DOI:
Original papers
DOI: 10.11152/mu-1223Sensitivity and specificity of ultrasound for the diagnosis of acute pulmonary edema: a systematic review and meta-analysis
Yan Wang, Zhiyang Shen, Xuefeng Lu, Yanhua Zhen, Huixia Li
Ultrasound Department, the Second Affiliated Hospital of Zhengzhou University, Zhengzhou, Henan 450014, China
Received 17.08.2017 Accepted 22.10.2017 Med Ultrason
2018, Vol. 20, No 1, 32-36
Corresponding author: Dr. Zhiyang Shen
Ultrasound Department, The Second Affiliated Hospital of Zhengzhou University,
2 Jingba Road, Jinshui, Zhengzhou, Henan 450014, China
Phone: +86-0371-63921691 E-mail: [email protected]
Introduction
Understanding the microvascular fluid exchange in the lung is important for knowing the causes of acute pulmonary edema (PE). In a normal lung the outflow of fluid occurs though gaps between the capillary endothe- lial cells. A sudden increase of hydrostatic pressure in the pulmonary capillaries leads to edema [1]. The common causes of PE include arterial hypertension, severe coro- nary occlusion, cerebral diseases, pulmonary and heart diseases, infections, and shock [2].
Acute PE is produced by accumulation of the fluid in alveoli and pulmonary interstitial spaces, impairing in this way the diffusion of gases [3]. PE is one of the
common causes of acute dyspnea. Selection of more sensitive and specific diagnostic approach of acute PE is a critical issue that continues to gain attention from medical staff. Accurate and rapid determination of the nature of acute dyspnea is an important and challenging issue in the intensive care unit (ICU) and the emergency department (ED) [4]. The common diagnostic methods that are used to determine the cause of acute dyspnea includes B-type natriuretic peptide (BNP) test, N-termi- nal (NT) proBNP test, X-ray, ultrasound, and thoracic computed tomography (CT) scan. Although chest radi- ography is the routine examination, CT scan remains the gold standard examination for pulmonary diseases.
The chest radiography has some disadvantages, includ- ing inapplicability to pregnant women, the non-specific findings, the difficulties in acquiring the posteroanterior and laterolateral projections [4]. On the other hand, CT scan also has limitations such as high dosage of radia- tion that is required, lack of CT scan facility in certain hospitals, and patient the needs for moving the patient in the radiology room [5].
Diagnosis of acute PE with the non-invasive ultra- sound method has been gaining popularity in the past Abstract
Aims: This study aimed to determine the sensitivity and specificity of ultrasound for the diagnosis of acute pulmonary edema by meta-analysis. Materials and methods: A systematic search was conducted through the following databases:
Cochrane, PubMed, EMBASE and Ovid MEDLINE. Prospective cohort and prospective case-control studies that reported sensitivity and specificity of lung ultrasound in diagnosis of acute pulmonary edema were selected. An independent review of citations was carried out for inclusion and data extraction. Quality assessment was conducted using the QUADAS-2 tool.
Sensitivity and specificity were taken from the studied articles and then calculated with the contingency tables. A total of 984 articles were identified but only eight studies (1301 patients) were included in this meta-analysis. One study was a case-control study and seven studies were prospective cohort study. Results: The overall sensitivity of ultrasound for the diagnosis of acute pulmonary edema is 97% (95% CI: 96%–98%) and the overall specificity was 98% (95% CI: 97%–99%). Conclusion: The diagnostic test accuracy suggests that lung ultrasound using B-lines is a useful and reliable diagnostic tool for critically ill patients with acute pulmonary edema.
Keywords: acute pulmonary edema; diagnosis; dyspnea; lung; ultrasound
decades. Lung ultrasound is a useful diagnosis imaging technique, particularly in a situation when a CT scan can- not be used, and allows a rapid bedside examination and immediate interpretation of scanning report by trained physicians [6]. Besides, the patient will not subject to any form of radiation. The main purpose of this study is to calculate the sensitivity and specificity of lung ultrasound from up-to-date studies regarding the diagnosis of acute pulmonary edema.
Materials and methods Search strategy
This meta-analysis was carried out by following the PRISMA guidelines and the Cochrane Handbook for Diagnostic Test Accuracy Reviews [7,8]. Systematic search of published literature was carried out for dates prior to 17th July 2016 without limitation of start time in the following databases: Cochrane, PubMed, EMBASE and Ovid MEDLINE. Keywords that used to search in titles and abstracts were included “lung ultrasonogra- phy” or “lung ultrasound” and “acute pulmonary ede- ma”. Endnote software (version 7) was used to manage the literature.
Inclusion and exclusion criteria for article selection
Our outcome of interest in this study was the diagno- sis of acute PE using the lung ultrasound with B-lines.
Prospective case-control and prospective cohort stud- ies that involved lung ultrasound B-lines in diagnosis of acute PE were included in this study. Commentaries, letters, reviews, and case reports were excluded from this analysis. Studies that enrolled patients with clinical suspicion of acute PE and acute dyspnea were included.
Studies without acute PE and asymptomatic pulmonary diseases were excluded. No restriction was applied to the ultrasound scanning protocol that was used for diagno- sis. Lung ultrasound procedure has to be performed by trained personnel at the patient’s bed-side.
Titles and abstracts of the literature that were identi- fied were independently reviewed by 2 reviewers. Then, the full-texts in the filtered list of references were re- viewed by the same reviewers. If both reviewers had dis- agreements, a discussion was carried out. The inclusion and exclusion criteria were followed to find out suitable studies for the meta-analysis. All data were extracted by the same reviewers. For quality assessment, the same reviewers independently reviewed the included studies using the QUADAS-2 tool [9]. The QUADAS-2 qual- ity assessment was structured to evaluate the four key points, including patient selection, index test, reference standard, and flow and timing.
Data analysis
Data analyses were performed by using the Stata (version 14) statistical software. Results for sensitivity and specificity of all the included studies were plotted on a forest plot for heterogeneity assessment.
Results
A total of 984 studies were retrieved from PubMed, Ovid MEDLINE, EMBASE and Cochrane databases (fig 1). After removing the duplicate studies and the studies that did not fulfill the inclusion criteria, 8 studies were included in the meta-analysis (1301 patients; Table I) [10-17]. Two studies were performed in the ICU, three studies in the ED, two studies in the ward and one study in a pre-hospital setting and ED. The ultrasonographers of all studies were blinded to the results of the reference standard. Only one study was a case-control study, while the others were prospective cohort studies. Four stud- ies followed the procedure reported by Volpicelli et al [10,14-16,18]. Two studies diagnosed the PE by detect- ing B-line in anterior and lateral chest and three or more B-lines in at least two zones on each hemithorax [12,13].
Another two studies followed comet-score scanning protocol for diagnosis of PE [11,17]. The QUADAS-2 quality assessment revealed that quality of the included studies was from moderate to high (Table II). Low risk of bias showed that the quality evaluation of the studies is high.
The ultrasound examinations were carried out by trained physicians from the ED and ICU, nurses and medical students (Table III). In three studies the inter- rater reliability was reported. Sensitivity and specificity of included studies are presented in a forest plot (fig 2).
The overall sensitivity of lung ultrasound using B-lines for the diagnosis of acute pulmonary edema is 97% (95%
CI: 96%–98%), and 98% overall specificity (95% CI:
97%–99%).
Fig 1. Study selection flow diagram.
Table I. Characteristics of the included studies in the meta-analysis First au-
thor, year of publication
Lichten- stein, 1998 [12]
Lichten- stein, 2008 [13]
Gargani, 2008 [11]
Liteplo, 2009 [14]
Prosen, 2011 [16]
Vitturi, 2011 [17]
Cibinel, 2012 [10]
Mumoli, 2016 [15]
Journal Intens Care
Med Chest Eur J Heart
Fail Acad Emerg
Med Crit Care J Ultras Intern Emerg
Med Medicine
Study coun-
try France France Italy USA Slovenia Italy Italy Italy
Place of
study ICU ICU CPD ED Prehospital
and ED Internal medicine ward
ED ED
Number of
patients 146 260 149 94 218 152 56 226
Ultrasound
machine Hitachi-405,
ADR 400 Hitachi-405 Philips Sonos
7500 Sonosite Sonosite Toshiba
Aplio XV GE Electric
LOGIQ 3 GE Vivid S5 Probe type 3.5 MHz
cardiac 5.0 MHz micro-con- vex
2.5–3.5 MHz
cardiac 2.5 MHz
curved-array Not specified 3.5 MHz
convex 3.5 MHz
convex 2.0–5.5 MHz curved-array
Protocol Lichtenstein Lichtenstein Comet score Volpicelli Volpicelli Comet score Volpicelli Volpicelli Reference
standard Blinded Blinded Blinded Blinded Final hospital
diagnostic Blinded Final hospital
diagnostic Final hospital diagnostic Examination
time <1 min <3 min <5 min <5 min <1 min <3 min <5 min <6 min
Study type Case-control Cohort Cohort Cohort Cohort Cohort Cohort Cohort
Sensitivity,
% (95% CI) 100 97 76.2 58 (36–77) 100
(98–100) 97 93.6 95.3
(92.6–98.1) Specificity,
% (95% CI) 97 95 88 85 (69–95) 95 (91–100) 79 84 88.2
(84.0–92.4) Conflict of
interest Unclear Unclear Unclear Unclear None None None None
ICU – intensive care unit; ED – emergency department; CPD – cardiology and pneumology division; CI – confidence interval
Table II. QUADAS-2 assessment of methodologic quality of the included studies
Study Risk of bias Applicability concerns
Patient
selection Index test Criterion
standard Flow and
timing Patient
selection Index test Criterion standard
Lichtenstein, 1998 [12] High Unknown Unknown High Low Low Low
Lichtenstein, 2008 [13] Unknown Low Low Unknown Low Low Low
Gargani, 2008 [11] Low Unknown Low Low Low Low Low
Liteplo, 2009 [14] Unknown Low Low Low Low Low Low
Prosen, 2011 [16] Unknown Low Low Low Low Low Low
Vitturi, 2011 [17] Low Low Low Unknown High High Low
Cibinel, 2012 [10] Low Low Low Low Low Low Low
Mumoli, 2016 [15] Low Low Low Low Low Low Low
Discussions
A rapid diagnosis using non-invasive methods en- sures an appropriate and timely treatment. In the present study, the lung ultrasound using B-lines were evaluated for sensitivity and specificity in the diagnosis of acute PE. Three studies that employed scanning protocol of Volpicelli et al [18] showed different values of sensitivity and specificity as different methodologies were applied in the mentioned studies. The ultrasound scanning in the Liteplo et al study was performed by trained medical stu- dents, Mumoli et al study was conducted by trained nurs- es, and Cibinel et al study was implemented by attending physicians [10,14,15]. Operation of scanning by different methodologies and people with different training levels leads to inconsistent results.
The comet tail artifacts B-line was used for the detec- tion of extravascular lung water and it enables differen- tiation of acute PE from chronic obstructive pulmonary disease [12]. The ultrasound scanning protocol described by Volpicelli et al involved a single scan on eight zones at anterior and lateral of each lung [18] and they were interpreted as abnormal when two or more zones pre- sented B-lines in both hemithoraces. The protocol of Li- chtenstein et al involved scanning of comet-tail artifact produced from pleural line [12]. Comet score scanning
protocol as described by Picano et al [19] was applied by Vitturi et al [17] and Gargani et al [11] in determination of extravascular lung water; the test was considered as abnormal when the number of B-lines was greater than five [11] or eight [17]. Subgroup analysis was not carried out in these studies and this contributed to the heteroge- neity of our data.
The sensitivity of chest radiography, BNP, and NT proBNP tests in the diagnosis of acute dyspnea ranged from 56%–93%, 86%–99% and 92%–97%, respectively, while the specificity ranged from 51%–98%, 74%–99%
and 44%–93%, respectively [11,15-17,20-25]. Although high sensitivity and specificity of the mentioned diagnos- tic methods were reported, real-time valuation of the out- come is almost impossible and the assessments are not available in the pre-hospital setting. Besides, the BNP test is not accessible to all clinicians in hospitals.
Limitations in this study included incomplete retriev- al of identified research, publishing bias, reporting bias, and inconsistencies of the ultrasonic inspection method.
Besides, patients included in analysis were from different populations. A larger number of patients presenting with acute PE should be included in future studies. In addi- tion, standardization of ultrasound scanning protocol and qualification of ultrasonographers should be applied to minimize heterogeneity of the analysis.
Fig 2. Forest plot of sensitivity and specificity of the included studies. CI: confidence interval.
Table III. Characteristics of ultrasonographers in the included studies
Study Number of ultra-
sonographer Training level of ultrasonographer Inter-rater reliability
Lichtenstein, 1998 [12] 2 ICU physicians Not reported
Lichtenstein, 2008 [13] 2 ICU physicians Not reported
Gargani, 2008 [11] Not reported Not reported Not reported
Liteplo, 2009 [14] 7 2 Emergency physicians, 5 trained medical students 0.82
Prosen, 2011 [16] 10 Emergency physicians Not reported
Vitturi, 2011 [17] 2 Not reported 0.98
Cibinel, 2012 [10] Not reported Emergency physicians 0.92
Mumoli, 2016 [15] 5 Trained nurses Not reported
Conclusions
The diagnostic test accuracy suggests that lung ultra- sound using B-lines is one of the best tools for the diag- nosis of acute PE especially for the critically ill patients.
Lung ultrasound provides high sensitivity and specificity diagnosis for moderate to severe acute PE. In addition, negative lung ultrasound helps to exclude the PE.
Conflict of interest: none
References
1. Ware LB, Matthay MA. Clinical practice. Acute pulmonary edema. New Engl J Med 2005;353:2788-2796.
2. Luisada AA, Cardi L. Acute pulmonary edema; pathol- ogy, physiology and clinical management. Circulation 1956;13:113-135.
3. Sciscione AC, Ivester T, Largoza M, Manley J, Shlossman P, Colmorgen GH. Acute pulmonary edema in pregnancy.
Obstet Gynecol 2003;101:511-515.
4. Nielsen LS, Svanegaard J, Wiggers P, Egeblad H. The yield of a diagnostic hospital dyspnoea clinic for the primary health care section. J Intern Med 2001;250:422-428.
5. Lee CI, Haims AH, Monico EP, Brink JA, Forman HP. Di- agnostic CT scans: assessment of patient, physician, and radiologist awareness of radiation dose and possible risks.
Radiology 2004;231:393-398.
6. Yu CJ, Yang PC, Chang DB, Luh KT. Diagnostic and thera- peutic use of chest sonography: value in critically ill pa- tients. AJR Am J Roentgenol 1992;159:695-701.
7. Macaskill P, Gatsonis P, Deeks JJ, Harbord R, Takwoingi Y.
Handbook for Systematic Reviews of Diagnostic Test Ac- curacy: The Cochrane Collaboration 2010.
8. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg 2010;8:336-341.
9. Whiting PF, Rutjes AW, Westwood ME, et al. QUADAS-2:
a revised tool for the quality assessment of diagnostic ac- curacy studies. Ann Intern Med 2011;155:529-536.
10. Cibinel GA, Casoli G, Elia F, et al. Diagnostic accuracy and reproducibility of pleural and lung ultrasound in discrimi- nating cardiogenic causes of acute dyspnea in the emergen- cy department. Intern Emerg Med 2012;7:65-70.
11. Gargani L, Frassi F, Soldati G, Tesorio P, Gheorghiade M, Picano E. Ultrasound lung comets for the differential di- agnosis of acute cardiogenic dyspnoea: A comparison with natriuretic peptides. Eur J Heart Fail 2008;10:70-77.
12. Lichtenstein DA, Meziere GA. A lung ultrasound sign al- lowing bedside distinction between pulmonary edema and COPD: the comet-tail artifact. Intensive Care Med 1998;24:1331-1334.
13. Lichtenstein DA, Mezière GA. Relevance of lung ultra- sound in the diagnosis of acute respiratory failure: The BLUE Protocol. Chest 2008;134:117-125.
14. Liteplo AS, Marill KA, Villen T, et al. Emergency thoracic ultrasound in the differentiation of the etiology of shortness of breath (ETUDES): sonographic B-lines and N-terminal pro-brain-type natriuretic peptide in diagnosing congestive heart failure. Acad Emerg Med 2009;16:201-210.
15. Mumoli N, Vitale J, Giorgi-Pierfranceschi M, et al. Accu- racy of nurse-performed lung ultrasound in patients with acute dyspnea: A prospective observational study. Medicine (Baltimore) 2016;95:e2925.
16. Prosen G, Klemen P, Strnad M, Grmec S. Combination of lung ultrasound (a comet-tail sign) and N-terminal pro- brain natriuretic peptide in differentiating acute heart fail- ure from chronic obstructive pulmonary disease and asthma as cause of acute dyspnea in prehospital emergency setting.
Crit Care 2011;15:R114.
17. Vitturi N, Soattin M, Allemand E Simoni F, Realdi G. Tho- racic ultrasonography: A new method for the work-up of patients with dyspnea. J Ultrasound 2011;14:147-151.
18. Volpicelli G, Mussa A, Garofalo G, et al. Bedside lung ultrasound in the assessment of alveolar-interstitial syn- drome. Am J Emerg Med 2006;24:689-696.
19. Picano E, Frassi F, Agricola E, Gligorova S, Gargani L, Mottola G. Ultrasound lung comets: a clinically useful sign of extravascular lung water. J Am Soc Echocardiogr 2006;19:356-363.
20. Baker K, Mitchell G, Thompson AG, Stieler G. Compari- son of a basic lung scanning protocol against formally re- ported chest x-ray in the diagnosis of pulmonary oedema.
Australas J Ultrasound Med 2013;16:183-189.
21. Brown LM, Calfee CS, Howard JP, Craig TR, Matthay MA, McAuley DF. Comparison of thermodilution measured ex- travascular lung water with chest radiographic assessment of pulmonary oedema in patients with acute lung injury.
Ann Intensive Care 2013;3:25.
22. Cardinale L, Priola AM, Moretti F, Volpicelli G. Effective- ness of chest radiography, lung ultrasound and thoracic computed tomography in the diagnosis of congestive heart failure. World J Radiol 2014;6:230-237.
23. Fonseca C, Mota T, Morais H, et al. The value of the elec- trocardiogram and chest X-ray for confirming or refuting a suspected diagnosis of heart failure in the community. Eur J Heart Fail 2004;6:807-812.
24. Mant J, Doust J, Roalfe A, et al. Systematic review and individual patient data meta-analysis of diagnosis of heart failure, with modelling of implications of different diag- nostic strategies in primary care. Health Technol Assess 2009;13:1-207.
25. Morrison LK, Harrison A, Krishnaswamy P, Kazanegra R, Clopton P, Maisel A. Utility of a rapid B-natriuretic peptide assay in differentiating congestive heart failure from lung disease in patients presenting with dyspnea. J Am Coll Car- diol 2002;39:202-209.