**5. Noninvasive imaging modalities**

Venous duplex ultrasound uses the ability to detect venous blood flow and real-time B-mode images to identify clots in both upper and lower limbs [36]. The specific diagnostic DVT criteria include the following, lack of venous segment collapse on pressure (more specific), respiration induced loss of phase changes, a weak venous response to Valsalva, echogenic substance in the lumen, loss of increase in flow due to compression, and loss of flow or decreased flow on color Doppler [1]. In symptomatic patients, duplex ultrasonography sensitivity and specificity in diagnosing DVT are higher than in asymptomatic patients (lesser accuracy). While 30–40% of PE patients are clinically symptomatic for proximal DVT, the venous duplex can detect proximal DVT in 60–80% of PE patients [37]. In postoperative orthopedic patients, the performance of venous duplex ultrasound was comparable to contrast venography in asymptomatic proximal DVT detection [38]. Asymptomatic DVT in contralateral LE was seen in 5–10% of patients with acute symptomatic DVT [39]. Duplex ultrasonography accuracy in identifying deep calf vein limited DVT, and asymptomatic proximal vein DVT is limited in high-risk populations [40]. After an initially negative result in suspected DVT patients, serial duplex ultrasonography can detect the proximal extension [41].

The sensitivity of detecting PE via TTE (transthoracic echocardiogram) is only 50%, so it is a poor imaging modality for acute PE diagnosis [42]. RV pressure or volume overload suggestive of PE can guide PE diagnostic imaging without other differentials [1]. TTE can reveal the McConnell sign (RV mid-free wall lack of movement with no apex involvement) [43]. On rare occasions, emboli can be visualized in the right heart on TTE. TTE based risk assessment helps in guiding acute PE therapy. Patients with RV dysfunction on TTE in a normotensive patient indicate adverse outcomes or early mortality [44, 45]. An appropriately done TEE (transesophageal echocardiogram) can detect central PE (Pulmonary artery and its branches) with a true positive and negative rate > 90% [46]. It is an excellent modality to consider in a speculated massive PE patient hemodynamically unstable for transport or has contrast contraindication. TTE helps ward off other differential diagnoses, including infective endocarditis, pericardial effusion or tamponade, aortic dissection, and RV myocardial infarction. Significant changes seen on a TTE are mentioned below in **Table 3** [47].

A meta-analysis assessed multiple echo studies and consistently showed that TTE had a greater specificity and sensitivity for PE diagnosis and is a definitive rule in test at the bedside for suspected patients [47]. As per ACEP guidelines, the presence of an RV dysfunction on TTE in an unstable patient can suggest acute PE and an indication for thrombolytic therapy [48].

A positive or negative transthoracic lung ultrasound (TLS) can cause an increment or decrement in PE probability by 30% in a moderate risk population, which can change the diagnostic workup [49]. TLS can detect smaller PE in the periphery of the lungs [50]. Most emboli are observed in the lower lungs, which can be easily accessed by TLS [51]. An endobronchial ultrasound (EBUS) can detect central PE and immensely help PE patients with AKI, contrast contraindication, pregnancy, and hemodynamically unstable patients with diagnosis [52–54]. Simultaneously it can measure the acutely elevated pulmonary hypertension in patients with PE. Endobronchial ultrasound findings can supplement TLS in acute PE detection [49].


#### **Table 3.**

*Distinct TTE signs seen in acute PE.*

However, the resources required (regular bronchoscope, trained nursing staff) to perform a bedside bronchoscopy with EBUS makes it challenging to achieve in the emergency department or the MICU on an as-needed basis. Performing a bronchoscopy in a hemodynamically unstable patient may worsen the patient's overall cardiopulmonary status and increase his high risk of adverse outcomes.

A meta-analysis revealed that cardiopulmonary ultrasound (CPUS) sensitivity was 91% and specificity was 81% for PE diagnosis in comparison to CT pulmonary angiography (CTPA) [55]. The BLUE (bedside lung ultrasound in emergency) protocol was made to diagnose PE based on a DVT positive venous duplex combined with TLS. It was 99% specific and 81% sensitive for PE diagnosis and ruled out other acute respiratory failure differentials [56]. BLUE protocol and TTE consistently revealed a greater specificity than sensitivity due to the lack of ruling PE out with no CTPA [56, 57]. The combination may help in decreasing the excessive CTPA done currently [58]. In resource-limited settings such as in developing countries or the absence of CTPA availability, CPUS may have a role in managing PE [55].

Chest X-ray can either be normal or abnormal. Most often, a chest X-ray reveals nonspecific abnormal findings such as effusion, infiltrates, or atelectasis. Certain signs with interesting names that have been observed on chest radiograph imaging are mentioned in **Table 4** [59–61].

The occurrence of Westermark's sign and Palla's sign suggests embolic obstruction of either the lobar/segmental pulmonary artery/widespread small arterial involvement [60]. A patient with acute shortness of breath, respiratory distress, or hypoxia and a benign chest radiograph is suspicious of possible PE. A chest radiograph also helps in ruling out other causes such as empyema, pneumonia, and pneumothorax.


**Table 4.** *Interesting findings noted on chest-X-ray.*

### *Diagnosis of Pulmonary Embolism DOI: http://dx.doi.org/10.5772/intechopen.101047*

After CTPA, V/Q (ventilation/perfusion) scintigraphy is an alternative utilized in diagnosis. V/Q scintigraphy negative or high-probability result is of critical value in PE diagnosis [62, 63]. A negative V/Q scan result is as efficacious as a pulmonary angiogram, ruling out acute PE and slightly better than a CTPA [64]. A normal perfusion scan sensitivity in ruling out PE is exceptional, observed even with a higher pretest PE probability in severely sick patients [62, 65, 66]. A meta-analysis validates this observation by revealing a minimal 0.3% PE incidence in individuals with an intact perfusion result [67]. Similarly, high-risk V/Q scintigraphy (multisegmental mismatch defects) correlates with acute PE in 87% of patients, and the positive predictive value (PPV) is increased to 96% by a higher pretest probability [62]. However, most suspected acute PE or PE patients do not have V/Q scan findings suggestive of a high probability scan. Also, most patients with no PE did not have a normal V/Q result. A clinically significant portion of patients (33% = moderate risk and 10% = low risk) had positive angiograms. The prospective trial PIOPED stressed on the number of perfusion defects and size along with a concurrent image to identify V/Q mismatch defects [62]. In the PISA-PED study, greater emphasis was placed on the perfusion defect shape than the number and size or ventilation image correlation [65]. The PISA-PED study confirmed that perfusion in combination with pretest probability in the absence of ventilation image could diagnose acute PE without angiography [65]. A majority of intermediate V/Q imaging results are observed in obstructive lung disease patients [68]. V/Q scintigraphy is favored in patients with renal failure, contrast allergy and offers similar diagnostic efficacy in pregnancy [1, 69]. The severely sick patients can undergo bedside perfusion imaging to avoid transportation-associated risks [1].

SPECT-V/Q scan, a new scintigraphy process that generates 3-dimensional images than a planar image seen in V/Q scans. Advantages associated are better visualization of all perfusion defects in different lung areas, less radiation exposure than CTPA, fewer nondiagnostic test results (0.5–3%) [70–75]. SPECT-V/Q efficiency is similar to that of CTPA in suspected acute PE patients [76]. SPECT-V/Q true positive and negative rates were noted to be in the range of 95–100% [76, 77]. However, it cannot replace CTPA as a test of choice in acute PE due to the lack of vigorous extensive testing to verify its validity in suspected patients [78]. Specific clinical scenarios might be appropriate for using a SPECT-VQ scan, including a nondiagnostic CTPA study and post-discharge evaluation of lingering perfusion abnormalities.

The imaging modality of choice to exclude acute PE in suspected patients is CTPA. CTPA sensitivity is 83%, specificity is 96%, with an NPV of 97% plus a PPV of 86% in suspected patients [17]. CTPA sensitivity and specificity are greater than 95% in central PE (pulmonary artery and lobar branches) [79]. The sensitivity and specificity decline gradually when the emboli involve segmental or subsegmental pulmonary arteries. In a study of CTPA for subsegmental artery, involvement sensitivity was noted in the range of 71–84% [80]. PE involving only the subsegmental arteries of the pulmonary circulation is seen in around 30% of PE patients [81, 82]. CTPA evaluation is diagnostic when emboli involve the main or lobar pulmonary arteries and is considered suggestive if the segmental and subsegmental pulmonary arteries are occluded. CTPA predictive value is critically hampered in discordance with clinical evaluation, and further imaging tests merit consideration in this scenario [1]. CTPA has its limitations which are significant to be ignored. Intravenous iodinated contrast given during CTPA can cause AKI. CTPA cannot effectively diagnose emboli in the subsegmental pulmonary arteries and cannot supplement the V/Q scan, which also lacks in this particular territory [62, 81]. CTPA is not able to identify whether the emboli is acute, subacute, or chronic. CTPA occurrence results in significant radiation exposure to the patient, especially in young females (breast and lungs) [83]. Clinical observations have suggested significant overuse of CTPA resulting in overdiagnosis of pulmonary embolism; however, it cannot determine whether the positive CTPA identifies acute PE, subacute PE, or chronic PE [63]. A clinical study observed no substantial difference in results when anticoagulation was withheld in suspected individuals with a normal CTPA plus negative leg duplex ultrasonography and negative or non-high probable V/Q imaging with a negative leg venous duplex [63]. The study was performed on relatively stable patients, so this observation cannot be utilized in severely sick patients or patients with inadequate cardiopulmonary reserve [84].

A clinical trial PIOPED III evaluated a magnetic resonance angiogram for PE diagnosis [85]. Overall, the sensitivity to diagnose a PE involving the main and lobar pulmonary artery was 79%. It may be an ideal test for patients with intravenous contrast allergy or to avoid radiation exposure, such as pregnancy.

### **6. Invasive imaging modalities**

CT venography (CTV) has similar diagnostic accuracy as venous duplex ultrasound for the LE in diagnosing or ruling out DVT; however, the test is invasive and comes with exposure to contrast and radiation [86]. In addition to the LE venous system, vena cava and pelvic veins are visualized. CTPA and CTV combined revealed a mild improvement in diagnostic outcome with a substantial increase in cost and pelvic exposure to radiation.

Contrast venography is the best imaging study for substantiating LE venous thrombosis. Its diagnostic criteria include a persistent venous filling defect observed in ≥ two views. It is an expensive, invasive test requiring clinical expertise and accurate interpretation with significant exposure to intravenous iodine contrast. Due to the above reasons, Venous duplex has replaced it as the test of choice to diagnose acute DVT.

Before CTPA, a pulmonary angiogram was the best imaging study to diagnose acute PE. It requires appropriate clinical expertise to perform the invasive procedure and interpret it. Three factors determine the result, including the location of the emboli, image quality, and interpreter's experience [1]. Diagnosis of PE is indicated by either a filling defect and/or abrupt vessel cutoff. Flow defects can be avoided by ensuring good vascular opacification and obtaining multiple sequences of films. The PIOPED trial revealed that it was diagnostic in 97% of patients and associated with 1% complications, including a mortality rate of 0.5% [87]. Adverse outcomes were significantly seen in MICU patients transported for an angiogram. A pulmonary angiogram is considered in a tiny patient subset when PE diagnosis cannot be determined by noninvasive imaging studies, significant discordance between imaging study and clinical evaluation, and chronic thromboembolic disease.

### **7. Diagnostic workup**

The current clinical approach for acute PE or DVT diagnosis utilizes a Bayesian analysis. Here pretest probability of the clinical condition is measured exclusive of the test outcome wire clinical means or a consistent prediction rule such as Well's or Geneva score. Then a posttest probability of the clinical condition is generated by utilizing pretest probability combined with a test's likelihood ratio. The posttest probability is used as guidance for clinical decision making that confirms or to excludes the disease with a degree of probability or helps in deciding additional imaging studies. Clinical predictive rules for acute DVT include the Well's, revised Well's, and Geneva scores. Well's score classifies suspected acute DVT patients into three subclassification's unlikely (3%), moderate (17%), and likely (75%). With the help of Well's score

*Diagnosis of Pulmonary Embolism DOI: http://dx.doi.org/10.5772/intechopen.101047*

**Figure 1.**

*Diagnostic testing of acute PE based on clinical probability.*

and LE venous duplex ultrasound, diagnosis of acute DVT can be established in likely individuals with a positive LE venous duplex result. In unlikely individuals, a negative LE venous duplex result excludes acute DVT [88]. The revised Well's score divided the patients into likely and unlikely categories. The unlikely group with a negative D-dimer excluded an acute DVT without needing a LE venous duplex ultrasound [89, 90].

Clinical predictive tools for acute PE include Well's criteria, revised Geneva score, Pisa model, PERC (PE rule-out criteria), and Charlotte rule [91–95]. Among the clinical predictor rules or scores for acute PE diagnosis, Well's score fared better than the revised Geneva score [96]. Most diagnostic algorithms for acute PE use either CTPA or V/Q scintigraphy as a first test. A diagnostic algorithm based on clinical probability has been described in **Figure 1** below for acute PE.

Clinical predictive rules or scores are not superior to clinical assessment but offset the variation observed with physician judgment and experience by standardization [1]. These rules were framed for patients seen in outpatient settings and applicable in primary care and emergency departments; however, they fare poorly and lack clinical validity in hospitalized patients. In hospitalized patients, the clinical predictive scores or rules and D-dimer are of minimal help to make a clinical decision. As a result, most of these patients need an imaging study to rule in or exclude acute PE diagnosis [1].

### **8. COVID-19 infection and venous thromboembolism**

Although acutely ill patients are at a higher risk of acute PE, COVID -19 (coronavirus disease 2019) infected patients suffer from in situ immunothrombosis characterized by small and medium pulmonary artery numerous thrombi occurring at a greater frequency (24%) than in H1N1 influenza patients [97–101]. PE phenotype in COVID-19 patients in comparison to others correlates with peripheral thrombotic lesions with a lesser clot burden [102]. Pulmonary localized immunothrombosis in COVID-19 is a minimal addition to the overall procoagulant state resulting in DVT seen only in 42.4% of PE patients in COVID-19 than the 60% occurrence seen frequently in other PE patients [103]. D-dimer elevation in COVID-19 could be due to the acute infection-induced inflammation causing a procoagulant state or a localized micro thrombosis in the pulmonary vasculature.

D-dimer levels elevation >500 μg/L & 1000 μg/L were associated with a higher sensitivity (> 90%) but a lower specificity (< 30%) in PE diagnosis. D-dimer performance in COVID-19 is similar to that seen in other prothrombotic conditions [104]. COVID-19 patients on statin therapy before admission had a lower risk of PE occurrence [105]. COVID-19 patients with significant inflammation measured by elevated C-reactive protein and D-dimer were at a higher risk of acute PE [106]. PE was detected frequently in the periphery than in the central pulmonary arteries [107]. The most frequent site was the segmental, followed by the lobar, central, and subsegmental PE [106]. Acute PE was observed at a greater prevalence in COVID-19 patients with an increased Body mass index (BMI) [106].

A systematic review and meta-analysis on PE and DVT in COVID-19 patients disclosed a higher pooled PE incidence of 24.7% in ICU patients. This percentage is substantially higher than the proportion seen in other viral pneumonia admitted to ICU in the presence or absence of acute respiratory distress syndrome (1.3%–7.5%) [108, 109]. In contrast, it was 10.5% in non-ICU patients (higher than the usual) [107]. Overall the incidence rates of PE and DVT in COVID-19 patients were 16.5% and 14.8% [107]. COVID -19 infection severity and CTPA universal screening correlated with a greater frequency of PE diagnosis. DVT was a concurrent finding in 42.4% of patients with acute PE. An elevated BMI (> 30) correlated substantially with a 2.7 times higher frequency of an acute PE [106]. As observed in a recent study, obese patients with COVID-19 suffer from a more severe disease [110]. A meta-analysis revealed an increased prevalence of venous thromboembolism and acute PE with increasing age in COVID-19 patients [111].

### **9. Conclusion**

Acute PE is an emergency, and ongoing research will reveal newer biochemical assays and better imaging studies for accurate earlier detection of acute PE in the upcoming few years. SPECT V/Q scan is currently undergoing evaluation at multiple centers where it is being compared to planar V/Q study and CTPA in suspected PE patients for accuracy. Physicians must understand the fallacies of each biochemical test and imaging study for their appropriate utilization in a clinical scenario for the patient's best outcome. Critical ECG findings and bedside echocardiogram findings should be stressed upon admission and utilized for prognostification. Physicians should be aware that these clinical scores or prediction rules play no role in hospitalized patients and should not be used for decision-making in this particular patient subset.

### **Acknowledgements**

"No external funds have been utilized in the preparation of this manuscript."

### **Conflict of interest**

"We declare no conflict of interest."
