Table 1 (continued)
Study
|
Country
|
Design
|
Setting
|
Recruitment periods a
|
Severity of illness (%)
|
Sample size
|
Female (%)
|
Age (years) c
|
Mortality (%)
|
Tang N et al. (b) [27]
|
China
|
SC, RS
|
Tongji Hospital
|
Jan 1-Feb 3
|
SI (100)
|
183
|
46.4
|
54.1 ± 16.2
|
11.5
|
Wang L et al. (a) [28]
|
China
|
SC, RS
|
Renmin Hospital of Wuhan University
|
Jan 31-Feb 5
|
SI and CI (61.9)
|
202
|
56.4
|
63 (51-70)
|
16.3
|
Wang L et al. (b) [29]
|
China
|
SC, RS
|
Renmin Hospital of Wuhan University
|
Jan 1- Feb 6
|
MI (29.5), SI (46.9), and CI (23.6)
|
339
|
51.0
|
69 (65-76)
|
19.2
|
Wu C et al. [30]
|
China
|
SC, RS
|
Wuhan Jinyintan Hospital
|
Dec 25 b -Jan 26
|
ICU (26.4)
|
201
|
36.3
|
51 (43-60)
|
21.9
|
Xie J et al. [31]
|
China
|
SC, RS
|
Union Hospital
|
Jan 28-Feb 28
|
MI (30.7), SI (52.1), and CI (17.1)
|
140
|
48.6
|
60 (47-68)
|
25.7
|
Xu B et al. [32]
|
China
|
SC, RS
|
Hubei Provincial Hospital of Traditional Chinese and Western Medicine
|
Dec 26 b-Mar 1
|
SI (24.1), CI (24.1)
|
187
|
44.9
|
62 (48.5-71)
|
15.0
|
Yang X et al. (a) [33]
|
China
|
SC, RC
|
Wuhan Jin Yintan hospital
|
Dec b -Feb 25
|
NA
|
1476
|
47.4
|
57 (47-67)
|
16.1
|
Yang X et al. (b) [34]
|
China
|
SC, RC
|
Wuhan Jin Yintan hospital
|
Dec 24 b -Jan 26
|
CI (100)
|
52
|
32.7
|
59.7 ± 13.3
|
61.5
|
Yao Q et al. [35]
|
China
|
SC, RC
|
Huanggang Central Hospital
|
Jan 30-Feb 11
|
SI (23.1)
|
108
|
60.2
|
52 (37-58)
|
11.1
|
Zhang J et al. [36]
|
China
|
SC, RS
|
Renmin Hospital of Wuhan University
|
Jan 11-Feb 6
|
MI (37.8), SI (47.5), and CI (14.2)
|
663
|
51.6
|
55.6 (44-69)
|
3.77
|
Zhang L et al. [37]
|
China
|
SC, RC
|
Wuhan Asia General Hospital
|
Jan 12-Mar 15
|
NA
|
343
|
50.7
|
62 (48-69)
|
3.8
|
Zhou F et al. [38]
|
China
|
MC, RC
|
Jinyintan Hospital and Wuhan Pulmonary Hospital
|
Dec 29 b -Jan 31
|
ICU (26.2), SI (35), and CI (28)
|
191
|
37.7
|
56 (46, 67)
|
28.3
|
Abbreviations: CI, critically ill; ICU, intensive care unit; MC, multicenter; MI, moderately ill, NA, not available; PS, prospective study; RS, retrospective study, SI, severely ill.
a The default year is 2020 unless otherwise noted.
b The year is 2019.
c Values are expressed as medians (interquartile ranges) or means ± standard deviations.
Table 2 Continuous variables and risk of mortality in patients with COVID-19
Variables
|
No. of patients (studies)
|
MD (95% Cl)
|
P value
|
I2 (%)
|
P value of Egger test
|
Age, years
|
1829 (8)
|
13.9 [8.95, 18.9]
|
< 0.001
|
89
|
0.20
|
Time from illness onset to hospital admission, days
|
1036 (6)
|
0.62 [-0.04, 1.27]
|
0.07
|
0
|
0.20
|
Respiratory rate, breaths per min
|
922 (5)
|
2.94 [1.10, 4.79]
|
0.002
|
70
|
0.84
|
Heart rate, beat per min
|
974 (6)
|
3.91 [1.09, 6.72]
|
0.006
|
9
|
0.28
|
Partial pressure of oxygen, mm Hg
|
358 (3)
|
-28.7 [-52.7, -4.65]
|
0.02
|
89
|
0.46
|
Partial pressure of carbon dioxide, mm Hg
|
358 (3)
|
-5.03 [-6.95, -3.12]
|
< 0.001
|
0
|
0.36
|
Peripheral oxygen saturation
|
491 (3)
|
-7.45 [-15.9, 0.96]
|
0.08
|
96
|
0.52
|
Partial pressure of oxygen: fraction of inspired oxygen
|
410 (4)
|
-122.6 [-198.2, -47.0]
|
0.001
|
92
|
0.12
|
Acute Physiology and Chronic Health Evaluation II score
|
77 (2)
|
3.86 [1.90, 5.81]
|
< 0.001
|
0
|
Not available
|
Sequential Organ Failure Assessment score
|
216 (2)
|
3.48 [3.06, 3.90]
|
< 0.001
|
0
|
Not available
|
Laboratory findings
|
|
|
|
|
|
White blood cell count, ×109/L
|
1178 (7)
|
3.53 [2.53, 4.54]
|
< 0.001
|
48
|
0.61
|
Neutrophil count, ×109/L
|
1178 (7)
|
3.86 [2.74, 4.99]
|
< 0.001
|
61
|
0.32
|
Lymphocyte count, ×109/L
|
1178 (7)
|
-0.34 [-0.47, -0.21]
|
< 0.001
|
78
|
0.18
|
Monocyte count, ×109/L
|
919 (4)
|
-0.06 [-0.11, -0.01]
|
0.02
|
62
|
0.95
|
Platelet count, ×109/L
|
1029 (5)
|
-34.0 [-57.1, -10.9]
|
0.004
|
59
|
0.44
|
Hemoglobin, g/L
|
829 (4)
|
0.54 [-2.49, 3.58]
|
0.73
|
8
|
0.32
|
Albumin, g/L
|
694 (5)
|
-4.05 [-6.51, -1.59]
|
0.001
|
87
|
0.51
|
Total bilirubin, μmol/L
|
639 (4)
|
4.01 [3.04, 4.99]
|
< 0.001
|
0
|
0.12
|
Alanine aminotransferase, U/L
|
1178 (7)
|
3.58 [-0.72, 7.87]
|
0.10
|
60
|
0.43
|
Aspartate aminotransferase, U/L
|
1153 (6)
|
14.8 [9.56, 20.1]
|
< 0.001
|
57
|
0.68
|
Creatinine, μmol/L
|
1178 (7)
|
18.0 [10.1, 25.9]
|
< 0.001
|
64
|
0.42
|
Blood urea nitrogen, mmol/L
|
419 (2)
|
3.94 [1.97, 5.90]
|
< 0.001
|
80
|
Not available
|
Urea, mmol/L
|
500 (2)
|
3.41 [1.81, 5.02]
|
< 0.001
|
71
|
Not available
|
Prothrombin time, seconds
|
1273 (5)
|
0.98 [0.74, 1.22]
|
< 0.001
|
0
|
0.68
|
Activated partial thromboplastin time, seconds
|
953 (4)
|
-0.47 [-3.31, 2.37]
|
0.75
|
91
|
0.71
|
D-dimer, mg/L
|
1353 (7)
|
5.26 [3.58, 6.93]
|
< 0.001
|
31
|
0.92
|
C-reactive protein, mg/L
|
1178 (7)
|
55.9 [27.3, 84.5]
|
< 0.001
|
92
|
0.39
|
Procalcitonin, ng/mL
|
1029 (6)
|
0.27 [0.14, 0.40]
|
< 0.001
|
76
|
0.27
|
Ferritin, μg/L
|
435 (2)
|
912.1 [705.2, 1119.0]
|
< 0.001
|
25
|
Not available
|
Lactate dehydrogenase, U/L
|
829 (4)
|
225.6 [153.8, 297.3]
|
< 0.001
|
81
|
0.45
|
Creatine kinase, U/L
|
1029 (6)
|
59.3 [26.2, 92.5]
|
< 0.001
|
75
|
0.66
|
γ-glutamyl transpeptidase, U/L
|
453 (2)
|
4.91 [-16.5, 26.4]
|
0.65
|
92
|
Not available
|
Erythrocyte sedimentation rate, mm/h
|
490 (3)
|
9.01 [3.85, 14.2]
|
< 0.001
|
0
|
0.58
|
Creatine kinase-MB, U/L
|
670 (4)
|
3.11 [0.84, 5.37]
|
0.007
|
74
|
0.17
|
N-terminal pro-brain natriuretic peptide, pg/mL
|
476 (2)
|
865.7 [690.2, 1041.2]
|
< 0.001
|
0
|
Not available
|
Hypersensitive cardiac troponin I, ng/mL
|
804 (3)
|
68.6 [29.3, 107.9]
|
< 0.001
|
87
|
0.32
|
Myoglobin, ng/mL
|
381 (2)
|
170.2 [91.4, 249.1]
|
< 0.001
|
0
|
Not available
|
Cystatin C, mg/L
|
186 (2)
|
0.28 [0.15, 0.42]
|
< 0.001
|
0
|
Not available
|
Interleukin-6, pg/mL
|
974 (5)
|
47.8 [10.6, 85.0]
|
0.01
|
96
|
0.09
|
CD3+ cell count, /μL
|
306 (2)
|
-315.1 [-362.2, -268.0]
|
< 0.001
|
0
|
Not available
|
CD4+ cell count, /μL
|
645 (3)
|
-168.5 [-196.1, -140.8]
|
< 0.001
|
20
|
0.02
|
CD8+ cell count, /μL
|
645 (3)
|
-116.3 [-139.2, -93.5]
|
< 0.001
|
47
|
0.62
|
Compared with survivors, the mean age of the deceased patients was significantly higher (MD, 13.9; 95% CI, 8.95 to 18.9). Advanced age (age ≥60-70 years) was associated with a significant increase in the risk of mortality (RR, 2.31; 95% CI, 1.99 to 2.67). Male sex was also correlated with a higher death rate than female sex (RR, 1.30; 95% CI, 1.17 to 1.44).
Patients with any comorbidity had a 2.85-fold higher risk of mortality than patients without comorbidities (RR, 2.85; 95% CI, 1.47 to 5.51). The highest mortality risk was observed in patients with chronic kidney disease (RR, 8.37; 95% CI, 3.94 to 17.8), followed by cerebrovascular disease (RR, 7.66; 95% CI, 3.87 to 15.2), chronic respiratory disease (RR, 3.84; 95% CI, 1.81 to 8.16), cardiovascular disease (RR, 3.16; 95% CI, 2.19 to 4.56), diabetes mellitus (RR, 2.21; 95% CI, 1.37 to 3.56) and hypertension (RR, 2.11; 95% CI, 1.49 to 2.99). However, other comorbidities were not associated with survival after pooled analyses.
Symptoms of dyspnea (RR, 1.98; 95% CI, 1.70 to 2.30) were more common in the nonsurvivor group than in survivor group. No associations between the remaining clinical symptoms and mortality were observed. Antiviral agent and immunoglobulin therapies were not associated with mortality. However, patients who received glucocorticoids (RR, 1.79; 95% CI, 1.25 to 2.55) or antibiotics (RR, 1.20; 95% CI, 1.02 to 1.40) were more likely to die than those who did not.
The values of the following laboratory parameters were significantly higher in the deceased patients than in survivors: white blood cells (WBCs), neutrophils (NEUs), total bilirubin (TBIL), aspartate aminotransferase (AST), creatinine (Cr), blood urea nitrogen (BUN), urea, prothrombin time (PT), D-dimer, C-reactive protein (CRP), procalcitonin (PCT), ferritin, lactate dehydrogenase (LDH), creatine kinase (CK), the erythrocyte sedimentation rate (ESR), creatine kinase-MB (CK-MB), N-terminal pro-brain natriuretic peptide (NT-proBNP), hypersensitive cardiac troponin I (hs-cTnI), myoglobin, cystatin C, and interleukin-6 (IL-6). The lymphocyte (LYM), monocyte (MON), platelet (PLT), albumin (ALB), CD3+, CD4+, and CD8+ cell counts were significantly lower in the nonsurvivors than in the survivors. Moreover, the following abnormal laboratory parameters were more prevalent in the nonsurvivor group than in the survivor group: WBCs ≥10 × 109/L, LYMs <0.8 × 109/L, PLTs <125 × 109/L, alanine aminotransferase (ALT) >40 U/L, AST >40 U/L, Cr ≥133 μmol/L, PCT ≥0.5 ng/mL (or ≥0.1 ng/mL and ≥0.05 ng/mL) and D-dimer ≥1.0 μmol/L.
Patients suffering from any comorbidity had a significantly increased risk of mortality, with RRs ranging from 2.59 to 12.6. The highest mortality risk was associated with shock (RR, 12.6; 95% CI, 1.25 to 127.1), followed by acute infection (RR, 9.78; 95% CI, 2.05 to 46.7), acute kidney injury (AKI, RR, 9.64; 95% CI, 6.01 to 15.4), acute cardiac injury (RR, 8.22; 95% CI, 4.95 to 13.7), acute respiratory distress syndrome (ARDS, RR, 6.82; 95% CI, 2.56 to 18.2), arrhythmia (RR, 4.86; 95% CI, 1.24 to 19.0), heart failure (RR, 4.18; 95% CI, 2.37 to 7.36), acute liver injury (RR, 3.78; 95% CI, 1.18 to 12.1) and sepsis (RR, 2.59; 95% CI, 2.11 to 3.17).
Oxygen treatment was applied more often in nonsurvivors than in survivors (high flow nasal cannula, RR, 10.6; 95% CI, 5.97 to 18.8; noninvasive ventilation, RR, 5.12; 95% CI, 3.98 to 6.57; invasive ventilation, RR, 29.3; 95% CI, 21.5 to 39.9). Renal replacement therapy was also observed to be more prevalent in the deceased patients than in the surviving patients (RR, 53.5; 95% CI, 22.4 to 127.3). No correlation between the use of extracorporeal membrane oxygenation (ECMO) and mortality was observed.
A significant difference between nonsurvivors and survivors was also observed for the following variables: respiratory rate, heart rate, respiratory rate >30 (or 24) breaths per min, partial pressure of oxygen (PaO2), partial pressure of carbon dioxide (PaCO2), peripheral oxygen saturation (SpO2), ratio of partial pressure of oxygen to fraction of inspired oxygen (PaO2/FiO2), bilateral pneumonia, Acute Physiology and Chronic Health Evaluation II (APACHEII) score, and Sequential Organ Failure Assessment (SOFA) score.
Significant heterogeneity was observed in the analyses of 54 tested predictors. Sensitivity analyses did not change the conclusions about the most tested variables while the results were inconsistent in the following 14 tested predictors: preexisting malignancy, symptoms of anorexia, antiviral therapy, immunoglobulin therapy, CRP ≥10 mg/L, shock, acute liver injury, MON and PLT count, APTT, γ-glutamyl transpeptidase, PaO2, SpO2, and IL-6 (Supplementary Table S108-S109). No significant publication bias was observed for any risk factors except for two predictors, current smoking status and antibiotic therapy (Table 2 and Figure 2).
Qualitative analysis of adjusted estimates
Adjusted data regarding mortality due to COVID-19 infection were available in 16 studies [4,5,7,9,10,12,17-19,21,26,28,29,31,37,38] (Table 3). A multivariate Cox regression model was used in ten studies, and a multivariate logistic regression model was applied in six studies. Demographic characteristics highlighted as predictors for increased risk of mortality were advanced age [4,9,18,26,28,29,38], male sex [5,17] and presence of a comorbidity [4,5,9,12,18,21,29,31], such as cardio-cerebrovascular disease [4,9,29], chronic obstructive pulmonary disease (COPD) [12,29], diabetes [12], hypertension [12], and malignancy [12]. In two studies, the symptom of dyspnea was correlated with a greater risk of death [4,31].