Using Non-Invasive Respiratory Monitoring for COVID-19 Pulmonary Embolism Diagnosis

Abstract

With the high incidence rate of pulmonary embolism (PE) and pneumonia reported in hospitalized patients with COVID-19, the ability to determine the dominant etiology for severe respiratory distress quickly and accurately is crucial to a patient’s well-being. Traditionally, D-dimer blood tests and diagnostic imaging studies would be utilized to determine the presence of a PE or a venous thromboembolism. However, COVID-19 places patients in a prothrombotic state and performing diagnostic imaging studies on all patients with COVID-19 would be impractical, making the need for a simple and reliable method to determine the likelihood of PE or venous thromboembolism a priority for emergency departments. The authors believe the use of non-invasive respiratory monitoring technology to assess lung function in hospitalized patients with COVID-19 can aid in discerning the dominant hypoxia etiology and tailoring of their treatment. Here, the authors outline a case and method of using non-invasive respiratory monitoring of lung function in the successful diagnosis of a PE in a 62-year-old patient with COVID-19.

Introduction

COVID-19–related pulmonary embolism (PE) and pneumonia are two severe respiratory consequences linked with a high risk of Intensive Care Unit admission, various morbidities, and reported mortality rates for severe PE and pneumonia ranging between 5% and 50%, depending on severity. ^1,2 Timely differentiation between the two is of the upmost importance, as the treatment for each differ greatly and long-term consequences are associated with acute respiratory distress in hospitalized patients with COVID-19. However, attempting to determine the dominant etiology of COVID-19–related respiratory distress is difficult if solely based on a clinical picture of shortness of breath (SOB), hypoxia, labored breathing, and fever.

Recommendations suggest treating hospitalized patients with COVID-19 with anticoagulants to improve mortality; however, dosage protocols are varied, with some utilizing elevated D-dimer levels as a guide to dosing. ³ However, because patients with COVID-19 are in both an inflammatory and a prothrombotic state, both of which may be responsible for elevations in D-dimer levels, this laboratory test may not be reliable as an indicator for PE or venous thromboembolism (VTE) in patients with COVID-19. ⁴ This can inadvertently lead to the overuse of anticoagulants and the undertreatment of systemic issues in patients with COVID-19. Though diagnostic imaging with computed tomography (CT) or ultrasonography can aid in the diagnosis of PE, performing imaging on all patients with COVID-19 would be costly and potentially increase the risk of transmission. ⁵ With the reported incidence of VTE complication exceeding 30% and the incidence of PE between the range of 1.9% and 8.9%, it is crucial to have a simple, reliable, and non-invasive method to determine the likelihood of PE in hospitalized patients with COVID-19. ^6–8

Here, the authors report on an older, but generally healthy, man admitted for COVID-19 with respiratory distress and hypoxia while maintaining normal respiratory dynamics when measured by a non-invasive respiratory volume monitor (RVM), the ExSpiron 1xi. The patient was subsequently diagnosed with a PE via CT pulmonary angiogram (CTPA) scan.

Case Report

A 62-year-old man presenting with SOB was admitted to the emergency department (ED) of an urban community hospital. Fifteen days before presentation to the ED, the patient tested positive for COVID-19, and subsequently developed cough, fever, malaise, and diarrhea 4 days after. Ten days after first developing symptoms, the patient experienced SOB, which prompted him to seek medical attention by his primary care physician the following day. He was prescribed prednisone and colchicine, which he began taking the next day, but was subsequently admitted to the ED on the same day. He denied having chest pain, vomiting, melena, hematochezia, bruising, lower extremity swelling, or orthopnea upon presentation.

The patient had a relatively benign medical history that included controlled hypertension, a body mass index of 28, and a remote history of a cholecystectomy. Medications taken at home before admission to the ED included Losartan, vitamins C and D, and melatonin.

The patient was hypoxemic, with an oxygen saturation (SpO2) of 72% on room air, and tachypneic, with a respiratory rate (RR) of 24 breaths/min. The rest of his vital signs were within normal limits, with a blood pressure of 140/84 mmHg, a pulse of 98 bpm, and an oral temperature of 36.7°C (98°F). His SpO2 corrected to 92% when placed on a non-rebreather face mask at 15 L/min. A non-invasive RVM, the ExSpiron 1xi, was utilized to monitor his respiratory mechanics and reported an average RR of 23 breaths/min, tidal volume (TV) of 200–350 mL, a minute ventilation (MV) of 3.5 to 4.7 L/min (calculated by the device to be between 50% and 100% of the patient’s predicted minute ventilation [MV_pred]). MV_pred is reported as a percentage of a calculated MV based on height, weight, and sex, with a value of 80% or greater considered within normal limits and a value of greater than 200% being considered hyperventilation. ⁹ This report of relatively normal respiratory dynamics conflicted with the patient’s hypoxia and subsequently raised suspicions of PE as a likely etiology for the patient’s respiratory distress.

A chest X-ray was then performed, which found focal parenchymal opacities in the right upper lobe and right infrahilar region and a faint parenchymal opacity in the left lower lung. No pleural effusions were seen, and the cardiomediastinal silhouette size was within the upper limits of normal. These readings were consistent with the patient’s diagnosis of COVID-19 pneumonia; however, the degree of hypoxemia appeared out of proportion with the X-ray findings and the relatively normal respiratory monitor readings. In addition, a D-dimer test showed a value greater than 20, which raised the suspicion of a vascular shunt. The decision was made to perform a CTPA, which showed bilateral PE (Figures 1 and 2). No subsequent testing was performed this day.

Figure 1:

Computed tomography pulmonary angiogram demonstrating patient’s right lower lobe and left interlobar artery pulmonary emboli.

Figure 2:

Computed tomography pulmonary angiogram demonstrating patient’s left upper lobe segmental pulmonary emboli.

The patient was started on a therapeutic dose of Lovenox in addition to Remdesivir, Anakinra, and dexamethasone. The patient’s SpO2 remained stable at 95% for 4 days on a non-rebreather mask at 15 L/min. The patient was then downgraded to a ventimask at 40% fraction of inspired oxygen. On hospital day 3, the patient’s D-dimer results showed a decrease in fibrin degradation products, which suggested a decrease in the clot size, though this decline could also indicate a decrease in active inflammation due to dexamethasone. On day 6, the patient’s SpO2 was 95% at rest and 91% while ambulating, and he was discharged home with home oxygen at 4 L/min. Eventually, the patient was weaned off oxygen 5 days after discharge and had a complete recovery by day 15 post-discharge.

Discussion

At the time of an ED visit or hospitalization, patients with viral pulmonary infections can present with a range of symptoms, such as SOB, labored breathing, hypoxia, and, at times, fever. ¹⁰ However, difficulty arises in differentiation between PE and pneumonia without testing given the similarities in both presentation and symptoms. Chest X-rays are often used to diagnose pneumonia in patients presenting with viral pulmonary symptoms; however, X-rays are unable to visualize pulmonary clots, and as such, basing a diagnosis solely on X-ray imaging may lead to the inaccurate treatment for only viral pneumonia. CTPA imaging is often performed in order to rule out PE, but given the increased risk of transmission of COVID-19, EDs have been advised to minimize the number of imaging studies performed. COVID-19 has also been reported to cause coagulation abnormalities in hospitalized patients, including elevated von Willebrand Factor, factor VIII, D-dimer, fibrinogen, neutrophil extracellular traps, prothrombotic microparticles, and anionic phospholipids. ¹¹ This fact limits the accuracy of blood tests, such as D-dimer tests, and may result in inaccurate detection of VTE in patients with COVID-19.

Prior to the COVID-19 pandemic, a clinical decision-making tool known as the Pulmonary Embolism Rule-out Criteria (PERC) made use of a set of criteria to define a low-risk subset of patients in whom the diagnosis of PE was being considered. The use of the PERC effectively and substantially reduced the number of imaging studies performed. Risk factors indicated within the PERC included age ≥ 50 years, heart rate ≥ 100 bpm, SpO2 < 95%, unilateral leg swelling, hemoptysis, recent surgery or trauma, prior PE or deep vein thrombosis, and hormone use, of which any positive indicated the necessity for further testing. Though considered generally accurate, a study on the criteria’s use in patients with COVID-19 indicates that the PERC does not perform well under the circumstances and suggests that the criteria be modified to include COVID-19 infection as an independent risk factor. ¹² As such, any individual presenting with COVID-19–related respiratory distress would not qualify for PE rule-out under a modified PERC and would subsequently require PE testing.

As a result of the need to rapidly determine the risk of serious respiratory distress in patients with COVID-19, a study proposed the implementation of a criteria called the Quick COVID-19 Severity Index (qCSI), which uses standard vital readings, such as RR, SpO2, and supplemental oxygen flow, to determine the general risk of a patient for severe respiratory distress within the first 24 h of a patient encounter. ¹³ Unfortunately, the qCSI does not specify the risk of developing specific types of respiratory distress, making its use limited in the context of rapid detection of PE. Looking at the parameters given in the qCSI, a potential substitute for RR as a measure of a patient’s lung’s health could be MV_pred, which could enable the criteria to be expanded to consider PE specifically, though further study would need to be conducted to determine this system’s efficacy in practice.

Consequently, there is a need to develop a novel decision-making algorithm specifically for COVID-19–related PE. The authors propose that if a hospitalized patient with COVID-19 presents with adequate levels of hypoxemia (≤ 93%), the use of respiratory monitoring should be implemented. The authors further suggest the use of RVM, such as the ExSpiron 1xi, as it circumvents the need for intubation, and therefore limits risk of COVID-19 transmission, by measuring electrical currents across the lungs and calculating the volume of air entering and exiting with an accuracy of 90% or greater for MV, TV, and RR. ^9,14 PE can cause hypoxia without seriously inhibiting respiratory dynamics by decreasing blood flow through the pulmonary arteries without obstructing airflow into the lungs. By contrast, pneumonia directly restricts airflow within the lungs due to inflammation of the air sacs. This then can cause an increase in RR to compensate for the loss of volume, subsequently causing an abnormal MV_pred reading ( < 150%). Therefore, if respiratory monitoring shows the patient has relatively normal respiratory dynamics while concurrently presenting with hypoxemia, the probability of PE being the dominant etiology is elevated and a diagnostic imaging study should be performed for confirmation (Figure 3).

Figure 3:

Flow chart for pulmonary embolism (PE) diagnosis in hospitalized patients with COVID-19. COPD = chronic obstructive pulmonary disease; CHF = congestive heart failure; CT = computed tomography; MVpred = predictive minute ventilation percentage; ILD = interstitial lung disease.

Conclusion

The ExSpiron 1xi enables non-invasive monitoring of lung function in hospitalized patients with COVID-19 by providing real-time, continuous measurements of TV, accurate RR and MV, and calculating a MV_pred for the patient based on height, weight, and sex. This data, combined with hypoxemia as measured by standard pulse oximetry, can serve as a guide for respiratory diagnosis and can help practitioners improve resource utilization and diagnostic accuracy of PE in patients with COVID-19, as well as reduce transmission, during this pandemic.