Brief cardiovascular imaging with pocket-size ultrasound devices improves the accuracy of the initial assessment of suspected pulmonary embolism

Brief cardiovascular imaging with pocket-size ultrasound devices improves the accuracy of the... Pulmonary embolism onset is frequently neglected due to the non-specific character of its symptoms. Pocket-size imaging devices (PSID) present an opportunity to implement imaging diagnostics into conventional physical examination. The aim of this study was to test the hypothesis that supplementation of the initial bedside assessment of patients with suspected pulmonary embolism (PE) with four-point compression venous ultrasonography (CUS) and right ventricular size assessment with the use of PSID equipped with dual probe could positively influence the accuracy of clinical predictions. A single- centre, prospective analysis was conducted on 100 patients (47 men, mean age 68 ± 13 years) with suspected PE. Clinical assessment on the basis of Wells and revised Geneva score and physical examination were supplemented with CUS and RV measurements by PSID. The mean time of PSID scanning was 4.9 ± 0.8 min and was universally accepted by the patients. Fifteen patients had deep venous thrombosis and RV enlargement was observed in 59 patients. PE was confirmed in 24 patients. If the both CUS was positive and RV enlarged, the specificity was 100% and sensitivity 54%, ROC AUC 0.771 [95% CI 0.68–0.85]. The Wells rule within our study population had the specificity of 86% and sensitivity of 67%, ROC AUC 0.776 (95% CI 0.681–0.853, p < 0.0001). Similar values calculated for the revised Geneva score were as follows: specificity 58% and sensitivity 63%, ROC AUC 0.664 (95% CI 0.563–0.756, p = 0.0104). Supplementing the revised Geneva score with additional criteria of CUS result and RV measurement resulted in significant improvement of diagnostic accuracy. The difference between ROC AUCs was 0.199 (95% Cl 0.0893–0.308, p = 0.0004). Similar modification of Wells score increased ROC AUC by 0.133 (95% CI 0.0443–0.223, p = 0.0034). Despite the well-acknowledged role of the PE clinical risk assess- ment scores the diagnostic process may benefit from the addition of basic bedside ultrasonographic techniques. Keywords Pocket-size imaging devices · Pulmonary embolism · Compression ultrasound test · Wells rule · Revised Geneva score Introduction is estimated that over 600,000 cases of PE are diagnosed in United States every year [2]. Majority of death related Pulmonary embolism (PE) is a common, oftentimes mis- cases of PE occur when the condition was not diagnosed diagnosed emergency cardiovascular state burdened with in time [3]. potentially fatal consequences. Symptoms are non-specific According to current guidelines [1, 4–6], the first step in varying from chest pain, shortness of breath, cough and the diagnostic protocol of PE is determination of its clini- hemoptysis to syncope and cardiac arrest [1]. Annually cal probability. For this purpose several scoring systems are over 10 million patients in the United States seek medical suggested. The Wells rule and the modified Geneva score help with the complaints of dyspnea, chest pain or both. It are among the most validated scales used for PE clinical risk assessment. In order to reduce the number of unnecessary procedures, guidelines recommend first to identify patients * Dominika Filipiak-Strzecka with a very low probability of the condition, so that referral dominika.filipiak@gmail.com for further diagnostics can safely be withheld. The remain- 1 ing patients are categorized into one of the recommended Department of Cardiology, Bieganski Hospital, Medical diagnostic paths; which in most cases involves further blood University of Lodz, Kniaziewicza 1/5, 91-347 Lodz, Poland Vol.:(0123456789) 1 3 The International Journal of Cardiovascular Imaging testing or—perceived as a golden standard—Multi-Slice CT further diagnostic procedures. All patients underwent clini- Angiography. Undoubtedly, unjustified referral for diagnos- cal assessment in the emergency room on the basis of the tic procedures involving a radiographic exposure results in Wells rule and the revised Geneva score, including the eval- excessive cost of care, increased use of hospital resources, uation of medical history (previous episode of pulmonary significant risk of contrast-associated acute renal failure and embolism or deep vein thrombosis, recent surgery), current exposure-dependent malignancies [7]. On the contrary, in medical condition (active cancer, heart rate, clinical signs certain cases even with calculated low clinical risk (evalu- of DVT) and reported symptoms (hemoptysis) (Table 1). ated by established scoring systems) postponing of final Subsequently, the regular physical examination was sup- diagnosis and late introduction of proper treatment may plemented with short, focused bedside ultrasonographic cause fatal consequences. For this reason a persistent search assessment consisting of four-point compression venous for new diagnostic modalities improving initial assessment ultrasonography and measurements of the right ventricle of patient condition, which could increase the accuracy of performed by cardiology resident with the use of PSID. grading clinical risk of PE is continued. However, a novel The operator’s training in echocardiography was included as diagnostic protocol needs to meet certain requirements in part of her residency program and was based on six-month order to gain clinical acceptance: it has to be widely avail- rotation in the echo lab. It included conducting and ana- able, quick to implement and should be cost effective. A lyzing transthoracic echocardiographic examinations under method/strategy where all patients suspected of PE should specialist supervision, as well as a basic vascular examina- undergo ultrasonographic examination as part of initial eval- tions such as compression ultrasound test. This study was uation proves challenging in the urgency of the emergency conducted in accordance with the amended Declaration of department. Certain constraints such as urgent transportation Helsinki. Informed consent was obtained from individual to echocardiographic laboratory, use of high-end equipment participants included in the study. The study protocol was and need for trained echocardiographer generate high costs. approved by bioethics committee of our institution (Decision The trend to miniaturize echocardiographic devices has led No. RNN/8/10/KE with the Supplement No. KE/2011/15). to the creation of pocket-size imaging devices (PSID). With the ultraportability being their biggest advantage PSIDs may be used in almost every clinical setting, including the Pocket‑size imaging device emergency room. Moreover, with an interface optimized for ease of operation, PSIDs present easy access to a wide The pocket-size imaging device used in this study was range of users. It has been shown, that with targeted training V-Scan (GE Vingmed Ultrasound, Horten, Norway) basic information can be obtained with the use of PSIDs by equipped with dual probe, combining two transducers in one non-echocardiographers [8–10]. We hypothesized, that in probe- the phased array (frequency range of 1.7–3.8 MHz, patients with suspected pulmonary embolism augmentation image sector limited to 75°, depth range 6–24 cm) and the of initial emergency room assessment with four-point com- linear probe (frequency range of 3.4–8.0  MHz, aperture pression venous ultrasonography (CUS) and right ventricu- width of 2.9 cm, maximum depth of 8 cm). The four-point lar size assessment with the use of PSID equipped with dual compression ultrasound tests were performed using linear probe can improve the diagnostic accuracy of established probe and the vascular preset, whereas the right ventricle clinical prediction rules. assessment with the phased array probe and cardiac preset. Materials and methods Compression ultrasonography This was a prospective study conducted between February The examination was performed in the supine position. The femoral artery was assessed from the level just distal 2015 and May 2016. 100 consecutive patients (47 males, mean age 68 ± 13 years) who were referred to our depart- to the inguinal ligament to the 2 cm distal to the junction of the common femoral vein and the greater saphenous vein. ment during office hours (8 a.m.–4 p.m.) or during the 24-h medical shift of the resident performing examination (D.S.) The collapsing of common and deep femoral veins was evaluated. The popliteal vein was assessed from the level of were included in our analysis. The inclusion criterion was the suspicion of pulmonary embolism based on the medical popliteal fossa up to the level of its trifurcation. The direct pressure with the use of transducer was applied in order to history and basic physical examination only. All patients reported dyspnea as their main symptom, in some cases completely compress the vein. If the vein was compressed completely, then a DVT at this site was ruled out. The lack with concomitant chest pain (49%), cough (17%) or tachy- cardia (27%). If such a diagnosis was suggested, in all cases of possibility to completely compress the vein was treated as a positive test result (Fig. 1). the investigator (D.S.) was informed and continued with 1 3 The International Journal of Cardiovascular Imaging Table 1 The original and modified versions of Wells rule and revised Geneva score used in the study Wells rule Original version Modified version Points % of patients Previous PE or DVT Previous PE or DVT 1.5 7 HR > 100 b.p.m HR > 100 b.p.m 1.5 27 Surgery or immobilization within the past Surgery or immobilization within the past 4 weeks 1.5 9 4 weeks Hemoptysis Hemoptysis 1 0 Active cancer Active cancer 1 2 Clinical signs of DVT Positive CUS result 3/4 9 15 Alternative diagnosis less likely than PE Alternative diagnosis less likely than PE 3 16 – RV enlargement 1 59 Basal diameter (4CH) ≤ 47 mm 2 35 Basal diameter (4CH) > 47 mm 19 Revised Geneva score Original version Modified version Points % of patients Previous PE or DVT Previous PE or DVT 3 7 HR HR 3 15 75–94 b.p.m 75–94 b.p.m 5 27 ≥ 95 b.p.m ≥ 95 b.p.m Surgery or fracture within the past month Surgery or fracture within the past month 2 9 Hemoptysis Hemoptysis 2 0 Active cancer Active cancer 2 2 Unilateral lower limb pain Positive CUS result 3 10 15 Pain on lower limb deep venous palpation and 4 6 unilateral oedema Age > 65 years Age > 65 years 1 63 – RV enlargement 1 59 Basal diameter (4CH) ≤ 47 mm 2 35 Basal diameter (4CH) > 47 mm 19 Fig. 1 Compression ultrasound test of common femoral vein per- baseline; panel b, d—compression. Panel b—vein completely com- formed with the use of PSID in two patients: without thrombosis pressed; panel d—abnormal study indicating venous thrombosis; A (panel a, b) and with venous thrombosis (panel c, d). Panel a, c— femoral artery; V femoral vein Assessment of the right ventricle ventricle-focused 4-chamber apical view and proximal right ventricle outflow diameter measured in long axis parasternal Two linear measurements of the right ventricle were view (Fig. 2). Right ventricle enlargement was defined as the performed: right ventricular basal diameter in the right 1 3 The International Journal of Cardiovascular Imaging right ventricular basal diameter > 41 mm and/or proximal Statistical analysis right ventricle outflow diameter > 35 mm [11]. Continuous and categorical variables are expressed as Modified clinical prediction rules mean ± SD and as percentages (%), respectively. To assess the diagnostic value of different tests the ROC analysis on The points “clinical signs of DVT” in the Wells rule and the basis of DeLong at al. methodology was performed. “pain on lower limb deep venous palpation and unilateral Sensitivity, specificity and overall diagnostic accuracy were oedema” in the revised Geneva score were changed into compared with the use of N-1 Chi square test. A difference ‘positive CUS result’. Additional points were also given, was considered statistically significant when p < 0.05. Cal- when the RV enlargement was detected. Subsequently, we culations were performed with the use of MedCalc Software tested the hypothesis that incorporating the results of brief version 12.2.1.0. ultrasonographic assessment into the risk scale improves its diagnostic value. The score of this additional criterion was retrospectively determined on the basis of our population, Results to achieve the best ROC AUC. (Table 1). The assessment by recommended clinical prediction Diagnosis of pulmonary embolism rules Final diagnosis was established in accordance with the algo- Pulmonary embolism was eventually confirmed by con- rithms recommended by the ESC guidelines, on the basis of trast-enhanced chest computed tomography in 24 patients. clinical gold-standard including all necessary examinations The patients’ characteristic is presented in Table 1. None [1]. Two patients with the high clinical probability of PE of the patients had been given vasopressors on admission. (according to the Wells score and revised Geneva score) In one patient the systolic blood pressure was < 90 mmHg. immediately underwent CT-angiography. The remaining 98 17 patients had the history of chronic lung disease, in two patients with low or intermediate clinical probability had patients atrial septal defect was present, three patients had d -dimer plasma level initially assessed. In 47 patients with the tricuspid regurgitation diagnosed. Among final diagno- normal d -dimer plasma level (cut-off value: 500 µg/L) the ses other than pulmonary embolism coronary artery dis- pulmonary embolism was ruled out. In the remaining 51 ease (25 patients,) chronic heart failure exacerbation (22 patients CT angiography was performed (Fig. 3). patients), pneumonia (11 patients), heart rhythm disorders (seven patients), valve disease (six patients) were most often detected. According to the three-category Wells rule the clinical risk of PE was estimated as low in 74 patients, among which ten were eventually diagnosed with PE, as Fig. 2 RV linear measure- ments: a RV focused 4-chamber apical view, basal dimension; c parasternal long-axis view, proximal RV outflow diameter 1 3 The International Journal of Cardiovascular Imaging Fig. 3 Diagram presenting the diagnostic path of patients included in the study intermediate in 24 patients (12 cases of confirmed PE), as Modified clinical prediction rules high in two patients (PE confirmed). In compliance with revised Geneva score, 54 patients had low clinical risk of On the basis of our study population we have established PE (in nine patients PE was confirmed), 44 patients-inter - that in case of positive lower extremity ultrasound the best mediate (13 cases of PE); two patients-high (PE confirmed) ROC AUC were achieved for the values 4 points for Wells (Figs. 4 and 5). The diagnostic accuracy of Wells rule and score and 9 points for revised Geneva score. For the RV revised Geneva score based on our study population is pre- enlargement we tested 1-point and 2-points scoring. The sented in Table 2. highest ROC AUC was obtained with two grade model: 1 point was added, if the basal diameter measured in the Scanning with the use of PSID four chamber view was within the range of 42–47 mm, 2 points were given when this diameter exceeded 47  mm. The mean time of scanning with the use of PSID was Supplementing the revised Geneva score with additional 4.9 ± 0.8 (95% CI 4.7–5.0) min and was universally accepted criteria of positive CUS test and RV enlargement resulted by patients. One patient was excluded from the analysis, as in significant improvement of diagnostic accuracy of this the unilateral visualisation of popliteal vessels with the use score- difference between areas 0.212 (95% Cl 0.100–0.325, of PSID was impossible. In two patients with the history of p < 0.0001), as presented on the graph. The overall diagnos- thoracotomy obtaining parasternal view proved impossible tic accuracy improved from 59 to 94% (p = 0.02). Similar and RV size was determined in apical view exclusively. Fif- modification of Wells score increased ROC AUC by 0.138 teen patients had the deep vein thrombosis (five cases proxi- (95% CI 0.0429–0.223, p = 0.0045), the overall diagnostic mal) detected in compression ultrasound test, whereas RV accuracy from 81 to 93% (p = 0,012). Modification of both enlargement was observed in 59 patients. Table 3 presents scales resulted in statistically significant improvement of the diagnostic accuracy of the PSID scanning for identifica - specificity but not sensitivity. (Table  2; Figs. 6 and 7). tion of patients with PE calculated for various criteria of test positivity. 1 3 The International Journal of Cardiovascular Imaging Fig. 4 Clinical risk of PE esti- mated on the basis of three cat- egory Wells rule (Panel a) and revised Geneva score (Panel c); for modified Wells rule (Panel b) the following threshold were established: low 0–1, intermedi- ate 2–6, high ≥ 7; for modified revised Geneva score (Panel d) the following threshold were established: low 0–3, intermedi- ate 4–10, high ≥ 11 Table 2 The diagnostic accuracy of Wells rule, revised Geneva score, modified Wells rule and modified revised Geneva score (supplemented with the PSID test results) based on the study population Sensitivity Specificity (%) PPV (%) NPV (%) Overall diagnostic ROC AUC (%) (95% CI) (95% CI) accuracy (%) (95% CI, p) Wells rule ≥ 2 66.7 85.5 59 89 81 0.776 (44.7–84.4) (75.6–92.5) (39–78) (80–95) (0.681 to 0.853, p < 0.0001) Revised Geneva score ≥ 4 62.5 57.9 35 82 59 0.664 (0.563 to 0.756, p = 0.0104) (40.6–81.2) (46.0–69.1) (20–54) (70–90) Modified Wells rule ≥ 5 70.8 98.7 94 92 93 0.914 (48.9–87.4) (92.9–100.0) (73–100) (83–97) (0.841 to 0.961, p < 0.0047) Modified revised Geneva 75.0 88.2 67 92 94 0.877 (0.796 to 0.934, p < 0.0001) score ≥ 7 (53.3–90.2) (78.7–94.4) (46–84) (83–97) Thresholds: Wells rule ≥ 2, revised Geneve score ≥ 4, modified Wells rule ≥ 5, modified revised Geneve score ≥ 7; determined from the ROC curves ROC AUC area under receiver operating characteristic curve, NPV negative predictive value, PPV positive predictive value 1 3 The International Journal of Cardiovascular Imaging Fig. 5 Clinical risk of PE esti- mated on the basis of two cat- egory Wells rule (Panel a) and revised Geneva score (Panel c); for modified Wells rule (Panel b) the following threshold were established: PE unlikely 0–3,5, PE likely ≥ 4; for modified revised Geneva score (Panel d) the following threshold were established: PE unlikely 0–6,5, PE likely ≥ 7 Table 3 The diagnostic accuracy of the PSID scanning for identification of patients with PE Sensitivity (95% CI) Specificity (95% CI) PPV (95% CI) NPV (95% CI) ROC AUC (95% CI, p) Positive CUS 54% (33–74) 97% (91–100) 87% (58–99) 87% (78–93) 0.758 (0.662 to 0.838, p < 0.0001) Proximal DVT diagnosed in 21% (7–42) 100% (95–100) 100% (48–100) 80% (71–88) 0.604 (0.501 to 0.701, CUS p = 0.0139) RV enlargement 92% (73–99) 51% (40–63) 37% (25–51) 95% (84–99) 0.715 (0.616 to 0.801, p < 0.0001) RV enlargement (basal 4CH) 92% (73–99) 51% (40–63) 37% (25–51) 95% (84–99) 0.746 (0.649 to 0.828, 2-Point scoring (42–47; > 47) p < 0.0001) CUS and RV enlargement 54% (33–74) 100% (95–100) 100% (74–100) 87% (79–94) 0.771 (0.676 to 0.849, p < 0.0001) CUS or RV enlargement 92% (73–99) 49% (37–60) 36% (24–50) 95% (83–99) 0.702 (0.602 to 0.789, p < 0.0001) ROC AUC ar ea under receiver operating characteristic curve, CUS compression ultrasound test, DVT deep vein thrombosis, NPV negative pre- dictive value, PPV positive predictive value, RV right ventricle 1 3 The International Journal of Cardiovascular Imaging venous thrombosis (DVT) are considered as a continuum of the same clinical entity, namely venous thromboembo- lism. Even up to 90% of pulmonary emboli may arise from lower limbs or pelvis deep venous thrombosis [14]. The application of compression ultrasound in detection of the deep vein thrombosis in patients with suspected PE has been previously proposed and mentioned in recent guide- lines as a method of reducing the number of CT angiog- Fig. 6 ROC curves comparison between revised Geneva score and raphy in appropriate patients. An early Dutch study based modified revised Geneva score with additional criteria of CUS and on the patients with suspected PE indicated that the use RV enlargement of compression US reduced the need for other imaging by 22%, at the expense of 2–4% of patients being unnecessar- ily treated for venous thromboembolism [15]. Data elic- ited from the metaanalysis indicate that CUS sensitivity ranged from 23 to 58%, while its specificity ranged from 89 to 99% [16]. Additionally CUS is commonly accepted as a non-invasive diagnostic modality, which may prove particularly relevant in patients with relative contraindica- tions to CT, such as chronic kidney disease and creatinine clearance below 30 mL/min, allergy to iodinated contrast dye, pregnant women or younger patients, in which the reduction of irradiation is desirable. The argument of limited cost-effectiveness was raised Fig. 7 ROC curves comparison between Wells rule and modified Wells rule with additional criteria of CUS and RV enlargement against the implementation of CUS in routine diagnostic process as it would require patient transportation to the ultra- sound examination lab, specialized workforce and equip- ment. Adversely, according to ESC, examination with the Discussion use of PSID should not be treated as a separate procedure but rather as an augmentation of physical examination. PSID To the best of our knowledge our study is the first to report examination can be performed at the point of care. In con- the diagnostic potential of brief scanning with the use of temporary clinical reality in which the concept of FOCUS PSID in the initial assessment of patients with suspected is increasingly recognized and basics of ultrasonographic pulmonary embolism. The main findings can be summa - examination are becoming a vital part of numerous medical rized as follows: (i) expanding the initial patient assess- professionals’ training, one can expect that such diagnostic ment in the ER with the elements of ultrasonographic approach may not require additional personnel apart from imaging did not excessively prolong physical examination the attending physician. and was universally accepted by the patients (ii) simulta- In agreement with the current clinical guidelines we neous RV enlargement and positive CUS result identified have assumed in our group positive proximal CUS result with the use of PSID has a very high positive predictive would allow for instantaneous (on the level of ER examina- value for PE (iii) the risk of PE assessed as low in accord- tion) identification of PE in 5 (5%) patients. Among them ance with the Wells and Geneva modified scores does not in four patients with either low or intermediate clinical risk rule out the possibility of PE diagnosis (iv) the diagnostic as estimated by means of Wells and revised Geneva scale value of the Well’s rule and the revised Geneva score can PSID examination would help to avoid a prolonged wait for be significantly improved by implementing the results of d -dimer test results and subsequent CT. In one patient with PSID examination in the clinical prediction rules criteria. high clinical risk, immediate PE diagnosis would eliminate Prompt diagnosis of pulmonary embolism still poses a the need for CT scanning along with the potentially danger- challenge to clinicians, in spite of all the available mod- ous need for transportation. What is more, a DVT detected ern diagnostic procedures. Based on the autopsy findings, in CUS (also distal) accompanied by the RV enlargement diagnosis of PE is missed in up to 30–50% of patients [12]. in 100% of cases was related with the presence of thrombi 2/3 of deaths associated with PE occurs within the first in pulmonary circulation. One may hypothesize that also hours from the symptoms manifestation [13]. For this rea- in the group of patients with the confirmed distal DVT, an son any improvement in diagnostic accuracy and hastening additional screening for the RV enlargement could eliminate of the whole diagnostic process is essential.PE and deep the need for blood testing/CT. 1 3 The International Journal of Cardiovascular Imaging Although, in accordance with the guidelines, echocar- very high; only one patient with diagnosed PE could not be diography does not play an essential role in the diagnos- diagnosed with RV pathology. tic process of PE, it can undoubtedly prove useful as a Nazerian et al. [33] presented an appealing approach, in method of treatment progress assessment. Furthermore, it which they proved that enhancing the Wells rule with lung was previously confirmed that focused echocardiographic and lower limb venous ultrasound improved the diagnostic assessment, as a part of multiorgan bedside ultrasonogra- value of this scale. However, their approach still involved phy can improve clinical evaluation of patients with sus- patient transportation, the use of high-end equipment and pected PE prior to definitive imaging [ 17]. Pathologies required involvement of a specialist in ultrasonography. One described by the incorrect values of the ratio of RV to left of the most highly regarded advantages of clinical prediction ventricular end-diastolic diameter; RV systolic pressure, rules is the easiness of their application and the immediate tricuspid annular plane systolic excursion and inferior result. PSID examination shares these features as it can be vena cava collapsibility were confirmed to be related with performed at any point of care, including the ER. Further- the increased mortality during the course of acute PE [18]. more PSIDs can be operated by less experienced medical Parameters related with RV dysfunction proved to have professionals who after the completion of the short training relatively high specificity while being burdened with low should be able to perform a reliable, specifically-aimed ultra- sensitivity [19]. In the study conducted by Kuznicka et al. sographic screening [8–10, 34–36]. Importantly, according [20] aimed at the assessment of the frequency of patholog- to the ESC guidelines PSID screening should be integrated ical findings in echocardiographic examination in patients into the routine physical examination rather than treated as with the confirmed PE, RV enlargement was relatively a separate diagnostic procedure. common, particularly in the high-risk patients. However, Latest and most advanced PSIDs are equipped with a incorrect ventricular ratio criterion was not fulfilled in all dual-probe, which shares the advantages of a linear and such cases. Apart from that, RV enlargement as a quantita- sector probe in one ultraportable tool. Older generations tive parameter is in our opinion easier to objectively assess of PSID were not perfectly suited for the vascular imag- than qualitative criteria such as free wall hypokinesis or ing and significant shortcomings in this area were present. paradox movement of intraventricular septum. Due to the Clinicians had to overcome the obstacles of the minimal above mentioned rationale RV enlargement was chosen for depth, image sector size and insufficient probe frequency the purpose of RV function evaluation. [37]. PSID equipped with dual-probe appears to be capable RV dilatation as diagnosed during echocardiographic of being successfully utilized in new clinical applications. examination in patients with pulmonary embolism has been Importantly, practicality of ultraportable ultrasound may suf- previously demonstrated to be related with the permanent fer from limited imaging capabilities. Although in all 100 RV dysfunction, RV failure, recurrent pulmonary embolism patients studied we were able to visualise RV at least partly, and death [19–22]. Although evaluating RV size and sys- and sufficiently for measurements, this may not be possible tolic function is not sufficient to make a direct diagnosis in all-comers population in clinical setting. Lower extrem- of PE it could provide an additionally valuable evidence in ity vein assessment may also exceed the capabilities of the some patients. Should the dilated RV be detected during the device in some patients (in our study group ca. 1%). For this bedside echocardiographic examination in a high-risk PE reason the supplementary role of PSID examination needs patient, the proper treatment introduction should be hastened to be re-emphasized; final diagnosis should be obtained on with the improvement of morbidity and mortality [13]. the basis of the complete set of clinical data. In our study population, RV enlargement was a relatively common finding. Although such diagnosis may be associ- ated with worse prognosis in patient with the suspected PE, Conclusion it is important to point out that the majority of causes of RV enlargement in the study group was not related with PE. Despite the well-established value of the PE clinical predic- Thus, RV enlargement alone should not alter the diagnostic tion rules, the diagnostic process of patients with suspected process and trigger prompt CT-scan. Nevertheless, the dis- PE benefits from the addition of brief assessment with the cussed parameter is useful as one of the factors reflected in use of PSID. risk scale. It was previously established that trained emergency Limitations physician was able to perform a reliable evaluation of the RV dysfunction during a bedside examination [13, 23–29]. This is a single center study with relatively small study The efficacy of PSID screening in the assessment of the RV population. All of the examinations with the use of PSID dilatation was also confirmed [30– 32]. In our study popula- were performed by the same cardiology resident, and for this tion the prevalence of RV enlargement in PE patients was reason the calculation of the inter-rater agreement index was 1 3 The International Journal of Cardiovascular Imaging 9. DeCara JM, Lang RM, Koch R, Bala R, Penzotti J, Spencer KT not possible. RV analysis was limited to two linear meas- (2003) The use of small personal ultrasound devices by internists urements and its function was not evaluated. Compressive without formal training in echocardiography. Eur J Echocardiogr ultrasonography does not offer the possibility to diagnose 4(2):141–147 pelvic deep venous thrombosis thus singular cases of proxi- 10. Filipiak-Strzecka D, John B, Kasprzak JD, Michalski B, Lipiec P (2013) Pocket-size echocardiograph–a valuable tool for nonex- mal DVT could have remained undetected. perts or just a portable device for echocardiographers? Adv Med Sci 58(1):67–72 Compliance with ethical standards 11. Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L et al (2015) Recommendations for cardiac chamber Conflict of interest The authors declare that they have no conflict of quantification by echocardiography in adults: an update from interest. the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr Ethical approval This article does not contain any studies with animals 28(1):1–39.e14 performed by any of the authors. All procedures performed in stud- 12. White RH (2003) The epidemiology of venous thromboembolism. ies involving human participants were in accordance with the ethical Circulation 107(23 Suppl 1):I4–I8 standards of the institutional and national research committee and with 13. Dresden S, Mitchell P, Rahimi L, Leo M, Rubin-Smith J, Bibi S the 1964 Helsinki declaration and its later amendments or comparable et al (2014) Right ventricular dilatation on bedside echocardiog- ethical standards. raphy performed by emergency physicians aids in the diagnosis of pulmonary embolism. Ann Emerg Med 63(1):16–24 Informed consent Informed consent was obtained from all individual 14. Da Costa Rodrigues J, Alzuphar S, Combescure C, Le Gal G, participants included in the study. Perrier A (2016) Diagnostic characteristics of lower limb venous compression ultrasonography in suspected pulmonary embolism: a meta-analysis. J Thromb Haemost 14(9):1765–1772 Open Access This article is distributed under the terms of the Crea- 15. Becattini C, Vedovati MC, Agnelli G (2010) Right ventricle dys- tive Commons Attribution 4.0 International License (http://creat iveco function in patients with pulmonary embolism. Intern Emerg Med mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- 5(5):453–455 tion, and reproduction in any medium, provided you give appropriate 16. Kasper W, Konstantinides S, Geibel A, Tiede N, Krause T, Just H credit to the original author(s) and the source, provide a link to the (1997) Prognostic significance of right ventricular afterload stress Creative Commons license, and indicate if changes were made. detected by echocardiography in patients with clinically suspected pulmonary embolism. Heart 77(4):346–349 17. Nazerian P, Vanni S, Volpicelli G, Gigli C, Zanobetti M, Bar- tolucci M et  al (2014) Accuracy of point-of-care multiorgan ultrasonography for the diagnosis of pulmonary embolism. Chest References 145(5):950–957 18. Khemasuwan D, Yingchoncharoen T, Tunsupon P, Kusunose K, 1. Konstantinides SV, Torbicki A, Agnelli G, Danchin N, Fitzmau- Moghekar A, Klein A et al (2015) Right ventricular echocardio- rice D, Galiè N et al (2014) ESC guidelines on the diagnosis graphic parameters are associated with mortality after acute pul- and management of acute pulmonary embolism. Eur Heart J monary embolism. J Am Soc Echocardiogr 28:355–362 35(43):3033–3073; 3033–3069, 3069a–3069k 19. Fields JM, Davis J, Girson L, Au A, Potts J, Morgan CJ et al 2. Aydoğdu M, Topbaşi Sinanoğlu N, Doğan NO, Oğuzülgen IK, (2017) Transthoracic echocardiography for diagnosing pulmo- Demircan A, Bildik F et al (2014) Wells score and pulmonary nary embolism: a systematic review and meta-analysis. J Am Soc embolism rule out criteria in preventing over investigation of Echocardiogr 30:714–723.e4 pulmonary embolism in emergency departments. Tuberk Toraks 20. Kurnicka K, Lichodziejewska B, Goliszek S, Dzikowska-Diduch 62(1):12–21.3 O, Zdonczyk O, Kozlowska M et al (2016) Echocardiographic 3. Dalen JE (2002) Pulmonary embolism: what have we learned pattern of acute pulmonary embolism: analysis of 511 consecutive since Virchow? Natural history, pathophysiology, and diagnosis. patients. J Am Soc Echocardiogr 29:907–913 Chest 122(4):1440–1456 21. Wolde L, Söhne M, Quak E, Mac Gillavry MR, Büller HR (2004) 4. Agnelli G, Becattini C (2010) Acute pulmonary embolism. N Engl Prognostic value of echocardiographically assessed right ventricu- J Med 363(3):266–274 lar dysfunction in patients with pulmonary embolism. Arch Intern 5. Jaff MR, McMurtry MS, Archer SL, Cushman M, Goldenberg Med 164(15):1685–1689 N, Goldhaber SZ et al (2011) Management of massive and sub- 22. Sanchez O, Trinquart L, Colombet I, Durieux P, Huisman MV, massive pulmonary embolism, iliofemoral deep vein thrombosis, Chatellier G et al (2008) Prognostic value of right ventricular and chronic thromboembolic pulmonary hypertension: a scien- dysfunction in patients with haemodynamically stable pulmonary tific statement from the American Heart Association. Circulation embolism: a systematic review. Eur Heart J 29(12):1569–1577 123(16):1788–1830 23. Beraud A-S, Rizk NW, Pearl RG, Liang DH, Patterson AJ (2013) 6. Konstantinides SV, Barco S, Lankeit M, Meyer G (2016) Man- Focused transthoracic echocardiography during critical care medi- agement of pulmonary embolism: an update. J Am Coll Cardiol cine training: curriculum implementation and evaluation of profi- 67(8):976–990 ciency. Crit Care Med 41:e179–e181 7. Brenner DJ, Hall EJ (2007) Computed tomography-an increasing 24. Mozzini C, Garbin U, Fratta Pasini AM, Cominacini L (2015) source of radiation exposure. N Engl J Med 357(22):2277–2284 Short training in focused cardiac ultrasound in an Internal Medi- 8. Decara JM, Kirkpatrick JN, Spencer KT, Ward RP, Kasza K, Fur- cine department: what realistic skill targets could be achieved? long K et al (2005) Use of hand-carried ultrasound devices to aug- Intern Emerg Med 10:73–80 ment the accuracy of medical student bedside cardiac diagnoses. 25. Taylor RA, Davis J, Liu R, Gupta V, Dziura J, Moore CL (2013) J Am Soc Echocardiogr 18(3):257–263 Point-of-care focused cardiac ultrasound for prediction of pulmo- nary embolism adverse outcomes. J Emerg Med 45:392–399 1 3 The International Journal of Cardiovascular Imaging 26. Hulett CS, Pathak V, Katz JN, Montgomery SP, Chang LH (2014) 32. Biais M, Carrié C, Delaunay F, Morel N, Revel P, Janvier G (2012) Development and preliminary assessment of a critical care ultra- Evaluation of a new pocketechoscopic device for focused cardiac sound course in an adult pulmonary and critical care fellowship ultrasonography in an emergency setting. Crit Care 16:R82 program. Ann Am Thorac Soc 11:784–788 33. Nazerian P, Volpicelli G, Gigli C, Becattini C, Sferrazza Papa 27. See KC, Ong V, Ng J, Tan RA, Phua J (2014) Basic critical care GF, Grifoni S et al (2017) Diagnostic performance of Wells score echocardiography by pulmonary fellows: learning trajectory and combined with point-of-care lung and venous ultrasound in sus- prognostic impact using a minimally resourced training model. pected pulmonary embolism. Acad Emerg Med 24(3):270–280 Crit Care Med 42:2169–2177 34. Michalski B, Kasprzak JD, Szymczyk E, Lipiec P (2012) Diagnos- 28. Townsend NT, Kendall J, Barnett C, Robinson T (2016) An effec- tic utility and clinical usefulness of the pocket echocardiographic tive curriculum for focused assessment diagnostic echocardiogra- device. Echocardiography 29(1):1–6 phy: establishing the learning curve in surgical residents. J Surg 35. Ruddox V, Stokke TM, Edvardsen T, Hjelmesæth J, Aune E, Educ 73:190–196 Bækkevar M, Norum IB, Otterstad JE (2013) The diagnostic 29. Vignon P, Mucke F, Bellec F, Marin B, Croce J, Brouqui T, Palo- accuracy of pocket-size cardiac ultrasound performed by unse- bart C, Senges P, Truffy C, Wachmann A, Dugard A, Amiel J-B lected residents with minimal training. Int J Cardiovasc Imaging (2011) Basic critical care echocardiography: validation of a cur- 29(8):1749–1757 riculum dedicated to noncardiologist residents. Crit Care Med 36. Lipiec P, Bąk J, Braksator W, Fijałkowski M, Gackowski A, 39:636–642 Gąsior Z et al (2018) Transthoracic echocardiography in adults— 30. Galderisi M, Santoro A, Versiero M, Lomoriello VS, Esposito R, guidelines of the Working Group on Echocardiography of the Pol- Raia R, Farina F, Schiattarella PL, Bonito M, Olibet M, de Simone ish Cardiac Society. Kardiol Pol 76(2):488–493 G (2010) Improved cardiovascular diagnostic accuracy by pocket 37. Filipiak-Strzecka D, Michalski B, Kasprzak JD, Lipiec P (2014) size imaging device in non-cardiologic outpatients: the NaUSiCa Pocket-size imaging devices allow for reliable bedside screening (Naples Ultrasound Stethoscope in Cardiology) study. Cardiovasc for femoral artery access site complications. Ultrasound Med Biol Ultrasound 8(1):51 40(12):2753–2758 31. Andersen GN, Haugen BO, Graven T, Salvesen O, Mjølstad OC, Dalen H (2011) Feasibility and reliability of point-of-care pocket- sized echocardiography. Eur J Echocardiogr 12:665—70 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The International Journal of Cardiovascular Imaging Springer Journals

Brief cardiovascular imaging with pocket-size ultrasound devices improves the accuracy of the initial assessment of suspected pulmonary embolism

Free
11 pages
Loading next page...
 
/lp/springer_journal/brief-cardiovascular-imaging-with-pocket-size-ultrasound-devices-iRqTFIIbro
Publisher
Springer Netherlands
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Cardiology; Imaging / Radiology; Cardiac Imaging
ISSN
1569-5794
eISSN
1573-0743
D.O.I.
10.1007/s10554-018-1382-5
Publisher site
See Article on Publisher Site

Abstract

Pulmonary embolism onset is frequently neglected due to the non-specific character of its symptoms. Pocket-size imaging devices (PSID) present an opportunity to implement imaging diagnostics into conventional physical examination. The aim of this study was to test the hypothesis that supplementation of the initial bedside assessment of patients with suspected pulmonary embolism (PE) with four-point compression venous ultrasonography (CUS) and right ventricular size assessment with the use of PSID equipped with dual probe could positively influence the accuracy of clinical predictions. A single- centre, prospective analysis was conducted on 100 patients (47 men, mean age 68 ± 13 years) with suspected PE. Clinical assessment on the basis of Wells and revised Geneva score and physical examination were supplemented with CUS and RV measurements by PSID. The mean time of PSID scanning was 4.9 ± 0.8 min and was universally accepted by the patients. Fifteen patients had deep venous thrombosis and RV enlargement was observed in 59 patients. PE was confirmed in 24 patients. If the both CUS was positive and RV enlarged, the specificity was 100% and sensitivity 54%, ROC AUC 0.771 [95% CI 0.68–0.85]. The Wells rule within our study population had the specificity of 86% and sensitivity of 67%, ROC AUC 0.776 (95% CI 0.681–0.853, p < 0.0001). Similar values calculated for the revised Geneva score were as follows: specificity 58% and sensitivity 63%, ROC AUC 0.664 (95% CI 0.563–0.756, p = 0.0104). Supplementing the revised Geneva score with additional criteria of CUS result and RV measurement resulted in significant improvement of diagnostic accuracy. The difference between ROC AUCs was 0.199 (95% Cl 0.0893–0.308, p = 0.0004). Similar modification of Wells score increased ROC AUC by 0.133 (95% CI 0.0443–0.223, p = 0.0034). Despite the well-acknowledged role of the PE clinical risk assess- ment scores the diagnostic process may benefit from the addition of basic bedside ultrasonographic techniques. Keywords Pocket-size imaging devices · Pulmonary embolism · Compression ultrasound test · Wells rule · Revised Geneva score Introduction is estimated that over 600,000 cases of PE are diagnosed in United States every year [2]. Majority of death related Pulmonary embolism (PE) is a common, oftentimes mis- cases of PE occur when the condition was not diagnosed diagnosed emergency cardiovascular state burdened with in time [3]. potentially fatal consequences. Symptoms are non-specific According to current guidelines [1, 4–6], the first step in varying from chest pain, shortness of breath, cough and the diagnostic protocol of PE is determination of its clini- hemoptysis to syncope and cardiac arrest [1]. Annually cal probability. For this purpose several scoring systems are over 10 million patients in the United States seek medical suggested. The Wells rule and the modified Geneva score help with the complaints of dyspnea, chest pain or both. It are among the most validated scales used for PE clinical risk assessment. In order to reduce the number of unnecessary procedures, guidelines recommend first to identify patients * Dominika Filipiak-Strzecka with a very low probability of the condition, so that referral dominika.filipiak@gmail.com for further diagnostics can safely be withheld. The remain- 1 ing patients are categorized into one of the recommended Department of Cardiology, Bieganski Hospital, Medical diagnostic paths; which in most cases involves further blood University of Lodz, Kniaziewicza 1/5, 91-347 Lodz, Poland Vol.:(0123456789) 1 3 The International Journal of Cardiovascular Imaging testing or—perceived as a golden standard—Multi-Slice CT further diagnostic procedures. All patients underwent clini- Angiography. Undoubtedly, unjustified referral for diagnos- cal assessment in the emergency room on the basis of the tic procedures involving a radiographic exposure results in Wells rule and the revised Geneva score, including the eval- excessive cost of care, increased use of hospital resources, uation of medical history (previous episode of pulmonary significant risk of contrast-associated acute renal failure and embolism or deep vein thrombosis, recent surgery), current exposure-dependent malignancies [7]. On the contrary, in medical condition (active cancer, heart rate, clinical signs certain cases even with calculated low clinical risk (evalu- of DVT) and reported symptoms (hemoptysis) (Table 1). ated by established scoring systems) postponing of final Subsequently, the regular physical examination was sup- diagnosis and late introduction of proper treatment may plemented with short, focused bedside ultrasonographic cause fatal consequences. For this reason a persistent search assessment consisting of four-point compression venous for new diagnostic modalities improving initial assessment ultrasonography and measurements of the right ventricle of patient condition, which could increase the accuracy of performed by cardiology resident with the use of PSID. grading clinical risk of PE is continued. However, a novel The operator’s training in echocardiography was included as diagnostic protocol needs to meet certain requirements in part of her residency program and was based on six-month order to gain clinical acceptance: it has to be widely avail- rotation in the echo lab. It included conducting and ana- able, quick to implement and should be cost effective. A lyzing transthoracic echocardiographic examinations under method/strategy where all patients suspected of PE should specialist supervision, as well as a basic vascular examina- undergo ultrasonographic examination as part of initial eval- tions such as compression ultrasound test. This study was uation proves challenging in the urgency of the emergency conducted in accordance with the amended Declaration of department. Certain constraints such as urgent transportation Helsinki. Informed consent was obtained from individual to echocardiographic laboratory, use of high-end equipment participants included in the study. The study protocol was and need for trained echocardiographer generate high costs. approved by bioethics committee of our institution (Decision The trend to miniaturize echocardiographic devices has led No. RNN/8/10/KE with the Supplement No. KE/2011/15). to the creation of pocket-size imaging devices (PSID). With the ultraportability being their biggest advantage PSIDs may be used in almost every clinical setting, including the Pocket‑size imaging device emergency room. Moreover, with an interface optimized for ease of operation, PSIDs present easy access to a wide The pocket-size imaging device used in this study was range of users. It has been shown, that with targeted training V-Scan (GE Vingmed Ultrasound, Horten, Norway) basic information can be obtained with the use of PSIDs by equipped with dual probe, combining two transducers in one non-echocardiographers [8–10]. We hypothesized, that in probe- the phased array (frequency range of 1.7–3.8 MHz, patients with suspected pulmonary embolism augmentation image sector limited to 75°, depth range 6–24 cm) and the of initial emergency room assessment with four-point com- linear probe (frequency range of 3.4–8.0  MHz, aperture pression venous ultrasonography (CUS) and right ventricu- width of 2.9 cm, maximum depth of 8 cm). The four-point lar size assessment with the use of PSID equipped with dual compression ultrasound tests were performed using linear probe can improve the diagnostic accuracy of established probe and the vascular preset, whereas the right ventricle clinical prediction rules. assessment with the phased array probe and cardiac preset. Materials and methods Compression ultrasonography This was a prospective study conducted between February The examination was performed in the supine position. The femoral artery was assessed from the level just distal 2015 and May 2016. 100 consecutive patients (47 males, mean age 68 ± 13 years) who were referred to our depart- to the inguinal ligament to the 2 cm distal to the junction of the common femoral vein and the greater saphenous vein. ment during office hours (8 a.m.–4 p.m.) or during the 24-h medical shift of the resident performing examination (D.S.) The collapsing of common and deep femoral veins was evaluated. The popliteal vein was assessed from the level of were included in our analysis. The inclusion criterion was the suspicion of pulmonary embolism based on the medical popliteal fossa up to the level of its trifurcation. The direct pressure with the use of transducer was applied in order to history and basic physical examination only. All patients reported dyspnea as their main symptom, in some cases completely compress the vein. If the vein was compressed completely, then a DVT at this site was ruled out. The lack with concomitant chest pain (49%), cough (17%) or tachy- cardia (27%). If such a diagnosis was suggested, in all cases of possibility to completely compress the vein was treated as a positive test result (Fig. 1). the investigator (D.S.) was informed and continued with 1 3 The International Journal of Cardiovascular Imaging Table 1 The original and modified versions of Wells rule and revised Geneva score used in the study Wells rule Original version Modified version Points % of patients Previous PE or DVT Previous PE or DVT 1.5 7 HR > 100 b.p.m HR > 100 b.p.m 1.5 27 Surgery or immobilization within the past Surgery or immobilization within the past 4 weeks 1.5 9 4 weeks Hemoptysis Hemoptysis 1 0 Active cancer Active cancer 1 2 Clinical signs of DVT Positive CUS result 3/4 9 15 Alternative diagnosis less likely than PE Alternative diagnosis less likely than PE 3 16 – RV enlargement 1 59 Basal diameter (4CH) ≤ 47 mm 2 35 Basal diameter (4CH) > 47 mm 19 Revised Geneva score Original version Modified version Points % of patients Previous PE or DVT Previous PE or DVT 3 7 HR HR 3 15 75–94 b.p.m 75–94 b.p.m 5 27 ≥ 95 b.p.m ≥ 95 b.p.m Surgery or fracture within the past month Surgery or fracture within the past month 2 9 Hemoptysis Hemoptysis 2 0 Active cancer Active cancer 2 2 Unilateral lower limb pain Positive CUS result 3 10 15 Pain on lower limb deep venous palpation and 4 6 unilateral oedema Age > 65 years Age > 65 years 1 63 – RV enlargement 1 59 Basal diameter (4CH) ≤ 47 mm 2 35 Basal diameter (4CH) > 47 mm 19 Fig. 1 Compression ultrasound test of common femoral vein per- baseline; panel b, d—compression. Panel b—vein completely com- formed with the use of PSID in two patients: without thrombosis pressed; panel d—abnormal study indicating venous thrombosis; A (panel a, b) and with venous thrombosis (panel c, d). Panel a, c— femoral artery; V femoral vein Assessment of the right ventricle ventricle-focused 4-chamber apical view and proximal right ventricle outflow diameter measured in long axis parasternal Two linear measurements of the right ventricle were view (Fig. 2). Right ventricle enlargement was defined as the performed: right ventricular basal diameter in the right 1 3 The International Journal of Cardiovascular Imaging right ventricular basal diameter > 41 mm and/or proximal Statistical analysis right ventricle outflow diameter > 35 mm [11]. Continuous and categorical variables are expressed as Modified clinical prediction rules mean ± SD and as percentages (%), respectively. To assess the diagnostic value of different tests the ROC analysis on The points “clinical signs of DVT” in the Wells rule and the basis of DeLong at al. methodology was performed. “pain on lower limb deep venous palpation and unilateral Sensitivity, specificity and overall diagnostic accuracy were oedema” in the revised Geneva score were changed into compared with the use of N-1 Chi square test. A difference ‘positive CUS result’. Additional points were also given, was considered statistically significant when p < 0.05. Cal- when the RV enlargement was detected. Subsequently, we culations were performed with the use of MedCalc Software tested the hypothesis that incorporating the results of brief version 12.2.1.0. ultrasonographic assessment into the risk scale improves its diagnostic value. The score of this additional criterion was retrospectively determined on the basis of our population, Results to achieve the best ROC AUC. (Table 1). The assessment by recommended clinical prediction Diagnosis of pulmonary embolism rules Final diagnosis was established in accordance with the algo- Pulmonary embolism was eventually confirmed by con- rithms recommended by the ESC guidelines, on the basis of trast-enhanced chest computed tomography in 24 patients. clinical gold-standard including all necessary examinations The patients’ characteristic is presented in Table 1. None [1]. Two patients with the high clinical probability of PE of the patients had been given vasopressors on admission. (according to the Wells score and revised Geneva score) In one patient the systolic blood pressure was < 90 mmHg. immediately underwent CT-angiography. The remaining 98 17 patients had the history of chronic lung disease, in two patients with low or intermediate clinical probability had patients atrial septal defect was present, three patients had d -dimer plasma level initially assessed. In 47 patients with the tricuspid regurgitation diagnosed. Among final diagno- normal d -dimer plasma level (cut-off value: 500 µg/L) the ses other than pulmonary embolism coronary artery dis- pulmonary embolism was ruled out. In the remaining 51 ease (25 patients,) chronic heart failure exacerbation (22 patients CT angiography was performed (Fig. 3). patients), pneumonia (11 patients), heart rhythm disorders (seven patients), valve disease (six patients) were most often detected. According to the three-category Wells rule the clinical risk of PE was estimated as low in 74 patients, among which ten were eventually diagnosed with PE, as Fig. 2 RV linear measure- ments: a RV focused 4-chamber apical view, basal dimension; c parasternal long-axis view, proximal RV outflow diameter 1 3 The International Journal of Cardiovascular Imaging Fig. 3 Diagram presenting the diagnostic path of patients included in the study intermediate in 24 patients (12 cases of confirmed PE), as Modified clinical prediction rules high in two patients (PE confirmed). In compliance with revised Geneva score, 54 patients had low clinical risk of On the basis of our study population we have established PE (in nine patients PE was confirmed), 44 patients-inter - that in case of positive lower extremity ultrasound the best mediate (13 cases of PE); two patients-high (PE confirmed) ROC AUC were achieved for the values 4 points for Wells (Figs. 4 and 5). The diagnostic accuracy of Wells rule and score and 9 points for revised Geneva score. For the RV revised Geneva score based on our study population is pre- enlargement we tested 1-point and 2-points scoring. The sented in Table 2. highest ROC AUC was obtained with two grade model: 1 point was added, if the basal diameter measured in the Scanning with the use of PSID four chamber view was within the range of 42–47 mm, 2 points were given when this diameter exceeded 47  mm. The mean time of scanning with the use of PSID was Supplementing the revised Geneva score with additional 4.9 ± 0.8 (95% CI 4.7–5.0) min and was universally accepted criteria of positive CUS test and RV enlargement resulted by patients. One patient was excluded from the analysis, as in significant improvement of diagnostic accuracy of this the unilateral visualisation of popliteal vessels with the use score- difference between areas 0.212 (95% Cl 0.100–0.325, of PSID was impossible. In two patients with the history of p < 0.0001), as presented on the graph. The overall diagnos- thoracotomy obtaining parasternal view proved impossible tic accuracy improved from 59 to 94% (p = 0.02). Similar and RV size was determined in apical view exclusively. Fif- modification of Wells score increased ROC AUC by 0.138 teen patients had the deep vein thrombosis (five cases proxi- (95% CI 0.0429–0.223, p = 0.0045), the overall diagnostic mal) detected in compression ultrasound test, whereas RV accuracy from 81 to 93% (p = 0,012). Modification of both enlargement was observed in 59 patients. Table 3 presents scales resulted in statistically significant improvement of the diagnostic accuracy of the PSID scanning for identifica - specificity but not sensitivity. (Table  2; Figs. 6 and 7). tion of patients with PE calculated for various criteria of test positivity. 1 3 The International Journal of Cardiovascular Imaging Fig. 4 Clinical risk of PE esti- mated on the basis of three cat- egory Wells rule (Panel a) and revised Geneva score (Panel c); for modified Wells rule (Panel b) the following threshold were established: low 0–1, intermedi- ate 2–6, high ≥ 7; for modified revised Geneva score (Panel d) the following threshold were established: low 0–3, intermedi- ate 4–10, high ≥ 11 Table 2 The diagnostic accuracy of Wells rule, revised Geneva score, modified Wells rule and modified revised Geneva score (supplemented with the PSID test results) based on the study population Sensitivity Specificity (%) PPV (%) NPV (%) Overall diagnostic ROC AUC (%) (95% CI) (95% CI) accuracy (%) (95% CI, p) Wells rule ≥ 2 66.7 85.5 59 89 81 0.776 (44.7–84.4) (75.6–92.5) (39–78) (80–95) (0.681 to 0.853, p < 0.0001) Revised Geneva score ≥ 4 62.5 57.9 35 82 59 0.664 (0.563 to 0.756, p = 0.0104) (40.6–81.2) (46.0–69.1) (20–54) (70–90) Modified Wells rule ≥ 5 70.8 98.7 94 92 93 0.914 (48.9–87.4) (92.9–100.0) (73–100) (83–97) (0.841 to 0.961, p < 0.0047) Modified revised Geneva 75.0 88.2 67 92 94 0.877 (0.796 to 0.934, p < 0.0001) score ≥ 7 (53.3–90.2) (78.7–94.4) (46–84) (83–97) Thresholds: Wells rule ≥ 2, revised Geneve score ≥ 4, modified Wells rule ≥ 5, modified revised Geneve score ≥ 7; determined from the ROC curves ROC AUC area under receiver operating characteristic curve, NPV negative predictive value, PPV positive predictive value 1 3 The International Journal of Cardiovascular Imaging Fig. 5 Clinical risk of PE esti- mated on the basis of two cat- egory Wells rule (Panel a) and revised Geneva score (Panel c); for modified Wells rule (Panel b) the following threshold were established: PE unlikely 0–3,5, PE likely ≥ 4; for modified revised Geneva score (Panel d) the following threshold were established: PE unlikely 0–6,5, PE likely ≥ 7 Table 3 The diagnostic accuracy of the PSID scanning for identification of patients with PE Sensitivity (95% CI) Specificity (95% CI) PPV (95% CI) NPV (95% CI) ROC AUC (95% CI, p) Positive CUS 54% (33–74) 97% (91–100) 87% (58–99) 87% (78–93) 0.758 (0.662 to 0.838, p < 0.0001) Proximal DVT diagnosed in 21% (7–42) 100% (95–100) 100% (48–100) 80% (71–88) 0.604 (0.501 to 0.701, CUS p = 0.0139) RV enlargement 92% (73–99) 51% (40–63) 37% (25–51) 95% (84–99) 0.715 (0.616 to 0.801, p < 0.0001) RV enlargement (basal 4CH) 92% (73–99) 51% (40–63) 37% (25–51) 95% (84–99) 0.746 (0.649 to 0.828, 2-Point scoring (42–47; > 47) p < 0.0001) CUS and RV enlargement 54% (33–74) 100% (95–100) 100% (74–100) 87% (79–94) 0.771 (0.676 to 0.849, p < 0.0001) CUS or RV enlargement 92% (73–99) 49% (37–60) 36% (24–50) 95% (83–99) 0.702 (0.602 to 0.789, p < 0.0001) ROC AUC ar ea under receiver operating characteristic curve, CUS compression ultrasound test, DVT deep vein thrombosis, NPV negative pre- dictive value, PPV positive predictive value, RV right ventricle 1 3 The International Journal of Cardiovascular Imaging venous thrombosis (DVT) are considered as a continuum of the same clinical entity, namely venous thromboembo- lism. Even up to 90% of pulmonary emboli may arise from lower limbs or pelvis deep venous thrombosis [14]. The application of compression ultrasound in detection of the deep vein thrombosis in patients with suspected PE has been previously proposed and mentioned in recent guide- lines as a method of reducing the number of CT angiog- Fig. 6 ROC curves comparison between revised Geneva score and raphy in appropriate patients. An early Dutch study based modified revised Geneva score with additional criteria of CUS and on the patients with suspected PE indicated that the use RV enlargement of compression US reduced the need for other imaging by 22%, at the expense of 2–4% of patients being unnecessar- ily treated for venous thromboembolism [15]. Data elic- ited from the metaanalysis indicate that CUS sensitivity ranged from 23 to 58%, while its specificity ranged from 89 to 99% [16]. Additionally CUS is commonly accepted as a non-invasive diagnostic modality, which may prove particularly relevant in patients with relative contraindica- tions to CT, such as chronic kidney disease and creatinine clearance below 30 mL/min, allergy to iodinated contrast dye, pregnant women or younger patients, in which the reduction of irradiation is desirable. The argument of limited cost-effectiveness was raised Fig. 7 ROC curves comparison between Wells rule and modified Wells rule with additional criteria of CUS and RV enlargement against the implementation of CUS in routine diagnostic process as it would require patient transportation to the ultra- sound examination lab, specialized workforce and equip- ment. Adversely, according to ESC, examination with the Discussion use of PSID should not be treated as a separate procedure but rather as an augmentation of physical examination. PSID To the best of our knowledge our study is the first to report examination can be performed at the point of care. In con- the diagnostic potential of brief scanning with the use of temporary clinical reality in which the concept of FOCUS PSID in the initial assessment of patients with suspected is increasingly recognized and basics of ultrasonographic pulmonary embolism. The main findings can be summa - examination are becoming a vital part of numerous medical rized as follows: (i) expanding the initial patient assess- professionals’ training, one can expect that such diagnostic ment in the ER with the elements of ultrasonographic approach may not require additional personnel apart from imaging did not excessively prolong physical examination the attending physician. and was universally accepted by the patients (ii) simulta- In agreement with the current clinical guidelines we neous RV enlargement and positive CUS result identified have assumed in our group positive proximal CUS result with the use of PSID has a very high positive predictive would allow for instantaneous (on the level of ER examina- value for PE (iii) the risk of PE assessed as low in accord- tion) identification of PE in 5 (5%) patients. Among them ance with the Wells and Geneva modified scores does not in four patients with either low or intermediate clinical risk rule out the possibility of PE diagnosis (iv) the diagnostic as estimated by means of Wells and revised Geneva scale value of the Well’s rule and the revised Geneva score can PSID examination would help to avoid a prolonged wait for be significantly improved by implementing the results of d -dimer test results and subsequent CT. In one patient with PSID examination in the clinical prediction rules criteria. high clinical risk, immediate PE diagnosis would eliminate Prompt diagnosis of pulmonary embolism still poses a the need for CT scanning along with the potentially danger- challenge to clinicians, in spite of all the available mod- ous need for transportation. What is more, a DVT detected ern diagnostic procedures. Based on the autopsy findings, in CUS (also distal) accompanied by the RV enlargement diagnosis of PE is missed in up to 30–50% of patients [12]. in 100% of cases was related with the presence of thrombi 2/3 of deaths associated with PE occurs within the first in pulmonary circulation. One may hypothesize that also hours from the symptoms manifestation [13]. For this rea- in the group of patients with the confirmed distal DVT, an son any improvement in diagnostic accuracy and hastening additional screening for the RV enlargement could eliminate of the whole diagnostic process is essential.PE and deep the need for blood testing/CT. 1 3 The International Journal of Cardiovascular Imaging Although, in accordance with the guidelines, echocar- very high; only one patient with diagnosed PE could not be diography does not play an essential role in the diagnos- diagnosed with RV pathology. tic process of PE, it can undoubtedly prove useful as a Nazerian et al. [33] presented an appealing approach, in method of treatment progress assessment. Furthermore, it which they proved that enhancing the Wells rule with lung was previously confirmed that focused echocardiographic and lower limb venous ultrasound improved the diagnostic assessment, as a part of multiorgan bedside ultrasonogra- value of this scale. However, their approach still involved phy can improve clinical evaluation of patients with sus- patient transportation, the use of high-end equipment and pected PE prior to definitive imaging [ 17]. Pathologies required involvement of a specialist in ultrasonography. One described by the incorrect values of the ratio of RV to left of the most highly regarded advantages of clinical prediction ventricular end-diastolic diameter; RV systolic pressure, rules is the easiness of their application and the immediate tricuspid annular plane systolic excursion and inferior result. PSID examination shares these features as it can be vena cava collapsibility were confirmed to be related with performed at any point of care, including the ER. Further- the increased mortality during the course of acute PE [18]. more PSIDs can be operated by less experienced medical Parameters related with RV dysfunction proved to have professionals who after the completion of the short training relatively high specificity while being burdened with low should be able to perform a reliable, specifically-aimed ultra- sensitivity [19]. In the study conducted by Kuznicka et al. sographic screening [8–10, 34–36]. Importantly, according [20] aimed at the assessment of the frequency of patholog- to the ESC guidelines PSID screening should be integrated ical findings in echocardiographic examination in patients into the routine physical examination rather than treated as with the confirmed PE, RV enlargement was relatively a separate diagnostic procedure. common, particularly in the high-risk patients. However, Latest and most advanced PSIDs are equipped with a incorrect ventricular ratio criterion was not fulfilled in all dual-probe, which shares the advantages of a linear and such cases. Apart from that, RV enlargement as a quantita- sector probe in one ultraportable tool. Older generations tive parameter is in our opinion easier to objectively assess of PSID were not perfectly suited for the vascular imag- than qualitative criteria such as free wall hypokinesis or ing and significant shortcomings in this area were present. paradox movement of intraventricular septum. Due to the Clinicians had to overcome the obstacles of the minimal above mentioned rationale RV enlargement was chosen for depth, image sector size and insufficient probe frequency the purpose of RV function evaluation. [37]. PSID equipped with dual-probe appears to be capable RV dilatation as diagnosed during echocardiographic of being successfully utilized in new clinical applications. examination in patients with pulmonary embolism has been Importantly, practicality of ultraportable ultrasound may suf- previously demonstrated to be related with the permanent fer from limited imaging capabilities. Although in all 100 RV dysfunction, RV failure, recurrent pulmonary embolism patients studied we were able to visualise RV at least partly, and death [19–22]. Although evaluating RV size and sys- and sufficiently for measurements, this may not be possible tolic function is not sufficient to make a direct diagnosis in all-comers population in clinical setting. Lower extrem- of PE it could provide an additionally valuable evidence in ity vein assessment may also exceed the capabilities of the some patients. Should the dilated RV be detected during the device in some patients (in our study group ca. 1%). For this bedside echocardiographic examination in a high-risk PE reason the supplementary role of PSID examination needs patient, the proper treatment introduction should be hastened to be re-emphasized; final diagnosis should be obtained on with the improvement of morbidity and mortality [13]. the basis of the complete set of clinical data. In our study population, RV enlargement was a relatively common finding. Although such diagnosis may be associ- ated with worse prognosis in patient with the suspected PE, Conclusion it is important to point out that the majority of causes of RV enlargement in the study group was not related with PE. Despite the well-established value of the PE clinical predic- Thus, RV enlargement alone should not alter the diagnostic tion rules, the diagnostic process of patients with suspected process and trigger prompt CT-scan. Nevertheless, the dis- PE benefits from the addition of brief assessment with the cussed parameter is useful as one of the factors reflected in use of PSID. risk scale. It was previously established that trained emergency Limitations physician was able to perform a reliable evaluation of the RV dysfunction during a bedside examination [13, 23–29]. This is a single center study with relatively small study The efficacy of PSID screening in the assessment of the RV population. All of the examinations with the use of PSID dilatation was also confirmed [30– 32]. In our study popula- were performed by the same cardiology resident, and for this tion the prevalence of RV enlargement in PE patients was reason the calculation of the inter-rater agreement index was 1 3 The International Journal of Cardiovascular Imaging 9. DeCara JM, Lang RM, Koch R, Bala R, Penzotti J, Spencer KT not possible. RV analysis was limited to two linear meas- (2003) The use of small personal ultrasound devices by internists urements and its function was not evaluated. Compressive without formal training in echocardiography. Eur J Echocardiogr ultrasonography does not offer the possibility to diagnose 4(2):141–147 pelvic deep venous thrombosis thus singular cases of proxi- 10. Filipiak-Strzecka D, John B, Kasprzak JD, Michalski B, Lipiec P (2013) Pocket-size echocardiograph–a valuable tool for nonex- mal DVT could have remained undetected. perts or just a portable device for echocardiographers? Adv Med Sci 58(1):67–72 Compliance with ethical standards 11. Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L et al (2015) Recommendations for cardiac chamber Conflict of interest The authors declare that they have no conflict of quantification by echocardiography in adults: an update from interest. the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr Ethical approval This article does not contain any studies with animals 28(1):1–39.e14 performed by any of the authors. All procedures performed in stud- 12. White RH (2003) The epidemiology of venous thromboembolism. ies involving human participants were in accordance with the ethical Circulation 107(23 Suppl 1):I4–I8 standards of the institutional and national research committee and with 13. Dresden S, Mitchell P, Rahimi L, Leo M, Rubin-Smith J, Bibi S the 1964 Helsinki declaration and its later amendments or comparable et al (2014) Right ventricular dilatation on bedside echocardiog- ethical standards. raphy performed by emergency physicians aids in the diagnosis of pulmonary embolism. Ann Emerg Med 63(1):16–24 Informed consent Informed consent was obtained from all individual 14. Da Costa Rodrigues J, Alzuphar S, Combescure C, Le Gal G, participants included in the study. Perrier A (2016) Diagnostic characteristics of lower limb venous compression ultrasonography in suspected pulmonary embolism: a meta-analysis. J Thromb Haemost 14(9):1765–1772 Open Access This article is distributed under the terms of the Crea- 15. Becattini C, Vedovati MC, Agnelli G (2010) Right ventricle dys- tive Commons Attribution 4.0 International License (http://creat iveco function in patients with pulmonary embolism. Intern Emerg Med mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- 5(5):453–455 tion, and reproduction in any medium, provided you give appropriate 16. Kasper W, Konstantinides S, Geibel A, Tiede N, Krause T, Just H credit to the original author(s) and the source, provide a link to the (1997) Prognostic significance of right ventricular afterload stress Creative Commons license, and indicate if changes were made. detected by echocardiography in patients with clinically suspected pulmonary embolism. Heart 77(4):346–349 17. Nazerian P, Vanni S, Volpicelli G, Gigli C, Zanobetti M, Bar- tolucci M et  al (2014) Accuracy of point-of-care multiorgan ultrasonography for the diagnosis of pulmonary embolism. Chest References 145(5):950–957 18. Khemasuwan D, Yingchoncharoen T, Tunsupon P, Kusunose K, 1. Konstantinides SV, Torbicki A, Agnelli G, Danchin N, Fitzmau- Moghekar A, Klein A et al (2015) Right ventricular echocardio- rice D, Galiè N et al (2014) ESC guidelines on the diagnosis graphic parameters are associated with mortality after acute pul- and management of acute pulmonary embolism. Eur Heart J monary embolism. J Am Soc Echocardiogr 28:355–362 35(43):3033–3073; 3033–3069, 3069a–3069k 19. Fields JM, Davis J, Girson L, Au A, Potts J, Morgan CJ et al 2. Aydoğdu M, Topbaşi Sinanoğlu N, Doğan NO, Oğuzülgen IK, (2017) Transthoracic echocardiography for diagnosing pulmo- Demircan A, Bildik F et al (2014) Wells score and pulmonary nary embolism: a systematic review and meta-analysis. J Am Soc embolism rule out criteria in preventing over investigation of Echocardiogr 30:714–723.e4 pulmonary embolism in emergency departments. Tuberk Toraks 20. Kurnicka K, Lichodziejewska B, Goliszek S, Dzikowska-Diduch 62(1):12–21.3 O, Zdonczyk O, Kozlowska M et al (2016) Echocardiographic 3. Dalen JE (2002) Pulmonary embolism: what have we learned pattern of acute pulmonary embolism: analysis of 511 consecutive since Virchow? Natural history, pathophysiology, and diagnosis. patients. J Am Soc Echocardiogr 29:907–913 Chest 122(4):1440–1456 21. Wolde L, Söhne M, Quak E, Mac Gillavry MR, Büller HR (2004) 4. Agnelli G, Becattini C (2010) Acute pulmonary embolism. N Engl Prognostic value of echocardiographically assessed right ventricu- J Med 363(3):266–274 lar dysfunction in patients with pulmonary embolism. Arch Intern 5. Jaff MR, McMurtry MS, Archer SL, Cushman M, Goldenberg Med 164(15):1685–1689 N, Goldhaber SZ et al (2011) Management of massive and sub- 22. Sanchez O, Trinquart L, Colombet I, Durieux P, Huisman MV, massive pulmonary embolism, iliofemoral deep vein thrombosis, Chatellier G et al (2008) Prognostic value of right ventricular and chronic thromboembolic pulmonary hypertension: a scien- dysfunction in patients with haemodynamically stable pulmonary tific statement from the American Heart Association. Circulation embolism: a systematic review. Eur Heart J 29(12):1569–1577 123(16):1788–1830 23. Beraud A-S, Rizk NW, Pearl RG, Liang DH, Patterson AJ (2013) 6. Konstantinides SV, Barco S, Lankeit M, Meyer G (2016) Man- Focused transthoracic echocardiography during critical care medi- agement of pulmonary embolism: an update. J Am Coll Cardiol cine training: curriculum implementation and evaluation of profi- 67(8):976–990 ciency. Crit Care Med 41:e179–e181 7. Brenner DJ, Hall EJ (2007) Computed tomography-an increasing 24. Mozzini C, Garbin U, Fratta Pasini AM, Cominacini L (2015) source of radiation exposure. N Engl J Med 357(22):2277–2284 Short training in focused cardiac ultrasound in an Internal Medi- 8. Decara JM, Kirkpatrick JN, Spencer KT, Ward RP, Kasza K, Fur- cine department: what realistic skill targets could be achieved? long K et al (2005) Use of hand-carried ultrasound devices to aug- Intern Emerg Med 10:73–80 ment the accuracy of medical student bedside cardiac diagnoses. 25. Taylor RA, Davis J, Liu R, Gupta V, Dziura J, Moore CL (2013) J Am Soc Echocardiogr 18(3):257–263 Point-of-care focused cardiac ultrasound for prediction of pulmo- nary embolism adverse outcomes. J Emerg Med 45:392–399 1 3 The International Journal of Cardiovascular Imaging 26. Hulett CS, Pathak V, Katz JN, Montgomery SP, Chang LH (2014) 32. Biais M, Carrié C, Delaunay F, Morel N, Revel P, Janvier G (2012) Development and preliminary assessment of a critical care ultra- Evaluation of a new pocketechoscopic device for focused cardiac sound course in an adult pulmonary and critical care fellowship ultrasonography in an emergency setting. Crit Care 16:R82 program. Ann Am Thorac Soc 11:784–788 33. Nazerian P, Volpicelli G, Gigli C, Becattini C, Sferrazza Papa 27. See KC, Ong V, Ng J, Tan RA, Phua J (2014) Basic critical care GF, Grifoni S et al (2017) Diagnostic performance of Wells score echocardiography by pulmonary fellows: learning trajectory and combined with point-of-care lung and venous ultrasound in sus- prognostic impact using a minimally resourced training model. pected pulmonary embolism. Acad Emerg Med 24(3):270–280 Crit Care Med 42:2169–2177 34. Michalski B, Kasprzak JD, Szymczyk E, Lipiec P (2012) Diagnos- 28. Townsend NT, Kendall J, Barnett C, Robinson T (2016) An effec- tic utility and clinical usefulness of the pocket echocardiographic tive curriculum for focused assessment diagnostic echocardiogra- device. Echocardiography 29(1):1–6 phy: establishing the learning curve in surgical residents. J Surg 35. Ruddox V, Stokke TM, Edvardsen T, Hjelmesæth J, Aune E, Educ 73:190–196 Bækkevar M, Norum IB, Otterstad JE (2013) The diagnostic 29. Vignon P, Mucke F, Bellec F, Marin B, Croce J, Brouqui T, Palo- accuracy of pocket-size cardiac ultrasound performed by unse- bart C, Senges P, Truffy C, Wachmann A, Dugard A, Amiel J-B lected residents with minimal training. Int J Cardiovasc Imaging (2011) Basic critical care echocardiography: validation of a cur- 29(8):1749–1757 riculum dedicated to noncardiologist residents. Crit Care Med 36. Lipiec P, Bąk J, Braksator W, Fijałkowski M, Gackowski A, 39:636–642 Gąsior Z et al (2018) Transthoracic echocardiography in adults— 30. Galderisi M, Santoro A, Versiero M, Lomoriello VS, Esposito R, guidelines of the Working Group on Echocardiography of the Pol- Raia R, Farina F, Schiattarella PL, Bonito M, Olibet M, de Simone ish Cardiac Society. Kardiol Pol 76(2):488–493 G (2010) Improved cardiovascular diagnostic accuracy by pocket 37. Filipiak-Strzecka D, Michalski B, Kasprzak JD, Lipiec P (2014) size imaging device in non-cardiologic outpatients: the NaUSiCa Pocket-size imaging devices allow for reliable bedside screening (Naples Ultrasound Stethoscope in Cardiology) study. Cardiovasc for femoral artery access site complications. Ultrasound Med Biol Ultrasound 8(1):51 40(12):2753–2758 31. Andersen GN, Haugen BO, Graven T, Salvesen O, Mjølstad OC, Dalen H (2011) Feasibility and reliability of point-of-care pocket- sized echocardiography. Eur J Echocardiogr 12:665—70 1 3

Journal

The International Journal of Cardiovascular ImagingSpringer Journals

Published: May 30, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off