Handheld Lung Ultrasound to Detect COVID-19 Pneumonia in Inpatients: A Prospective Cohort Study

Abstract Background: Chest imaging, including chest X-ray (CXR) and computed tomography (CT), can be a helpful adjunct to nucleic acid test (NAT) in the diagnosis and management of Coronavirus Disease 2019 (COVID-19). Lung point of care ultrasound (POCUS), particularly with handheld devices, is an imaging alternative that is rapid, highly portable, and more accessible in low-resource settings. A standardized POCUS scanning protocol has been proposed to assess the severity of COVID-19 pneumonia, but it has not been sufficiently validated to assess diagnostic accuracy for COVID-19 pneumonia. Purpose: To assess the diagnostic performance of a standardized lung POCUS protocol using a handheld POCUS device to detect patients with either a positive NAT or a COVID-19-typical pattern on CT scan. Methods: Adult inpatients with confirmed or suspected COVID-19 and a recent CT were recruited from April to July 2020. Twelve lung zones were scanned with a handheld POCUS machine. Images were reviewed independently by blinded experts and scored according to the proposed protocol. Patients were divided into low, intermediate, and high suspicion based on their POCUS score. Results: Of 79 subjects, 26.6% had a positive NAT and 31.6% had a typical CT pattern. The receiver operator curve for POCUS had an area under the curve (AUC) of 0.787 for positive NAT and 0.820 for a typical CT. Using a two-point cutoff system, POCUS had a sensitivity of 0.90 and 1.00 compared to NAT and typical CT pattern, respectively, at the lower cutoff; it had a specificity of 0.90 and 0.89 compared to NAT and typical CT pattern at the higher cutoff, respectively. Conclusions: The proposed lung POCUS protocol with a handheld device showed reasonable diagnostic performance to detect inpatients with a positive NAT or typical CT pattern for COVID-19. Particularly in low-resource settings, POCUS with handheld devices may serve as a helpful adjunct for persons under investigation for COVID-19 pneumonia.

adjunct, e.g., for patients with a pending NAT test or an initial negative NAT but high clinical concern for COVID-19 [10,11].Both the World Health Organization and the Fleischner Society have recommended chest imaging for patients with moderate or severe symptoms of suspected COVID-19, even if NAT is negative [12,13].Several validated reporting systems are used by radiologists in assessing likelihood of COVID-19 pneumonia on chest CT.A commonly used consensus guideline from the Radiological Society of North America (RSNA) classifies CT morphologies as "typical" (highest suspicion), "indeterminate", "atypical", or "negative" for COVID-19 pneumonia [14].For example, typical CT findings include peripheral bilateral ground-glass opacities [15].Despite its utility, CT scan has drawbacks; it requires significant healthcare resources and time, exposes the patient to ionizing radiation, poses infection control risks, and may be unsafe for unstable, hypoxemic patients.Chest X-ray (CXR) reduces some of these drawbacks but has a low sensitivity for detecting COVID-19, particularly early in the disease course (55-83%) [16].Thus, some hospitals rely on more than one category of diagnostic test, as well as patient history and risk factors, to attempt to rule out COVID-19 infection and decide on isolation precautions.In our quaternary hospital system, both CXR and CT are used in an evidence-based algorithm for admitted persons under investigation (PUIs); for example, CT is often required to allow a clinician to "rule out" COVID-19 (and remove a patient from isolation) if initial CXR findings are concerning [11].Thus, in daily clinical practice, CXR and CT are sometimes used as helpful adjuncts to help rule out and rule in COVID-19 [8,17].
Point of care ultrasound (POCUS) has been increasingly studied as a viable alternative diagnostic modality.POCUS is inexpensive, rapid, ionizing radiation-free, widely available, and does not require travel to and possible infectious exposure of a radiology suite.Ultraportable, handheld machines are typically less expensive and easier to disinfect than traditional, larger ultrasound machines.In low-resource settings, POCUS may be available when CXR or CT are not.Importantly, COVID-19 tends to affect the lung periphery, which is visualized well by POCUS.Numerous studies have described characteristic features for COVID-19 on lung ultrasound, including bilateral B-lines (particularly "confluent" B-lines) and an irregular (or "serrated") pleural line.[15 18-20].B-lines, similar in appearance to movie premier searchlights, appear as hyperechoic laser-like reverberation artifacts that arise from the pleural line, obliterate A-lines, and extend to the bottom of the screen (Figure 1).They can coalesce into "confluent" B-lines, which appear as a broader beam of light (appearing less as a laser and more as a broad flashlight in the fog); this artifact has also been called "light beam artifact," "waterfall" B-line, or "white lung."Compared to a normal pleural line, which is distinct and unbroken (something that could be drawn in one stroke by an imaginary white pencil), an irregular or "serrated" pleural line appears jagged and/or broken.Of note, while these ultrasound findings raise suspicion for COVID-19 pneumonia, they are not specific and can be seen in other pulmonary disease processes such as acute respiratory distress syndrome, cardiogenic pulmonary edema, interstitial lung disease, or pneumonia from other microbial etiologies.Of course, some CXR or CT findings suggestive of COVID-19 are similarly non-specific.three points per zone; the total number of points would correlate with the severity of COVID-19 pneumonia [21].The same group analyzed protocols with fewer lung zones (e.g., four to twelve zones), given that an abbreviated protocol could help reduce exposure time of the scanner with infectious patients; the study concluded that a 12-zone protocol (including 4 posterior zones) was optimal to assess COVID-19 severity [22].The same group has validated their protocol for prognosis of COVID -19 pneumonia [23], but to our knowledge none has attempted to validate it for diagnosis of COVID-19 pneumonia (particularly for with a 12-zone protocol following Soldati's score).Notably, there are several valuable studies suggesting the utility of POCUS for possible diagnosis of COVID-19 [24][25][26][27][28][29].However, these studies have limitations which hinder their generalizability, such as using a single operator, relying on the scanner's overall subjective gestalt for COVID-19 pneumonia, not comparing to CT scan, or excluding patients with heart failure (which may have similar POCUS findings as COVID-19 pneumonia).In addition, few studies have gathered all data using handheld, ultraportable devices.
Our aim was to validate a 12-zone protocol following Soldati et al. with a handheld device, assessing the diagnostic accuracy of lung POCUS performed with a handheld ultrasound machine to detect either (1) a highsuspicion ("typical") pattern for COVID-19 on CT scan or (2) a positive COVID-19 NAT.We hypothesized that handheld lung POCUS has superior sensitivity and specificity compared to CXR for both outcomes.

Study design and setting
This prospective cohort study took place in a 1,000-bed quaternary care hospital in the Northeastern U.S., from April to July 2020 (the first wave of the pandemic).We included a convenience sample of adult patients who were admitted to either the medical ward or intensive care unit (ICU).We calculated that we would need to recruit 70 patients, estimating a prevalence of 50% of cases with typical CT pattern (see Supplementary Material).A query of the electronic medical record system, roughly twice a week based on scanner availability, identified patients who had either confirmed or suspected COVID-19 infection (i.e., "PUI" status and awaiting further testing) and who had completed or were planned to complete a chest CT within 24 hours of the POCUS scan.Patients with a history of interstitial lung disease (ILD) were excluded, given these patients are relatively rare, typically know their underlying diagnosis, and at baseline will have quite abnormal lung POCUS scans.Otherwise, we attempted to recruit all consecutive patients.The protocol was approved by the local Partners Healthcare Institutional Review Board.

Selection of participants, interventions, and measurements
After obtaining assent from the treating team and verbal consent from the patient or proxy, a research physician scanned 12 lung zones (Figure 2), similar to prior protocols [24].The patient wore a surgical mask, and the scanning physician wore hospital-recommended personal protective equipment (N-95 respirator, eye protection, gown, and gloves; Appendix Figure S1).The probe was held longitudinally and perpendicular to the ribs to obtain the "bat sign" view [25].Scanning physicians attempted to capture at least two intercostal spaces for each zone (and at least three intercostal spaces for each of the four posterior zones).For posterior lung zones, patients either sat upright or lay in lateral decubitus position.Scanning was completed using a handheld Butterfly iQ ultrasound machine (Butterfly Network, Inc., Guilford, CT) connected to an iPhone (Apple Inc., Cupertino, CA).The lung preset was used for all zones; a second clip was obtained using the abdominal preset for zones R4 and L4 (to investigate for pleural effusion).Single-use gel packets were used for ultrasound gel, to avoid cross-contamination.After scanning, machines were sanitized with hospitalapproved disinfectant wipes.Scanning physicians included four internal medicine and two emergency   For both CT and CXR, the mode of the three interpretations was considered the consensus.For cases in which the third reader assigned a grade different from the primary two readers, the grade corresponding to the value closest to the median of the three observer grades was used as the consensus.

Analysis
For all imaging modalities, we calculated a Cohen's kappa to compare the interrater reliability between the first two readers (not including the "tiebreaker").An ordinal scale was used for CT, CXR, and POCUS interpretations.For POCUS scans, we recorded the duration of time spent scanning (from the start of the first clip to the final clip).
We assessed the accuracy of the 12-zone POCUS protocol to detect either (1) typical CT pattern or (2) positive NAT (by the time of discharge).The Soldati protocol assigned a score between zero to three for each of the lung zones (thus, for our 12 zones, the maximum score for a patient was 36) [21].
To calculate individual test characteristics (sensitivity, specificity, positive and negative predictive value) for our simplified POCUS score, we created two score cutoffs.These cutoffs allowed us to create low, intermediate, and high COVID-suspicion categories from the ordinal POCUS score (i.e., 0-36 points); these categories would mirror the RSNA CT and BSTI CXR categories.Data analysis was performed in Python 3.8.3(Python Software Foundation, Beaverton, OR) [27].
Receiver operating characteristic (ROC) curves were created to compare the performance of POCUS and CXR for our first outcome: a typical pattern for COVID-19 on CT.Additional ROC curves compared Soldati score, CXR, and CT against our second outcome: positive NAT.Areas under the curve (AUCs) were generated with the algorithm suggested by DeLong, DeLong, and Clark-Pearson [28].AUCs were compared with the test of equality of receiver operating areas in STATA IC 14.2 (Stata Corp LLC, College Station, TX).

Results
Ninety patients were initially screened, and 79 were scanned and included in the final analysis.The remaining eleven were excluded based on lack of consent, presence of exclusion criteria (namely ILD), or CT scan not completed (Appendix Figure S3).Patient characteristics at the time of scan are given in Table 1 (based on scoring system, below).Interrater reliability between the two readers for each of the imaging modalities was as follows: κ = 0.822 for CT scan, κ = 0.559 for CXR, κ = 0.704 for POCUS.The median time between POCUS exam and CT scan completion was 13 hours (95% confidence interval [CI]: 11.1, 16.8).For ten patients, >24 hours elapsed between scans, typically from delays in obtaining CT scan after initial order.The POCUS exam took a median of 10 (95% CI: 9.4; 10.8) minutes.There was no statistically significant difference in the POCUS scan time between NAT-positive and negative patients (p=0.845).
The ROC curves with corresponding AUCs for CXR, CT, and POCUS are given in Figure 3.The AUC for POCUS was numerically higher than the AUC for CXR for the outcome of typical pattern on CT scan, but this difference did not reach statistical significance (p = 0.1201).To mirror the consensus categories used for CT and CXR, we created two cutoffs for the POCUS protocol, to divide patients into low, intermediate, and high suspicion for COVID.We set low suspicion at 6 or fewer points; intermediate at >6 and <24 points, and high suspicion at Abbreviations: CT, computed tomography; NAT, nucleic acid test; NPV, negative predictive value; PPV, positive predictive practical method so that this protocol could be used to "rule-in" and "rule-out"; namely, to maximize sensitivity and NPV below the lower cutoff and maximize specificity and PPV above the higher cutoff--while also including a sizeable number of patients within each category.Operating characteristics are summarized in Table 2.
The entire protocol is delineated in Table 3.

Discussion
A 12-zone lung POCUS protocol adapted from Soldati et al. and using handheld devices had reasonable diagnostic performance to detect either a typical CT pattern or a positive NAT for COVID-19.For patients with a higher POCUS score (high risk of COVID-19), the specificity was approximately 90% for detecting either typical CT pattern or positive NAT.For patients above the lower cut-off point, the sensitivities and negative predictive values for typical CT pattern and for positive NAT were all quite high (88-100%).There was no statistically significant difference in the diagnostic accuracy among POCUS, CT, and CXR.Thus, arguably patients with a high POCUS score could be considered high risk for having COVID-19, and further imaging with a CT might be avoided.Similarly, one might extrapolate that patients below the low cut-off could be considered to have a low risk for COVID-19 pneumonia (and might be able to forgo a CT scan that might be ordered to check for typical signs of COVID-19 pneumonia).
The study results add to the literature that POCUS can assist in the diagnosis or risk stratification of PUIs for COVID-19 [29][30][31][32][33]. Namely, handheld lung POCUS may be a helpful adjunct to history and risk factors both to "rule in" or "rule out" COVID-19 pneumonia.Certainly, some of the more typical POCUS findings of COVID-19 can also be seen in other pulmonary conditions; but the same is true of typical CXR findings.The interrater reliability for readers for the POCUS score was overall good, notably higher than CXR (κ = 0.704 vs. κ = 0.559).
Handheld POCUS, as used in this study, is relatively inexpensive and easy to disinfect, and may be particularly helpful in resource-limited settings, where access to CT or NAT may be limited or results may be delayed.In addition, the vast majority of scans were completed by internists with extra training in POCUS: notably, internists represent the largest medical specialty group in the United States [34].Finally, lung POCUS is rapid.Our 12-zone POCUS mean scanning time was ten minutes.However, this is not an insignificant amount of exposure time with an infectious patient, and it is a limitation of using POCUS.We plan to perform a subgroup analysis to assess whether a protocol with fewer lung zones would still perform well.
Our study is subject to several limitations.First, this was a single-center study, using a convenience sample of inpatients based on scanning physician availability.unknown whether this scanning protocol would yield similar results for outpatients or ED patients.Fourth, our sample size had a relatively narrow BMI range.Fifth, the relatively high positivity rate (26.6%) could lead to a spectrum effect.However, this incidence is lower than some other studies, and our sample did include many patients with low-suspicion CXR and CT results.Sixth, although scanning physicians were blinded to CT results, they were not blinded to NAT status (positive COVID status is displayed prominently in our hospital electronic record); it is possible that sampling bias of the lung zones was introduced.However, COVID-positive and PUI patients had similar POCUS scan times.Seventh, there is certainly a degree of subjectivity in interpreting different POCUS findings as described in the protocol from Soldati et al. (or any lung protocol).Our interrater reliability for POCUS readers was excellent, but the subjectivity could make the protocol less generalizable worldwide.Finally, none of the patients with an initial negative NAT subsequently had a positive NAT.However, this was an inpatient study, and initial NAT is always performed in the Emergency Department in our institution.Incidence of positive NAT after initial negative NAT is very low [2].Thus, it would be extremely challenging to capture an appreciable number of NATpositive patients if we had limited our study to only those patients with an initial negative NAT.

Conclusions
Handheld lung POCUS could be a valuable tool in assessing the likelihood of COVID-

Figure 2 .
Figure 2. Lung Ultrasound Zones.Anterior lung zones (R1, R2, L1, L2) were scanned along the mid-clavicular line; lateral zones (R3, R3, L3, L4) were scanned along the mid-axillary line; the fourth rib separated the upper and lower anterior-lateral zones.Posterior lung zones (R5, R6, L5, L6) were scanned between the spine and the scapula; upper and lower posterior zones were separated by the inferior border of the scapula.Pictured is one of our physicians.
medicine physicians, all of whom had completed formal POCUS training, including a minimum of 25 lung scans reviewed by a POCUS expert.All scanners were blinded to CT results, though not to NAT results (positive results are prominently displayed in the chart), at the time of scanning.Patient demographics, vital signs, amount of supplemental oxygen, and lab values were recorded by non-scanning research staff.POCUS clips were reviewed by two EmergencyUltrasound Fellowship-trained physicians, who were blinded to CT and NAT results.Scans were scored along numerous criteria, including categories of B-lines, pleural line irregularity, and consolidation, selecting the most severe pathology in each of these 3 categories for each lung zone (see Appendix FigureS2).Examples are pictured (Figure1).Similarly, CTs and CXRs were reviewed by two board-certified radiologists specialized in thoracic imaging and blinded to clinical and POCUS information.Radiologists were blinded to CT images when reading CXRs, and vice versa.Radiologists gave each CT and CXR a COVID-19 suspicion grade, following consensus criteria (namely, RSNA criteria for CT scans and British Society of Thoracic Imaging (BSTI) criteria for CXRs): namely, (1) typical/high suspicion, (2) indeterminate, (3) atypical/low suspicion, or (4) negative[14,26].A third radiologist provided an interpretation in the case of discordant interpretations (as a tiebreaker).

Table 1 .
Patient characteristics at the time of POCUS scan

Table 2 .
Test characteristics of POCUS protocol Scan 12zones with probe perpendicular to ribs, recording at least 2 intercostal spaces per zone (and 3 intercostal spaces per posterior zone) -Scoring: select the highest-scoring pathology for each zone.Points for the 12 zones are added to create a total score for each patient (maximum score = 36).
[11]ction[11].Whether the test characteristics described herein apply to current variants, vaccinated patients, and generally lower prevalence and disease severity is questionable and should be explored in newer cohorts.Third, our study examined only inpatients (mostly from the ward, with a small number of ICU patients); it is -
19 pneumonia.A 12zone protocol following Soldati et al. and using handheld devices performed well to detect patients with either a high-suspicion CT pattern for COVID-19 or a positive NAT.In patients under investigation for COVID-19, particularly in low-resource settings, a lung POCUS exam with an inexpensive, handheld machine could potentially replace other chest imaging as a helpful adjunct for clinicians to decide whether a diagnosis of COVID-19 pneumonia should be further entertained or no.