Int J Clin Exp Med 2015;8(8):14416-14423 www.ijcem.com /ISSN:1940-5901/IJCEM0003623
Original Article Diagnosis efficiency for pulmonary embolism using magnetic resonance imaging method: a meta-analysis Kuitao Yue Department of Medical Imaging, Weifang Medical University, Weifang 261053, Shandong, China Received November 5, 2014; Accepted January 10, 2015; Epub August 15, 2015; Published August 30, 2015 Abstract: PE (Pulmonary embolism, PE) is a common disease, usually caused by blockage of pulmonary artery and its branches due to exogenous or endogenous embolic obstruction. PE always be misdiagnosed in clinical. The aim of this study is to calculate the sensitivity and specificity of magnetic resonance imaging (MRI) in assessing the resectability of PE. In this study, a meta-analysis of the reported sensitivity and specificity of each study with 95% confidence intervals (CI) was performed. Five studies were included in the meta-analysis. The results indicated that the quality assessment scores ranged from 11 to 13, with a mean study quality score of 12. The sensitivity and specificity values including 95% CI at the patient level were calculated. The sensitivities ranged from 78% to 100%, and the specificity ranged from 99% to 100%. The pooled sensitivity value including 95% CI was 0.83 (0.78-0.88), and with inconsistency (I2) of 62.8%. The pooled specificity value including 95% CI was 0.99 (0.98-1.00), with inconsistency (I2) of 0.0%. Pooled positive likelihood ratio (PLR) (95% CI) was 70.22 (29.04-169.76), and the pooled negative likelihood ratio (NLR) (95% CI) was 0.19 (0.14-0.25). The overall diagnostic odds ratio (DOR) (95% CI) was 448.98 (163.47-1233.18). The summary receiver operating characteristic (SROC) data illustrated that the area under the curve (AUC) was 0.9852. In conclusion, the MRI method may be acts as a potential and assistant method for the PE diagnosis. Keywords: Pulmonary embolism, magnetic resonance imaging, diagnosis, meta-analysis
Introduction PE (Pulmonary embolism, PE) is a common disease, usually caused by blockage of pulmonary artery and its branches due to exogenous or endogenous embolic obstruction [1]. The PE could also result in pulmonary circulation disorders and blood flow interruption in lung tissue, and leading to lung tissue necrosis pathological changes, which is also called pulmonary infarction [2]. PE misdiagnosis often appears clinically. In recent years, CT (computed tomography) technology has been very mature one for the diagnosis of pulmonary embolism with good sensitivity and specificity [3]. Meanwhile, Magnetic Resonance Imaging (MRI) is also commonly used in the diagnosis of pulmonary embolism [4-6]. Compared to CT, MRI has the advantage that there is no ionizing radiation, nor the use of iodinated contrast media, which has small hazard to the patient’s health [7]. This meta-analy-
sis aims to investigate the diagnostic validity of MRI in PE. Methods Search strategy We searched PUBMED, EMBASE and Springer for relevant citations. The search was completed in June 2014 and covered published literature since January 1990 (Figure 1). The terms in the search were “Pulmonary Embolism” AND (“MR imaging” OR “MR” OR “MRI” OR “Magnetic Resonance Imaging” OR “Magnetic Resonance”). The reference lists of known previous reviews and of all of the primary studies included were examined to identify cited articles not found by the electronic searches (Figure 1). There were no restrictions regarding publication language. Eligibility criteria Titles and abstracts identified by the electronic search were analyzed by two independent
Diagnosis for pulmonary embolism with MRI nosed as diseased according to the index test). Two reviewers independently selected articles on the basis of the title and abstract; if one or both reviewers considered the study potentially eligible, the full article was evaluated by both reviewers. Two reviewers abstracted the data from each article and recorded data using a standardized form. Disagreement was resolved by means of consensus.
Figure 1. Flow chart of the study selection procedure.
reviewers (A.V. and V.G.), and full text versions of all citations that were likely to meet the selection criteria were obtained. Disagreements were resolved by consensus or after discussion with a third reviewer (J.Z.). Between-reviewer agreement was assessed by the kappa coefficient. Studies were selected for the review if 1) they were designed as diagnostic experimental research; 2) the research subjects were ≥18 years old suspected PE patients; 3) the diagnostic gold standard was CT; 4) the data of (true positive, TP), (false positive, FP), (true negative, TN), (false negative, FN) were provided or can be calculated. In cases of suspected duplicate publication, the most recent and complete version was selected. We also excluded studies for 1) there were no MRI check 48 hrs after CT diagnostic; 2) non-English documentations; 3) reviews, abstracts, isolated cases, commentaries, editorials or letters. Data extraction For each study, the following count data were entered into a 2×2 table: (a) true-positive results (disease cases in which findings of the reference standard and the index test were positive for the studied target condition), (b) false-negative results (diseased cases that were misdiagnosed as non-diseased according to the index test), (c) true-negative results (nondiseased cases that were correctly identified with the index test), and (d) false-positive results (non-diseased cases that were misdiag-
14417
Quality assessment Methodological quality was assessed independently by the same reviewers based on the Quality Assessment of Studies of Diagnostic Accuracy included on Systematic Reviews (QUADAS) guidelines [8] disagreement was resolved by consensus (Table 1). This evidencebased tool was developed specifically to assess the quality of diagnostic accuracy studies and includes 14 quality items. The 14 items, phrased as questions, are scored as “yes,” “no”, or “unclear”. The quality assessment score can range from 0 to 14, where 14 is the maximum attainable score (Table 1). A more detailed description of each item, together with a guide on how to score each item, is provided by Whiting et al. in 2003 [8]. Additionally, we noted whether the study design was prospective or retrospective. Statistical analysis This meta-analysis was performed with MetaDisc (1.4.0.0) statistical softwares [9]. Primary outcomes of this meta-analysis were summary receiver operating characteristic (SROC) curve, sensitivity (sensitivity, Sen), specificity (specificity, Spe), positive likelihood ratio (positive likelihood ratios, PLR), negative likelihood likelihood ratio (negative likelihood ratios, NLR), diagnostic tests odds ratio (diagnostic odds ratio, DOR) and their 95% confidence intervals (confidence intervals, CI). DOR ranges from 0 to infinity, the greater the DOR values, the powerful MRI diagnosis of pul-
Int J Clin Exp Med 2015;8(8):14416-14423
Diagnosis for pulmonary embolism with MRI Table 1. Quality assessment of the included articles Reference number of the included studies
QUADAS list item 7
13
14
15
16
1. Did the spectrum of patients represent the patients who will receive the test in practice?
+
+
+
+
+
2. Were selection criteria clearly described?
+
+
+
+
+
3. Is the reference standard likely to correctly classify the target condition?
+
+
+
+
+
4. Is the period between the reference standard and index test short enough to be reasonably sure that the target condition did not change between the 2 tests?
+
+
+
+
+
5. Did the entire sample or a random selection of the sample receive verification using a reference standard of diagnosis?
+
+
+
+
+
6. Did patients receive the same reference standard regardless of index test result?
+
+
+
+
+
7. Was the reference standard independent of the index test (i.e., index test did not form part of the reference standard)?
+
+
+
+
+
8. Was execution of the index test described in sufficient detail to permit replication of the test?
+
+
+
+
+
9. Was execution of the reference standard described in sufficient detail to permit its replication?
+
+
+
+
+
10. Were index test results interpreted without knowledge of results of the reference standard?
+
-
+
0
0
11. Were reference standard results interpreted without knowledge of results of the index test?
-
+
0
0
+
12. Were the same clinical data available when test results were interpreted as would be available when the test is used in practice?
+
0
0
0
0
13. Were uninterruptable/intermediate test results reported?
+
+
-
+
-
14. Were withdrawals from the study explained?
+
+
+
+
+
QUADAS: Quality Assessment of Diagnostic Accuracy Studies. +: YES; -: NO; 0: not clear.
14418
Int J Clin Exp Med 2015;8(8):14416-14423
Diagnosis for pulmonary embolism with MRI Table 2. Characteristics of each included study Author reference
year
Country
1
Kluge [5]
2006
Germany
1.5T
3-D FLASH
65
2
Revel [12]
2012
French
1.5T
3-D FSPGR /SSFP/GRE
274
3
Zhang [14]
2013
China
3T
3-D FLASH
27
18/9
38.9±14.4
4
Stein [13]
2010
USA
1.5T/3T
NA
279
NA
≥18
5
Pleszewski [11]
2006
Canada
1.5T
3-D SEGP
34
20/28
22-84
9
id
Magnetic MRI scan sequence field intensity
n
M/F
Age
TP
TN
FP
34/31
60.9±15.7
127/147
59.8±19.0
FN
15
43
0
4
87
170
1
16
24
3
0
0
59
201
2
17
23
0
2
3-D: 3-dimensional; FLASH: fast low-angle shot gradient-echo sequence; FSPGR: fast spoiled gradient echo; SSFP: Steady-state-free precession; GRE: gradient recalled echo; SEGP: spoiled echo gradient pulse; NA: Not available.
Results Literature search
Figure 2. Sensitivity and specificity of MRI detection for the diagnosis of PE. A. The sensitivity of MRI detection for the diagnosis of PE. B. The specificity of MRI detection for the diagnosis of PE. PE: Pulmonary embolism; CI: confidential interval; MRI: Magnetic Resonance Imaging.
The literature search identified 1542 articles (Figure 1). After screening of the titles and abstracts, 1231 items was remained because the duplication. Among the 1231 items, 1214 items were excluded because one or more inclusion criteria were not fulfilled. The remaining 17 articles were evaluated by reading the full text. Of these, 12 were excluded (the reasons are given in Figure 1). The remaining 5 studies were included in the meta-analysis [7, 13-16]. The study characteristics are shown in Table 2. Quality assessment
monary embolism recognition (discriminatory ability) [10]. SROC curve is based on the sensitivity and specificity, and area under the curve closer to 1, the higher the diagnostic validity [11]. Heterogeneity was tested by using the Higgins and Thompson test to calculate the I2 statistic [12]. This statistic uses the conventional Cochran Q statistic to calculate the percentage of tool variation heterogeneity on a scale ranging from 0% (no heterogeneity) to 100% (all variance due to heterogeneity). In contrast to the Cochran Q, the I2 is less affected by the number of studies included in a meta-analysis. If no or moderate heterogeneity is found (2≤50%), pooling is justified. 14419
The quality assessment scores of this study ranged from 11 to 13, with a mean study quality score of 12 (Table 2). All of the studies with the higher quality. MRI detection for sensitivity and specificity The results showed that all of the MRI detection with the relative higher sensitivity and specificity (Figure 2). The sensitivity and specificity values including 95% CI at the patient level were calculated. The sensitivities ranged from 78% to 100%, and the specificity ranged from 99% to 100%. The pooled sensitivity value including 95% CI was 0.83 (0.78-0.88), Q =10.75, df =4 (P=0.0295), inconsistency (I2) =62.8% (Figure 2A). The pooled specificity value including 95%
Int J Clin Exp Med 2015;8(8):14416-14423
Diagnosis for pulmonary embolism with MRI with the MRI method [19]. The method acts as a diagnostic tool depends on the sensitivity and the specificity for the disease [20]. The present meta-analysis study involves the studies of Magnetic Resonance Imaging (MRI) for diagnosis of PE. By analyzing the data in these studies, we found that MRI is a highly accurate diagnostic efficiency with a pooled sensitivity of 83% and a pooled specificity of 99%. Furthermore, the positive likelihood ratios (PLR), negative likelihood ratios (NLR), diagnostic odds ratio Figure 3. PLR and NLR of MRI detection for the diagnosis of PE. A. PLR of MRI (DOR) and 95% confidendetection for the diagnosis of PE. B. NLR of MRI detection for the diagnosis of PE. ce intervals (CI, 29.04 to PE: Pulmonary embolism; PLR: positive likelihood ratio; NLR: negative likelihood 169.76) were also higher ratio; CI: confidential interval; MRI: Magnetic Resonance Imaging. enough for the method applies as a diagnostic CI was 0.99 (0.98-1.00), Q =1.21, df =4 tool in clinical. For the area under the curve (P=0.8762), inconsistency (I2) =0.0% (Figure analysis, the AUC level was also close to 1, 2B). which suggests that MRI method with a better diagnostic efficiency. Pooled PLR and NLR The MRI method has been used in the assisOur results indicated that the pooled positive tant diagnosis for many other diseases. Nael et likelihood ratio (PLR) (95% CI) =70.22 (29.04al. [21] used the six-minute magnetic reso169.76), Q =3.49, df =4 (P=0.4802), inconsisnance imaging protocol to evaluate the acute tency (I2) =0.0% (Figure 3A). The pooled negaischemic stroke, which results in significant tive likelihood ratio (NLR) (95% CI) =0.19 (0.14reduction in scan time rivaling that of the multi0.25), Q =3.86, df =4 (P=0.4258), inconsistenmodal computed tomographic protocol. Puskas cy (I2) =0.0% (Figure 3B). et al. [22] indicated that the diffusion-weighted magnetic resonance imaging with conventional DOR and SROC curves sequences is a useful and promising functional imaging modality in the early diagnosis of Overall diagnostic odds ratio (DOR) and summyeloma multiple. Thompson et al. [23] also mary receiver operating characteristic (SROC) found that the MRI plays a useful role in the curves for the diagnostic performance of MRI diagnosis and management of the prostate detection was drew. The DOR (95% CI) =448.98 cancer, and appears to be of value in planning (163.47-1233.18), Q=1.01, df =4 (P=0.9077), dosimetry in men undergoing radiotherapy, and 2 inconsistency (I ) =0.0% (Figure 4A). The SROC in guiding selection for and monitoring on active data illustrated that the AUC was 0.9852 surveillance. Filippi et al. [24] also analyzed the (Figure 4B). application and exploring the insights of the magnetic resonance imaging for some other Discussion diseases, and pointed out that all of the conventional and non-conventional MRI techniques Clinically, the diagnosis of PE always be perwill be benefit from the use of high-field MRI formed by using the CT method. However, the systems (3.0T or more). CT may cause some side-effects compared
14420
Int J Clin Exp Med 2015;8(8):14416-14423
Diagnosis for pulmonary embolism with MRI
Figure 4. Overall DOR and SROC curves for the diagnostic performance of MRI detection in PE. A. Overll DOR curves for the diagnostic performance of MRI detection. B. SROC curves for the diagnostic performance of MRI detection. DOR: diagnostic odds ratio; SROC: summary receiver operating characteristic; AUC: area under the curve.
Though the MRI has been used in many diseases, there are also some disadvantages as the followings: (1) The samples are relative small, which may affect the statistical efficiency for the diagnosis; (2) Though there are not the threshold effect and no-threshold effects, the heterogenicity may be exist among the researches; (3) This study only included the published studies, the un-published documents may be omitted, which may also cause the
14421
false positive and amplify the diagnostic efficiency of MRI. In conclusion, the data of this meta-analysis suggests that the MRI method for PE diagnosis with a higher sensitivity and specificity, and also with lower misdiagnosis rate and missed diagnosis rate. The MRI method may be acts as a potential and assistant method for the PE diagnosis.
Int J Clin Exp Med 2015;8(8):14416-14423
Diagnosis for pulmonary embolism with MRI Disclosure of conflict of interest None. Address correspondence to: Dr. Kuitao Yue, Department of Medical Imaging, Weifang Medical University, 7166 Baotong Jie (West), Weifang 261053, Shandong, China. Tel: +86-0536-2603780; Fax: +86-0536-2603780; E-mail: yuektwf@yeah. net
[12]
[13]
References [1]
Li X, Li Q, Miao Y, Xu H, Liu Y, Qiu X, Wang EH. A case of renal antiomyolipoma with intracardiac extension and asymptomatic pulmonary embolism. Int J Clin Exp Pathol 2013; 6: 11801186. [2] Moutzouris JP, Chow V, Chiang Yong AS, Chung T, Naganathan V, Kritharides L, Chwan AC. Acute pulmonary embolism in individuals aged 80 and older. J Am Geriatr Soc 2014; 62: 2004-2006. [3] Schibany N, Fleischmann D, Thallinger C, Schibany A, Hahne J, Ba-Ssalamah A, Herold CJ. Equipment availability and diagnostic strategies for suspected pulmonary embolism in Austria. Eur Radiol 2001; 11: 2287-2294. [4] Yuan C, Zhu H, Song D, Wei W, Zhu R, Mei X, Zou J, Yang H. Int J Clin Exp Med 2014; 7: 163169. [5] Remy-Jardin M, Remy J, Baghaie F, Fribourg M, Artaud D, Duhamel A. Clinical value of thin collimation in the diagnostic workup of pulmonary embolism. Am J Roentgenol 2000; 175: 407411. [6] Schoepf UJ, Costello P. CT angiography for diagnosis of pulmonary embolism: State of the art. Radiology 2004; 230: 329-337. [7] Kluge A, Luboldt W, Bachmann G. Acute pulmonary embolism to the subsegmental level: diagnostic accuracy of three MRI techniques compared with 16-MDCT. AJR Am J Roentgenol 2006; 187: W7-14. [8] Whiting P, Rutjes AW, Reitsma JB, Bossuyt PM, Kleijnen J. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003; 3: 25. [9] Zamora J, Abraira V, Muriel A, Khan K, Coomarasamy A. Meta-DiSc: a software for metaanalysis of test accuracy data. BMC Med Res Methodol 2006; 6: 31. [10] Glas AS, Lijmer JG, Prins MH, Bonsel GJ, Bossuyt PM. The diagnostic odds ratio: a single indicator of test performance. J Clin Epidemiol 2003; 56: 1129-1135. [11] Moses LE, Shapiro D, Littenberg B. Combining independent studies of a diagnostic test into a
14422
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
summary ROC curve: Data-analytic approaches and some additional considerations. Stat Med 1993; 12: 1293-1316. Devillé WL, Buntinx F, Bouter LM, Montori VM, de Vet HC, van der Windt DA. Conducting systematic reviews of diagnostic studies: didactic guidelines. BMC Med Res Methodol 2002; 2: 9. Pleszewski B, Chartrand-Lefebvre C, Qanadli SD, Dery R, Perreault P, Oliva VL. Gadoliniumenhanced pulmonary magnetic resonance angiography in the diagnosis of acute pulmonary embolism: a prospective study on 48 patients. Clin Imaging 2006; 30: 166-172. Revel MP, Sanchez O, Couchon S, Planquette B, Hernigou A, Niarra R, Meyer G, Chatellier G. Diagnostic accuracy of magnetic resonance imaging for an acute pulmonary embolism: results of the ‘IRM-EP’ study. J Thromb Haemost 2012; 10: 743-750. Stein PD, Chenevert TL, Fowler SE, Goodman LR, Gottschalk A, Hales CA, Hull RD, Jablonski KA, Leeper KV Jr, Naidich DP, Sak DJ, Sostman HD, Tapson VF, Weg JG, Woodard PK; PIOPED III (Prospective Investigation of Pulmonary Embolism Diagnosis III) Investigators. Gadolinium-enhanced magnetic resonance angiography for pulmonary embolism: a multicenter prospective study (PIOPED III). Ann Intern Med 2010; 152: 434-443, W142-1433. Zhang LJ, Luo S, Yeh BM, Zhou CS, Tang CX, Zhao Y, Li L, Zheng L, Huang W, Lu GM. Diagnostic accuracy of three-dimensional contrast-enhanced MR angiography at 3-T for acute pulmonary embolism detection: Comparison with multidetector CT angiography. Int J Cardiol 2013; 168: 4775-4783. Raznatovic ZJ, Zaric ND, Galun DA, Lekic NS, Micev M, Djordjevic VR, Djurasic LM, Kerkez MD. Multiple port-site metastasis of incidental gallbladder carcinoma after laparoscopic cholecystectomy. Acta Chir Lugosl 2012; 59: 105109. Wright BE, Lee CC, Iddings DM, Kavanagh M, Bilchik AJ. Management of T2 gallbladder cancer: are practice patterns consistent with national recommendations? Am J Surg 2007; 194: 820-825. Partovi S, Kohan AA, Zipp L, Faulhaber P, Kosmas C, Ros PR, Robbin MR. Hybrid PET/MR imaging in two sarcoma patients: clinical benefits and implic ations for future trials. Int J Clin Exp Med 2014; 7: 640-648. Lin H, Shen YC, Long HY, Wang H, Luo ZY, Wei ZX, Hu SQ, Wen FQ. Performance of osteopontin in the diagnosis of malignant pleural mesothelioma: a meta-analysis. Int J Clin Exp Med 2014; 7: 1289-1296. Nael K, Khan R, Choudhary G, Meshksar A, Villablanca P, Tay J, Drake K, Coull BM, Kidwell
Int J Clin Exp Med 2015;8(8):14416-14423
Diagnosis for pulmonary embolism with MRI CS. Six-minute magnetic resonance imaging protocol for evaluation of acute ischemic stroke: pushing the boundaries. Stroke 2014; 45: 1985-1991. [22] Puskas T, Henits I. Use of whole-body diffusionweighted magnetic resonance imaging for the diagnosis of and therapeutic response in multiple myeloma. Orv Hetil 2014; 155: 12411245.
14423
[23] Thompson J, Lawrentschuk N, Frydenberg M, Thompson L, Stricker P; USANZ. The role of magnetic resonance imaging in the diagnosis and management of prostate cancer. BJU Int 2013; 112: 6-20. [24] Filippi M, Charil A, Rovaris M, Absinta M, Rocca MA. Insights from magnetic resonance imaging. Handb Clin Neurol 2014; 122: 115-149.
Int J Clin Exp Med 2015;8(8):14416-14423
Original Article Diagnosis efficiency for pulmonary embolism using magnetic resonance imaging method: a meta-analysis Kuitao Yue Department of Medical Imaging, Weifang Medical University, Weifang 261053, Shandong, China Received November 5, 2014; Accepted January 10, 2015; Epub August 15, 2015; Published August 30, 2015 Abstract: PE (Pulmonary embolism, PE) is a common disease, usually caused by blockage of pulmonary artery and its branches due to exogenous or endogenous embolic obstruction. PE always be misdiagnosed in clinical. The aim of this study is to calculate the sensitivity and specificity of magnetic resonance imaging (MRI) in assessing the resectability of PE. In this study, a meta-analysis of the reported sensitivity and specificity of each study with 95% confidence intervals (CI) was performed. Five studies were included in the meta-analysis. The results indicated that the quality assessment scores ranged from 11 to 13, with a mean study quality score of 12. The sensitivity and specificity values including 95% CI at the patient level were calculated. The sensitivities ranged from 78% to 100%, and the specificity ranged from 99% to 100%. The pooled sensitivity value including 95% CI was 0.83 (0.78-0.88), and with inconsistency (I2) of 62.8%. The pooled specificity value including 95% CI was 0.99 (0.98-1.00), with inconsistency (I2) of 0.0%. Pooled positive likelihood ratio (PLR) (95% CI) was 70.22 (29.04-169.76), and the pooled negative likelihood ratio (NLR) (95% CI) was 0.19 (0.14-0.25). The overall diagnostic odds ratio (DOR) (95% CI) was 448.98 (163.47-1233.18). The summary receiver operating characteristic (SROC) data illustrated that the area under the curve (AUC) was 0.9852. In conclusion, the MRI method may be acts as a potential and assistant method for the PE diagnosis. Keywords: Pulmonary embolism, magnetic resonance imaging, diagnosis, meta-analysis
Introduction PE (Pulmonary embolism, PE) is a common disease, usually caused by blockage of pulmonary artery and its branches due to exogenous or endogenous embolic obstruction [1]. The PE could also result in pulmonary circulation disorders and blood flow interruption in lung tissue, and leading to lung tissue necrosis pathological changes, which is also called pulmonary infarction [2]. PE misdiagnosis often appears clinically. In recent years, CT (computed tomography) technology has been very mature one for the diagnosis of pulmonary embolism with good sensitivity and specificity [3]. Meanwhile, Magnetic Resonance Imaging (MRI) is also commonly used in the diagnosis of pulmonary embolism [4-6]. Compared to CT, MRI has the advantage that there is no ionizing radiation, nor the use of iodinated contrast media, which has small hazard to the patient’s health [7]. This meta-analy-
sis aims to investigate the diagnostic validity of MRI in PE. Methods Search strategy We searched PUBMED, EMBASE and Springer for relevant citations. The search was completed in June 2014 and covered published literature since January 1990 (Figure 1). The terms in the search were “Pulmonary Embolism” AND (“MR imaging” OR “MR” OR “MRI” OR “Magnetic Resonance Imaging” OR “Magnetic Resonance”). The reference lists of known previous reviews and of all of the primary studies included were examined to identify cited articles not found by the electronic searches (Figure 1). There were no restrictions regarding publication language. Eligibility criteria Titles and abstracts identified by the electronic search were analyzed by two independent
Diagnosis for pulmonary embolism with MRI nosed as diseased according to the index test). Two reviewers independently selected articles on the basis of the title and abstract; if one or both reviewers considered the study potentially eligible, the full article was evaluated by both reviewers. Two reviewers abstracted the data from each article and recorded data using a standardized form. Disagreement was resolved by means of consensus.
Figure 1. Flow chart of the study selection procedure.
reviewers (A.V. and V.G.), and full text versions of all citations that were likely to meet the selection criteria were obtained. Disagreements were resolved by consensus or after discussion with a third reviewer (J.Z.). Between-reviewer agreement was assessed by the kappa coefficient. Studies were selected for the review if 1) they were designed as diagnostic experimental research; 2) the research subjects were ≥18 years old suspected PE patients; 3) the diagnostic gold standard was CT; 4) the data of (true positive, TP), (false positive, FP), (true negative, TN), (false negative, FN) were provided or can be calculated. In cases of suspected duplicate publication, the most recent and complete version was selected. We also excluded studies for 1) there were no MRI check 48 hrs after CT diagnostic; 2) non-English documentations; 3) reviews, abstracts, isolated cases, commentaries, editorials or letters. Data extraction For each study, the following count data were entered into a 2×2 table: (a) true-positive results (disease cases in which findings of the reference standard and the index test were positive for the studied target condition), (b) false-negative results (diseased cases that were misdiagnosed as non-diseased according to the index test), (c) true-negative results (nondiseased cases that were correctly identified with the index test), and (d) false-positive results (non-diseased cases that were misdiag-
14417
Quality assessment Methodological quality was assessed independently by the same reviewers based on the Quality Assessment of Studies of Diagnostic Accuracy included on Systematic Reviews (QUADAS) guidelines [8] disagreement was resolved by consensus (Table 1). This evidencebased tool was developed specifically to assess the quality of diagnostic accuracy studies and includes 14 quality items. The 14 items, phrased as questions, are scored as “yes,” “no”, or “unclear”. The quality assessment score can range from 0 to 14, where 14 is the maximum attainable score (Table 1). A more detailed description of each item, together with a guide on how to score each item, is provided by Whiting et al. in 2003 [8]. Additionally, we noted whether the study design was prospective or retrospective. Statistical analysis This meta-analysis was performed with MetaDisc (1.4.0.0) statistical softwares [9]. Primary outcomes of this meta-analysis were summary receiver operating characteristic (SROC) curve, sensitivity (sensitivity, Sen), specificity (specificity, Spe), positive likelihood ratio (positive likelihood ratios, PLR), negative likelihood likelihood ratio (negative likelihood ratios, NLR), diagnostic tests odds ratio (diagnostic odds ratio, DOR) and their 95% confidence intervals (confidence intervals, CI). DOR ranges from 0 to infinity, the greater the DOR values, the powerful MRI diagnosis of pul-
Int J Clin Exp Med 2015;8(8):14416-14423
Diagnosis for pulmonary embolism with MRI Table 1. Quality assessment of the included articles Reference number of the included studies
QUADAS list item 7
13
14
15
16
1. Did the spectrum of patients represent the patients who will receive the test in practice?
+
+
+
+
+
2. Were selection criteria clearly described?
+
+
+
+
+
3. Is the reference standard likely to correctly classify the target condition?
+
+
+
+
+
4. Is the period between the reference standard and index test short enough to be reasonably sure that the target condition did not change between the 2 tests?
+
+
+
+
+
5. Did the entire sample or a random selection of the sample receive verification using a reference standard of diagnosis?
+
+
+
+
+
6. Did patients receive the same reference standard regardless of index test result?
+
+
+
+
+
7. Was the reference standard independent of the index test (i.e., index test did not form part of the reference standard)?
+
+
+
+
+
8. Was execution of the index test described in sufficient detail to permit replication of the test?
+
+
+
+
+
9. Was execution of the reference standard described in sufficient detail to permit its replication?
+
+
+
+
+
10. Were index test results interpreted without knowledge of results of the reference standard?
+
-
+
0
0
11. Were reference standard results interpreted without knowledge of results of the index test?
-
+
0
0
+
12. Were the same clinical data available when test results were interpreted as would be available when the test is used in practice?
+
0
0
0
0
13. Were uninterruptable/intermediate test results reported?
+
+
-
+
-
14. Were withdrawals from the study explained?
+
+
+
+
+
QUADAS: Quality Assessment of Diagnostic Accuracy Studies. +: YES; -: NO; 0: not clear.
14418
Int J Clin Exp Med 2015;8(8):14416-14423
Diagnosis for pulmonary embolism with MRI Table 2. Characteristics of each included study Author reference
year
Country
1
Kluge [5]
2006
Germany
1.5T
3-D FLASH
65
2
Revel [12]
2012
French
1.5T
3-D FSPGR /SSFP/GRE
274
3
Zhang [14]
2013
China
3T
3-D FLASH
27
18/9
38.9±14.4
4
Stein [13]
2010
USA
1.5T/3T
NA
279
NA
≥18
5
Pleszewski [11]
2006
Canada
1.5T
3-D SEGP
34
20/28
22-84
9
id
Magnetic MRI scan sequence field intensity
n
M/F
Age
TP
TN
FP
34/31
60.9±15.7
127/147
59.8±19.0
FN
15
43
0
4
87
170
1
16
24
3
0
0
59
201
2
17
23
0
2
3-D: 3-dimensional; FLASH: fast low-angle shot gradient-echo sequence; FSPGR: fast spoiled gradient echo; SSFP: Steady-state-free precession; GRE: gradient recalled echo; SEGP: spoiled echo gradient pulse; NA: Not available.
Results Literature search
Figure 2. Sensitivity and specificity of MRI detection for the diagnosis of PE. A. The sensitivity of MRI detection for the diagnosis of PE. B. The specificity of MRI detection for the diagnosis of PE. PE: Pulmonary embolism; CI: confidential interval; MRI: Magnetic Resonance Imaging.
The literature search identified 1542 articles (Figure 1). After screening of the titles and abstracts, 1231 items was remained because the duplication. Among the 1231 items, 1214 items were excluded because one or more inclusion criteria were not fulfilled. The remaining 17 articles were evaluated by reading the full text. Of these, 12 were excluded (the reasons are given in Figure 1). The remaining 5 studies were included in the meta-analysis [7, 13-16]. The study characteristics are shown in Table 2. Quality assessment
monary embolism recognition (discriminatory ability) [10]. SROC curve is based on the sensitivity and specificity, and area under the curve closer to 1, the higher the diagnostic validity [11]. Heterogeneity was tested by using the Higgins and Thompson test to calculate the I2 statistic [12]. This statistic uses the conventional Cochran Q statistic to calculate the percentage of tool variation heterogeneity on a scale ranging from 0% (no heterogeneity) to 100% (all variance due to heterogeneity). In contrast to the Cochran Q, the I2 is less affected by the number of studies included in a meta-analysis. If no or moderate heterogeneity is found (2≤50%), pooling is justified. 14419
The quality assessment scores of this study ranged from 11 to 13, with a mean study quality score of 12 (Table 2). All of the studies with the higher quality. MRI detection for sensitivity and specificity The results showed that all of the MRI detection with the relative higher sensitivity and specificity (Figure 2). The sensitivity and specificity values including 95% CI at the patient level were calculated. The sensitivities ranged from 78% to 100%, and the specificity ranged from 99% to 100%. The pooled sensitivity value including 95% CI was 0.83 (0.78-0.88), Q =10.75, df =4 (P=0.0295), inconsistency (I2) =62.8% (Figure 2A). The pooled specificity value including 95%
Int J Clin Exp Med 2015;8(8):14416-14423
Diagnosis for pulmonary embolism with MRI with the MRI method [19]. The method acts as a diagnostic tool depends on the sensitivity and the specificity for the disease [20]. The present meta-analysis study involves the studies of Magnetic Resonance Imaging (MRI) for diagnosis of PE. By analyzing the data in these studies, we found that MRI is a highly accurate diagnostic efficiency with a pooled sensitivity of 83% and a pooled specificity of 99%. Furthermore, the positive likelihood ratios (PLR), negative likelihood ratios (NLR), diagnostic odds ratio Figure 3. PLR and NLR of MRI detection for the diagnosis of PE. A. PLR of MRI (DOR) and 95% confidendetection for the diagnosis of PE. B. NLR of MRI detection for the diagnosis of PE. ce intervals (CI, 29.04 to PE: Pulmonary embolism; PLR: positive likelihood ratio; NLR: negative likelihood 169.76) were also higher ratio; CI: confidential interval; MRI: Magnetic Resonance Imaging. enough for the method applies as a diagnostic CI was 0.99 (0.98-1.00), Q =1.21, df =4 tool in clinical. For the area under the curve (P=0.8762), inconsistency (I2) =0.0% (Figure analysis, the AUC level was also close to 1, 2B). which suggests that MRI method with a better diagnostic efficiency. Pooled PLR and NLR The MRI method has been used in the assisOur results indicated that the pooled positive tant diagnosis for many other diseases. Nael et likelihood ratio (PLR) (95% CI) =70.22 (29.04al. [21] used the six-minute magnetic reso169.76), Q =3.49, df =4 (P=0.4802), inconsisnance imaging protocol to evaluate the acute tency (I2) =0.0% (Figure 3A). The pooled negaischemic stroke, which results in significant tive likelihood ratio (NLR) (95% CI) =0.19 (0.14reduction in scan time rivaling that of the multi0.25), Q =3.86, df =4 (P=0.4258), inconsistenmodal computed tomographic protocol. Puskas cy (I2) =0.0% (Figure 3B). et al. [22] indicated that the diffusion-weighted magnetic resonance imaging with conventional DOR and SROC curves sequences is a useful and promising functional imaging modality in the early diagnosis of Overall diagnostic odds ratio (DOR) and summyeloma multiple. Thompson et al. [23] also mary receiver operating characteristic (SROC) found that the MRI plays a useful role in the curves for the diagnostic performance of MRI diagnosis and management of the prostate detection was drew. The DOR (95% CI) =448.98 cancer, and appears to be of value in planning (163.47-1233.18), Q=1.01, df =4 (P=0.9077), dosimetry in men undergoing radiotherapy, and 2 inconsistency (I ) =0.0% (Figure 4A). The SROC in guiding selection for and monitoring on active data illustrated that the AUC was 0.9852 surveillance. Filippi et al. [24] also analyzed the (Figure 4B). application and exploring the insights of the magnetic resonance imaging for some other Discussion diseases, and pointed out that all of the conventional and non-conventional MRI techniques Clinically, the diagnosis of PE always be perwill be benefit from the use of high-field MRI formed by using the CT method. However, the systems (3.0T or more). CT may cause some side-effects compared
14420
Int J Clin Exp Med 2015;8(8):14416-14423
Diagnosis for pulmonary embolism with MRI
Figure 4. Overall DOR and SROC curves for the diagnostic performance of MRI detection in PE. A. Overll DOR curves for the diagnostic performance of MRI detection. B. SROC curves for the diagnostic performance of MRI detection. DOR: diagnostic odds ratio; SROC: summary receiver operating characteristic; AUC: area under the curve.
Though the MRI has been used in many diseases, there are also some disadvantages as the followings: (1) The samples are relative small, which may affect the statistical efficiency for the diagnosis; (2) Though there are not the threshold effect and no-threshold effects, the heterogenicity may be exist among the researches; (3) This study only included the published studies, the un-published documents may be omitted, which may also cause the
14421
false positive and amplify the diagnostic efficiency of MRI. In conclusion, the data of this meta-analysis suggests that the MRI method for PE diagnosis with a higher sensitivity and specificity, and also with lower misdiagnosis rate and missed diagnosis rate. The MRI method may be acts as a potential and assistant method for the PE diagnosis.
Int J Clin Exp Med 2015;8(8):14416-14423
Diagnosis for pulmonary embolism with MRI Disclosure of conflict of interest None. Address correspondence to: Dr. Kuitao Yue, Department of Medical Imaging, Weifang Medical University, 7166 Baotong Jie (West), Weifang 261053, Shandong, China. Tel: +86-0536-2603780; Fax: +86-0536-2603780; E-mail: yuektwf@yeah. net
[12]
[13]
References [1]
Li X, Li Q, Miao Y, Xu H, Liu Y, Qiu X, Wang EH. A case of renal antiomyolipoma with intracardiac extension and asymptomatic pulmonary embolism. Int J Clin Exp Pathol 2013; 6: 11801186. [2] Moutzouris JP, Chow V, Chiang Yong AS, Chung T, Naganathan V, Kritharides L, Chwan AC. Acute pulmonary embolism in individuals aged 80 and older. J Am Geriatr Soc 2014; 62: 2004-2006. [3] Schibany N, Fleischmann D, Thallinger C, Schibany A, Hahne J, Ba-Ssalamah A, Herold CJ. Equipment availability and diagnostic strategies for suspected pulmonary embolism in Austria. Eur Radiol 2001; 11: 2287-2294. [4] Yuan C, Zhu H, Song D, Wei W, Zhu R, Mei X, Zou J, Yang H. Int J Clin Exp Med 2014; 7: 163169. [5] Remy-Jardin M, Remy J, Baghaie F, Fribourg M, Artaud D, Duhamel A. Clinical value of thin collimation in the diagnostic workup of pulmonary embolism. Am J Roentgenol 2000; 175: 407411. [6] Schoepf UJ, Costello P. CT angiography for diagnosis of pulmonary embolism: State of the art. Radiology 2004; 230: 329-337. [7] Kluge A, Luboldt W, Bachmann G. Acute pulmonary embolism to the subsegmental level: diagnostic accuracy of three MRI techniques compared with 16-MDCT. AJR Am J Roentgenol 2006; 187: W7-14. [8] Whiting P, Rutjes AW, Reitsma JB, Bossuyt PM, Kleijnen J. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003; 3: 25. [9] Zamora J, Abraira V, Muriel A, Khan K, Coomarasamy A. Meta-DiSc: a software for metaanalysis of test accuracy data. BMC Med Res Methodol 2006; 6: 31. [10] Glas AS, Lijmer JG, Prins MH, Bonsel GJ, Bossuyt PM. The diagnostic odds ratio: a single indicator of test performance. J Clin Epidemiol 2003; 56: 1129-1135. [11] Moses LE, Shapiro D, Littenberg B. Combining independent studies of a diagnostic test into a
14422
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
summary ROC curve: Data-analytic approaches and some additional considerations. Stat Med 1993; 12: 1293-1316. Devillé WL, Buntinx F, Bouter LM, Montori VM, de Vet HC, van der Windt DA. Conducting systematic reviews of diagnostic studies: didactic guidelines. BMC Med Res Methodol 2002; 2: 9. Pleszewski B, Chartrand-Lefebvre C, Qanadli SD, Dery R, Perreault P, Oliva VL. Gadoliniumenhanced pulmonary magnetic resonance angiography in the diagnosis of acute pulmonary embolism: a prospective study on 48 patients. Clin Imaging 2006; 30: 166-172. Revel MP, Sanchez O, Couchon S, Planquette B, Hernigou A, Niarra R, Meyer G, Chatellier G. Diagnostic accuracy of magnetic resonance imaging for an acute pulmonary embolism: results of the ‘IRM-EP’ study. J Thromb Haemost 2012; 10: 743-750. Stein PD, Chenevert TL, Fowler SE, Goodman LR, Gottschalk A, Hales CA, Hull RD, Jablonski KA, Leeper KV Jr, Naidich DP, Sak DJ, Sostman HD, Tapson VF, Weg JG, Woodard PK; PIOPED III (Prospective Investigation of Pulmonary Embolism Diagnosis III) Investigators. Gadolinium-enhanced magnetic resonance angiography for pulmonary embolism: a multicenter prospective study (PIOPED III). Ann Intern Med 2010; 152: 434-443, W142-1433. Zhang LJ, Luo S, Yeh BM, Zhou CS, Tang CX, Zhao Y, Li L, Zheng L, Huang W, Lu GM. Diagnostic accuracy of three-dimensional contrast-enhanced MR angiography at 3-T for acute pulmonary embolism detection: Comparison with multidetector CT angiography. Int J Cardiol 2013; 168: 4775-4783. Raznatovic ZJ, Zaric ND, Galun DA, Lekic NS, Micev M, Djordjevic VR, Djurasic LM, Kerkez MD. Multiple port-site metastasis of incidental gallbladder carcinoma after laparoscopic cholecystectomy. Acta Chir Lugosl 2012; 59: 105109. Wright BE, Lee CC, Iddings DM, Kavanagh M, Bilchik AJ. Management of T2 gallbladder cancer: are practice patterns consistent with national recommendations? Am J Surg 2007; 194: 820-825. Partovi S, Kohan AA, Zipp L, Faulhaber P, Kosmas C, Ros PR, Robbin MR. Hybrid PET/MR imaging in two sarcoma patients: clinical benefits and implic ations for future trials. Int J Clin Exp Med 2014; 7: 640-648. Lin H, Shen YC, Long HY, Wang H, Luo ZY, Wei ZX, Hu SQ, Wen FQ. Performance of osteopontin in the diagnosis of malignant pleural mesothelioma: a meta-analysis. Int J Clin Exp Med 2014; 7: 1289-1296. Nael K, Khan R, Choudhary G, Meshksar A, Villablanca P, Tay J, Drake K, Coull BM, Kidwell
Int J Clin Exp Med 2015;8(8):14416-14423
Diagnosis for pulmonary embolism with MRI CS. Six-minute magnetic resonance imaging protocol for evaluation of acute ischemic stroke: pushing the boundaries. Stroke 2014; 45: 1985-1991. [22] Puskas T, Henits I. Use of whole-body diffusionweighted magnetic resonance imaging for the diagnosis of and therapeutic response in multiple myeloma. Orv Hetil 2014; 155: 12411245.
14423
[23] Thompson J, Lawrentschuk N, Frydenberg M, Thompson L, Stricker P; USANZ. The role of magnetic resonance imaging in the diagnosis and management of prostate cancer. BJU Int 2013; 112: 6-20. [24] Filippi M, Charil A, Rovaris M, Absinta M, Rocca MA. Insights from magnetic resonance imaging. Handb Clin Neurol 2014; 122: 115-149.
Int J Clin Exp Med 2015;8(8):14416-14423
