Clinical or laboratory data of 2254 patients admitted to Group HM Hospitales due to COVID-19 infection during the study period, from March 1 to June 10, 2020, are shown. Clinical characteristics are summarized in Table 1 and laboratory findings in Tables 2 and 3. The median age of patients was 69 [57–80] and 59.5% were men. At the time of admission, all patients were assessed in the emergency room and a blood sample was taken for current analysis and subsequently all patients received standard approved treatment for COVID-19. Infection by SARS-CoV-2 was confirmed in one thousand nine hundred fifty four (93.8%) patients.
One hundred eighty five (8.29%) patients experienced severe acute respiratory failure and were admitted to the ICU and were compared with those who did not require ICU admission. Three hundred thirty one (14.7%) patients passed away, sixty one (33%) from the ICU group and two hundred seventy (13%) of non-ICU group (p < 0.0001).
At the time of hospital admission, clinical differences were observed between patients who were admitted to ICU and those who did not, including sex (74.6% vs 58.1% males, odds ratio [OR]: 2.11; 95% CI: 1.51-3; p < 0.0001), temperature above 38ºC (14.1% vs 6.3%, odds ratio [OR]: 2.77; 95% CI: 1.72–4.34; p < 0.0001) and SaO2 (SaO2 < 90% 22.7% vs 13.8%, odds ratio [OR]: 2.94; 95% CI: 1.92–4.48; p < 0.0001). Regarding presence of comorbidities, only hypertension was significantly more frequent among ICU-admitted (43.8% vs 35.8%, odds ratio [OR]: 1.41; 95% CI: 1.04–1.91; p = 0.026) (Table 1).
Table 1
Demographics and Clinical characteristics (% and median value (interquartile range)). ICU-admitted vs Non-ICU admitted. Abbreviations: ICU, Intensive Care Unit; HBP, High blood pressure; DM,
|
Total
(n = 2254)
|
ICU
(n = 185)
|
Non-ICU
(n = 2069)
|
P value
|
Univariable analysis
|
|
OR (95%CI)
|
P value
|
Demographics characteristics
|
Age (years)
|
69 (57–80)
|
68 (61–74)
|
70 (57–81)
|
0.057
|
0.99 (0.98–1.00)
|
0.18
|
Male (%)
|
59.5%
|
74.6%
|
58.1%
|
< 0.0001
|
2.11 (1.51–3.00)
|
< 0.0001
|
Comorbidities
|
HBP
|
36.4%
|
43.8%
|
35.8%
|
0.032
|
1.41 (1.04–1.91)
|
0.026
|
DM
|
18.1%
|
21.6%
|
17.7%
|
0.21
|
NA
|
NA
|
COPD
|
5.7%
|
4.3%
|
5.8%
|
0.52
|
NA
|
NA
|
CD
|
11.5%
|
11.9%
|
11.5%
|
0.95
|
NA
|
NA
|
Clinical Characteristics
|
Temperature >38ºC (%)
|
6.9%
|
14.1%
|
130 (6.3%)
|
< 0.0001
|
2.77 (1.72–4.34)
|
< 0.0001
|
Heart rate (bpm)
|
89 (78–101)
|
91 (84–101)
|
89 (78–101)
|
0.10
|
NA
|
NA
|
BP max (mm Hg)
|
131 (117–146)
|
130 (119–140)
|
131 (117–146)
|
0.36
|
NA
|
NA
|
BP min (mm Hg)
|
76 (67–84)
|
66–83)
|
76 (67–84)
|
0.56
|
NA
|
NA
|
Sat O2 (%) > 94%
|
50.4%
|
29.2%
|
52.2%
|
< 0.0001
|
NA
|
NA
|
Sat O2 (%) 90–94%
|
20.4%
|
21.6%
|
20.3%
|
< 0.0001
|
1.91 (1.24–2.91)
|
0.0029
|
Sat O2 (%) < 90%
|
14.6%
|
22.7%
|
13.8%
|
< 0.0001
|
2.94 (1.92–4.48)
|
< 0.0001
|
Exitus
|
14.7%
|
33%
|
13%
|
< 0.0001
|
NA
|
NA
|
In relation to blood test values, many significant differences were found (Tables 2 and 3). Assessing rates relating to different parameters of the blood count, patients requiring ICU admission had significantly higher rates at the time of hospital admission in the NLR (6.9 [4–11.7] vs 4.1 [2.6–7.6], p < 0.0001), PLR (2 [1.4–3.3] vs 1.9 [1.3–2.9], p = 0.023), NPR (3 [2.1–4.2] vs 2.3 [1.6–3.2], p < 0.0001) and SII (13 [6.5–25.7] vs 9 [4.9–17.5], p < 0.0001) than those who were not admitted in the ICU (Table 3).
Table 2
Laboratory findings at admission. Median value (interquartile range). *The variable is continuous, the OR is for each increment in a unit. ICU-admitted vs Non-ICU admitted. Abbreviations: : ICU, Intensive Care Unit; MCHC, mean corpuscular hemoglobin concentration; MCV, Mean corpuscular volume; MPV, Mean platelet volume; NLR, neutrophil-lymphocyte ratio; PLR, platelet-lymphocyte ratio; NPR, neutrophil-platelets ratio; SII, systemic immune-inflammation index; AST, Aspartate aminotransferase; ALT; lactate aminotransferase; LDH, lactate dehydrogenase.
|
Total
(n: 2254)
|
ICU
(n = 185)
|
Non-ICU
(n = 2069)
|
p value
|
Univariable analysis
|
|
OR (95%CI)
|
P value
|
Laboratory findings
|
|
|
|
|
|
|
White blood cells (10^9/L)
|
6.6 (5.0–8.9)
|
7.2 (5.4–10.3)
|
6.6 (5.0–8.8)
|
0.0034
|
1.04 (1.01–1.08)*
|
0.012
|
Red blood cells (10^12/L)
|
4.7 (4.2–5.0)
|
4.7 (4.2–5.0)
|
4.7 (4.2–5.0)
|
0.85
|
NA
|
NA
|
Neutrophils (10^9/L)
|
4.8 (3.3–6.9)
|
5.7 (4.0–8.6)
|
4.7 (3.3–6.7)
|
< 0.0001
|
1.07 (1.03–1.10)*
|
0.00014
|
Lymphocytes (10^9/L)
|
1.1 (0.8–1.5)
|
0.9 (0.6–1.2)
|
1.1 (0.8–1.5)
|
< 0.0001
|
0.59 (0.44–0.78)*
|
0.00036
|
Monocytes (10^9/L)
|
0.5 (0.3–0.7)
|
0.4 (0.3–0.6)
|
0.5 (0.3–0.7)
|
< 0.0001
|
0.27 (0.14–0.48)*
|
< 0.0001
|
Platelets (10^9/L)
|
205 (159–266)
|
186 (150–240)
|
208 (161–268)
|
0.0034
|
1.00 (1.00–1.00)
|
0.072
|
Hemoglobin (g/dL)
|
13.8 (12.5–14.9)
|
13.9 (12.5–15.1)
|
13.8 (12.5–14.9)
|
0.37
|
NA
|
NA
|
MCHC (g/dL)
|
33.7 (32.8–34.5)
|
34.0 (33.3–34.8)
|
33.6 (32.7–34.4)
|
< 0.0001
|
1.27 (1.14–1.42)*
|
< 0.0001
|
AST (U/L)
|
31.6 (22.3–49.2)
|
43.1 (29.0–70.7)
|
31.0 (21.8–47.5)
|
< 0.0001
|
1.01 (1.01–1.01)*
|
< 0.0001
|
ALT (U/L)
|
25.5 (16.1–42.4)
|
32.0 (20.3–49.0)
|
25.0 (15.7–41.4)
|
< 0.0001
|
1.00 (1.00–1.00)
|
0.095
|
Creatinine (mg/dL)
|
0.9 (0.7–1.1)
|
1.0 (0.8–1.2)
|
0.9 (0.7–1.1)
|
0.0042
|
1.10 (0.88–1.31)
|
0.30
|
LDH (U/L)
|
521 (397–677)
|
675 (532–931)
|
510 (390–655)
|
< 0.0001
|
1.00 (1.00–1.00)*
|
< 0.0001
|
C-reactive protein (mg/L)
|
64 (24–131)
|
117 (59–225)
|
61 (22–122)
|
< 0.0001
|
1.01 (1.00–1.01)*
|
< 0.0001
|
Urea (mg/dL)
|
34.7 (26.0–49.4)
|
36.3 (27.8–50.6)
|
34.5 (26.0–49.3)
|
0.18
|
NA
|
NA
|
Glucose (mg/dL)
|
114 (100–137)
|
125 (111–150)
|
113 (100–135)
|
< 0.0001
|
1.01 (1.00–1.01)*
|
< 0.0001
|
Partial thromboplastin time (s)
|
32 (30–35)
|
32 (30–34)
|
32 (30–35)
|
0.18
|
NA
|
NA
|
D-dimer (mg/L)
|
1 (0–1)
|
1 (1–2)
|
1 (0–1)
|
0.0032
|
1.03 (1.01–1.04)*
|
0.00030
|
Prothrombine time (s)
|
13.3 (12.3–14.5)
|
13.4 (12.6–14.7)
|
13.2 (12.3–14.5)
|
0.24
|
NA
|
NA
|
Table 3
Hemogram rates findings. Median value. (interquartile range). * The variable is continuous, the OR is for each increment in a unit. ICU-admitted vs Non-ICU admitted. The rate of change of the different inflammation rates was obtained with up to four consecutive blood cells measurements since hospital entry. The rate of change was defined as the slope of the linear fit of the relative rates versus time from hospital entry in days. A rate of change higher than 10% per day was considered as positive. Abbreviations: ICU, Intensive Care Unit; NLR, neutrophil-lymphocyte ratio; PLR, platelet-lymphocyte ratio; NPR, neutrophil-platelets ratio; SII, systemic immune-inflammation index.
|
Total
(n = 2254)
|
ICU
(n = 185)
|
Non-ICU
(n = 2069)
|
p value
|
Univariable analysis
|
|
OR (95%CI)
|
P value
|
Hemogram rates at admission
|
NLR
|
4.3 (2.7–8.0)
|
6.9 (4.0–11.7)
|
4.1 (2.6–7.6)
|
< 0.0001
|
1.03 (1.02–1.05)*
|
< 0.0001
|
PLR
|
1.9 (1.3–2.9)
|
2.0 (1.4–3.3)
|
1.9 (1.3–2.9)
|
0.023
|
1.13 (1.05–1.21)*
|
0.00057
|
NPR
|
2.3 (1.6–3.3)
|
3.0 (2.1–4.2)
|
2.3 (1.6–3.2)
|
< 0.0001
|
1.15 (1.09–1.22)*
|
< 0.0001
|
SII
|
9.2 (5.0–18.2)
|
13.0 (6.5–25.7)
|
9.0 (4.9–17.5)
|
< 0.0001
|
1.01 (1.01–1.02)*
|
0.00028
|
Positive rate of change (> 10 % · day^-1)
|
NLR
|
20.9%
|
37.8%
|
19.4%
|
< 0.0001
|
1.87 (1.33–2.65)
|
0.00036
|
PLR
|
27.2%
|
35.7%
|
26.5%
|
0.33
|
NA
|
NA
|
NPR
|
42.7%
|
60%
|
41.1%
|
0.0029
|
1.74 (1.26–2.42)
|
0.00087
|
SII
|
30%
|
44.9%
|
28.7%
|
0.003
|
1.55 (1.11–2.19)
|
0.011
|
Independent mortality prediction ability was shown for each hemogram-derived ratio (ROC curves are shown in Figure S2 and optimal cut-off values are shown in Table 4).
Table 4
Optimal cut-off values for the different immunoinflammatory ratios with their sensitivities and specificities and their corresponding 95% confidence interval.
Variable
|
Cut-off
|
Sensitivity
|
Specificity
|
NLR
|
4.93
|
0.68 (0.49–0.80)
|
0.58 (0.47–0.74
|
PLR
|
2.50
|
0.47 (0.19–0.85)
|
0.66 (0.27–0.89)
|
NPR
|
2.44
|
0.68 (0.54–0.77)
|
0.58 (0.54–0.71)
|
SII
|
12.26
|
0.55 (0.32–0.87)
|
0.64 (0.30–0.84)
|
As for the evolution of these rates on successive days, patients requiring ICU admission showed a significantly higher rate of ascent in the velocity of NLR (37.8% vs 19.4% odds ratio [OR]: 1.87; 95% CI: 1.33–2.65, p < 0.0001), NPR (60% vs 41.1% odds ratio [OR]: 1.74; 95% CI: 1.26–2.42, p = 0.0029) and SII (44.9% vs 28.7% odds ratio [OR]: 1.55; 95% CI: 1.11–2.19, p = 0.0032), but not in the velocity of PLR (35.7% vs 26.5%, p = 0.33) (Table 3).
The results of multivariable logistic regression models assessing the relation of the different hemogram rates and needing of ICU admission are shown in Table 5. Model A adjusted the hemogram rates OR for age, heart rate, temperature > 38ºC, systolic and diastolic blood pressure, NLR rate of change > 10 % per day, AST, D-Dimer and glucose. This adjustment made every inflammatory rate to lose its association with ICU admission but NPR (model A adjusted OR: 1.15, 95% CI: 1.05–1.25, p = 0.0018). NPR showed no dependence with SaO2 < 90% (model B adjusted OR: 1.14, 95% CI: 1.03–1.25, p = 0.0044) nor with LDH (model C adjusted OR: 1.14, 95% CI: 1.03–1.24, p = 0.0061), but did show an adjusted OR drop to 1.11 (95% CI: 1.03–1.24) with a border line significance with CRP (model D, p = 0.055) (Table 5).
Table 5
Multivariable adjusted models. Model A: Age, heart rate, temperature > 38ºC, systolic and dyastolic blood pressure, NLR rate of change > 10% per day, AST, D-dimer and glucose. Model B: Model A + Saturation O2. Model C: Model B + LDH. Model D: Model C + CRP. Abbreviations: NLR, neutrophil-lymphocyte ratio; PLR, platelet-lymphocyte ratio; NPR, neutrophil-platelets ratio; SII, systemic immune-inflammation index; BP, blood pressure; LDH, lactate dehydrogenase; CRP, C-reactive protein.
Model
|
NLR
|
PLR
|
NPR
|
SII
|
Unadjusted
|
OR (95% CI)
|
1.03 (1.02–1.05)
|
1.13 (1.05–1.21)
|
1.15 (1.09–1.22)
|
1.01 (1.01–1.02)
|
p value
|
< 0.0001
|
0.00057
|
< 0.0001
|
0.00028
|
Model A
|
OR (95% CI)
|
1.02 (0.993–1.05)
|
1.04 (0.903–1.17)
|
1.15 (1.05–1.25)
|
1.01 (0.995–1.02)
|
p value
|
0.11
|
0.58
|
0.0018
|
0.19
|
Model A + SatO2
|
OR (95% CI)
|
1.02 (0.985–1.05)
|
1.03 (0.888–1.18)
|
1.14 (1.03–1.25)
|
1.01 (0.993–1.02)
|
p value
|
0.24
|
0.65
|
0.0044
|
0.31
|
Model B + LDH
|
OR (95% CI)
|
1.02 (0.983–1.05)
|
1.03 (0.885–1.17)
|
1.14 (1.03–1.24)
|
1.01 (0.992–1.02)
|
p value
|
0.3
|
0.68
|
0.0061
|
0.37
|
Model C + CRP
|
OR (95% CI)
|
0.998 (0.961–1.03)
|
0.979 (0.831–1.13)
|
1.11 (0.986–1.22)
|
0.996 (0.98–1.01)
|
p value
|
0.93
|
0.78
|
0.055
|
0.64
|
Stratified analysis showed that increasing values of NPR significantly associates with ICU admission for age < 70 years (odds ratio [OR]: 1.34; 95% CI: 1.1.17–1.56, p < 0.001), sex male (odds ratio [OR]: 1.24; 95% CI: 1.10–1.41, p < 0.001), absence of hypertension (odds ratio [OR]: 1.22; 95% CI: 1.07–1.39, p = 0.0025), SatO2 > 90% (odds ratio [OR]: 1.23; 95% CI: 1.11–1.39, p < 0.001), LDH below median (< 677 U/L) (odds ratio [OR]: 1.2; 95% CI: 1.08–1.34, p < 0.001) and CPR below median (< 131 mg/L) (odds ratio [OR]: 1.16; 95% CI: 1.02–1.31, p = 0.0098), showing statistical significant interaction with age (p = 0.009) and SaO2 (p = 0.024) (Fig. 1A).
Higher values of NLR were significantly associated with ICU entry only for strata with SaO2 > 90% (odds ratio [OR]: 1.04; 95% CI: 1.00–1.08, p = 0.03), male sex (odds ratio [OR]: 1.04; 95% CI: 1.01–1.07, p = 0.015) and age < 70 years (odds ratio [OR]: 1.09; 95% CI: 1.04–1.14, p = < 0.001), this last one showing significant interaction (p = 0.06) (Figure S1B).
Higher values of SII were significantly associated with ICU entry for patients with CRP lower than 131 mg/L (median value) (odds ratio [OR]: 1.02; 95% CI: 1.00–1.04, p = 0.045) and for patients younger than 70 (odds ratio [OR]: 1.03; 95% CI: 1.01–1.05, p = 0.0019), with significant interaction found with age (p = 0.009) (Figure S1C).
PLR showed no significant association with ICU entry in the stratified analysis (Figure S1D).
When tested the dependency of the variables used for stratification by Chi-squared test we obtained that being older or younger than 70 years is related with sex (p < 0.0001), HBP (high blood pressure) (p < 0.0001), SaO2 < 90% (p < 0.0001), LDH < 677 U/L (p = 0.0004) and CRP < 131 mg/L (p = 0.0023).
Correlation analysis between the four hemogram ratios (Figure S3) shows that NLR is correlated with the other three independently from mortality (NLR vs PLR, ρ = 0.7, p < 0.001; NLR vs NPR, ρ = 0.669, p < 0.001; NLR vs SII, ρ = 0.894, p < 0.001). However, PLR is correlated with SII (ρ = 0.816, p < 0.001) but not with NPR (ρ = 0.013, p = 0.53). Finally, NPR and SII showed a significant but weak correlation (ρ = 0.424, p < 0.001) (Figure S3).
As expected, the hemogram ratios were correlated with some of the hemogram parameters like white blood cells (NLR: ρ = 0.536, p < 0.001; NPR, ρ = 0.622, p < 0.001; SII, ρ = 0.642, p < 0.001), neutrophils (NLR: ρ = 0.743, p < 0.001; NPR, ρ = 0.722, p < 0.001; SII, ρ = 0.792, p < 0.001), lymphocytes (NLR: ρ = -0.699, p < 0.001; PLR, ρ = -0.727, p < 0.001; SII, ρ = -0.502, p < 0.001) or platelets (SII, ρ = 0.51, p < 0.001). All the hemogram rates were significantly but weakly correlated with most of the different laboratory and demographic variables but NLR and CRP (ρ = 0.552, p < 0.001) which was the only case with a correlation higher than 0.5 (Figures S4A and S4B).