Demographics and baseline characteristics of patients with COVID-19
A total of 225 patients diagnosed as COVID-19 were included in this study. According to the definition described in the methods, 225 patients were divided into 155 patients in the hyperinflammation group and 70 patients in the non-hyperinflammation group. As shown in Table 1, the median age of 225 patients was 67 (IQR 57.0–74.0) years, and the mean age of hyperinflammation group was 68 (IQR 58.0–75.0) years, which was similar with non-hyperinflammation group (median [IQR],65 [57.0–73.0] years). About half (50.7%) of patients were female. Male was more predominant in hyperinflammation group (93, 60%) than in non-hyperinflammation group (18, 25.7%). All patients are residents of Wuhan, but no patients had direct exposure history of Huanan wet markets or wildlife animals. Overall, 55 (24.4%) patients had a history of close contact with previously confirmed COVID-19. Few patients had a current or former smoking history, which might be inaccurate due to the large number of patients and endangered state of most patients on admission. Hypertension, diabetes, and cardiovascular disease were the most common comorbidities among hyperinflammation group (72 [46.5%], 39 [25.5%], and 24 [15.5%]) and non-hyperinflammation group (26 [37.1%], 10 [14.3%], and 8 [11.4%]). Only six patients with COPD were identified and all were in hyperinflammation group.
Table 1
Characteristics of patients with COVID-19
| Total(n = 225) | Hyperinflammation |
| Yes(n = 155) | No(n = 70) | P value |
Age, years-median (IQR) | 67.0(57.0–74.0) | 68.0(58.0–75.0) | 65.0(57.0–73.0) | 0.1211 |
Sex-No. (%) | | | | |
Female | 114(50.7) | 62(40.0) | 52(74.3) | < 0.0001 |
Male | 111(49.3) | 93(60.0) | 18(25.7) | .. |
Exposure history-No. (%) | 55(24.4) | 41(26.5) | 14(20.0) | 0.3202 |
Smoked-n/N (%) | | | | |
Never | 210/222(94.6) | 142/152(93.4) | 68/70(97.1) | 0.5111* |
Current | 5/222(2.3) | 4/152(2.6) | 1/70(1.4) | .. |
Former | 7/222(3.2) | 6/152(3.9) | 1/70(1.4) | .. |
Comorbidity-No. (%) | | | | |
Hypertension | 98(38.4) | 72(46.5) | 26(37.1) | 0.2453 |
Diabetes | 49(19.2) | 39(25.2) | 10(14.3) | 0.0813 |
Cardiovascular disease | 32(12.5) | 24(15.5) | 8(11.4) | 0.5374 |
Malignancy† | 7(2.7) | 7(4.5) | 0(0.0) | 0.1020 |
Chronic kidney disease | 7(2.7) | 6(3.9) | 1(1.4) | 0.4398 |
Chronic obstructive lung disease | 6(2.4) | 6(3.9) | 0(0.0) | 0.1850 |
Asthma | 2(0.8) | 0(0.0) | 2(2.9) | 0.0958 |
Interstitial lung disease | 0(0.0) | 0(0.0) | 0(0.0) | > 0.9999 |
Tuberculosis | 4(1.6) | 4(2.6) | 0(0.0) | 0.3129 |
Bronchiectasis | 1(0.4) | 0(0.0) | 1(1.4) | 0.3111 |
Chronic hepatitis B | 3(1.2) | 3(1.9) | 0(0.0) | 0.5540 |
Others | 64(25.1) | 46(29.7) | 18(25.7) | 0.6327 |
Respiratory rate breath, per min-median (IQR) | 22.0(20.0–26.0) | 22.0(20.0–28.0) | 20.0(20.0–25.0) | 0.0054 |
< 24, n/N (%) | 134/222(60.4) | 85/152(55.9) | 49(70.0) | 0.0119* |
24–30, n/N (%) | 56/222(25.2) | 38/152(25.0) | 18(25.7) | .. |
≥ 30, n/N (%) | 32/222(14.4) | 29/152(19.1) | 3/70(4.3) | .. |
Heart rate, beat per min-median (IQR) | 85.0(78.0-101.0) | 87.0(78.0-103.0) | 82.0(76.0–96.0) | 0.0899 |
> 100, n/N (%) | 58/224(25.9) | 45/154(29.2) | 13/70(18.6) | 0.1019 |
Percutaneous oxygen saturation, %- median (IQR) | 95.0(90.0–97.0) | 93.0(85.0–97.0) | 96.0(92.0–98.0) | 0.0002 |
≤ 93%, n/N (%) | 90/224(40.2) | 72/154(46.8) | 18/70(25.7) | 0.0032 |
Systolic pressure, mmHg-median (IQR) | 134.0(120.0-150.0) | 134.0(121.0-152.0) | 134.0(120.0-144.3) | 0.2615 |
Diastolic pressure, mmHg-median (IQR) | 80.0(73.0–89.0) | 80.0(71.0–90.0) | 81.0(74.0–86.0) | 0.8798 |
COVID-19 = coronavirus disease 2019, IQR = interquartile range. |
Data are median (IQR), n (%), or n/N (%), where N is the total number of patients with available data. Percentages may not total 100 because of rounding. P values were calculated by unpaired 2-sided Student’s t test, Mann-Whitney U test, χ² test, or Fisher’s exact test, as appropriate. *χ² test comparing all subcategories. |
† Included in this category is any type of cancer. |
Respiratory rates, heart rates, systolic pressure, and diastolic pressure were similar in the two groups, but patients with hyperinflammation more frequently developed tightness/dyspnea (respiratory rate > 30 breaths per minute (19 [19.1%]vs 3 [4.3%] ). Percutaneous oxygen saturation on admission were lower in patients with hyperinflammation (median [IQR], 93.0 [85.0–97.0], %) than those with non-hyperinflammation (median [IQR], 96.0 [92.0–98.0], %). Seventy two (46.8%) hyperinflammatory patients and 18 (25.7%) non-hyperinflammatory patients had percutaneous oxygen saturation of 93% or below.
Clinical symptoms of patients with COVID-19
Symptoms of the patients on admission are shown in Table 2. The most common symptoms at onset of illness were fever (193 [85.8%]), followed by chest tightness/dyspnea (177 [78.7%]) and cough (170 [75.6%]). Less common symptoms were chills, sore throat, fatigue, myalgia, sputum, gastrointestinal symptoms (nausea or vomiting, anorexia, abdominal pain, diarrhea), headache, dizziness and unconscious. Fever were the most prevalent symptoms in both hyperinflammation (136 [87.7%]) and non-hyperinflammation group (57 [81.4%]), and the proportions in the two groups were comparable. No difference was identified for the occurrence rates of most symptoms between the two groups, but chest tightness/dyspnea were much more common in patients with hyperinflammation (131 [84.5%]) than those with non-hyperinflammation (46 [65.7%]), and sputum were more commonly observed in patients with non-hyperinflammation (13 [18.6%] vs 12 [7.7%]).
Table 2
Clinical symptoms of patients with COVID-19
Clinical symptoms-No. (%) | Total (n = 225) | Hyperinflammation |
Yes(n = 155) | No(n = 70) | P value |
Fever | 193(85.8) | 136(87.7) | 57(81.4) | 0.2209 |
Chills | 21(9.3) | 13(8.4) | 8(11.4) | 0.4665 |
Chest tightness/dyspnea | 177(78.7) | 131(84.5) | 46(65.7) | 0.0025 |
Sore throat | 11(4.9) | 6(3.9) | 5(7.1) | 0.3239 |
Cough | 170(75.6) | 116(74.8) | 54(77.1) | 0.7410 |
Chest pain | 8(3.6) | 4(2.6) | 4(5.7) | 0.2591 |
Sputum | 25(11.1) | 12(7.7) | 13(18.6) | 0.0221 |
Fatigue | 93(41.3) | 67(43.2) | 26(37.1) | 0.4650 |
Myalgia | 39(17.3) | 24(15.5) | 15(21.4) | 0.3414 |
Gastrointestinal symptoms | | | | |
Nausea or vomiting | 18(8.0) | 11(7.1) | 7(10.0) | 0.4397 |
Anorexia | 21(9.3) | 17(11.0) | 4(5.7) | 0.3216 |
Diarrhea | 61(27.1) | 44(28.4) | 17(24.3) | 0.6274 |
Abdominal pain | 6(2.7) | 4(2.6) | 2(2.9) | > 0.9999 |
Headache | 19(8.4) | 12(7.7) | 7(10.0) | 0.6082 |
Unconscious | 6(2.7) | 6(3.9) | 0(0.0) | 0.1085 |
Dizziness | 8(3.6) | 8(5.2) | 0(0.0) | 0.0601 |
Data are n (%). Percentages may not total 100 because of rounding. P values were calculated by Fisher’s exact test. |
Laboratory and radiologic findings on admission in patients with COVID-19
Hyperinflammation group presented with significantly higher white blood cell count (median [IQR], 7.3 [5.1–10.5] vs 5.9 [4.4–7.8], x109/L) and neutrophil counts (median [IQR], 5.7 [4-9.4] vs 3.9 [2.6–5.8], x109/L) and clearly lower lymphocyte counts (median [IQR], 0.7 [0.5–0.9 vs 1.2 [0.9–1.5], x109/L) and platelet (median [IQR], 180 [132.8-248.3 vs 253.5 [193.8-327.5], x109/L) than non-hyperinflammation group (Table 3). Hemoglobin in hyperinflammation group was slightly lower than in non-hyperinflammation group (median [IQR], 129.0 [117.0-139.0] vs 124.0 [113.0-134.0], g/L). Red blood cell counts of the two groups were similar. Hyper-inflammation group also had significantly longer prothrombin time (median [IQR], 14.5 [13.8–15.7] vs 13.6 [13.2–14.4], seconds), longer activated partial thromboplastin time (median [IQR], 40.2 [36–44] vs 37.5 [34.8–41.5], seconds), and a significant higher level of D-dimer (median [IQR], 2.5 [1.1–21] vs 1.4 [0.6–2.3], µg/ml).
Table 3
Laboratory and radiologic findings on admission in patients with COVID-19
Findings (normal range) | Total(n = 225) | Hyperinflammation |
Yes(n = 155) | No(n = 70) | P value |
Blood routine test-median (IQR) | | | | |
White blood cell, x109/L (3.5–9.5) | 6.7(4.9–9.9) | 7.3(5.1–10.5) | 5.9(4.4–7.8) | 0.0009 |
Neutrophil, x109/L (1.8–6.3) | 5.3(3.7–8.4) | 5.7(4.0-9.4) | 3.9(2.6–5.8) | < 0.0001 |
Lymphocyte, x109/L (1.1–3.2) | 0.8(0.6–1.2) | 0.7(0.5–0.9) | 1.2(0.9–1.5) | < 0.0001 |
Red blood cell, x1012/L (3.8–5.1) | 4.1(3.7–4.5) | 4.2(3.8–4.6) | 4.0(3.6–4.3) | 0.1218 |
Haemoglobin, g/L (130–175) | 126.0(115.0-137.5) | 129.0(117.0-139.0) | 124.0(113.0-134.3) | 0.0303 |
Platelet, x109/L (125–350) | 209.5(148.0-280.5) | 180.0(132.8-248.3) | 253.5(193.8-327.5) | < 0.0001 |
Coagulation function-median (IQR) | | | | |
Prothrombin time, s (11.5–14.5) | 14.3(13.5–15.3) | 14.5(13.8–15.7) | 13.6(13.2–14.4) | < 0.0001 |
Activated partial thromboplastin time, s (29.0–42.0) | 38.9(35.8–43.2) | 40.2(36–44) | 37.5(34.8–41.5) | 0.0116 |
D-dimer, µg/mL (< 0.5) | 1.8(0.9–7.8) | 2.5(1.1–21.0) | 1.4(0.6–2.3) | < 0.0001 |
Biochemical test-median (IQR) | | | | |
Albumin, g/L (35.0–52.0) | 32.1(29.5–35.2) | 31.4(28.5–33.9) | 35(31.6–37.9) | < 0.0001 |
Globulin, g/L (20.0–35.0) | 35.3(32.2–39.0) | 35.7(32.6–39.3) | 34.4(31.1–37.7) | < 0.0001 |
Aspartate aminotransferase, U/L (≤ 40) | 33.0(22.0-48.5) | 37.0(26.5–56.5) | 24.0(18.0–33.0) | < 0.0001 |
Alanine aminotransferase, U/L (≤ 41) | 29.0(17.0–44.0) | 31.0(19.0–49.0) | 23.5(14.0-36.3) | 0.0088 |
Total-bilirubin, µmol/L (≤ 26) | 10.6(7.5–15.3) | 11.7(8.3–17.1) | 8.5(6.1–12.0) | < 0.0001 |
Direct-bilirubin, umol/L (≤ 8) | 4.7(3.3–7.3) | 5.5(3.9–8.8) | 3.5(2.6–4.9) | < 0.0001 |
Creatinine, µmol/L (59–104) | 76.0(61.0–93.0) | 82.0(66.0-95.3) | 62.5(54.0–80.0) | < 0.0001 |
Urea nitrogen, mmol/L (3.1-8.0) | 5.6(4.1–8.4) | 6.4(4.7–9.6) | 4.6(3.2–5.6) | < 0.0001 |
Positive urinary protein-n/N (%) | 96/190(50.5) | 85/122(69.7) | 11/68(16.2) | < 0.0001 |
Positive urinary occult blood-n/N (%) | 67/190(35.3) | 55/122(45.1) | 12/68(17.6) | 0.0001 |
Infection-related biomarkers- median (IQR) | | | | |
Procalcitonin, ng/mL (0.02–0.05) | 0.09(0.04–0.28) | 0.15(0.07–0.45) | 0.03(0.02–0.05) | < 0.0001 |
Erythrocyte sedimentation rate, mm/h (0–15) | 43.5(25.0-69.8) | 47(27-72.3) | 33.5(22-62.5) | 0.0810 |
Ferritin, µg/L (30–400) | 1093.0(548.0-1804.0) | 1222.0(730.5–2014.0) | 435.0(306.3-633.9) | < 0.0001 |
High sensitivity C-reactive protein, mg/L (< 1) | 47.5(16.5-112.1) | 75.5(38.3-140.2) | 8.75(2.4–27.8) | < 0.0001 |
> 10 mg/L, No. (%) | 182(80.9) | 149(96.1) | 33(47.1) | < 0.0001 |
Lactate dehydrogenase, U/L (135–225) | 379.0(264.0-504.0) | 442.0(305.0-601.0) | 264.0(225.0-368.3) | < 0.0001 |
Cytokines-median (IQR) or No. (%) | | | | |
Interleukin 1β ≥ 5 pg/mL | 21(9.3) | 18(11.6) | 3(4.3) | 0.0887 |
Interleukin 2 receptor, U/mL (223–710) | 778(470–1175) | 996(632–1341) | 487.5(364.5-696.8) | < 0.0001 |
≥ 710 U/L | 124(55.1) | 109(70.3) | 15(21.4) | < 0.0001 |
Interleukin 6, pg/mL (< 7) | 18.8(4.0-55.6) | 37.2(16.5–93.2) | 2.7(1.5–4.3) | < 0.0001 |
≥ 7 pg/mL | 150(66.7) | 145(93.5) | 5(7.1) | < 0.0001 |
Interleukin 8, pg/mL (< 62) | 12.6(6.1–26.6) | 20.6(10.7–35.6) | 5.2(5–10.0) | < 0.0001 |
≥ 62 pg/mL | 20(8.9) | 20(12.9) | 0(0.0) | 0.0006 |
Interleukin 10 ≥ 9.1 pg/mL | 65(28.9) | 63(40.6) | 2(2.9) | < 0.0001 |
Tumour necrosis factor α, pg/mL (< 8.1) | 8.3(5.7–11.5) | 9.9(7.7–13.5) | 5.3(4-6.7) | < 0.0001 |
≥ 8.1 pg/mL | 118(52.4) | 112(72.3) | 6(8.6) | < 0.0001 |
Myocardial enzymes-median (IQR) | | | | |
Creatine kinase, U/L (≤ 190) | 89.0(49.0-191.0) | 107.0(54.8-222.5) | 49.0(41.0-66.5) | 0.0001 |
N-terminal pro-brain natriuretic peptide, pg/mL (< 285) | 244.5(109.3-965.8) | 402.0(155.5–1597.0) | 179.0(75.0-467.0) | 0.0001 |
Hypersensitive cardiac troponin I, pg/mL (≤ 15.6) | 9.5(3.8–28.7) | 12.1(5.6–53.2) | 4.5(2.3–11.2) | < 0.0001 |
Myoglobin, ng/mL (≤ 106) | 69.2(41.4-151.2) | 87.0(51.6-175.1) | 42.7(29.2–75.9) | 0.0005 |
SARS-CoV-2 identified-median (IQR) | | | | |
IgM antibody, AU/ml (≤ 10) | 44.9(17.1-124.2) | 49.2(23.6-124.3) | 37.8(15.2–124) | 0.3894 |
IgG antibody, AU/ml (≤ 10) | 182.3(159.2-216.6) | 187.7(161.2-220.1) | 177.7(157.3-210.2) | 0.4156 |
RT-PCR assay (+)-n/N (%) | 173/213(81.2) | 119/143(83.2) | 54/70(77.1) | 0.3504 |
Pathogen antibody-n/N (%) | 55/107(51.4) | 42/81(51.9) | 13/26(50.0) | > 0.9999* |
Influenza A virus IgM (+) | 46/107(43.0) | 36/81(44.4) | 10/26(38.5) | .. |
Influenza B virus IgM (+) | 4/107(3.7) | 3/81(3.7) | 1/26(3.8) | .. |
Mycoplasma pneumoniae IgM (+) | 10/107(9.3) | 8/81(9.9) | 2/26(7.7) | .. |
Chlamydia pneumoniae IgM (+) | 4/107(3.7) | 3/81(3.7) | 1/26(3.8) | .. |
Adenovirus IgM (+) | 1/107(0.9) | 1/81(1.2) | 0/26(0.0) | .. |
Bilateral involvement on CT-n/N (%) | 175/217(80.6) | 118/147(80.3) | 57/70(81.4) | > 0.9999 |
COVID-19 = coronavirus disease 2019; IQR = interquartile range; CT = chest computed tomography; SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2; RT-PCR = real-time polymerase chain reaction; IgM = Immunoglobulin M; IgG = Immunoglobulin G. |
Data are median (IQR), n (%), or n/N (%), where N is the total number of patients with available data. Percentages may not total 100 because of rounding. P values were calculated by unpaired 2-sided Student’s t test, Mann-Whitney U test, χ² test, or Fisher’s exact test, as appropriate. *χ² test comparing all subcategories. |
Data were missing for erythrocyte sedimentation rate in 29 patients (12.9%), for creatine kinase in 98 patients (43.6%), for N-terminal pro-brain natriuretic peptide in 21 patients (9.3%), for hypersensitive cardiac troponin I in 14 patients (6.2%), for coagulation function in 8 patients (3.6%), for ferritin in 74 patients (32.9%). |
Data were available for myoglobin in 100 patients, for anti- SARS-CoV-2 antibody in 113 patients. |
+The antibody was positive. |
Compared with non-hyperinflammation group, hyperinflammation group were more likely to have multiple organ function damage including the liver, kidney and heart. Hyperinflammation group had significantly higher levels of aspartate aminotransferase (median [IQR], 37.0 [26.6–56.6] vs 24.0 [18.0–33.0], U/L), alanine aminotransferase (median [IQR], 31.0 [19.0–49.0] vs 23.5 [14.0-36.3], U/L), total-bilirubin (median [IQR], 11.7 [8.3–17.1] vs 8.5 [6.1–12.0], µmol/L), and direct-bilirubin (median [IQR], 5.5[3.9–8.8] vs 3.5 [2.6–4.9], µmol/L), and lower levels of albumin concentrations(median [IQR], 31.4[28.5–33.9] vs 35.0 [31.6–37.9], g/L, P < 0.0001). Hyperinflammation group also had evidence of higher levels of creatinine (median [IQR], 82.0 [66.0-95.3] vs 62.5 [54.0–80.0], µmol/L), urea nitrogen (median [IQR], 6.4 [4.7–9.6] vs 4.6 [3.2–5.6], mmol/L), higher proportions of positive urinary protein (85/122 [69.7%] vs 11/68 [16.2%]) and positive urinary occult blood (55/122 [45.1%] vs 12/68 [17.6%]). The levels of biomarkers indicating cardiac injury in patients with hyperinflammation were significantly higher, as hypersensitive cardiac troponin I (median [IQR], 12.1 [5.6–53.2] vs 4.5 [2.3–11.2], pg/mL) and N-terminal pro-brain natriuretic peptide (NT-proBNP) (median [IQR], 402.0 (155.5–1597.0] vs 179.0 [75.0-467.0], pg/mL). Of patients with available data, concentrations of creatine kinase (median [IQR], 107.0 [54.8-222.5] vs 49.0 [41.0-66.5], g/L) and myoglobin (median [IQR], 87.0 [51.6-175.1] vs 42.7 [29.2–75.9], ng/L) were significantly higher in patients with hyperinflammation than cases with non-hyperinflammation.
The infection-related biomarkers, including procalcitonin (median [IQR], 0.15 [0.07–0.45] vs 0.03 [0.02–0.05], ng/mL), ferritin (median [IQR], 1222.0 [730.5–2014.0] vs 435.0 [306.3-633.9], µg/L), lactate dehydrogenase (median [IQR], 442.0[305.0-601.0] vs 264.0 [225.0-368.3], U/L), and globulin (median [IQR], 35.7 [32.6–39.3] vs 34.4 [31.1–37.7], g/L), high-sensitivity C-reactive protein (median [IQR], 75.5[38.3-140.2] vs 8.75 [2.4–27.8], mg/L) were significantly higher in hyperinflammation group. 149 (96.1%) patients with hyperinflammation and 33 (47.1%) who with non-hyperinflammation had increased concentration of hsCRP (> 10 mg/L). But erythrocyte sedimentation rate did not differ between the 2 groups.
Assessment of serum cytokines on admission revealed that levels of IL-2 receptor (IL-2R), IL-6, IL-8, IL-10, and TNF-α were significantly higher in hyperinflammation group. The concentration of IL-1β was undetectable (< 5 pg/mL) in nearly all COVID-19 patients with either hyperinflammation or non-hyperinflammation. The abnormal IL-2R (≥ 710U/L, 109 (70.3%) vs 15 (21.4%)), IL-6 (≥ 7 pg/mL, 145 (94.3%) vs 5 (7.1%)), IL-8 (≥ 62 pg/mL, 20 (12.9%) vs 0 (0%)), IL-10 (≥ 9.1 pg/mL, 63 (40.6%) vs 2 (2.9%)), and TNF-α (≥ 8.1 pg/mL, 112 (72.3%) vs 6 (8.6%)) were more common in hyperinflammation group than in non- hyperinflammation group. Besides SARS-CoV-2, nine respiratory pathogens were also detected within some patients. About half of the patients infected other pathogens, Influenza A virus was the most common (46/107, 43%). There was no difference in the risk of co-infection with other pathogens between hyperinflammation group and non-hyperinflammation group. Of the 217 patients with chest CT scan on admission, the majority (175, 80.6%) had bilateral involvement, showing typical images that were bilateral multiple ground glass opacities or consolidation (Table 3).
Treatment and outcomes in patients with COVID-19
Hypoxemia was more difficult to correct in hyperinflammation group than non-hyperinflammation group (high-flow nasal cannula oxygen therapy (35 (22.6%) vs 4 (5.7%), non-invasive mechanical ventilation (66 (42.6%) vs 4 (5.7%), invasive mechanical ventilation (65 (30.2%) vs 3 (4.3%) (Table 4). Four patients in hyperinflammation group were treated with extracorporeal membrane oxygenation (ECMO). 85 (37.8%) patients admission to intensive care unit (ICU), of whom 81 (95.3%) were with hyperinflammation, with a median time from illness onset to ICU admission was 16.0 (IQR 11.0–22.0) days for hyperinflammation group and 23.0 (IQR 17.0–26.0) days for non-hyperinflammation group.
Table 4
Treatment and outcomes in patients with COVID-19
| Total(n = 225) | Hyperinflammation |
Yes(n = 155) | No(n = 70) | P value |
Respiratory support-No. (%) | | | | |
High-flow nasal cannula oxygen therapy | 39(17.3) | 35(22.6) | 4(5.7) | 0.0020 |
Non-invasive mechanical ventilation | 70(31.1) | 66(42.6) | 4(5.7) | < 0.0001 |
Invasive mechanical ventilation | 68(30.2) | 65(41.9) | 3(4.3) | < 0.0001 |
ECMO | 4(1.8) | 4(2.6) | 0(0.0) | 0.3129 |
Outcomes-median (IQR) | | | | |
ICU admission- No. (%) | 85(37.8) | 81(52.3) | 4(5.7) | < 0.0001 |
ICU length of stay, days ‡ | 9.0(5.0–20.0) | 10.0(5.0–21.0) | 7.0(2.0–17.0) | 0.2903 |
Onset of symptom to admission, days | 12.0(8.0–16.0) | 11.0(7.0–15.0) | 13.0(10.0–17.0) | 0.0160 |
Hospital length of stay, days ‡ | 27.0(18.0–37.0) | 25.0(15.0–37.0) | 29.0(24.0–37.0) | 0.0181 |
Time from illness onset to ICU admission, days | 16.0(12.0–23.0) | 16.0(11.0–22.0) | 23.0(17.0–26.0) | 0.1379 |
Time from illness onset to end-point events, days ¶ | 39.0(28.0–48.0) | 36.0(25.0–48.0) | 42.0(35.0–49.0) | 0.0066 |
Viral shedding | 173/213(81.2) | 119/143(83.2) | 54/70(77.1) | 0.3504 |
Duration of viral shedding from illness onset, days ‡ | 26.0(21.0–36.0) | 25.0(19.0–34.0) | 30.0(23.0–38.0) | 0.0048 |
Severity-No. (%) | | | | |
Severe | 119(52.9) | 99(63.9) | 20(28.6) | < 0.0001 |
Non-severe | 106(47.1) | 56(36.1) | 50(71.4) | .. |
Status at data cutoff -No. (%) | | | | |
Hospitalization | 13(5.8) | 12(7.7) | 1(1.4) | < 0.0001* |
Discharge | 139(61.8) | 74(47.7) | 65(92.9) | .. |
Death | 73(32.4) | 69(44.5) | 4(5.7) | .. |
Data are median (IQR), n (%), or n/N (%), where N is the total number of patients with available data. Percentages may not total 100 because of rounding. P values were calculated by unpaired 2-sided Student’s t test, Mann-Whitney U test, χ² test, or Fisher’s exact test, as appropriate. *χ² test comparing all subcategories. |
ICU = intensive care unit; ECMO = extracorporeal membrane oxygenation. |
¶The end-point events were discharge or death. |
‡Detectable until death. |
The median time from onset of symptoms to hospital admission was 11.0 (IQR 7.0–15.0) days, which tended to be shorter in hyperinflammation group compared with non-hyperinflammation (median [IQR], 13.0 [10.0–17.0], days). And the median time from illness onset to end-point event was shorter in hyperinflammation group (median [IQR], 36.0 [25.0–48.0] vs 42.0 [35.0–49.0], days). Median length of hospital stay was also shorter in hyperinflammation group than non-hyperinflammation group (median [IQR], 25.0 [15.0–37.0] vs 29.0 [24.0–37.0], days). For patients with hyperinflammation, the median duration of viral shedding was 25 days (IQR, 19.0–34.0) from illness onset,while it was longer in non-hyperinflammatory patients (median [IQR], 30.0 [23.0–38.0] ). The virus was continuously detectable until death in non-survivors of the two groups. Of all 225 patients, 119 (52.9%) were severe pneumonia based on the 2019 American Thoracic Society / Infectious Disease Society of America guidelines[34], and 69.3% (99/155) patients with hyperinflammation. Whereas 92.9% (65/70) patients with non-hyperinflammation were classified as non-severe pneumonia.
None of the 225 patients were lost to follow-up during the study. Up to 5 April 2020, a primary composite end-point event occurred in 212 patients (94.2%), including 32.4% who died, 61.8% who discharged from hospital. Among the 155 patients with hyperinflammation, a primary composite end-point event occurred in 133 patients (92.3%) and 44.5% (69/155) died. The majority of patients with non-hyperinflammation are discharged (65/70, 92.9%).
Mortality of severe/non-severe patients with/without hyperinflammation
Further subanalysis suggested that most (60/73, 82.2%) of the non-survivors were severe patients with hyperinflammation. Four of 99 cases with hyperinflammation and severe COVID-19 died within 10 days after the onset of the illness, and about 73.3% (44/60) of non-survivors in hyperinflammation and severe COVID-19 group died within 30 days (Table 5). Among 225 patients, 4.0% (2 of 50) in non-severe patients without hyperinflammation, 10.0% (2 of 20) in severe patients without hyperinflammation, 16.1% (9 of 56) in non-severe patients with hyperinflammation, and 60.6% (60 of 99) severe patients with hyperinflammation with died during hospitalization (Fig. 1). Of note, the mortality of severe patients with hyperinflammation (60/99, 60.6%) was significantly higher than without hyperinflammation (2/20, 10.0%). Non-severe patients with hyperinflammation even tended to have higher mortality (9/56, 16.1%) than those in severe cases without hyperinflammation (2/20, 10%).
Table 5
Number of deaths involved and mortality of patients with severe/non-severe COVID-19 and with/without hyperinflammation
Time since onset of symptoms to end-point events, days ¶ | 10 | 20 | 30 | 40 | 50 | 60 | 70 |
No. of death involved | | | | | | | |
Severe patients with hyperinflammation (n = 99) | 4 | 22 | 44 | 55 | 59 | 60 | 60 |
Non-severe patients with hyperinflammation (n = 56) | 0 | 2 | 5 | 7 | 8 | 8 | 9 |
Severe patients without hyperinflammation (n = 20) | 0 | 0 | 2 | 2 | 2 | 2 | 2 |
Non-severe patients without hyperinflammation (n = 50) | 0 | 0 | 2 | 2 | 2 | 2 | 2 |
Mortality, % | | | | | | | |
Severe patients with hyperinflammation (n = 99) | 4.0 | 22.2 | 44.4 | 55.6 | 59.6 | 60.6 | 60.6 |
Non-severe patients with hyperinflammation (n = 56) | 0.0 | 3.6 | 8.9 | 12.5 | 14.3 | 14.3 | 16.1 |
Severe patients without hyperinflammation (n = 20) | 0.0 | 0.0 | 10.0 | 10.0 | 10.0 | 10.0 | 10.0 |
Non-severe patients without hyperinflammation (n = 50) | 0.0 | 0.0 | 4.0 | 4.0 | 4.0 | 4.0 | 4.0 |
¶The end-point events were discharge or death |