M. J.H.A. Kruip, M. G.L. Leclercq, C. v. d. Heul, M. H. Prins, and H. R. Buller Diagnostic Strategies for Excluding Pulmonary Embolism in Clinical Outcome Studies: A Systematic Review Ann Intern Med, June 17, 2003; 138(12): 941 - 951. [Abstract] [Full Text] [PDF]

				C. Kearon Diagnosis of pulmonary embolism Can. Med. Assoc. J., January 21, 2003; 168(2): 183 - 194. [Abstract] [Full Text] [PDF]

				F. Verschuren, F. Thys, and M. Rodger Steady-State End-Tidal Alveolar Dead Space Measure and D-dimer Chest, April 1, 2002; 121(4): 1373 - 1373. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Free Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Article Archive Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (18) Citing Articles via Google Scholar Google Scholar Articles by Rodger, M. A. Articles by Wells, P. S. Search for Related Content PubMed PubMed Citation Articles by Rodger, M. A. Articles by Wells, P. S.