• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Respir Med. Author manuscript; available in PMC Nov 27, 2007.
Published in final edited form as:
PMCID: PMC2096631
NIHMSID: NIHMS33916

Prediction of pulmonary hypertension in idiopathic pulmonary fibrosis[star]

Summary

Background

Reliable, noninvasive approaches to the diagnosis of pulmonary hypertension in idiopathic pulmonary fibrosis are needed. We tested the hypothesis that the forced vital capacity to diffusing capacity ratio and room air resting pulse oximetry may be combined to predict mean pulmonary artery pressure (MPAP) in idiopathic pulmonary fibrosis.

Methods

Sixty-one idiopathic pulmonary fibrosis patients with available right-heart catheterization were studied. We regressed measured MPAP as a continuous variable on pulse oximetry (SpO2) and percent predicted forced vital capacity (FVC) to percent-predicted diffusing capacity ratio (% FVC/% DLco) in a multivariable linear regression model.

Results

Linear regression generated the following equation: MPAP = −11.9+0.272 × SpO2+0.0659 × (100−SpO2)2+3.06 × (% FVC/% DLco); adjusted R2 = 0.55, p<0.0001. The sensitivity, specificity, positive predictive and negative predictive value of model-predicted pulmonary hypertension were 71% (95% confidence interval (CI): 50–89%), 81% (95% CI: 68–92%), 71% (95% CI: 51–87%) and 81% (95% CI: 68–94%).

Conclusions

A pulmonary hypertension predictor based on room air resting pulse oximetry and FVC to diffusing capacity ratio has a relatively high negative predictive value. However, this model will require external validation before it can be used in clinical practice.

Keywords: Pressure, Pulmonary artery, Hypertension, Pulmonary, Pulmonary fibrosis, Oximetry

Introduction

Pulmonary hypertension (PH) is common in patients with idiopathic pulmonary fibrosis (IPF) and its presence has a significant adverse impact on survival.1,2 Echocardiography is commonly used to diagnose PH; however, echocardiography is frequently inaccurate in patients with interstitial lung disease (ILD).3 Right-heart catheterization (RHC) is the gold standard for diagnosis of PH in IPF patients.3,4 However, it is an invasive method with significant risks for complications.

Noninvasive approaches to the diagnosis of PH in IPF would improve patient safety and reduce cost. The ability to predict which IPF patients have PH using noninvasive measures could guide the selection of patients for RHC to confirm its presence.

In ILD patients, the diffusing capacity for carbon monoxide (DLco) falls because of fibrosis, emphysema and pulmonary vascular disease.5,6 In these patients, a reduction in DLco out of proportion to the reduction in lung volumes might indicate underlying pulmonary vascular disease. For instance, in patients with scleroderma, when there is a mixture of both fibrosis and pulmonary vasculopathy, the forced vital capacity (FVC) is moderately decreased but the DLco is even lower and the FVC to DLco ratio is often greater than 1.8.712 We therefore hypothesized that a high FVC/DLco ratio might be a marker for increased pulmonary artery pressures in IPF.

Chronic hypoxia causes pulmonary vasoconstriction through a diversity of actions on pulmonary artery endothelium and smooth muscle cells, including downregulation of endothelial nitric oxide synthase and reduced production of the voltage-gated potassium channel alpha subunit.13,14 Although initially reversible, the pathologic changes induced by hypoxia-induced vasoconstriction ultimately result in irreversible vascular remodeling.15,16

The independent associations of FVC/DLco ratio and chronic hypoxia with PH in ILD patients, raise the possibility that these factors may be combined to improve the prediction of PH in IPF patients.

Methods

Study sample

We retrospectively reviewed the medical records of all IPF patients who were seen at our institution between July 1999 and June 2006. During their initial visit, all patients provided written informed consent to use their clinical and demographic information for research purposes. All patients met accepted diagnostic criteria for IPF and the majority (61%) had histopathologic evidence of usual interstitial pneumonia.17 Two hundred and ninety-eight IPF patients were candidates for inclusion in this study. To be included in the study, participants had to have had RHC and have pulmonary function test (PFT) and resting pulse oximetry data while breathing room air (SpO2) within 1 month of the RHC. Fifty five patients met this entry criterion. Six other patients had RHC and PFTs and were known to require supplemental oxygen; however, their actual SpO2 was not available. We included these six patients and used the imputed value of 85% for their SpO2. We used 85% because it was the mean SpO2 in the 15 patients requiring supplemental oxygen who had measurements of SpO2. Thus, our study sample consisted of 61 patients.

Measurements

RHC data included measurements of pulmonary arterial pressures with the patient at rest. We defined PH as a resting mean pulmonary artery pressure (MPAP) of >25 mmHg.18

SpO2 measurements were conducted in agreement with a clinical protocol: after at least 5 min of rest, SpO2 was measured on room air. All SpO2 measurements were done with the same oximeter (Digital Handheld Pulse Oximeter, Nonin Medical, Inc.).

Standard methodology was used for obtaining PFT, ABG, 6 MWD, and RVSP from Doppler echocardiography.1925

Statistical analysis

We compared patients in the study sample with those excluded with respect to the variables of interest, using standard tests for comparing means (Student's t-test), medians (Wilcoxon rank-sum test) and proportions (Chisquare test or the Fisher's exact test, if cell sizes are small).

We regressed the MPAP (obtained from RHC) as a continuous variable on SpO2 and % FVC/% DLco, both as continuous predictors in a multivariable linear regression model. Examination of the model residual indicated that the model underestimated MPAP in patients with high MPAP and overestimated MPAP in patients with low MPAP. We therefore added a quadratic term (100−SpO2)2 to the model. Given the moderately small sample size, no attempt was made to further refine the model, to avoid overly optimistic results.26

We compared this final model's prediction ability with alternate models that included other predictors or alternate predictors. MPAP prediction ability was assessed by model R2 in each case. Specifically, we examined the impact of adding the following variables to the model: age, gender, RVSP, 6MWD and % DLco/% VA. Further, we tried replacing SpO2 in the model by PaO2 and replacing % FVC/% DLco by DLco (absolute and % predicted).

The final model was internally validated using bootstrapping, which is superior to methods of validation by sample splitting.27,28 Using RHC-defined PH as the gold standard, we calculated the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of diagnosing PH based on model-predicted MPAP >25 mmHg in 1000 bootstrapped samples, and the empirical bootstrap distribution of each measure was used to determine the 95% confidence interval (CI) for that measure. Since some clinicians may be interested in a test with maximum sensitivity, so as to minimize the chance of missing a case of PH, we examined several alternate cut points for diagnosing PH from the model-predicted MPAP, using RHC-defined PH as the gold standard, and determined the cutoff that maximizes sensitivity with the least compromise in specificity.

We calculated the sensitivity, specificity, PPV and NPV of diagnosing PH based on echocardiography (RVSP>40 mmHg),29 using RHC-defined PH as the gold standard (MPAP>25 mmHg). We also determined the cutoff that maximizes sensitivity with the least compromise in specificity.

All tests were two-tailed, and p values of <0.05 were assumed to represent statistical significance. All statistical analyses were performed using SAS version 9.1 (SAS Institute Inc., Cary, NC, USA).

Results

The study sample (n = 61) was systematically different from the 237 patients who were excluded as shown in Table 1. Patients in the study sample were younger with more advanced IPF. As shown in Table 2, age, gender and FVC did not differ between those with or without PH. A trend towards lower absolute DLco, and higher % FVC/% DLco was seen in patients with PH. % DLco, SpO2 and PaO2 were significantly lower in those with PH. There were strong correlations in the expected directions between MPAP and the primary predictors (Table 3). There were also strong correlations in the expected directions between MPAP and other variables (Table 4): RVSP, PaO2, 6 MWD, % DLco/% VA and % DLco. There was no correlation between MPAP and FVC, FEV1, FEV1/FVC, TLC or the RV; data not shown.

Table 1
Descriptive statistics for major characteristics.
Table 2
Patient characteristics based on the presence or absence of PH by RHC.
Table 3
Pearson correlation coefficients between the mean pulmonary artery pressure (MPAP) of the primary predictors.
Table 4
Pearson correlation coefficients between the mean pulmonary artery pressure (MPAP) and other potential predictor variables.

Linear regression of MPAP on SpO2 (linear and quadratic terms) and % FVC/% DLco generated the following equation for MPAP in mmHg:

MPAP=11.9+0.272×SpO2+0.0659×(100SpO2)2+3.06×(%FVC/%DLCO).
(1)

The 95% CI for the model parameter estimates are listed in Table 5. This model explained 55% of the variance of MPAP (adjusted R2 = 0.55, p<0.0001). A scatter plot of predicted versus measured MPAP is shown in Fig. 1. Sequential and partial sums of squares associated with the primary predictors demonstrated that SpO2 provided the majority of the predictive information.

Figure 1
MPAP, mean pulmonary artery pressure; SpO2, resting room air pulse oximetry; % FVC/% DLco, percent predicted forced vital capacity/percent predicted diffusing capacity ratio.
Table 5
Parameter estimates for MPAP by candidate predictors.

We compared the prediction ability (using R2) of the model in Eq. (1) with alternate models. Regression parameter estimates for the primary predictors changed by <10% when potential confounders age and gender were added to the model (data not shown). Moreover, model fit did not improve with the addition of these demographic characteristics to the model (adjusted R2 changed from 0.55 to 0.53). We added RVSP, 6MWD and % DLco/% VA, as additional predictors to the model. Addition of these variables did not improve model fit. Further, we tried PaO2 in place of SpO2 and DLco (absolute and % predicted) in place of % FVC/% DLco. Neither improved model fit (data not shown).

To assess the impact of imputing 85% for SpO2 in the six patients missing the measurement, we added a 0/1 indicator variable that flagged those with imputed SpO2; the R2 did not change and the regression parameter estimate for the indicator variable was 1.5 (p = 0.6), indicating that MPAP in these patients was not biased by the imputation. We then replaced the measured SpO2 in everyone whose measured value was ≤88% and in everyone who was missing the measurement because he/she was on supplemental oxygen by a constant value of 88% in the model; the R2 fell to 0.29, implying that the extent of desaturation below 88% adds valuable information to the model.

The ability of the model to predict PH (using predicted MPAP>25 mmHg) in the study sample was examined (Table 6). To internally validate Eq. (1), we examined the sensitivity, specificity, PPV and NPV of the model-predicted PH in 1000 bootstrapped samples, and used the 2.5th and 97.5th percentile values of the resulting distributions to construct 95% CI. With predicted MPAP>25 mmHg as the definition of model-predicted PH, the sensitivity, specificity, PPV and NPV of model-predicted PH (together with bootstrapped 95% CI) were 71% (95% CI: 50–89%), 81% (95% CI: 68–92%), 71% (95% CI: 51–87%) and 81% (95% CI: 68–94%). We also examined the performance of the predictor with a more liberal definition of model-predicted PH: predicted MPAP>21 mmHg; sensitivity and NPV both increased to 100% and specificity and PPV fell to 40% and 52%, respectively.

Table 6
Performance of the model (compared to conventional echocardiogram) in establishing or excluding a diagnosis of pulmonary hypertension (PH), using right-heart catheterization (RHC) as the gold standard.

Estimation of RVSP by echocardiography was possible in 33 (54%) of the 61 patients. The sensitivity, specificity, PPV and NPV of echocardiography-diagnosed PH (RVSP>40 mmHg) were 76% (95% CI, 50–92%), 38% (95% CI, 16–64%), 56% (95% CI, 35–76%) and 60% (95% CI, 27–86%), respectively. We determined the cutoff that maximizes sensitivity with the least compromise in specificity. Using RVSP>37 mmHg, sensitivity and NPV increased to 93% and 75%, respectively, and specificity and PPV fell to 17% and 48%, respectively.

Discussion

We tested the hypothesis that the FVC/DLco ratio and SpO2 may be combined to predict MPAP in IPF patients. Sequential and partial sums of squares associated with the primary predictors demonstrated that SpO2 provided the majority of the predictive information. When we replaced measured SpO2 in everyone who required supplemental oxygen by a constant value of 88%, the R2 fell to 0.29, implying that the actual extent of desaturation below 88% adds valuable information to the model. In our study, DLco did not contribute to MPAP prediction above and beyond SpO2; however, the FVC/DLco did. Our finding that if not normalized by FVC, DLco adds no additional information beyond SpO2 is consistent with findings in patients with scleroderma and pulmonary fibrosis.712 Investigators examined DLco relative to lung volume in patients with scleroderma as a way of identifying patients with disproportionately low DLco who had PH. In their series, this ratio was greater than 1.4 in 70% of patients with PH.10 Similarly, in a study of patients with scleroderma who had moderate fibrosis with PH out of proportion to the degree of fibrosis,7 the FVC was moderately decreased but the DLco was even lower and the ratio was often greater than 1.8. We used FVC/DLco as a way of quantifying disproportionately low DLco and identifying patients with pulmonary vascular disease, and found that this ratio added valuable information to the prediction of MPAP.

The prediction model will require external validation before it can be used to guide clinical decision making regarding whether or not to proceed to RHC. The model's high NPV is not because of low prevalence of PH in our study sample. In fact, PH was present in 41% of our patients, a figure comparable to that reported by others.1,30 If externally validated, the model's high NPV could be used clinically to avoid RHC in selected patients (e.g., before a surgical lung biopsy, which is always safer in non-PH patients, or as exclusion criteria in clinical trials evaluating vasodilator therapy in patients with IPF). Using a predicted MPAP>25 mm as the cut point to select individuals for RHC would miss one in five cases with PH, a figure unacceptable to many clinicians; however, using a lower threshold such as predicted MPAP>21 mm to select individuals for RHC nearly eliminates the possibility of missing an individual with PH at the cost of having to do RHC in a few more individuals.

The performance characteristics of echo-based diagnosis of PH in a homogeneous sample of IPF patients have not been reported previously. Our data show that DE-estimated RVSP predicted PH in IPF patients with 76% sensitivity, 38% specificity, 56% PPV and 60% NPV. In a separate study, echocardiography predicted PH in patients with various ILD with 85% sensitivity, 17% specificity, 60% PPV and 44% NPV.3 When compared with echocardiography, our model's specificity and NPV are substantially better. More importantly, our model could allow the prediction of MPAP and PH in almost every IPF patient while echocardiographic estimation of RVSP is possible in only 44–54% of patients.3 Nevertheless, echocardiography provides additional important information (other than RVSP) such as right atrial size, right ventricular size and function, which may be pertinent to the identification of PH. The predictive ability of these measures is the topic of an ongoing study.

In a previous study, the need for supplemental oxygen together with a DLco<40% identified the presence of PH in IPF patients.1 In that study, PH was present in 31.6% of patients; however, the predicted prevalence of PH was 15.2%, suggesting that a prediction based on DLco alone and the need for supplemental oxygen (yes/no) would miss half the PH cases. By contrast, both the predicted and observed prevalence of PH in our study were 41%. In our study, DLco did not contribute to MPAP prediction above and beyond SpO2; however, the FVC/DLco ratio did. Furthermore, the extent of desaturation below 88% added valuable information to the model. Our study validates the notion that the need for supplemental oxygen together with a reduced DLco identifies the presence of PH in IPF patients, and it improves on it by using the extent of desaturation in place of need for oxygen (yes/no) and the FVC/DLco ratio in place of DLco to increase sensitivity and NPV.

There are limitations of our study. This was a retrospective review of patients evaluated at a single center. Most of our patients underwent evaluation for lung transplantation, reflecting the presence of younger patients with more advanced IPF. In clinical practice, RHC are almost exclusively performed as part of the pre-transplant evaluation in advanced IPF patients. Unfortunately, RHC data are not available from early IPF patients because there are no clinical indications to catheterize these patients. Prospective work including patients with the complete spectrum of disease severity will be required to avoid selection bias from studying only those with more advanced illness. However, this study and others have found that PH is more prevalent in patients with severe IPF defined by a reduced DLco1,2,31; hence, this population is the one in whom identification of PH is more critical. Since the RHC in this study was not done with exercise, we could not assess for exercise-induced PH. More importantly, our model may not be reproducible in other data sets; the prediction model will require external validation before it can be used to guide clinical decision making regarding whether or not to proceed to RHC.

In summary, a predictor based on SpO2 and FVC/DLco has a high NPV, which compares favorably to available noninvasive diagnostic assessments of PH in IPF. In our study, DLco alone did not contribute to MPAP prediction above and beyond SpO2; however, the FVC/DLco did. Furthermore, the extent of desaturation below 88% added valuable information to the model. Using the extent of desaturation in place of need for oxygen (yes/no) and the FVC/DLco ratio in place of DLco improved sensitivity and NPV considerably. This MPAP and PH prediction model will require external validation before it can be used in clinical practice.

Acknowledgments

This work was supported, in part, by grants from the NIH (5U10HL080411 and 5P50HL67665 to D.A.Z.; HL080206 and HL086491 to J.A.B.).

Abbreviations

ABG
arterial blood gas
CI
confidence interval
DE
Doppler echocardiography
DLco
diffusing capacity for carbon monoxide
FVC
forced vital capacity
FEV1
forced expiratory volume in 1 s
ILD
interstitial lung disease
IPF
idiopathic pulmonary fibrosis
IRB
institutional review board
MPAP
mean pulmonary artery pressure
NPV
negative predictive value
PH
pulmonary hypertension
PaO2
arterial blood oxygen tension
PFT
pulmonary function tests
PPV
positive predictive value
RHC
right-heart catheterization
RV
residual volume
RVSP
right ventricular systolic pressure from echocardiography
SpO2
resting room air pulse oximetry
TLC
total lung capacity
VA
alveolar volume
6 MWD
6-min walk distance

Footnotes

[star]All the work was performed at the David Geffen School of Medicine at UCLA.

Conflict of interest statement: David A. Zisman received research grants from Actelion Pharmaceuticals and Cotherix Pharmaceuticals to do multi-center studies. Dr. Zisman is funded by the National Institutes of Health IPF Clinical Research Network, which includes participation in a pulmonary hypertension study with sildenafil.

References

1. Lettieri CJ, Nathan SD, Barnett SD, Ahmad S, Shorr AF. Prevalence and outcomes of pulmonary arterial hypertension in advanced idiopathic pulmonary fibrosis. Chest. 2006;129(3):746–52. [PubMed]
2. Nadrous HF, Pellikka PA, Krowka MJ, Swanson KL, Chaowalit N, Decker PA, et al. Pulmonary hypertension in patients with idiopathic pulmonary fibrosis. Chest. 2005;128(4):2393–9. [PubMed]
3. Arcasoy SM, Christie JD, Ferrari VA, Sutton MS, Zisman DA, Blumenthal NP, et al. Echocardiographic assessment of pulmonary hypertension in patients with advanced lung disease. Am J Respir Crit Care Med. 2003;167(5):735–40. [PubMed]
4. Runo JR, Loyd JE. Primary pulmonary hypertension. Lancet. 2003;361(9368):1533–44. [PubMed]
5. Bonay M, Bancal C, de Zuttere D, Arnoult F, Saumon G, Camus F. Normal pulmonary capillary blood volume in patients with chronic infiltrative lung disease and high pulmonary artery pressure. Chest. 2004;126(5):1460–6. [PubMed]
6. Aduen J, Zisman D, Mobin S, Venegas C, Alvarez F, Biewend M, et al. Retrospective study of pulmonary function tests in patients presenting with isolated reduction in single-breath diffusion capacity: implications for the diagnosis of combined obstructive and restrictive lung disease. Mayo Clin Proc. 2007;82:48–54. [PubMed]
7. Chang B, Wigley FM, White B, Wise RA. Scleroderma patients with combined pulmonary hypertension and interstitial lung disease. J Rheumatol. 2003;30(11):2398–405. [PubMed]
8. Kono H, Inokuma S. Visualization and functional consequence of pulmonary vascular impairment in patients with rheumatic diseases. Chest. 2003;124(1):255–61. [PubMed]
9. Pronk LC, Swaak AJ. Pulmonary hypertension in connective tissue disease. Report of three cases and review of the literature. Rheumatol Int. 1991;11(2):83–6. [PubMed]
10. Steen VD, Graham G, Conte C, Owens G, Medsger TA., Jr Isolated diffusing capacity reduction in systemic sclerosis. Arthritis Rheum. 1992;35(7):765–70. [PubMed]
11. Stupi AM, Steen VD, Owens GR, Barnes EL, Rodnan GP, Medsger TA., Jr Pulmonary hypertension in the CREST syndrome variant of systemic sclerosis. Arthritis Rheum. 1986;29(4):515–24. [PubMed]
12. Trad S, Amoura Z, Beigelman C, Haroche J, Costedoat N, Boutin le TH, et al. Pulmonary arterial hypertension is a major mortality factor in diffuse systemic sclerosis, independent of interstitial lung disease. Arthritis Rheum. 2006;54(1):184–91. [PubMed]
13. Wang J, Juhaszova M, Rubin LJ, Yuan XJ. Hypoxia inhibits gene expression of voltage-gated K+ channel alpha subunits in pulmonary artery smooth muscle cells. J Clin Invest. 1997;100(9):2347–53. [PMC free article] [PubMed]
14. McQuillan LP, Leung GK, Marsden PA, Kostyk SK, Kourembanas S. Hypoxia inhibits expression of eNOS via transcriptional and posttranscriptional mechanisms. Am J Physiol. 1994;267(5 Part 2):H1921–7. [PubMed]
15. Vender RL. Chronic hypoxic pulmonary hypertension. Cell biology to pathophysiology. Chest. 1994;106(1):236–43. [PubMed]
16. Tozzi CA, Poiani GJ, Harangozo AM, Boyd CD, Riley DJ. Pressure-induced connective tissue synthesis in pulmonary artery segments is dependent on intact endothelium. J Clin Invest. 1989;84(3):1005–12. [PMC free article] [PubMed]
17. American Thoracic Society. Idiopathic pulmonary fibrosis: diagnosis and treatment. International consensus statement. American Thoracic Society (ATS), and the European Respiratory Society (ERS) Am J Respir Crit Care Med. 2000;161(2 Pt 1):646–64. [PubMed]
18. Rich S, Dantzker DR, Ayres SM, Bergofsky EH, Brundage BH, Detre KM, et al. Primary pulmonary hypertension. A national prospective study. Ann Intern Med. 1987;107(2):216–23. [PubMed]
19. Crapo RO, Morris AH, Gardner RM. Reference spirometric values using techniques and equipment that meet ATS recommendations. Am Rev Respir Dis. 1981;123(6):659–64. [PubMed]
20. Miller A, Thornton JC, Warshaw R, Anderson H, Teirstein AS, Selikoff IJ. Single breath diffusing capacity in a representative sample of the population of Michigan, a large industrial state. Predicted values, lower limits of normal, and frequencies of abnormality by smoking history. Am Rev Respir Dis. 1983;127(3):270–7. [PubMed]
21. Macintyre N, Crapo RO, Viegi G, Johnson DC, van der Grinten CP, Brusasco V, et al. Standardisation of the single-breath determination of carbon monoxide uptake in the lung. Eur Respir J. 2005;26(4):720–35. [PubMed]
22. ATS statement: guidelines for the six-minute walk test. Am J Respir Crit Care Med. 2002;166(1):111–7. [PubMed]
23. Chan KL, Currie PJ, Seward JB, Hagler DJ, Mair DD, Tajik AJ. Comparison of three Doppler ultrasound methods in the prediction of pulmonary artery pressure. J Am Coll Cardiol. 1987;9(3):549–54. [PubMed]
24. Currie PJ, Seward JB, Chan KL, Fyfe DA, Hagler DJ, Mair DD, et al. Continuous wave Doppler determination of right ventricular pressure: a simultaneous Doppler-catheterization study in 127 patients. J Am Coll Cardiol. 1985;6(4):750–6. [PubMed]
25. Yock PG, Popp RL. Noninvasive estimation of right ventricular systolic pressure by Doppler ultrasound in patients with tricuspid regurgitation. Circulation. 1984;70(4):657–62. [PubMed]
26. Harrell FE, Jr, Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med. 1996;15(4):361–87. [PubMed]
27. Efron B. Estimating the error rate of a prediction rule: improvement on cross validation. J Am Stat Assoc. 1983;78:316–87.
28. Brunelli A, Rocco G. Internal validation of risk models in lung resection surgery: bootstrap versus training-and-test sampling. J Thorac Cardiovasc Surg. 2006;131(6):1243–7. [PubMed]
29. Simonneau G, Galie N, Rubin LJ, Langleben D, Seeger W, Domenighetti G, et al. Clinical classification of pulmonary hypertension. J Am Coll Cardiol. 2004;43(12) S:5S–12S. [PubMed]
30. Leuchte HH, Neurohr C, Baumgartner R, Holzapfel M, Giehrl W, Vogeser M, et al. Brain natriuretic peptide and exercise capacity in lung fibrosis and pulmonary hypertension. Am J Respir Crit Care Med. 2004;170(4):360–5. [PubMed]
31. Weitzenblum E, Ehrhart M, Rasaholinjanahary J, Hirth C. Pulmonary hemodynamics in idiopathic pulmonary fibrosis and other interstitial pulmonary diseases. Respiration. 1983;44(2):118–27. [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...