High Fever, Wide Distribution of Viral Pneumonia, and Pleural Effusion are More Critical Findings at the First Visit in Predicting the Prognosis of COVID-19: A Single Center, retrospective, Propensity Score-Matched Case–Control Study

Figures & data

Figure 1 The annotation tool developed by Fujitsu Limited to annotate shadows manually for automatic calculation of volume.

Table 1 Baseline Characteristics in Unmatched and Matched Patients

	Unmatched Patients			Matched Patients
	Non-Severe Respiratory Failure Group	Severe Respiratory Failure Group	p-value	Non-Severe Respiratory Failure Group	Severe Respiratory Failure Group	ASD	p-value
	n=50	n=16		n=27	n=9
Age, years
Mean (SD)	47.9 (16.1)	62.7 (16.0)	0.0021	51.1 (18.0)	59.7 (18.0)		0.2263
Median [IQR]	48.5 [37–59]	61.5 [49–75.5]	0.0057	50 [37–62]	50 [49–70]		0.3218
Over 60, n (%) of patients	10 (20.0)	8 (50.0)	0.019	12 (44.4)	4 (44.4)	0	1.000
Sex, n (%) of patients
Male	27 (54.0)	11 (68.7)	0.2988	21 (77.8)	8 (88.9)	0.291	0.4658
Female	23 (46.0)	5 (31.3)		6 (22.2)	1 (11.1)
Body mass index, kg/m^2
Mean (SD)	23.9 (3.78)	27.3 (5.00)	0.0063	27.1 (2.35)	29.2 (5.31)		0.1070
Median [IQR]	23 [21.4–25.9]	28.1 [23.8–29.6]	0.0135	26.7 [25.8–27.6]	28.7 [28.4–30.0]		0.1112
Over 25, n (%) of patients	14 (28.0)	10 (62.5)	0.013	21 (77.8)	7 (77.8)	0	1.000
Comorbidity, n (%) of patients
Hypertension	11 (22.0)	3 (18.8)	0.7819	6 (22.2)	2 (22.2)	0	1.0000
Dyslipidemia	1 (2.0)	1 (6.3)	0.3880	0	0	-
Diabetes mellitus	4 (8.0)	6 (37.5)	0.0042	6 (22.2)	1 (11.1)	0.291	0.4658
Chronic renal failure	0	1 (6.3)	0.0749	0	0	-

Notes: A two-tailed p-value < 0.05 was considered to indicate statistical significance.

Abbreviations: ASD, absolute standardized difference; IQR, interquartile range; SD, standard deviation.

Table 2 Outcomes of Comparison Between Non-Severe and Severe Respiratory Failure Groups in Unmatched and Matched Patients

	Unmatched Patients			Matched Patients
	Non-Severe Respiratory Failure Group	Severe Respiratory Failure Group	p-value	Non-Severe Respiratory Failure Group	Severe Respiratory Failure Group	p-value
	n=50	n=16		n=27	n=9
Maximum body temperature until diagnosis, °C
Mean (SD)	38.5 (0.76)	38.9 (0.91)		38.4 (0.39)	39.3 (0.87)
Median [IQR]	38.5 [38–39]	39 [38.2–39.7]	0.0819	38.5 [38.2–38.5]	39.5 [39–39.9]	0.0043
CRP, mg/dL
Mean (SD)	2.12 (2.00)	8.19 (7.08)		3.28 (2.39)	8.69 (9.03)
Median [IQR]	1.57 [0.44–3.34]	6.29 [2.96–10.0]	<0.0001	2.16 [1.74–4.29]	5.84 [3.14–7.67]	0.0886
LDH, U/L
Mean (SD)	218 (60.4)	348 (171)		234 (46.6)	335 (158)
Median [IQR]	201 [179–254]	275 [218–460]	0.0009	224 [203–266]	269 [211–420]	0.0615
White blood cell count, /μL
Mean (SD)	4652 (1558)	5356 (2230)		5367 (1698)	5933 (2090)
Median [IQR]	4200 [3800–5200]	5100 [4000–6500]	0.1593	4400 [4000–6400]	5500 [4400–7000]	0.4395
Lymphocyte count, /μL
Mean (SD)	1172 (551.5)	940.8 (484.4)		1182 (458.7)	974.7 (484.7)
Median [IQR]	1067 [823.2–1310]	952.5 [594.4–1291]	0.2143	1170 [1100–1216]	959 [616–1270]	0.4752
Platelet count, ×10^4/μL
Mean (SD)	18.8 (5.18)	19.8 (7.26)		19.2 (4.47)	18.6 (5.78)
Median [IQR]	18.4 [14.9–20.9]	17.0 [14.8–24.3]	0.9582	19.7 [17.8–20.8]	16.7 [14.6–23.1]	0.6205
Uric acid, mg/dL
Mean (SD)	4.35 (1.55)	5.69 (2.03)		4.40 (1.64)	5.73 (1.29)
Median [IQR]	4.30 [3.1–5.2]	5.70 [3.8–6.6]	0.0297	4.07 [3.80–5.60]	6.30 [4.65–6.50]	0.0585
Fibrinogen, mg/dL
Mean (SD)	429 (133)	516 (132)		458 (163)	556 (161)
Median [IQR]	409 [327–488]	493 [443–570]	0.0217	443 [289–641]	570 [443–742]	0.1501
D-dimer, μg/mL
Mean (SD)	0.49 (0.59)	3.76 (12.1)		0.41 (0.41)	0.66 (0.71)
Median [IQR]	0.1 [0.1–0.8]	0.4 [0.1–1.6]	0.1386	0.10 [0.10–0.90]	0.4 [0.10–1.15]	0.4297
CT findings
Unilateral	10 (20.0)	1 (6.2)	0.1990	6 (22.2)	0	0.1213
Bilateral	40 (80.0)	15 (93.8)		21 (77.8)	9 (100)
Lobe, n
	3.68 (1.89)	5.19 (1.38)		4.11 (1.83)	5.33 (1.41)
	3.5 [2–6]	6 [4.5–6]	0.0038	5 [3–6]	6 [6–6]	0.0434
Ground glass opacity
To total lung field ratio, %
Mean (SD)	3.27 (3.39)	21.3 (23.7)	<0.0001	3.13 (2.34)	20.3 (21.1)
Median [IQR]	1.98 [0.81–4.54]	11.1 [3.38–30.3]	0.0002	2.89 [2.08–3.37]	20.7 [3.51–26.5]	0.0071
To upper lung field ratio, %
Mean (SD)	8.38 (11.0)	31.0 (18.1)		9.30 (11.3)	33.2 (12.0)
Median [IQR]	3.09 [0–16.23]	27.2 [16.5–41.9]	<0.0001	3.12 [2.55–8.94]	31.54 [27.1–36.1]	0.0001
To middle lung field ratio, %
Mean (SD)	36.8 (25.4)	33.0 (13.0)		44.3 (30.0)	40.5 (8.55)
Median [IQR]	35.7 [17.9–51.1]	35.9 [24.5–39.5]	0.7307	47.0 [17.3–55.1]	39.2 [36.3–43.5]	0.4100
To lower lung field ratio, %
Mean (SD)	54.8 (28.2)	36.0 (22.2)		46.4 (32.6)	26.3 (16.7)
Median [IQR]	57.3 [32.8–74.0]	36.9 [13.9–48.6]	0.0181	42.4 [21.7–73.8]	22.1 [12.4–37.2]	0.1112
Consolidation/total lung field, %
To total lung field ratio, %
Mean (SD)	0.65 (1.21)	1.38 (2.86)	0.1464	0.28 (0.37)	1.74 (3.63)
Median [IQR]	0.21 [0.05–0.89]	0.21 [0.07–1.46]	0.4906	0.15 [0.11–0.26]	0.19 [0.08–1.59]	0.2484
To upper lung field ratio, %
Mean (SD)	7.56 (18.4)	29.8 (33.7)		9.24 (15.3)	36.6 (32.8)
Median [IQR]	0 [0–3.17]	11.7 [7.18–52.6]	0.0001	0 [0–16.8]	27.2 [11.5–53.6]	0.0036
To middle lung field ratio, %
Mean (SD)	32.1 (31.4)	27.3 (22.8)		39.6 (31.1)	31.7 (24.8)
Median [IQR]	23.6 [0–56.0]	29.2 [4.9–36.7]	0.6286	38.2 [19.6–59.7]	29.8 [25.8–36.5]	0.5437
To lower lung field ratio, %
Mean (SD)	46.3 (35.7)	30.5 (30.5)		40.0 (30.7)	20.6 (26.5)
Median [IQR]	44.2 [11.4–81.0]	28.0 [0–59.8]	0.1312	40.3 [22.6–59.6]	5.28 [0–37.4]	0.1341
Subpleural curvilinear lines/fibrosis stripes	17 (34.0)	4 (25.0)	0.5011	9 (33.3)	2 (22.2)	0.5309
Traction bronchiectasis	0	1 (6.25)	0.0749	0	0
Vascular enlargement	1 (2.00)	1 (6.25)	0.3880	3 (11.1)	1 (11.1)	1.0000
Lymphadenopathy	22 (44.0)	12 (75.0)	0.0308	9 (33.3)	6 (66.7)	0.0790
Pleural thickening	10 (20.0)	1 (6.25)	0.1990	3 (11.1)	1 (11.1)	1.0000
Pleural effusion	0	2 (12.5)	0.0111	0	2 (22.2)	0.0117
Pericardial effusion	0	0		0	0

Notes: A two-tailed p-value < 0.05 was considered to indicate statistical significance.

Abbreviations: LDH, lactate dehydrogenase isozyme; CRP, C-reactive protein; IQR, interquartile range; SD, standard deviation.

Table 3 Association with Maximum Body Temperature Until Diagnosis, CT Findings (Lobe, Ground Glass Opacity-to-Total Lung Field Ratio, Ground Glass Opacity-to-Upper Lung Field Ratio, and Consolidation-to-Upper Lung Field Ratio) on Multivariate Linear Regressions in the Matched Cohort

	Matched Patients (n=36)
	Maximum body temperature until diagnosis
	β coefficient, 95% CI	p-value
Non-severe respiratory failure	0
Severe respiratory failure	0.93, 0.50–1.36	<0.0001
	CT finding
	Lobe
	β coefficient, 95% CI	p-value
Non-severe respiratory failure	0
Severe respiratory failure	1.38, 0.42–2.34	0.006
	Ground glass opacity to total lung field ratio
	β coefficient, 95% CI	p-value
Non-severe respiratory failure	0
Severe respiratory failure	16.7, 9.12–24.4	<0.0001
	Ground glass opacity to upper lung field ratio
	β coefficient, 95% CI	p-value
Non-severe respiratory failure	0
Severe respiratory failure	24.9, 17.7–32.2	<0.0001
	Consolidation to upper lung field ratio
	β coefficient, 95% CI	p-value
Non-severe respiratory failure	0
Severe respiratory failure	30.0, 16.0–43.9	<0.0001

Notes: Adjusted for patient characteristics used for the construction of propensity scores (age, sex, BMI, diabetes, hypertension, dyslipidemia, and chronic renal failure). A two-tailed P-value < 0.05 was considered to indicate statistical significance.

Abbreviation: CI, confidence interval.

Figure 2 ROC curve for GGO/total lung field and sensitivity/specificity as a function of probability cut-off. The optimal cutoff on logistic regression probabilities was the intersection point reflecting the optimal balance between sensitivity and specificity (GGOs/total lung field 14.6%).