ORIGINAL RESEARCH

published: 15 June 2021

doi: 10.3389/fmed.2021.651556

Differential Diagnosis of COVID-19

Pneumonia From Influenza A (H1N1)
Pneumonia Using a Model Based on
Clinicoradiologic Features
Wei-Ya Shi 1† , Shao-Ping Hu 2† , Hao-Ling Zhang 3 , Tie-Fu Liu 4 , Su Zhou 5 , Yu-Hong Tang 6 ,
Xin-Lei Zhang 1 , Yu-Xin Shi 1 , Zhi-Yong Zhang 1 , Nian Xiong 2* and Fei Shan 1*
1
Edited by: Department of Radiology, Shanghai Public Health Clinical Center, Fudan University, Shanghai, China, 2 Department of
Reza Lashgari, Radiology, Wuhan Union Red Cross Hospital, Wuhan, China, 3 Department of Radiology, Zhongshan Hospital, Fudan
Institute for Research in Fundamental University, Shanghai, China, 4 Department of Scientific Research, Shanghai Public Health Clinical Center, Fudan University,
Sciences, Iran Shanghai, China, 5 Department of Interventional Radiology, Shanghai Public Health Clinical Center, Fudan University,
Shanghai, China, 6 Department of Research and Development, Winning Health Technology Group Co., Ltd., Shanghai, China
Reviewed by:
Zhichao Feng,
Central South University, China Objectives: Both coronavirus disease 2019 (COVID-19) pneumonia and influenza A
Marcelo Brandao,
State University of Campinas, Brazil
(H1N1) pneumonia are highly contagious diseases. We aimed to characterize initial
*Correspondence:
computed tomography (CT) and clinical features and to develop a model for differentiating
Fei Shan COVID-19 pneumonia from H1N1 pneumonia.
[email protected]
Nian Xiong Methods: In total, we enrolled 291 patients with COVID-19 pneumonia from January
[email protected] 20 to February 13, 2020, and 97 patients with H1N1 pneumonia from May 24, 2009,
† These authors have contributed to January 29, 2010 from two hospitals. Patients were randomly grouped into a primary
equally to this work cohort and a validation cohort using a seven-to-three ratio, and their clinicoradiologic
data on admission were compared. The clinicoradiologic features were optimized by
Specialty section:
This article was submitted to the least absolute shrinkage and selection operator (LASSO) logistic regression analysis
Pulmonary Medicine, to generate a model for differential diagnosis. Receiver operating characteristic (ROC)
a section of the journal
Frontiers in Medicine
curves were plotted for assessing the performance of the model in the primary and
validation cohorts.
Received: 10 January 2021
Accepted: 26 April 2021 Results: The COVID-19 pneumonia mainly presented a peripheral distribution
Published: 15 June 2021
pattern (262/291, 90.0%); in contrast, H1N1 pneumonia most commonly presented a
Citation:
Shi W-Y, Hu S-P, Zhang H-L, Liu T-F,
peribronchovascular distribution pattern (52/97, 53.6%). In LASSO logistic regression,
Zhou S, Tang Y-H, Zhang X-L, Shi Y-X, peripheral distribution patterns, older age, low-grade fever, and slightly elevated
Zhang Z-Y, Xiong N and Shan F (2021)
aspartate aminotransferase (AST) were associated with COVID-19 pneumonia, whereas,
Differential Diagnosis of COVID-19
Pneumonia From Influenza A (H1N1) a peribronchovascular distribution pattern, centrilobular nodule or tree-in-bud sign,
Pneumonia Using a Model Based on consolidation, bronchial wall thickening or bronchiectasis, younger age, hyperpyrexia,
Clinicoradiologic Features.
Front. Med. 8:651556.
and a higher level of AST were associated with H1N1 pneumonia. For the primary and
doi: 10.3389/fmed.2021.651556 validation cohorts, the LASSO model containing above eight clinicoradiologic features

Frontiers in Medicine | www.frontiersin.org 1 June 2021 | Volume 8 | Article 651556

Shi et al. Differentiation of COVID-19 From H1N1

yielded an area under curve (AUC) of 0.963 and 0.943, with sensitivity of 89.7 and 86.2%,
specificity of 89.7 and 89.7%, and accuracy of 89.7 and 87.1%, respectively.
Conclusions: Combination of distribution pattern and category of pulmonary opacity
on chest CT with clinical features facilitates the differentiation of COVID-19 pneumonia
from H1N1 pneumonia.

Keywords: coronavirus disease 2019, influenza A (H1N1), computed tomography, multivariate analysis, differential
diagnosis

INTRODUCTION features of the two diseases and further establish a model based
on clinicoradiologic features so as to provide some guidance for
Coronavirus disease 2019 (COVID-19), caused by the novel their early identification.
severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2,
previously known as 2019-nCoV), has become a global health
concern that threaten human life and public health security.
METHODS
As of 3 January, 2021, more than 83 million cases and more Study Population
than 1.8 million deaths have been reported worldwide according A search of the medical records in two hospitals’ information
to World Health Organization (WHO) statistics (1). When system was conducted, and 303 COVID-19 patients from January
assessing COVID-19, it is noteworthy that influenza viruses 20 to February 13, 2020, and 224 patients with H1N1 pneumonia
occur in the same season. Influenza A (H1N1) is the most from May 24, 2009, to January 29, 2010, were identified. All
common influenza, and it caused a worldwide pandemic in 2009– patients were diagnosed according to the diagnostic criteria of
2010 with more than 18,449 deaths (2) and has now become the National Health Commission of China (22, 23) and were
an annual seasonal influenza, leading to a large number of confirmed by real-time Polymerase Chain Reaction (RT-PCR)
hospitalizations and deaths (3, 4). Because of the differences in detection of virus nuclear acid. Our exclusion criteria were (a)
therapy, prognosis, and protective measure between COVID- patients without CT scans; (b) patients with normal CT imaging;
19 and H1N1, it is important for clinicians and radiologists to and (c) COVID-19 patients with positive influenza A nuclear acid
identify these two respiroviral infections. tests. Finally, 291 patients with COVID-19 pneumonia and 97
Both COVID-19 and H1N1 pneumonias share similar clinical patients with H1N1 pneumonia were enrolled. All patients were
manifestations, such as mild to moderate flulike syndromes, and randomly grouped in seven-to-three ratio into a primary cohort
are often cured by symptomatic treatments (5, 6). But a few (n = 272) and a validation cohort (n = 116), respectively, by using
patients develop severe or even lethal acute respiratory distress computer-generated random numbers. The flow chart of patient
syndrome (ARDS), particularly in patients with comorbidities selection, grouping, and disease subtypes is shown in Figure 1.
(7–11), and their laboratory exams often display lymphopenia,
abnormalities in liver function, and myocardial zymogram (7, 8, Ethics Statement
10–14). This work was approved by the Ethics Committee of Shanghai
Significantly, the typical findings on computed tomography Public Health Clinical Center, China, and Wuhan Union
(CT) seem different in both pneumonias: bilateral, multifocal Red Cross Hospital, China, and informed consent for this
ground-glass opacities (GGOs) with or without consolidations retrospective study was waived (YJ-2020-S035-01).
or intralobular lines, in a predominant peripheral distribution
usually present in COVID-19 pneumonia (15–17); nodules and Clinical Data Collection
bud signs (18, 19), bronchiectasis (18, 20), and pleural effusion The clinical data on admission were retrospectively collected.
(20) are common in H1N1 pneumonia but rare in COVID-19 Particular attention was paid to the demographics, comorbidities,
pneumonia (16). Unlike the predominant peripheral distribution coinfection, symptoms, and laboratory findings. The
in COVID-19 pneumonia, H1N1 pneumonia presents as a comorbidities or underlying medical conditions mainly
predominant peribronchovascular (18, 21) or peripheral (18, 21) included chronic pulmonary, cardiac, renal and hepatic diseases,
or mixed distribution (18, 19). The differences in CT findings diabetes, cerebrovascular disease, hyperlipemia, malignancy,
in category and the distribution patterns of pulmonary opacity and immunosuppression.
suggest their significance in terms of a differential diagnosis of
both pneumonias. Imaging Acquiring
Although it is easy to identify these typical lesions, CT All patients underwent CT examinations at full inspiration from
manifestations of COVID-19 and H1N1 pneumonia are very the thoracic inlet to the costophrenic angle level. CT scans
diverse. To date, knowledge regarding the comprehensive were performed with one of two scanners (Hitachi Scenaria 64,
identification of both pneumonias still remains limited Hitachi Medical Systems, Tokyo, Japan; or Siemens Sensation
and cannot meet the urgent clinical needs. Therefore, this 16, Siemens Medical Systems, Forchheim, Germany) using
retrospective study aimed to assess initial CT and clinical automatic exposure control with the following parameters: tube

Frontiers in Medicine | www.frontiersin.org 2 June 2021 | Volume 8 | Article 651556

Shi et al. Differentiation of COVID-19 From H1N1

FIGURE 1 | Flow chart shows patient selection, grouping and disease subtypes. COVID-19, coronavirus disease 2019; RT-PCR, real-time Polymerase Chain Reaction.

voltage, 120 or 140 kV; tube current, 150–250 mA; detector Fleischner Society (24). The images were analyzed independently
width, 64 × 0.625 mm or 16 × 0.75 mm; pitch, 1.57 or 1; rotation by two radiologists (Weiya Shi and Fei Shan, with 12 and 19
time, 0.35 or 0.5 s; field of view (FOV), 350 mm; and matrix, 512 years of experience in chest radiology, respectively). In cases of
× 512. The reconstruction kernel used was lung smooth with disagreement, the results were determined by consensus.
a thickness of 1 mm and an interval of 0.8 mm. The following
windows were used: a mediastinal window with a window width Statistical Analysis
of 350 Hounsfield unit (Hu) and a window level of 40 Hu, and a The continuous data are expressed as the median and
lung window with a width of 1,200 Hu and a level of −600 Hu. interquartile range (IQR, 25th and 75th percentiles), because
a majority of them did not follow a normal distribution.
Image Interpretation The Fisher exact test and the chi-square test were used
CT images were assessed for the presence and distribution of for categorical variables, and the Wilcoxon rank sum test
pulmonary opacities, including pure GGOs, which manifested as was used for continuous variables when comparing the
a hazy opacity without obscuring the underlying vessels; GGO clinicoradiologic features of COVID-19 pneumonia with those of
with interlobular septal thickening or reticulation, which was H1N1 pneumonia.
defined as a crazy-paving sign; GGOs with consolidation, which In order to evaluate the interobserver agreement (IA) between
was defined as an area of opacification obscuring the underlying the two radiologists, the Cohen’s Kappa test was used for
vessels in GGO; consolidation; centrilobular nodule or tree-in- categorical variables and the intraclass correlation coefficient
bud sign, which was regarded present when centrilobular nodules (ICC) for continuous variables. The kappa coefficient (k) between
or nodular branching structures resembled a budding tree. 0.00 and 0.20; 0.21 and 0.40; 0.41 and 0.60; 0.61 and 0.80; and
The distribution pattern of pulmonary opacities was 0.81 and 1.00 indicated slight, fair, moderate, substantial, and
assessed as being in a predominant peripheral (outer third almost perfect agreement, respectively. The ICC values between
of the lungs), peribronchovascular, both of peripheral and 0.00 and 0.25; 0.26 and 0.40; 0.41 and 0.60; 0.61 and 0.75;
peribronchovascular, or diffuse distribution, or lacking a 0.75 and 1.00, indicated poor, low, fair, good, and excellent
specific distribution (Figure 2). The laterality (unilateral and agreement, respectively.
bilateral) and predominant involved pulmonary lobes (upper, The clinicoradiologic characteristics found to be significant
middle/lingula, lower, or diffuse) were also assessed. in univariate analysis were inputted into the least absolute
Bronchial wall thickening or bronchiectasis, focal pulmonary shrinkage and selection operator (LASSO) logistic regression
fibrosis (including reticulation and liner opacity), pleural analysis to identify the optimal subset of clinicoradiologic
effusion, and mediastinal lymphadenopathy (>1 cm in short- features in order to develop a model for identification. Receiver
axis diameter) were noted. The number of pulmonary segments operating characteristic (ROC) curves were plotted for assessing
involved was counted. All the terms were defined according to the the performance of the model in the primary and validation

Frontiers in Medicine | www.frontiersin.org 3 June 2021 | Volume 8 | Article 651556

Shi et al. Differentiation of COVID-19 From H1N1

FIGURE 2 | The distribution patterns of pneumonia due to COVID-19 or H1N1. (A) A 74-year-old male with COVID-19 pneumonia presents the onset symptom of
fever (37.5◦ C), and the CT shows GGO with consolidation and crazy-paving sign mainly along subpleural lungs, namely, as a peripheral distribution pattern. (B) A
29-year-old male with H1N1 pneumonia presents the onset symptoms of fever (39.5◦ C), cough, and shortness of breath, and the CT shows consolidation and small
centrilobular nodules mainly along bronchovascular bundles, namely, as a peribronchovascular distribution pattern. (C) A 59-year-old female with H1N1 pneumonia
presents the onset symptoms of fever (38.5◦ C), cough, and shortness of breath, and the CT shows consolidation with GGO along bronchovascular bundles and
subpleural lungs, namely, as a distribution pattern of both peripheral and peribronchovascular. (D) A 71-year-old male with COVID-19 pneumonia presents the onset
symptoms of fever (38◦ C) and dyspnea, and CT shows diffuse GGO with consolidation and crazy-paving sign in both lungs, namely, as a diffuse distribution pattern.
(E) A 50-year-old female with COVID-19 pneumonia presents the onset symptoms of fever (39◦ C) and cough, and CT shows very small non-rounded GGO located in
the middle lobe of the right lung lacking a specific distribution. COVID-19, coronavirus disease 2019; GGOs, ground-glass opacities.

Frontiers in Medicine | www.frontiersin.org 4 June 2021 | Volume 8 | Article 651556

Shi et al. Differentiation of COVID-19 From H1N1

TABLE 1 | Baseline clinical and laboratory characteristics of the patients with COVID-19 pneumonia vs. those with H1N1 pneumonia.

Characteristics COVID-19 (n = 291) H1N1 (n = 97) P-value

Age, years 51.0 (37.0–64.0) 31.0 (23.0–45.0) <0.001*

Sex
Male 156 (53.6%) 66 (68.0%) 0.013*
Female 135 (46.4%) 31 (32.0%)
Fever 239 (82.1%) 93 (95.9%) 0.001*
Body temperature, ◦ C 38.0 (37.5–38.5) 38.8 (38.0–39.5) <0.001*
Onset symptom to hospital admission, d 5.0 (3.0–8.0) 5.0 (2.0–7.0) 0.788
Comorbidities 95 (32.6%) 30 (30.9%) 0.754
Chronic pulmonary disease 7 (2.4%) 2 (2.1%) 0.601
Coinfection 13 (4.5%) 16 (16.5%) <0.001*
White blood cell count, ×109 /L; normal range: 3.5–9.5 4.7 (3.9–5.9) 4.3 (3.2–6.1) 0.082
Lymphocyte count, ×109 /L; normal range: 1.1–3.2 1.1 (0.8–1.5) 1.1 (0.8–1.6) 0.544
Alanine aminotransferase, U/L; normal range: 7–40 22.0 (15.0–33.5) 31.0 (21.0–56.0) <0.001*
AST, U/L; normal range: 13–35 24.0 (19.0–33.0) 41.0 (27.0–65.0) <0.001*
Lactate dehydrogenase, U/L; normal range: 120–250 232.0 (199.0–293.0) 263.5 (191.0–445.0) 0.032*
Total bilirubin, µmol/L; normal range: 3.4–20.5 8.1 (6.5–10.5) 9.0 (7.3–12.0) 0.011*
Albumin, g/L; normal range: 40.0–55.0 40.8 (37.9–43.1) 40.0 (35.2–45.4) 0.341
Blood urea nitrogen, mmol/L; normal range: 2.6–7.5 4.4 (3.6–5.5) 3.9 (3.1–5.1) 0.017*
Serum creatinine, µmol/L; normal range: Male, 53–106; Female, 44–97
Increased 21 (7.2%) 1 (1.0%) 0.023*
C-reactive protein, mg/L; normal range: <3.0 14.8 (5.1–33.0) 22.7 (11.2–51.6) 0.001*
CD4+ T counts, cell/µL; normal range: 410–1,590 417.0 (291.0–618.0) 365.0 (230.0–535.0) 0.010*

Continuous variables are presented as median (interquartile range), and dichotomous variables are presented as numbers of patients, with percentages in parentheses. *P < 0.05;
P-values are from Fisher’s exact test or Wilcoxon rank-sum test when comparing characteristics of COVID-19 pneumonia with those of H1N1 pneumonia; COVID-19, coronavirus
disease 2019; AST, aspartate aminotransferase.

cohorts. The cut-off values were defined based on the maximal 0.001, p < 0.001, p = 0.032, 0.011, 0.001, 0.010, 0.017, and
Youden index. 0.023, respectively).
Statistical analysis was performed with R version 3.6.1 (R
Project for Statistical Computing, Vienna, Austria). A two-tailed
CT Characteristics
α < 0.05 was considered statistically significant.
The IA between the two radiologists was almost perfect for all CT
findings except the tree-in-bud sign, for which it was substantial
RESULTS (k = 0.789).
The baseline CT characteristics of the 388 cases were shown in
Clinical and Laboratory Features Table 2. In terms of lesions’ distribution pattern, 90.0% (262/291)
The baseline clinical and laboratory characteristics of the 388 of COVID-19 pneumonia had a peripheral distribution pattern in
cases were shown in Table 1. Compared with H1N1 pneumonia contrast to only 20.6% (20/97) in H1N1 pneumonia (p < 0.001).
(97 patients), patients with COVID-19 pneumonia (291 patients) In the H1N1 pneumonia, a peribronchovascular distribution
were older (51.0 vs. 31.0 years, p < 0.001) and had a lower pattern (52/97, 53.6%) was the most common, though this is rare
proportion of men (156/291, 53.6% vs. 66/97, 68.0%, p = 0.013). in COVID-19 pneumonia (9/291, 3.1%, p < 0.001). The H1N1
They had lower fever incidence (239/291, 82.1% vs. 93/97, 95.9%, pneumonia was more likely to exist in a distribution pattern
p = 0.001) and lower body temperatures (38.0 vs. 38.8◦ C, p < of both peripheral and peribronchovascular (17/97, 17.5% vs.
0.001). H1N1 patients were more likely to have coinfection than 14/291, 4.8%, p < 0.001) or be lacking a specific distribution
COVID-19 patients (16/97, 16.5% vs. 13/291, 4.5%, p < 0.001). (6/97, 6.2% vs. 3/291, 1.0%, p = 0.003). The incidence of diffuse
Among them, all were bacterial infections except for 2 H1N1 distribution pattern was 1.0% (3/291) and 2.1% (2/97) in COVID-
cases with bacterial and fungal coinfections. 19 pneumonia and H1N1 pneumonia, respectively, and had no
The laboratory exams displayed lower serum levels of significant difference.
alanine aminotransferase, aspartate aminotransferase (AST), With respect to other CT characteristics, the COVID-
lactate dehydrogenase, total bilirubin, and c-reactive protein, 19 pneumonia was more likely to present a crazy-paving
higher level of CD4+ T counts and blood urea nitrogen, as well sign (206/291, 70.8% vs. 39/97, 40.2%, p < 0.001), GGO
as more frequency of increased serum creatinine in COVID- with consolidation (216/291, 74.2% vs. 49/97, 50.5%, p <
19 pneumonia in contrast to those in H1N1 pneumonia (p < 0.001), bilateral involvement (241/291, 82.8% vs. 71/97, 73.2%,

Frontiers in Medicine | www.frontiersin.org 5 June 2021 | Volume 8 | Article 651556

Shi et al. Differentiation of COVID-19 From H1N1

TABLE 2 | Baseline CT characteristics of the patients with COVID-19 pneumonia vs. those with H1N1 pneumonia.

Characteristics COVID-19 (n = 291) H1N1 (n = 97) P-value Agreement (k or ICC; 95% CI)

Main pulmonary opacities

pGGO 118 (40.5%) 37 (38.1%) 0.675 0.941 (0.903–0.973)
Crazy-paving sign 206 (70.8%) 39 (40.2%) <0.001* 0.884 (0.835–0.931)
GGO with consolidation 216 (74.2%) 49 (50.5%) <0.001* 0.915 (0.865–0.956)
Consolidation 91 (31.3%) 79 (81.4%) <0.001* 0.912 (0.870–0.948)
Centrilobular nodule or tree-in-bud sign 3 (1.0%) 43 (44.3%) <0.001* 0.789 (0.687–0.891)
NO. of pulmonary segments involved 9 (4–15) 12 (3–17) 0.183 0.998 (0.997–0.998)
Laterality
Bilateral involvement 241 (82.8%) 71 (73.2%) 0.039* 1.000 (1.000–1.000)
Involved pulmonary lobes 0.948 (0.916–0.972)
Upper 29 (10.0%) 10 (10.3%) 0.922
Middle/lingula 12 (4.1%) 12 (12.4%) 0.003*
Lower 172 (59.1%) 52 (53.6%) 0.342
Diffuse 78 (26.8%) 23 (23.7%) 0.548
Distribution pattern 0.877 (0.827–0.925)
Peripheral 262 (90.0%) 20 (20.6%) <0.001*
Peribronchovascular 9 (3.1%) 52 (53.6%) <0.001*
Peripheral + Peribronchovascular 14 (4.8%) 17 (17.5%) <0.001*
Diffuse 3 (1.0%) 2 (2.1%) 0.436
Lacking a specific distribution 3 (1.0%) 6 (6.2%) 0.003*
Other signs
Bronchial wall thickening or bronchiectasis 18 (6.2%) 28 (28.9%) <0.001* 0.805 (0.692–0.891)
Focal pulmonary fibrosis 103 (35.4%) 21 (21.6%) 0.012* 0.928 (0.886–0.965)
Pleural effusion 14 (4.8%) 27 (27.8%) <0.001* 1.000 (1.000–1.000)
Mediastinal lymphadenopathy 4 (1.4%) 1 (1.0%) 0.795 1.000 (1.000–1.000)

Continuous variables are presented as median (interquartile range), and dichotomous variables are presented as numbers of patients, with percentages in parentheses. *P < 0.05;
P-values are from Fisher’s exact test or Wilcoxon rank-sum test when comparing characteristics of COVID-19 pneumonia with those of H1N1 pneumonia. For categorical variables
and continuous variables, the interobserver agreement is assessed with the kappa coefficient (k) and the intraclass correlation coefficient (ICC), respectively. The k and ICC values are
reported with 95% confidence intervals (95% CI). COVID-19, coronavirus disease 2019; GGOs, ground-glass opacities.

p = 0.039), and focal pulmonary fibrosis (103/291, 35.4% vs. and a peribronchovascular distribution pattern were inversely
21/97, 21.6%, p = 0.012); H1N1 pneumonia, however, was more associated with COVID-19 pneumonia.
likely to present consolidation (79/97, 81.4% vs. 91/291, 31.3%, For the primary and validation cohorts, the LASSO model
p < 0.001), centrilobular nodule or tree-in-bud sign (43/97, containing above eight features yielded an area under curve
44.3% vs. 3/291, 1.0%, p < 0.001), predominant middle/lingula (AUC) of 0.963 (95% CI: 0.942–0.984) and 0.943 (95% CI: 0.900–
involvement (12/97, 12.4% vs. 12/291, 4.1%, p = 0.003), bronchial 0.986), with sensitivity of 89.7 and 86.2%, specificity of 89.7 and
wall thickening or bronchiectasis (28/97, 28.9% vs. 18/291, 6.2%, 89.7%, accuracy of 89.7 and 87.1%, positive predictive value of
p < 0.001), and pleural effusion (27/97, 27.8% vs. 14/291, 4.8%, 96.3 and 96.1%, and negative predictive value of 74.4 and 68.4%,
p < 0.001). respectively (Table 4 and Figure 4).

Performance of the Model for Differential DISCUSSION

Diagnosis
LASSO logistic regression analysis was applied to identify the RT-PCR detection of viral nuclear acid is widely used for
most valuable clinico-radiological features for differentiating diagnosis and conformation of COVID-19; however, its
COVID-19 pneumonia from H1N1 pneumonia when the sensitivity is largely affected by the disease phase, viral loading,
optimal value of log (λ) was −2.906 according to 10-fold cross- and sampling (25). Thus, the routine chest CT is a more
validation (Figure 3). The optimal features subset and their sensitive and rapid method (25), and the expert consensus of
coefficient values were shown in Table 3. LASSO-based feature the Radiological Society of North America (RSNA) has provided
selection revealed that age and peripheral distribution patterns guidance to report CT findings attributable to COVID-19
were positively associated with COVID-19 pneumonia, and pneumonia (16); however, the reported specificities of CT are
body temperature, AST, consolidation, centrilobular nodule or low so far, ranging from 25 to 53% (16). Because of the distinct
tree-in-bud sign, bronchial wall thickening or bronchiectasis, treatments and prognoses between COVID-19 and H1N1, we

Frontiers in Medicine | www.frontiersin.org 6 June 2021 | Volume 8 | Article 651556

Shi et al. Differentiation of COVID-19 From H1N1

FIGURE 3 | The selection of clinico-radiological features using LASSO logistic regression. (A) Optimal feature selection according to AUC value. (B) LASSO coefficient
profiles of the features. Vertical line is drawn at the selected value using 10-fold cross-validation, where optimal λ results in eight non-zero coefficients. LASSO, least
absolute shrinkage and selection operator; AUC, area under curve.

aimed to accurately identify these two diseases. In our study, (AUC, 0.943; sensitivity, 86.2%; specificity, 89.7%; accuracy,
CT manifestations of consolidation, centrilobular nodule or 87.1%) in the validation cohort, which provides guidance for
tree-in-bud sign, bronchial wall thickening or bronchiectasis, clinical diagnosis.
peripheral distribution pattern and peribronchovascular Several previous studies have found that COVID-19 typically
distribution pattern, together with the clinical features such presents GGO with or without consolidation in a peripheral
as age, body temperature and AST were identified the optimal distribution (15–17), which was endorsed by the Society of
features subset for differentiating COVID-19 pneumonia from Thoracic Radiology, the American College of Radiology, and
H1N1 pneumonia. Our model had high diagnostic efficiency RSNA (16) as guidance for COVID-19 diagnosis. Our study was

Frontiers in Medicine | www.frontiersin.org 7 June 2021 | Volume 8 | Article 651556

Shi et al. Differentiation of COVID-19 From H1N1

consistent with previous reports that the COVID-19 pneumonia is, coinfection (mainly bacterial infection) was more frequent in
mainly presented a peripheral distribution pattern (262/291, H1N1 pneumonia (16/97, 16.5%) than in COVID-19 pneumonia
90.0%); in contrast, H1N1 pneumonia most commonly presented (13/291, 4.5%).
a peribronchovascular distribution pattern (52/97, 53.6%). The Besides radiological features, clinical features such as age,
differences in CT imaging between these two pneumonias body temperature, and AST should be also taken into
may result from their distinct pathological changes in lungs. consideration for differentiation. Our cohort showed an older
The pathological findings of COVID-19 pneumonia include median age of the COVID-19 patients; however, it should be
exudative diffuse alveolar damage with alveolar and interstitial viewed cautiously, the statistics were based on the early stage
edema, alveolar fibrinous exudate with hyaline membranes, of the outbreak in Shanghai and Hubei province, China. With
and reactive pneumocytes (26), whereas H1N1 pneumonia, in the global spread, it has been found that the youth are also a
addition to diffuse alveolar damage, is usually accompanied susceptible population (30).
by necrotizing bronchiolitis and alveolar hemorrhage (27). Although the model had high diagnostic efficiency, it should
These pathologic lesion-dependent distribution patterns, which be noted that it had a negative predictive value of 68.4% in
are very conspicuous at the first glance of the images, are the validation cohort due to the misdiagnosis of 12 cases of
valuable indicators for differentiating COVID-19 pneumonia COVID-19 pneumonia as H1N1 pneumonia. The misdiagnosed
from H1N1 pneumonia. patients were relatively young (median, 38.5 years; IQR, 32.5–
The bronchiolitis causes central lobular nodules or tree-bud 58.5), with high fever (median, 39.0◦ C; IQR, 38.3–39.7) and
signs, and bronchial wall thickening or bronchiectasis; therefore, moderately elevated AST (median, 30.0 U/L; IQR, 22.5–48.0),
it is quite understandable that these two signs were more and 75.0% (9/12) of them presented consolidation and 58.3%
common in H1N1 pneumonia than in COVID-19 pneumonia (7/12) a non-peripheral distribution pattern (two cases with
(both p < 0.001), consisting with other studies (16, 18, 21). a peribronchovascular distribution pattern, four cases with a
We also found that consolidation was more frequent in H1N1 distribution pattern of both peripheral and peribronchovascular,
pneumonia (79/97, 81.4%) than in COVID-19 pneumonia and one case with a diffuse distribution pattern). Radiologists
(91/291, 31.3%), which is consistent with previous studies should pay more attention to these atypical clinicoradiologic
and is possibly associated with pathologic basis (8), disease manifestations of COVID-19 in young individuals so as to
progression or more severe disease (28), bacterial coinfection avoid misdiagnosis.
(29). The latter phenomenon was also found in our study, that However, our study had limitations. Firstly, there was an
imbalance between the sample sizes of COVID-19 pneumonia
and H1N1 pneumonia, and the proportion of severe and critically
ill H1N1 patients was greater than those of the COVID-
TABLE 3 | The coefficients of the elected features by LASSO logistic regression 19 cohort, which may have led to statistical disequilibrium.
analysis. Secondly, there is a bias in the laboratory tests because there
were several laboratory changes in COVID-19 patients compared
Feature selected Coefficient
to H1N1 patients, using only the tests common to both groups.
Age, years 0.013
Thirdly, patients comorbid with chronic pulmonary disease
Body temperature, ◦ C −0.523
were not excluded, which may lead to a bias in this study,
AST, U/L −2.998e-03
although the proportion of these patients was very low and
Consolidation (yes vs. no) −0.300
there was no statistically significant difference between both
Centrilobular nodule or tree-in-bud sign (yes vs. no) −0.984
groups. Fourthly, coinfection is common in patients of both
groups, especially in patients with N1H1. Due to the complex
Bronchial wall thickening or bronchiectasis (yes vs. no) −0.207
nature of the clinical situation, we believe that differential
Peripheral distribution pattern (yes vs. no) 1.505
diagnosis is also necessary for these patients to obtain effective
Peribronchovascular distribution pattern (yes vs. no) −0.712
subsequent treatment. Therefore, we did not exclude these
LASSO, least absolute shrinkage and selection operator; AST, aspartate patients when constructing our analysis model, but this may lead
aminotransferase. to a bias.

TABLE 4 | AUC values of the LASSO regression model for differentiating COVID-19 pneumonia from H1N1 pneumonia in the primary and validation cohorts.

LASSO regression model AUC value Sensitivity Specificity Accuracy PPV NPV
(95% CI) (%) (%) (%) (%) (%)

Primary cohort 0.963 89.7 89.7 89.7 96.3 74.4

(0.942–0.984)
Validation cohort 0.943 86.2 89.7 87.1 96.1 68.4
(0.900–0.986)

COVID-19, coronavirus disease 2019; AUC, area under curve; LASSO, least absolute shrinkage and selection operator; PPV, positive predictive value; NPV, negative predictive value;
95% CI, 95% confidence intervals.

Frontiers in Medicine | www.frontiersin.org 8 June 2021 | Volume 8 | Article 651556

Shi et al. Differentiation of COVID-19 From H1N1

FIGURE 4 | The performance of the LASSO logistic regression model for differentiating COVID-19 pneumonia from influenza A (H1N1) pneumonia. It was presented
by ROC curves in (A) the primary cohort (AUC, 0.963; 95% CI: 0.942–0.984) and (B) the validation cohort (AUC, 0.943; 95% CI: 0.900–0.986).

In conclusion, CT characteristics, including the distribution to participate in this study in accordance with the national
pattern and category of pulmonary opacity, combined legislation and the institutional requirements.
with clinical features, can help the early differentiation
of COVID-19 pneumonia from H1N1 pneumonia. CT AUTHOR CONTRIBUTIONS
manifestations of peripheral distribution patterns, together
with older age, low-grade fever, and slightly elevated AST, W-YS and FS participated in the study design and
indicate COVID-19 pneumonia; however, CT presentations conceptualization. W-YS, X-LZ, H-LZ, SZ, and S-PH participated
of peribronchovascular distribution patterns, centrilobular in the acquisition of data. W-YS, FS, and Y-HT participated in
nodule or tree-in-bud sign, consolidation, and bronchial analysis and interpretation of data. W-YS and Y-HT participated
wall thickening or bronchiectasis, together with younger in drafting of the manuscript and participated in the statistical
age, hyperpyrexia and higher level of AST, suggest analysis. FS and T-FL participated in critical revision of the
H1N1 pneumonia. manuscript for important intellectual content. Z-YZ, Y-XS, and
NX participated in administrative, technical, material support,
DATA AVAILABILITY STATEMENT and participated in study supervision. All authors contributed to
the article and approved the submitted version.
The original contributions presented in the study are included
in the article/supplementary material, further inquiries can be FUNDING
directed to the corresponding author/s.
This research was funded by the Novel Coronavirus Special
ETHICS STATEMENT Research Foundation of the Shanghai Municipal Science and
Technology Commission (Grant Number: 20441900600).
The studies involving human participants were reviewed and
approved by the Ethics Committee of Shanghai Public Health ACKNOWLEDGMENTS
Clinical Center, China, and Wuhan Union Red Cross Hospital,
China, and informed consent for this retrospective study was We thank all the doctors, nurses, disease control workers, and
waived (YJ-2020-S035-01). Written informed consent from researchers who have fought bravely and ceaseless against the
the participants’ legal guardian/next of kin was not required virus on the frontline during the COVID-19 epidemic.

Frontiers in Medicine | www.frontiersin.org 9 June 2021 | Volume 8 | Article 651556

Shi et al. Differentiation of COVID-19 From H1N1

REFERENCES and prognostic comparisons. J Comput Assist Tomogr. (2012) 36:285–

90. doi: 10.1097/RCT.0b013e31825588e6
1. WHO. Weekly Epidemiological Update - 5 January 2021. (2021). 19. Liu M, Zeng W, Wen Y, Zheng Y, Lv F, Xiao K. COVID-19 pneumonia: CT
Available online at: https://www.who.int/publications/m/item/weekly- findings of 122 patients and differentiation from influenza pneumonia. Eur
epidemiological-update---5-january-2021 (accessed January 7, 2021). Radiol. (2020) 30:5463–9. doi: 10.1007/s00330-020-06928-0
2. WHO. Pandemic (H1N1) 2009 - Update 112. (2010). Available online at: 20. Yin Z, Kang Z, Yang D, Ding S, Luo H, Xiao E. A comparison of clinical
https://www.who.int/csr/don/2010_08_06/en/ (accessed March 9, 2020). and chest CT findings in patients with influenza A (H1N1) virus infection
3. Dawood FS, Chung JR, Kim SS, Zimmerman RK, Nowalk MP, Jackson ML, and coronavirus disease (COVID-19). AJR Am J Roentgenol. (2020) 215:1065–
et al. Interim estimates of 2019-20 seasonal influenza vaccine effectiveness 71. doi: 10.2214/AJR.20.23214
- United States, February 2020. MMWR Morb Mortal Wkly Rep. (2020) 21. Li P, Su DJ, Zhang JF, Xia XD, Sui H, Zhao DH. Pneumonia in novel swine-
69:177–82. doi: 10.15585/mmwr.mm6907a1 origin influenza A (H1N1) virus infection: high-resolution CT findings. Eur J
4. Livingston E, Bucher K, Rekito A. Coronavirus disease 2019 and influenza Radiol. (2011) 80:e146–52. doi: 10.1016/j.ejrad.2010.05.029
2019-2020. JAMA. (2020) 323:1122. doi: 10.1001/jama.2020.2633 22. China National Health Commission. Diagnosis and Treatment of Pneumonitis
5. Adhikari SP, Meng S, Wu YJ, Mao YP, Ye RX, Wang QZ, et al. Epidemiology, Caused by New Coronavirus (Trial Version 7). Beijing: China National
causes, clinical manifestation and diagnosis, prevention and control of Health Commission (2020). Available online at: http://www.nhc.gov.cn/yzygj/
coronavirus disease (COVID-19) during the early outbreak period: a scoping s7653p/202003/46c9294a7dfe4cef80dc7f5912eb1989.shtml (accessed March
review. Infect Dis Poverty. (2020) 9:29. doi: 10.1186/s40249-020-00646-x 9, 2020).
6. Writing Committee of the WHOCoCAoPI, Bautista E, Chotpitayasunondh 23. China National Health Commission. Diagnosis and Treatment of Influenza
T, Gao Z, Harper SA, Shaw M, et al. Clinical aspects of pandemic 2009 A (H1N1) (Version 3). Beijing: China National Health Commission
influenza A (H1N1) virus infection. N Engl J Med. (2010) 362:1708– (2009). Available online at: http://www.nhc.gov.cn/yzygj/s3593g/201306/
19. doi: 10.1056/NEJMra1000449 5fc4b2d158d7475fa0da32e959f9a7ac.shtml (accessed March 9, 2020).
7. Wang D, Hu B, Hu C, Zhu F, Liu X, Zhang J, et al. Clinical characteristics of 24. Hansell DM, Bankier AA, MacMahon H, McLoud TC, Muller NL, Remy J.
138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Fleischner society: glossary of terms for thoracic imaging. Radiology. (2008)
Wuhan, China. JAMA. (2020) 323:1061–9. doi: 10.1001/jama.2020.1585 246:697–722. doi: 10.1148/radiol.2462070712
8. Tang X, Du RH, Wang R, Cao TZ, Guan LL, Yang CQ, et al. Comparison 25. Fang Y, Zhang H, Xie J, Lin M, Ying L, Pang P, et al. Sensitivity of chest
of hospitalized patients with ARDS caused by COVID-19 and H1N1. Chest. CT for COVID-19: comparison to RT-PCR. Radiology. (2020) 296:E115–
(2020) 158:195–205. doi: 10.1016/j.chest.2020.03.032 7. doi: 10.1148/radiol.2020200432
9. Jain S, Kamimoto L, Bramley AM, Schmitz AM, Benoit SR, Louie J, et al. 26. Xu Z, Shi L, Wang Y, Zhang J, Huang L, Zhang C, et al. Pathological
Hospitalized patients with 2009 H1N1 influenza in the United States, April- findings of COVID-19 associated with acute respiratory distress
June 2009. N Engl J Med. (2009) 361:1935–44. doi: 10.1056/NEJMoa0906695 syndrome. Lancet Respir Med. (2020) 8:420–2. doi: 10.1016/S2213-2600(20)
10. Huang C, Wang Y, Li X, Ren L, Zhao J, Hu Y, et al. Clinical features of 30076-X
patients infected with 2019 novel coronavirus in Wuhan, China. Lancet. 27. Nakajima N, Sato Y, Katano H, Hasegawa H, Kumasaka T, Hata
(2020) 395:497–506. doi: 10.1016/S0140-6736(20)30183-5 S, et al. Histopathological and immunohistochemical findings of 20
11. Chen N, Zhou M, Dong X, Qu J, Gong F, Han Y, et al. Epidemiological autopsy cases with 2009 H1N1 virus infection. Mod Pathol. (2012) 25:1–
and clinical characteristics of 99 cases of 2019 novel coronavirus 13. doi: 10.1038/modpathol.2011.125
pneumonia in Wuhan, China: a descriptive study. Lancet. (2020) 28. Marchiori E, Zanetti G, Fontes CA, Santos ML, Valiante PM, Mano CM,
395:507–13. doi: 10.1016/S0140-6736(20)30211-7 et al. Influenza A (H1N1) virus-associated pneumonia: high-resolution
12. Dominguez-Cherit G, Lapinsky SE, Macias AE, Pinto R, Espinosa-Perez L, de computed tomography-pathologic correlation. Eur J Radiol. (2011) 80:e500–
la Torre A, et al. Critically ill patients with 2009 influenza A(H1N1) in Mexico. 4. doi: 10.1016/j.ejrad.2010.10.003
JAMA. (2009) 302:1880–7. doi: 10.1001/jama.2009.1536 29. Coppola M, Porto A, De Santo D, De Fronzo S, Grassi R, Rotondo
13. Kumar A, Zarychanski R, Pinto R, Cook DJ, Marshall J, Lacroix J, et al. A. Influenza A virus: radiological and clinical findings of patients
Critically ill patients with 2009 influenza A(H1N1) infection in Canada. hospitalised for pandemic H1N1 influenza. Radiol Med. (2011) 116:706–
JAMA. (2009) 302:1872–9. doi: 10.1001/jama.2009.1496 19. doi: 10.1007/s11547-011-0622-0
14. Perez-Padilla R, de la Rosa-Zamboni D, Ponce de Leon S, Hernandez M, 30. Xinhua. WHO Europe Head Warns Young People as COVID-19 Infection Rates
Quinones-Falconi F, Bautista E, et al. Pneumonia and respiratory failure Soar. (2020). Available online at: http://en.people.cn/n3/2020/0821/c90000-
from swine-origin influenza A (H1N1) in Mexico. N Engl J Med. (2009) 9737171.html (accessed November 12, 2020).
361:680–9. doi: 10.1056/NEJMoa0904252
15. Shi H, Han X, Jiang N, Cao Y, Alwalid O, Gu J, et al. Radiological findings from Conflict of Interest: Y-HT was employed by Winning Health Technology Group
81 patients with COVID-19 pneumonia in Wuhan, China: a descriptive study. Co., Ltd., Shanghai, China.
Lancet Infect Dis. (2020) 20:425–34. doi: 10.1016/S1473-3099(20)30086-4
16. Simpson S, Kay FU, Abbara S, Bhalla S, Chung JH, Chung M, et al. The remaining authors declare that the research was conducted in the absence of
Radiological society of North America expert consensus statement on any commercial or financial relationships that could be construed as a potential
reporting chest CT findings related to COVID-19. Endorsed by the conflict of interest.
Society of Thoracic Radiology, the American College of Radiology,
and RSNA - secondary publication. J Thorac Imaging. (2020) 35:219– Copyright © 2021 Shi, Hu, Zhang, Liu, Zhou, Tang, Zhang, Shi, Zhang, Xiong and
27. doi: 10.1097/RTI.0000000000000524 Shan. This is an open-access article distributed under the terms of the Creative
17. Song F, Shi N, Shan F, Zhang Z, Shen J, Lu H, et al. Emerging 2019 Commons Attribution License (CC BY). The use, distribution or reproduction in
novel coronavirus (2019-nCoV) pneumonia. Radiology. (2020) 295:210– other forums is permitted, provided the original author(s) and the copyright owner(s)
7. doi: 10.1148/radiol.2020200274 are credited and that the original publication in this journal is cited, in accordance
18. Kang H, Lee KS, Jeong YJ, Lee HY, Kim KI, Nam KJ. Computed tomography with accepted academic practice. No use, distribution or reproduction is permitted
findings of influenza A (H1N1) pneumonia in adults: pattern analysis which does not comply with these terms.

Frontiers in Medicine | www.frontiersin.org 10 June 2021 | Volume 8 | Article 651556