key: cord-285946-1q4k5bnh
authors: Birk, Rubinder; Shaw, Dominick; Kennedy, Cheika; Higashi, Yutaro; Patel, Roma; Gupta, Ayushman; Au-Yong, Iain
title: Low detection rate of pulmonary embolism in patients presenting to the emergency department with suspected coronavirus disease 2019 (COVID-19): a single-centre UK study
date: 2020-09-23
journal: Curr Probl Diagn Radiol
DOI: 10.1067/j.cpradiol.2020.09.014
sha: 
doc_id: 285946
cord_uid: 1q4k5bnh

PURPOSE: Critically ill patients with COVID-19 are at increased risk of thrombosis. There are limited data on PE rates in COVID-19 patients at presentation to the emergency department (ED). In this study we evaluated the detection rates of PE in patients presenting to the ED with suspected and proven COVID-19. METHODS: A single-centre retrospective study was undertaken of 285 consecutive patients undergoing CT pulmonary angiogram (CTPA) in the Emergency Department at Nottingham University Hospitals NHS Trust in the UK between 25th March and 30th April 2020. At our institution, CTPA is performed in all patients undergoing CT for triage. The study group consisted of patients considered COVID-19 positive based on polymerase chain reaction (PCR) results and CTPA findings. The detection rate of PE in COVID-19 patients was compared to patients undergoing CTPA for suspected PE only and for suspected COVID-19 with no COVID CT findings and negative PCR (control group 1); and CTPAs prior to the coronavirus pandemic (control group 2). RESULTS: 1/48 patients in the study group had a PE (2%) compared to 25/215 (12%) in control group 1 and 10/50 (20%) in control group 2. Prevalence of PE in the study group was lower than in control group 1 (p=0.058) and compared to control group 2 (p=0.005). 11 patients undergoing CTPA had negative PCR but positive CT for COVID-19. CONCLUSION: Detection rate of pulmonary embolus is low in patients with COVID-19 undergoing CTPA on a triage pathway.

Coronavirus disease 2019 is caused by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2). Patients with severe COVID-19 have been shown to have coagulation dysfunction (1) (2) (3) (4) , which may lead to venous and arterial thromboembolism (4, 5) . This is thought to be due to several factors including increased inflammatory response and cytokine release, hypoxaemia and haemostatic abnormalities related to sepsis such as disseminated intravascular coagulation (3, 5, 6) . Raised D-dimer may indicate excessive coagulation activation and elevation is common in patients with COVID-19 (1) (2) (3) (4) (7) (8) (9) . Numerous studies show a clear association between COVID-19 and pulmonary embolism (PE) (5, 10) , with a high incidence of PE (4, 11, 12) in patients with severe disease admitted to intensive care (ITU). The prevalence of PE in these patients has been reported in studies as 14-23% (4, 5, 11) .

There is currently limited evidence on detection rates of PE at presentation to the emergency department (ED) in patients with COVID-19. These patients may have mild disease or may be earlier in the disease course. The lack of evidence may relate to variation in use of computed tomography (CT) in patients with COVID-19 at presentation to ED across the UK. The Nottingham University Hospitals (NUH) COVID-19 Adult Triage pathway for patients presenting to ED ( Figure 3 ) mirrors aspects of the British Society of Thoracic imaging (BSTI) decision tool for suspected COVID-19(13). Unwell patients with suspected COVID-19 based on symptoms (dry cough, fatigue, myalgia, fever, dyspnoea and anosmia) undergo a chest x-ray (CXR) and viral nasopharyngeal throat swab for polymerase chain reaction (PCR). If CXR appearances are normal or equivocal, CT is considered to assess for typical COVID-19 features if the PCR result is unavailable. A large meta analysis (14) of findings in COVID-19 reports that ground glass density is the most commonly encountered CT finding. Consolidation, interlobular septal thickening and air bronchograms are other commonly reported findings. Abnormality is most commonly bilateral and peripheral in distribution, without particular lobar or zonal predominance. Pleural effusion and lymphadenopathy are rarely encountered. Figures 1 and 2 demonstrate some illustrative CT findings in this cohort of patients. This meta-analysis suggests that CT is sensitive for the detection of COVID-19 but that there is overlap with the imaging findings seen in other infections, so the diagnosis should be confirmed with RT-PCR.

At NUH, patients were diagnosed with COVID-19 if the PCR was positive or if the CT had typical features of COVID-19 ( Figure 3 ). CTPA was performed on all patients on this pathway unless contraindicated. The NUH pathway considered several factors including staffing, scanner access (and cleaning) and a two-site service, with need to isolate and treat COVID-19 patients in a different hospital from where the ED is based. CTPA was specified because of emerging evidence regarding increased risk of thromboembolism and expected raised D-dimer in these patients, to prevent the need for reimaging further down the line when transfer to the scanner might be difficult. This however needed to be balanced against the potential disadvantages of administering intravenous contrast to patients with COVID-19, such as the potential risk of contrast-induced nephropathy and the potential for parenchymal changes on administration of contrast material. This provides us with a unique cohort of patients with COVID-19 presenting to ED that underwent CTPA, in that CTPA is normally only performed in patients with suspected PE. In our study all patients undergoing CT for triage were also scanned in the CTPA phase.

We conducted a retrospective observational study assessing the prevalence of PE in patients presenting to ED at Queens Medical Centre, Nottingham diagnosed with COVID-19, compared to two separate groups without COVID-19.

Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.

The study was approved by the local ethics committee.

Consecutive data were collected from the start of implementation of the NUH ED COVID-19 pathway on 25 th March, to 30 th April 2020. This pathway is illustrated in Figure 3 .

The study cohort included all patients attending ED who underwent CTPA on the NUH COVID-19 pathway in whom a diagnosis of COVID-19 was made by PCR on a nasopharyngeal swab or with high suspicion of COVID-19 on CT (irrespective of clinical presentation). The first control group (control group 1) included all patients who underwent CTPA in ED during the same time period to investigate PE only (without suspected COVID-19), based on clinical details provided in the radiology request, and patients undergoing CTPA on the COVID-19 pathway with a negative PCR and negative or equivocal CT for COVID-19.

Exclusion criteria included: nasopharyngeal swabs not sent, if a patient underwent unenhanced CT (due to renal failure or contrast allergy) and if the CTPA was non-diagnostic for PE. Diagnosis of PE and radiological suspicion of COVID-19 was determined on retrospective review of the radiology report. Reporting of suspected COVID-19 on CT followed the BSTI guidance(15). At our institution, CTPA is reported by Consultant radiologists specialising in body imaging.

Variable sensitivities of PCR have been reported in the literature (34-80%)(16), therefore negative PCR does not exclude COVID-19. It is possible that the first control group included patients with COVID-19 with false negative PCR and this may have affected the observed prevalence of PE. We therefore also studied a separate smaller cohort of 50 consecutive patients presenting to ED at our institution undergoing CTPA for suspected PE before the coronavirus pandemic in September 2019 (control group 2). 

Statistical analysis was performed using Stata version 15.1. The prevalence of PE in the study group was compared with the control groups. Statistical significance was assessed using Fisher's exact test if the expected value in any of the groups was <5 and Chi-squared was used if the expected value was ≥5. A p-value of <0.05 was considered to be statistically significant. Control group 1 was subdivided into patients undergoing CTPA on the COVID-19 pathway and those who had CTPA performed for suspected PE only. The prevalence of PE in the study group was compared with these control subgroups individually. The prevalence of PE in the study group was also compared with control group 2. Table 1 outlines the clinical characteristics of the study and control groups. 285 CTPAs (in 285 patients) were performed in Nottingham ED during the study period. 21 patients on the triage pathway did not have PCR and 1 patient had a non-diagnostic CTPA for PE, therefore were excluded. 174 of the remaining CTPAs were performed via the NUH COVID-19 pathway. Of these, 34 patients had positive PCR and 140 were PCR negative. 11/174 (6%) patients undergoing CTPA on the COVID pathway had negative PCR but positive CT for COVID-19. Three of these patients had a second negative PCR the same admission, none of the other patients had a repeat swab. 89 patients underwent CTPA outside of the pathway, for suspected PE, however 3 patients had positive CT for COVID-19 (all had negative PCR). Of interest one patient had an antibody test 3 months later suggesting that the PCR was false negative in this case.

The study group therefore consisted of 48 patients. Control group 1 consisted of 215 patients. Control group 2 consisted of 50 patients undergoing a CTPA in September 2019. Figure 4 is a study flowchart summarising the groupings. The mean age of all patients undergoing CTPA in the study period was 61 years (standard deviation 19). The COVID-19 positive patients were generally older than those in control group 1 and 2 (mean 65 years versus 60 years and 64 years respectively). A greater proportion of patients in all groups were female. (2) 25/215 (12) COVID pathway 12/129 (9) Non-COVID pathway 13/86 (15) 10/50 (20) Table 1 Clinical characteristics of the study and control groups of patients. The study group consisted of patients who were PCR positive or had CT changes consistent with COVID-19. Control group 1 consisted of patients who were PCR negative and did not have a positive CT. Control group 2 was a consecutive cohort undergoing CTPA at our institution before the pandemic started. Data are number/n (%) for gender, mean (SD) for age, or n/N (%), where N is the total number of patients in the cohort.

In the study group, 1 patient (2%) with PCR-positive COVID-19 had radiologically confirmed PE -a saddle embolus. The patient was an 81-year male patient with no risk factors for PE. Parenchymal lung features on CT were equivocal for COVID-19 in this patient as they could have represented pulmonary infarcts. In control group 1, 25 patients (12%) had confirmed PE. 12 of these patients had CTPA performed via the COVID-19 pathway but tested negative for COVID-19 on PCR or had negative or equivocal CT for COVID-19. Of the patients in control group 1 who had CTPA performed only for suspected PE, 15% had PE. The prevalence of PE in control group 2 was 20%.

There was no statistically significant difference in the prevalence of PE in the study group compared to control group 1 as a whole (p=0.058). This was also the case on subgroup analysis comparing the study group to patients in control group 1 undergoing CTPA on the COVID-19 pathway without COVID-19 (PCR and CT negative), p = 0.191. PE prevalence in the study group was significantly lower when compared to patients in control group 1 undergoing CTPA for suspected PE only who had no CT features of COVID-19, p = 0.018. PE prevalence was also significantly lower in the study group compared to control group 2 (p=0.005).

We also compared these figures to a consecutive group of inpatients over the same time period who were PCR-positive and underwent CTPA. 8/43 (19%) of these scans demonstrated PE.

The findings of this study demonstrate our experience of adding a CTPA examination when patients with suspected COVID-19 undergo CT at our institution for triage. The overall trend was of a lower detection rate of PE in COVID-19 patients based on PCR and CT scanned in ED than in the control groups. The prevalence of PE in patients with COVID-19 was significantly lower than in patients undergoing CTPA in Nottingham ED for suspected PE only, both during and before the coronavirus pandemic. Of note, if only the patients with PCR positivity for COVID-19 are considered to have the disease, and those with CT changes suggesting COVID-19 are not included in the study group, the prevalence of PE remains low at 1/34 (3%). The observed PE prevalence before the pandemic was slightly higher in the patients undergoing CTPA for suspected PE during the pandemic. The reasons for this are unclear but may in part relate to differences in imaging strategy during the different time periods. During the pandemic, CT was used at our institution to triage patients to appropriate parts of the hospital and the CTPA component was employed on every patient undergoing CT for triage to exclude PE (see Figure 3) . A large US multicentre study (17) cites the rate of PE normally encountered in the emergency department at 4%.

The findings of our study contrast with much of the published literature, as numerous published studies (1-4, 12, 18, 19) report high prevalence of PE in patients with COVID-19 undergoing CTPA. The high prevalence of PE has prompted some groups to suggest consideration of including CT pulmonary angiography to chest CT examinations in patients with suspected COVID-19, as we have done at our institution (20) . Coagulation dysfunction is however thought to be associated with late or severe disease, with most data from an ITU setting. It may be that the incidence of PE is lower early in the disease, as our cohort has been imaged early in their presentation in ED rather than ITU, as per the management flowchart ( Figure 3) . We found the PE detection rate in a PCR-positive comparative cohort locally to be 19%; some of these patients were imaged later on in their care and this figure is more comparable to the published literature.

Of interest, Cattaneo et al. report a low incidence of deep vein thrombosis (DVT) in their cohort and postulate that some of the findings on CTPA may relate to thrombosis in ITU patients rather than pulmonary embolus, given that PE often coexists with DVT (19) . The low rate of PE observed in our study early in patient management correlates with the findings of their study and it has been suggested that PE later on in the disease may relate to pulmonary arterial thrombosis. Pathological studies also suggest that patients with COVID-19 develop in situ vascular changes and thrombosis (21) . The study of Thomas et al. also demonstrates increasing incidence of PE in their cohort with duration of stay on ITU (22) , which would suggest incidence is lower earlier in the disease.

D-dimer was not routinely performed in the study group (only 6 out of 34 patients). Anticoagulation status is unknown, but it was not the policy of our ED at the time of admission to routinely commence thromboprophylaxis for patients with suspected or confirmed COVID-19.

The study has several limitations. Firstly it is a retrospective, relatively small sample single-centre study evaluating local practice. It does not assess all COVID-19 patients (Figure 3 ) as patients who did not require admission or patients who had a CXR suggestive of COVID-19 did not undergo CTPA. Given the nature of this pathway, we have probably excluded patients with mild disease who were discharged without imaging and patients with severe disease who had classic CXR changes for COVID-19 who did not require CT for triage. Given that all patients were scanned in ED prior to admission, the CT is assumed to have taken place early in their management. The limited sensitivity of PCR in detecting COVID-19 may have led to underestimation of the incidence of COVID-19 in the patient cohort. Although the initial sensitivity of PCR at NUH based on the first swab was quoted as 96% early on in the pandemic (unpublished data), 11 patients had COVID-19 diagnosed on CTPA with negative PCR. This was important because it potentially helped reduce nosocomial transmission by isolating patients in the correct ward environment and aided management and escalation decisions. This cohort is however thought to provide a unique insight into the detection rate of PE associated with adding a CTPA on every patient undergoing triage.

The detection rate of PE in patients presenting to ED diagnosed with COVID-19 is reported in this study at 3%. This is significantly lower than the both the PE prevalence in patients without suspected or diagnosed COVID-19 and the baseline PE prevalence outside of the coronavirus pandemic. It is also lower than reported rates of PE in COVID-19 patients in the literature, although many of these reports are in the ITU setting. This may reflect lower prevalence of thromboembolism early in the disease although this is a selected cohort. The low detection rate at this stage would suggest that clinicians should not necessarily be reassured by an initial negative study, which may not obviate the need for a repeat study further down the line, if this is clinically indicated. This may warrant further larger scale studies for clarification. The role of CTPA in evaluating patients with suspected COVID-19 at presentation remains unclear. We hope our experience will be useful for development of strategies for imaging in patients with COVID-19.

Clinical characteristics of 113 deceased patients with coronavirus disease 2019: retrospective study

Clinical characteristics of coronavirus disease 2019 in China

Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan

High risk of thrombosis in patients with severe SARS-CoV-2 infection: a multicenter prospective cohort study

Incidence of thrombotic complications in critically ill ICU patients with COVID-19

Prevalence of venous thromboembolism in patients with severe novel coronavirus pneumonia

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. The lancet

Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. The lancet

Abnormal coagulation parameters are associated with poor prognosis in patients with novel coronavirus pneumonia

Coexistent COVID-19 pneumonia and pulmonary embolism: challenges in identifying dual pathology

Acute pulmonary embolism associated with COVID-19 pneumonia detected by pulmonary CT angiography

Pulmonary embolism in COVID-19 patients: awareness of an increased prevalence

Coronavirus Disease 2019 (COVID-19) CT Findings: A Systematic Review and Meta-analysis

Imaging BSoT. **UPDATED** COVID-19 BSTI REPORTING TEMPLATES AND CODES 2020

Rapid tests for influenza, respiratory syncytial virus, and other respiratory viruses: a systematic review and metaanalysis

Multicenter evaluation of the YEARS criteria in emergency department patients evaluated for pulmonary embolism

Acute pulmonary embolism in COVID-19 patients on CT angiography and relationship to D-dimer levels

Pulmonary embolism or pulmonary thrombosis in COVID-19? Is the recommendation to use high-dose heparin for thromboprophylaxis justified? Thrombosis and Haemostasis

Pulmonary embolism in patients with COVID-19: Time to change the paradigm of computed tomography

Clinical pathology of critical patient with novel coronavirus pneumonia (COVID-19)

Thrombotic complications of patients admitted to intensive care with COVID-19 at a teaching hospital in the United Kingdom

Figure 1: Images from a CTPA demonstrate a right sided subpleural area of ground glass density (top image, arrowed) on lung window settings. The CTPA demonstrated extensive saddle embolus (bottom image). The patient had a positive PCR for COVID-19

We would like to thank all clinicians involved in designing and creating the NUH COVID-19 pathway.

R. Birk, I. Au-Yong, C. Kennedy, D. Shaw and Y. Higashi contributed to the study design and concept. R. Birk, R. Patel, C. Kennedy and I. Au-Yong collected the clinical data. Statistical analysis and interpretation was done by R. Birk and A. Gupta. The final manuscript has been read and approved by all authors.

None of the authors have any conflict of interest relating to this article.This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Study flowchart (* represents the study group, † combined represent control group 1 and ‡ represents control group 2). Please note that it was sometimes necessary to proceed to CTPA for triage when the PCR swab result was awaited.