Abstract
Background
We investigated the parameters of National Early Warning Score 2 (NEWS2) + lactate + D-dimer in predicting the intensive care unit (ICU) admission and in-hospital mortality in patients hospitalized with COVID-19.
Methods
Patients, who applied to the emergency department of a tertiary university hospital and were taken to the COVID-19 zone with suspected COVID-19 between March 2020 and June 2020, were retrospectively examined. In this study, 244 patients, who were hospitalized and had positive polymerase chain reaction test results, were included. NEWS2, lactate, and D-dimer levels of the patients were recorded. Patients were grouped by the states of in-hospital mortality and ICU admission.
Results
Of 244 patients who were included in the study, 122 (50%) were male, while their mean age was 53.76 ± 17.36 years. 28 (11.5%) patients were admitted to the ICU, while in-hospital mortality was seen in 14 (5.7%) patients. The levels of D-dimer, NEWS2, NEWS2 + lactate, NEWS2 + D-dimer, NEWS2 + lactate + D-dimer were statistically significantly higher in patients with in-hospital mortality and admitted to ICU ( p < 0.05). The area under the curve (AUC) values of D-dimer, lactate, NEWS2, NEWS2 + lactate, NEWS2 + D-dimer, NEWS2 + lactate + D-dimer in predicting ICU admission were as 0.745 (0.658–0.832), 0.589 (0.469–0.710), 0.760 (0.675–0.845), 0.774 (0.690–0.859), 0.776 (0.692–0.860), and 0.778 (0.694–0.862), respectively; while the AUC values of these parameters in predicting in-hospital mortality were found to be as 0.768 (0.671–0.865), 0.695 (0.563–0.827), 0.735 (0.634–0.836), 0.757 (0.647–0.867), 0.752 (0.656–0.848), and 0.764 (0.655–0.873), respectively.
Conclusions
Compared to using the NEWS2 value alone, a combination of NEWS2, lactate, and D-dimer was found to be more valuable in predicting in-hospital mortality and ICU admission.
Keywords: COVID-19 , hospital mortality , intensive care unit
Introduction
COVID-19, which was first reported in Wuhan, China, and has since spread worldwide, is a viral infectious disease affecting the respiratory tract. 1 The World Health Organization announced COVID-19 in the pandemic category in January 2020. 2 So far, about 36.2 million cases and 1.05 million deaths have been reported all around the world. 3 It has been reported that about one-fourth of the hospitalized cases were admitted to the intensive care unit (ICU) and had in-hospital mortality. 4 Because of the high risk of morbidity and mortality of COVID-19, difficulties have been experienced in the emergency management of COVID-19. There is a need for a scoring system that can be applied easily and quickly, which predicts ICU admission and in-hospital mortality in COVID-19 patients.
Hypoxia and hypercoagulopathy led by COVID-19 are thought to induce an increase in D-dimer and lactate levels. 5 Besides, a high lactate level is an early biomarker of tissue hypoxia. 6 There are studies in the literature reporting that lactate and D-dimer levels increased in patients who were admitted to ICU and who resulted in mortality with COVID-19. 7 , 8
National Early Warning Score (NEWS) is a scoring system developed to predict patients’ critical care needs and mortality. 9 NEWS2 is a scoring system prepared as an updated version of NEWS with minor changes. There are studies in the literature that NEWS2 predicts mortality and ICU admission of COVID-19. 10 , 11 The combination of NEWS, lactate, and D-dimer had a more powerful predictor of mortality compared to separate use in general emergency department patients and medical patients. 12 , 13 Therefore, we investigated the power of NEWS2, NEWS2 + lactate, NEWS2 + D-dimer, and NEWS2 + lactate + D-dimer in predicting ICU admission and in-hospital mortality in patients who applied to the emergency department and were hospitalized because of COVID-19.
Methods
Ethics committee approval for this single-center, retrospective, and observational study was obtained from the Necmettin Erbakan University Meram Medical Faculty Pharmaceutical and Non-Medical Device Studies Ethical Committee, decision number 2020/2710.
This study was conducted in a tertiary teaching hospital. Patients who were admitted to the emergency department, suspected of having COVID-19, hospitalized, and whose PCR test was positive were included in the study. Regardless of the number of tests submitted, any PCR results that appeared to be positive at least once were considered positive while other results as negative.
The following data of the patients, who were included in the study, were recorded from e-file of patients using the Hospital Information Management Systems program: NEWS2 value ( Table 1 ); lactate level; D-dimer level; his/her complaint (fever, cough, shortness of breath, and nasal discharge); comorbidity; information on ward/ICU admission; hospital outcomes (discharge, exitus in-hospital, discharged against medical advice (AMA), and referral); in-hospital mortality. These values were: NEWS2 + lactate value (the sum of the NEWS2 value and the lactate value); NEWS2 + D-dimer value (the sum of the NEWS2 value and the D-dimer value); and NEWS + lactate + D-dimer value (the sum of the NEWS2 value and the lactate and D-dimer values). Patients were grouped as a survivor group or a nonsurvivor group according to the in-hospital mortality and ward-ICU admission according to the emergency outcomes. The primary outcomes of the study are predicting in-hospital mortality and ICU admission.
Table 1 . NEWS2 .
|
Score |
3 |
2 |
1 |
0 |
1 |
2 |
3 |
|
Respiration rate (per minute) |
≤ 8 |
9–11 |
12–20 |
21–24 |
≥ 25 |
||
|
Hypercapnic respiratory failure (No) |
|||||||
|
SpO 2 (%) |
≤ 91 |
92–93 |
94–95 |
≥ 96 |
|||
|
Hypercapnic respiratory failure (Yes) |
|||||||
|
SpO 2 (%) |
≤ 83 |
84–85 |
86–87 |
88–92 ≥ 93 on air |
93–94 on oxygen |
95–96 on oxygen |
≥ 97 on oxygen |
|
Air or oxygen? |
Oxygen |
Air |
|||||
|
SBP (mmHg) |
≤ 90 |
91–100 |
101–110 |
111–219 |
≥ 220 |
||
|
Pulse (per minute) |
≤ 40 |
41–50 |
51–90 |
91–110 |
111–130 |
≥ 131 |
|
|
Consciousness |
Alert |
CVPU |
|||||
|
Temperature (ºC) |
≤ 35 |
35.1–36.0 |
36.1–38.0 |
38.1–39.0 |
≥ 39.1 |
The ICU admission decision of the patients was made according to the physician’s own experience and according to the statement published by the Ministry of Health 14 (respiration rate ≥ 30; signs of dyspnea and respiratory distress; cases with oxygen saturation below 90% despite nasal oxygen support of 5 L/min or more; cases with partial oxygen pressure below 70 mmHg despite nasal oxygen support of 5 L/min or more; PaO 2 /FiO 2 < 300; bilateral infiltrations or multi-lobar involvement on chest X-ray or tomography; hypotension (systolic blood pressure [SBP] < 90 mmHg, decrease in usual SBP > 40 mmHg, mean arterial pressure < 65 mmHg); skin perfusion disorder; organ dysfunction such as kidney function test, liver function test disorder, thrombocytopenia, confusion; the presence of immunosuppressive disease; the presence of uncontrolled comorbidity with multiple features; troponin elevation, arrhythmia).
Statistical analysis of the recorded data was made with the SPSS 20.0 (IBM Corp., Armonk, NY, USA) package program. Normality analysis of the data was made using histograms and the Kolmogorov–Smirnov test. Non-normally distributed quantitative data were expressed as median (25%–75% quartiles), while normally distributed quantitative data as mean ± standard deviation, and categorical variables as frequency (percentage). Differences between groups created by emergency outcomes and in-hospital mortality were investigated using the Mann–Whitney U -test for non-normally distributed quantitative variables and using Student’s t -test for normally distributed quantitative variables. We used the Bonferroni correction for multiple comparisons. Intragroup comparisons of categorical variables were made using the chi-square test. Receiver operating characteristic (ROC) analysis was performed to figure out the power of the levels of NEWS2, lactate, D-dimer, NEWS2 + lactate, NEWS2 + D-dimer, and NEWS2 + lactate + D-dimer for in-hospital mortality and ICU admission decision. The optimum cut-off levels of biochemical parameters were determined using Youden’s index (sensitivity + 1 − specificity). Sensitivity, specificity, positive predictive value, and negative predictive value of parameters were calculated for the optimum cut-off levels. The p -value < 0.05 was considered statistically significant.
Results
Of 658 patients with suspected COVID-19, who were hospitalized from emergency department COVID-19 zone; 414 patients were eliminated since they had negative PCR test results. In-hospital mortality was observed in 27 (6.5%) of these 414 patients. Of these 414 patients, 25 (6.0%) were admitted to ICU, and 6 (1.4%) died in the emergency department.
In the study, 244 patients were included, 122 (50%) were male, their mean age was 53.76 ± 17.36 years, and the median length of hospital stay was 6 (4–11) days. Median D-dimer value, median lactate value, median NEWS2 value, median NEWS2 + lactate value, median NEWS2 + D-dimer value, and median NEWS2 + lactate + D-dimer value were 0.14 (0.01–0.24) µg/mL, 1.60 (1.20–1.90) mEq/L, 2.00 (1.00–4.00), 3.95 (2.60–6.20), 2.39 (1.14–4.60), and 4.27 (2.81–6.54). In the medical history, 115 (47.1%) of the patients had at least one comorbidity while the most frequent comorbidity was hypertension with 65 (26.6%) patients. In the COVID-19 ward unit, 216 (88.5%) patients were hospitalized, and 28 (11.5%) were admitted to COVID-19 ICU. For the hospital outcome, 224 (91.8%) patients were discharged, 14 (5.7%) died, 1 (0.4%) requested to be discharged AMA, and 5 (2.0%) were referred to another hospital. In-hospital mortality was observed in 14 (5.7%) patients and was not observed in 230 (94.3%). Table 2 shows the detailed data of the cases.
Table 2 . Features of participants .
a There is at least one additional disease in his/her medical history.
|
Characteristic |
Value |
|
Number of persons, n (%) |
244 (100) |
|
Age |
53.76 ± 17.36 |
|
Gender |
|
|
Male, n (%) |
122 (50) |
|
Female, n (%) |
122 (50) |
|
Vital signs |
|
|
Fever (ºC) |
36.6 (36.2–37.0) |
|
Pulse (per minute) |
92.81 ± 15.50 |
|
SBP (mmHg) |
133.05 ± 21.40 |
|
DBP (mmHg) |
75.82 ± 11.67 |
|
MAP (mmHg) |
94.89 ± 13.02 |
|
Saturation (%) |
95 (93–97) |
|
Complaints |
|
|
Fever, n (%) |
165 (67.6) |
|
Cough, n (%) |
160 (65.6) |
|
Shortness of breath, n (%) |
93 (38.1) |
|
Nasal discharge, n (%) |
31 (12.7) |
|
Laboratory results |
|
|
D-dimer (µg/mL) |
0.14 (0.01–0.24) |
|
Lactate (mEq/L) |
1.6 (1.2–1.9) |
|
NEWS2 |
2 (1–4) |
|
NEWS2 + lactate |
3.95 (2.60–6.20) |
|
NEWS2 + D-dimer |
2.39 (1.14–4.60) |
|
NEWS2 + lactate + D-dimer |
4.27 (2.81–6.54) |
|
Medical history |
|
|
Comorbidity a , n (%) |
115 (47.1) |
|
Diabetes mellitus, n (%) |
58 (23.8) |
|
Hypertension, n (%) |
65 (26.6) |
|
Coronary artery disease, n (%) |
29 (11.9) |
|
Asthma–COPD, n (%) |
38 (15.6) |
|
CRF, n (%) |
4 (1.6) |
|
Malignancy, n (%) |
10 (4.1) |
|
Immune deficiency, n (%) |
12 (4.9) |
|
Length of hospital stay (day) |
6 (4–11) |
|
Emergency service outcome |
|
|
Ward unit admission, n (%) |
216 (88.5) |
|
ICU admission, n (%) |
28 (11.5) |
|
Hospital outcome |
|
|
Discharged, n (%) |
224 (91.8) |
|
Ex, n (%) |
14 (5.7) |
|
Discharged AMA, n (%) |
1 (0.4) |
|
Referral, n (%) |
5 (2.0) |
|
In-hospital mortality |
|
|
Survivor, n (%) |
230 (94.3) |
|
Nonsurvivor, n (%) |
14 (5.7) |
Table 2 . Features of participants (continued) .
|
Characteristic |
ICU admission (n = 28) |
Ward unit admission (n = 216) |
p a |
Survivor (n = 230) |
Nonsurvivor (n = 14) |
p b |
|
Age |
64.61 ± 15.10 |
52.35 ± 17.16 |
< 0.001 * |
52.84 ± 17.20 |
68.79 ± 12.71 |
0.001 * |
|
Fever (°C) |
36.65 (36.30–37.07) |
36.60 (36.20–37.00) |
0.555 |
36.60 (36.20–37.00) |
36.50 (36.30–37.02) |
0.969 |
|
Pulse (per minute) |
89.86 ± 19.07 |
93.19 ± 14.99 |
0.286 |
93.27 ± 14.89 |
85.14 ± 22.80 |
0.209 |
|
SBP (mmHg) |
128.64 ± 18.68 |
133.62 ± 21.70 |
0.248 |
133.77 ± 21.62 |
121.21 ± 12.97 |
0.004 * |
|
DBP (mmHg) |
73.25 ± 12.22 |
76.16 ± 11.58 |
0.216 |
76.15 ± 11.61 |
70.43 ± 11.74 |
0.075 |
|
MAP (mmHg) |
91.71 ± 13.03 |
95.31 ± 12.99 |
0.169 |
95.35 ± 13.00 |
87.35 ± 11.26 |
0.025 * |
|
Saturation (%) |
93.5 (89.0–96.0) |
95.0 (94.0–97.0) |
0.002 * |
95.0 (93.0–97.0) |
94.50 (89.75–96.25) |
0.219 |
|
D-dimer (µg/mL) |
0.28 (0.17–0.48) |
0.13 (0.08–0.22) |
< 0.001 * |
0.13 (0.09–0.23) |
0.28 (0.20–0.53) |
0.001 * |
|
Lactate (mEq/L) |
1.70 (1.20–2.20) |
1.50 (1.20–1.90) |
0.123 |
1.50 (1.20–1.90) |
1.75 (1.57–2.80) |
0.014 * |
|
NEWS2 |
4.50 (3.00–6.00) |
2.00 (1.00–4.00) |
< 0.001 * |
2.00 (1.00–4.00) |
4.00 (3.00–6.25) |
0.003 * |
|
NEWS2 + lactate |
6.55 (4.72–7.80) |
3.75 (2.42–5.70) |
< 0.001 * |
3.85 (2.50–5.92) |
6.40 (4.57–8.02) |
0.001 * |
|
NEWS2 + D-dimer |
5.05 (3.38–6.52) |
2.13 (1.11–4.16) |
< 0.001 * |
2.20 (1.12–4.37) |
4.28 (3.30–6.76) |
0.002 * |
|
NEWS2 + lactate + D-dimer |
6.90 (5.05–8.27) |
4.01 (2.63–5.98) |
< 0.001 * |
4.13 (2.69–6.06) |
6.90 (4.86–8.47) |
0.001 * |
|
Length of hospital stay (days) |
19.5 (14.0–29.5) |
6.0 (3.0–8.0) |
< 0.001 * |
6.0 (3.0–10.0) |
21.50 (13.50–28.50) |
< 0.001 * |
|
Gender |
||||||
|
Male, n (%) |
18 (64.3) |
104 (48.1) |
0.108 |
110 (47.8) |
12 (85.7) |
0.006 * |
|
Female, n (%) |
10 (35.7) |
112 (51.9) |
120 (52.2) |
2 (14.3) |
||
|
Fever (complaint), n (%) |
19 (67.9) |
146 (67.6) |
0.978 |
156 (67.8) |
9 (64.3) |
0.783 |
|
Cough, n (%) |
23 (82.1) |
137 (63.4) |
0.058 |
149 (64.8) |
11 (78.6) |
0.292 |
|
Shortness of breath, n (%) |
14 (50.0) |
79 (36.6) |
0.169 |
85 (37.0) |
8 (57.1) |
0.131 |
|
Nasal discharge, n (%) |
5 (17.9) |
26 (12.0) |
0.384 |
29 (12.6) |
2 (14.3) |
0.855 |
|
Comorbidity, n (%) |
19 (67.9) |
96 (44.4) |
0.02 * |
105 (45.7) |
10 (71.4) |
0.061 |
|
Diabetes mellitus, n (%) |
9 (32.1) |
49 (22.7) |
0.269 |
54 (23.5) |
4 (28.6) |
0.664 |
|
Hypertension, n (%) |
13 (46.4) |
52 (24.1) |
0.012 * |
59 (25.7) |
6 (42.9) |
0.157 |
|
Coronary artery disease, n (%) |
4 (14.3) |
25 (11.6) |
0.677 |
25 (10.9) |
4 (28.6) |
0.069 |
|
Asthma–COPD, n (%) |
5 (17.9) |
33 (15.3) |
0.723 |
36 (15.7) |
2 (14.3) |
0.891 |
|
CRF, n (%) |
1 (3.6) |
3 (1.4) |
— |
3 (1.3) |
1 (7.1) |
— |
|
Malignancy, n (%) |
1 (3.6) |
9 (4.2) |
— |
9 (3.9) |
1 (7.1) |
— |
|
Immune deficiency, n (%) |
3 (10.7) |
9 (4.2) |
0.132 |
9 (3.9) |
3 (21.4) |
0.024 * |
|
In-hospital mortality, n (%) |
14 (50) |
0 (0) |
< 0.001 * |
— |
— |
— |
The median D-dimer value of the patients who were admitted to ICU was statistically significantly higher than those who were hospitalized in ward unit (ICU 0.28 [0.17–0.48], ward unit 0.13 [0.08–0.22], p < 0.001). The median NEWS2 value of the patients who were admitted to ICU was statistically significantly higher than those who were hospitalized in ward unit (ICU 4.50 [3.00–6.00], ward unit 2.00 [1.00–4.00], p < 0.001). The median NEWS2 + lactate value of patients who were admitted to ICU was statistically significantly higher than those who were hospitalized in ward unit (ICU 6.55 [4.72–7.80], ward unit 3.75 [2.42–5.70], p < 0.001). The median NEWS2 + D-dimer value of the patients who were admitted to ICU was statistically significantly higher than those who were hospitalized in ward unit (ICU 5.05 [3.38–6.52], ward unit 2.13 [1.11–4.16], p < 0.001). The median NEWS2 + lactate + D-dimer value of the patients who were admitted to ICU was statistically significantly higher than those who were hospitalized in ward unit (ICU 6.90 [5.05–8.27], ward unit 4.01 [2.63–5.98], p < 0.001). Table 3 shows the detailed comparisons made by emergency outcomes.
Table 3 . Evaluation of participants by in-hospital mortality and ICU admission .
a The p values were obtained from the comparisons between ICU admission and ward unit admission groups.
|
Characteristic |
ICU admission (n = 28) |
Ward unit admission (n = 216) |
p a |
Survivor (n = 230) |
Nonsurvivor (n = 14) |
p b |
|
Age |
64.61 ± 15.10 |
52.35 ± 17.16 |
< 0.001 * |
52.84 ± 17.20 |
68.79 ± 12.71 |
0.001 * |
|
Fever (°C) |
36.65 (36.30–37.07) |
36.60 (36.20–37.00) |
0.555 |
36.60 (36.20–37.00) |
36.50 (36.30–37.02) |
0.969 |
|
Pulse (per minute) |
89.86 ± 19.07 |
93.19 ± 14.99 |
0.286 |
93.27 ± 14.89 |
85.14 ± 22.80 |
0.209 |
|
SBP (mmHg) |
128.64 ± 18.68 |
133.62 ± 21.70 |
0.248 |
133.77 ± 21.62 |
121.21 ± 12.97 |
0.004 * |
|
DBP (mmHg) |
73.25 ± 12.22 |
76.16 ± 11.58 |
0.216 |
76.15 ± 11.61 |
70.43 ± 11.74 |
0.075 |
|
MAP (mmHg) |
91.71 ± 13.03 |
95.31 ± 12.99 |
0.169 |
95.35 ± 13.00 |
87.35 ± 11.26 |
0.025 * |
|
Saturation (%) |
93.5 (89.0–96.0) |
95.0 (94.0–97.0) |
0.002 * |
95.0 (93.0–97.0) |
94.50 (89.75–96.25) |
0.219 |
|
D-dimer (µg/mL) |
0.28 (0.17–0.48) |
0.13 (0.08–0.22) |
< 0.001 * |
0.13 (0.09–0.23) |
0.28 (0.20–0.53) |
0.001 * |
|
Lactate (mEq/L) |
1.70 (1.20–2.20) |
1.50 (1.20–1.90) |
0.123 |
1.50 (1.20–1.90) |
1.75 (1.57–2.80) |
0.014 * |
|
NEWS2 |
4.50 (3.00–6.00) |
2.00 (1.00–4.00) |
< 0.001 * |
2.00 (1.00–4.00) |
4.00 (3.00–6.25) |
0.003 * |
|
NEWS2 + lactate |
6.55 (4.72–7.80) |
3.75 (2.42–5.70) |
< 0.001 * |
3.85 (2.50–5.92) |
6.40 (4.57–8.02) |
0.001 * |
|
NEWS2 + D-dimer |
5.05 (3.38–6.52) |
2.13 (1.11–4.16) |
< 0.001 * |
2.20 (1.12–4.37) |
4.28 (3.30–6.76) |
0.002 * |
|
NEWS2 + lactate + D-dimer |
6.90 (5.05–8.27) |
4.01 (2.63–5.98) |
< 0.001 * |
4.13 (2.69–6.06) |
6.90 (4.86–8.47) |
0.001 * |
|
Length of hospital stay (days) |
19.5 (14.0–29.5) |
6.0 (3.0–8.0) |
< 0.001 * |
6.0 (3.0–10.0) |
21.50 (13.50–28.50) |
< 0.001 * |
|
Gender |
||||||
|
Male, n (%) |
18 (64.3) |
104 (48.1) |
0.108 |
110 (47.8) |
12 (85.7) |
0.006 * |
|
Female, n (%) |
10 (35.7) |
112 (51.9) |
120 (52.2) |
2 (14.3) |
||
|
Fever (complaint), n (%) |
19 (67.9) |
146 (67.6) |
0.978 |
156 (67.8) |
9 (64.3) |
0.783 |
|
Cough, n (%) |
23 (82.1) |
137 (63.4) |
0.058 |
149 (64.8) |
11 (78.6) |
0.292 |
|
Shortness of breath, n (%) |
14 (50.0) |
79 (36.6) |
0.169 |
85 (37.0) |
8 (57.1) |
0.131 |
|
Nasal discharge, n (%) |
5 (17.9) |
26 (12.0) |
0.384 |
29 (12.6) |
2 (14.3) |
0.855 |
|
Comorbidity, n (%) |
19 (67.9) |
96 (44.4) |
0.02 * |
105 (45.7) |
10 (71.4) |
0.061 |
|
Diabetes mellitus, n (%) |
9 (32.1) |
49 (22.7) |
0.269 |
54 (23.5) |
4 (28.6) |
0.664 |
|
Hypertension, n (%) |
13 (46.4) |
52 (24.1) |
0.012 * |
59 (25.7) |
6 (42.9) |
0.157 |
|
Coronary artery disease, n (%) |
4 (14.3) |
25 (11.6) |
0.677 |
25 (10.9) |
4 (28.6) |
0.069 |
|
Asthma–COPD, n (%) |
5 (17.9) |
33 (15.3) |
0.723 |
36 (15.7) |
2 (14.3) |
0.891 |
|
CRF, n (%) |
1 (3.6) |
3 (1.4) |
— |
3 (1.3) |
1 (7.1) |
— |
|
Malignancy, n (%) |
1 (3.6) |
9 (4.2) |
— |
9 (3.9) |
1 (7.1) |
— |
|
Immune deficiency, n (%) |
3 (10.7) |
9 (4.2) |
0.132 |
9 (3.9) |
3 (21.4) |
0.024 * |
|
In-hospital mortality, n (%) |
14 (50) |
0 (0) |
< 0.001 * |
— |
— |
— |
Table 3 . Evaluation of participants by in-hospital mortality and ICU admission (continued) .
|
Variable |
D-dimer |
Lactate |
NEWS2 |
NEWS2 + lactate |
NEWS2 + D-dimer |
NEWS2 + lactate + D-dimer |
|
ICU admission |
||||||
|
AUC (95% CI) |
0.745 (0.658–0.832) |
0.589 (0.469–0.710) |
0.760 (0.675–0.845) |
0.774 (0.690–0.859) |
0.776 (0.692–0.860) |
0.778 (0.694–0.862) |
|
p value |
< 0.001 |
0.124 |
< 0.001 |
< 0.001 |
< 0.001 |
< 0.001 |
|
Cut-off level |
0.16 |
1.65 |
2.50 |
4.55 |
3.16 |
6.40 |
|
Sensitivity (%) |
85.7 |
57.1 |
85.7 |
82.1 |
85.7 |
67.9 |
|
Specificity (%) |
63.0 |
58.8 |
56.9 |
64.4 |
63.4 |
79.6 |
|
PPV (%) |
23.1 |
15.2 |
20.5 |
23.0 |
23.3 |
30.2 |
|
NPV (%) |
97.1 |
91.4 |
96.9 |
96.5 |
97.2 |
95.0 |
|
Mortality |
||||||
|
AUC (95% CI) |
0.768 (0.671–0.865) |
0.695 (0.563–0.827) |
0.735 (0.634–0.836) |
0.757 (0.647–0.867) |
0.752 (0.656–0.848) |
0.764 (0.655–0.873) |
|
p value |
0.001 |
0.067 |
0.003 |
0.001 |
0.002 |
0.001 |
|
Cut-off level |
0.17 |
1.65 |
2.5 |
4.15 |
3.16 |
4.54 |
|
Sensitivity (%) |
92.9 |
71.4 |
85.7 |
85.7 |
85.7 |
85.7 |
|
Specificity (%) |
61.7 |
58.7 |
54.3 |
55.7 |
60.4 |
57.8 |
|
PPV (%) |
12.9 |
9.5 |
10.3 |
10.5 |
11.7 |
11.0 |
|
NPV (%) |
99.3 |
97.1 |
98.4 |
98.5 |
98.6 |
98.5 |
The median D-dimer value of the patients in the non-survivor group was statistically significantly higher than that in the survivor group (nonsurvivor 0.28 [0.20–0.53], survivor 0.13 [0.09–0.23], p = 0.001). The median lactate value of the patients in the nonsurvivor group was statistically significantly higher than that in the survivor group (nonsurvivor 1.75 [1.57–2.80], survivor 1.50 [1.20–1.90], p = 0.014). The median NEWS2 value of the patients in the nonsurvivor group was statistically significantly higher than that in the survivor group (nonsurvivor 4.00 [3.00–6.25], survivor 2.00 [1.00–4.00], p = 0.003). The median NEWS2 + lactate value of the patients in the nonsurvivor group was statistically significantly higher than that in the survivor group (nonsurvivor 6.40 [4.57–8.02], survivor 3.85 [2.50–5.92], p = 0.001). The median NEWS2 + D-dimer value of the patients in the nonsurvivor group was statistically significantly higher than that in the survivor group (nonsurvivor 4.28 [3.30–6.76], survivor 2.20 [1.12–4.37], p = 0.002). The median NEWS2 + lactate + D-dimer value of the patients in the nonsurvivor group was statistically significantly higher than that in the survivor group (nonsurvivor 6.90 [4.86–8.47], survivor 4.13 [2.69–6.06], p < 0.001). Table 3 shows the detailed intragroup comparisons made by in-hospital mortality.
In the ROC analysis made to estimate the power of D-dimer, lactate, NEWS2, NEWS2 + lactate, NEWS2 + D-dimer, and NEWS2 + lactate + D-dimer levels in predicting critical care admission; AUC values were found as 0.745 (0.658–0.832), 0.589 (0.469–0.710), 0.760 (0.675–0.845), 0.774 (0.690–0.859), 0.776 (0.692-0.860), and 0.778 (0.694–0.862), respectively ( Table 4 ) ( Fig. 1 and Fig. 2 ). In the ROC analysis made to estimate the power of D-dimer, lactate, NEWS2, NEWS2 + lactate, NEWS2 + D-dimer, and NEWS2 + lactate + D-dimer levels in predicting in-hospital mortality; AUC values were 0.768 (0.671–0.865), 0.695 (0.563–0.827), 0.735 (0.634–0.836), 0.757 (0.647–0.867), 0.752 (0.656–0.848), and 0.764 (0.655–0.873), respectively ( Table 4 ) ( Fig. 3 and Fig. 4 ).
Table 4 . ROC analysis result by ICU admission and in-hospital mortality status .
|
Variable |
D-dimer |
Lactate |
NEWS2 |
NEWS2 + lactate |
NEWS2 + D-dimer |
NEWS2 + lactate + D-dimer |
|
ICU admission |
||||||
|
AUC (95% CI) |
0.745 (0.658–0.832) |
0.589 (0.469–0.710) |
0.760 (0.675–0.845) |
0.774 (0.690–0.859) |
0.776 (0.692–0.860) |
0.778 (0.694–0.862) |
|
p value |
< 0.001 |
0.124 |
< 0.001 |
< 0.001 |
< 0.001 |
< 0.001 |
|
Cut-off level |
0.16 |
1.65 |
2.50 |
4.55 |
3.16 |
6.40 |
|
Sensitivity (%) |
85.7 |
57.1 |
85.7 |
82.1 |
85.7 |
67.9 |
|
Specificity (%) |
63.0 |
58.8 |
56.9 |
64.4 |
63.4 |
79.6 |
|
PPV (%) |
23.1 |
15.2 |
20.5 |
23.0 |
23.3 |
30.2 |
|
NPV (%) |
97.1 |
91.4 |
96.9 |
96.5 |
97.2 |
95.0 |
|
Mortality |
||||||
|
AUC (95% CI) |
0.768 (0.671–0.865) |
0.695 (0.563–0.827) |
0.735 (0.634–0.836) |
0.757 (0.647–0.867) |
0.752 (0.656–0.848) |
0.764 (0.655–0.873) |
|
p value |
0.001 |
0.067 |
0.003 |
0.001 |
0.002 |
0.001 |
|
Cut-off level |
0.17 |
1.65 |
2.5 |
4.15 |
3.16 |
4.54 |
|
Sensitivity (%) |
92.9 |
71.4 |
85.7 |
85.7 |
85.7 |
85.7 |
|
Specificity (%) |
61.7 |
58.7 |
54.3 |
55.7 |
60.4 |
57.8 |
|
PPV (%) |
12.9 |
9.5 |
10.3 |
10.5 |
11.7 |
11.0 |
|
NPV (%) |
99.3 |
97.1 |
98.4 |
98.5 |
98.6 |
98.5 |
Fig. 1 . Receiver operating characteristic (ROC) curve by intensive care unit (ICU) admission.
Fig. 2 . Area under the curve (AUC) value for intensive care unit (ICU) admission.
Fig. 3 . Receiver operating characteristic (ROC) curve by in-hospital mortality.
Fig. 4 . Area under the curve (AUC) value for in-hospital mortality.
Discussion
In this study, we examined the parameters that will predict mortality and ICU admission to COVID-19. For this purpose, we used NEWS2, lactate, D-dimer, NEWS2 + lactate, NEWS2+D-dimer, and NEWS2+lactate+D-dimer parameters in patients hospitalized because of COVID-19. There are studies in the literature predicting COVID-19-led critical care admission and mortality using NEWS2. 10 , 11 In the literature, there are also studies evaluating the severity and mortality of COVID-19 by using D-dimer and lactate levels. 7 , 15 However, we did not find any study in the literature evaluating NEWS2 values together with lactate and D-dimer values in COVID-19 patients.
In this study, NEWS2 values were statistically significantly higher in patients who were admitted to ICU and had in-hospital mortality. Besides, in the ROC analysis, it predicted mortality with 0.735 AUC value and ICU admission with 0.760 AUC value. Gidari et al. 10 carried out a study with 68 COVID-19 patients (27 of them were admitted to ICU), and they found that NEWS2 reached 0.9 AUC, in the ROC analysis made to predict the ICU admission. Myrstad et al. 11 conducted a study with 66 hospitalized COVID-19 patients (13 of them were observed in-hospital mortality), and they found that NEWS2 reached 0.822 AUC value in predicting in-hospital mortality. There may be many reasons why NEWS2 values can predict mortality and ICU admission for COVID-19. That COVID-19 induces pneumonia and acute respiratory distress syndrome (ARDS) by affecting the lungs, and ARDS poses a risk in terms of mortality 16 can be regarded as a reason because there are many parameters (respiration rate, oxygen saturation, oxygen demand) among the NEWS2 parameters, as respiratory system indicator. Moreover, since respiratory failure developed in COVID-19 patients increases mortality, the NEWS2-scoring method, which is based on the respiratory system and vital signs, becomes compatible with COVID-19. In the study conducted by Zhang et al. 17 on 82 COVID-19 patients who were observed mortality, only 40% had mechanical ventilation support despite 100% needing oxygen. Besides, while pulmonary damage was observed in all patients, respiratory failure was shown as the cause of death in 69.5% of the patients. 17
In this study, D-dimer values were found to be statistically significantly higher in patients who were admitted to ICU and had in-hospital mortality. In the ROC analysis, while D-dimer level reached 0.768 AUC value in predicting mortality, it reached 0.745 AUC value in predicting ICU admission. Cai et al. 18 conducted a study with 432 COVID-19 patients (where 125 of them were in the severe group), the D-dimer level reached 0.65 AUC value in the ROC analysis made to determine the severity of the disease. Zhou et al. 4 carried out a study with 191 COVID-19 patients (where 54 died), and D-dimer level above 1 μg/mL reached an odds ratio of 20.04 in predicting mortality with reference to D-dimer level below 0.5 μg/mL. The fact that the D-dimer levels were high in COVID-19 patients who had in-hospital mortality and were admitted to ICU might be because COVID-19 leads to hypercoagulability. 19 The reason for the hypercoagulability resulting in high D-dimer levels can be shown as the hypoxia triggers thrombosis. 20
In this study, the values of NEWS2 + lactate, NEWS2 + D-dimer, and NEWS2 + lactate + D-dimer were statistically significantly higher in patients who were admitted to ICU and had in-hospital mortality. In the ROC analysis, NEWS2 + lactate, NEWS2 + D-dimer, and NEWS2 + lactate + D-dimer levels reached the AUC values of 0.757, 0.752, and 0.764, in predicting mortality; while these parameters reached 0.774, 0.776, and 0.778 in predicting ICU admission. In the literature, there was no COVID-19 study in which these values were used in combination. However, there are studies like this in different patient groups in the literature. In the study conducted by Dundar et al. 21 on geriatric patients applied to the emergency department, the lactate level alone reached 0.654 AUC value in predicting in-hospital mortality, while the NEWS + lactate level reached 0.714 AUC value. In the study conducted by Nickel et al. 13 on high-risk patients, the 30-day mortality was 0% in patients who had NEWS2 values below 3 and the D-dimer levels below 0.5 μg/mL, while 30-day mortality was 12.5% in patients who had NEWS values above 3 and the D-dimer level above 0.5 μg/mL. When studies using COVID-19 patients are examined, it was seen that studies use other parameters and combine with NEWS2. Carr et al. 22 carried out a study on a total of 452 COVID-19 patients (where 159 of them was observed mortality in the 14-day process), and they determined that NEWS2 reached the AUC value of 0.628 in predicting 14-day mortality, while NEWS2 reached the AUC value of 0.751 when they also add some lab parameters such as CRP, neutrophil, estimated glomerular filtration rate, and albumin. In this study, it was found that the use of NEWS2 with lactate and/or D-dimer was more valuable than NEWS2 in predicting in-hospital mortality and ICU admission. The most valuable parameter in predicting ICU admission was NEWS2 + lactate + D-dimer, while the most valuable parameter in predicting in-hospital mortality was D-dimer. There may be several reasons why NEWS2 is more valuable than D-dimer in predicting ICU admission, while D-dimer is more valuable than NEWS2 in predicting in-hospital mortality. This reason may be explained by the fact that the ICU admission decision was made according to vital parameters and the NEWS2 was designed according to vital parameters. The fact that micro thrombosis formed because of hypercoagulation and that COVID-19 causes death without disturbing vital parameters may explain the fact that D-dimer is a more valuable predictor of mortality.
In this study, length of hospital stay values were statistically significantly higher in patients who were admitted to ICU and had in-hospital mortality. This may be due to the fact that comorbidity is associated with the severity and mortality of the disease. 23 Because more comorbid patients are more likely to be exposed to conditions such as thromboembolism that will prolong the hospitalization period. 24 In our study, patients admitted to ICU had statistically high rates of comorbidities. In addition, although it was not statistically significant, a higher comorbidity rate was obtained in the non-survivor group (71.4%–45.7%). This resulted in increased levels of d-dimer and NEWS2 + D-dimer in more comorbid patients because of thromboembolism. It has been reported in the literature that D-dimer levels increase in COVID-19 patients with comorbidity. 25 Another reason for the high length of hospital stay in patients who were admitted to ICU and had in-hospital mortality is that D-dimer is an independent predictor of high length of hospital stay in COVID-19 patients (OR: 1.37). 26
In our study, the mean age of the patients was 56.04 ± 17.97 years, and the most frequent comorbidity was hypertension. Besides, the male sex ratio was statistically significantly higher in patients who had in-hospital mortality than in those who did not have in-hospital mortality. Richardson et al. 27 carried out a study with 5,700 hospitalized COVID-19 patients, and they found the median age of cases as 63, and the most frequent comorbidity was hypertension. In the study conducted by Palaiodimos et al. 28 with 200 COVID-19 patients; compared to the female gender, the male gender reached an odds ratio of 2.31 in predicting in-hospital mortality. The study data were compatible with the literature. In our study, the levels of saturation were statistically significantly lower in patients who were admitted to ICU, but there was no statistical difference between the non-survivor group and the survivor group. The fact that pneumonia and respiratory failure are indicators of poor prognosis has led to more attention to saturation levels in admission to the ICU.
The limitations of our study were as follows: the shortage in case numbers, being a retrospective and single-center study, failure to include discharged COVID-19 patients in the study, and failure to evaluate treatment protocols.
In conclusion, the parameters NEWS2, D-dimer, NEWS2 + lactate, NEWS2 + D-dimer, and NEWS2 + lactate + D-dimer have been found to be reliable in predicting critical care admission and in-hospital mortality in COVID-19 patients who applied to the emergency department. Besides, compared to using the NEWS2 value alone, a combination of NEWS2, lactate, and D-dimer was found to be more valuable in predicting in-hospital mortality and critical care admission. Multicenter prospective studies with more patient populations are needed.
References
- 1. Arslan Ş, Delice O, Kahraman M, Yılmaz Sİ, Aslan MH. Correlation of chest CT and RT-PCR testing in coronavirus disease 2019 (COVID-19) in Turkey. Ann Clin Anal Med . 2021;12:483-487. doi: 10.4328/ACAM.20310 [DOI]
- 2. Sohrabi C, Alsafi Z, O’Neill N, et al. World Health Organization declares global emergency: a review of the 2019 novel coronavirus (COVID-19). Int J Surg . 2020;76:71-76. doi: 10.1016/j.ijsu.2020.02.034 [DOI] [PMC free article] [PubMed]
- 3. None None
- 4. Zhou F, Yu T, Du R, et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study. Lancet . 2020;395:1054-1062. doi: 10.1016/S0140-6736(20)30566-3 [DOI] [PMC free article] [PubMed]
- 5. Iwasaki Y, Shiga T, Konno D, et al. Screening of COVID-19-associated hypercoagulopathy using rotational thromboelastometry. J Clin Anesth . 2020;67:109976. doi: 10.1016/j.jclinane.2020.109976 [DOI] [PMC free article] [PubMed]
- 6. Bakker J, Jansen TC. Don’t take vitals, take a lactate. Intensive Care Med . 2007;33:1863-1865. doi: 10.1007/s00134-007-0679-y [DOI] [PMC free article] [PubMed]
- 7. Li X, Wang L, Yan S, et al. Clinical characteristics of 25 death cases with COVID-19: a retrospective review of medical records in a single medical center, Wuhan, China. Int J Infect Dis . 2020;94:128-132. doi: 10.1016/j.ijid.2020.03.053 [DOI] [PMC free article] [PubMed]
- 8. Bansal A, Singh AD, Jain V, et al. A systematic review and meta-analysis of D-dimer levels in patients hospitalized with coronavirus disease 2019 (COVID-19). medRxiv 2020. . doi:10.1101/2020.06.24.20139600 doi: 10.1101/2020.06.24.20139600 [DOI]
- 9. Smith GB, Prytherch DR, Meredith P, Schmidt PE, Featherstone PI. The ability of the National Early Warning Score (NEWS) to discriminate patients at risk of early cardiac arrest, unanticipated intensive care unit admission, and death. Resuscitation . 2013;84:465-470. doi: 10.1016/j.resuscitation.2012.12.016 [DOI] [PubMed]
- 10. Gidari A, De Socio GV, Sabbatini S, Francisci D. Predictive value of National Early Warning Score 2 (NEWS2) for intensive care unit admission in patients with SARS-CoV-2 infection. Infect Dis (Lond) . 2020;52:698-704. doi: 10.1080/23744235.2020.1784457 [DOI] [PubMed]
- 11. Myrstad M, Ihle-Hansen H, Tveita AA, et al. National Early Warning Score 2 (NEWS2) on admission predicts severe disease and in-hospital mortality from COVID-19—a prospective cohort study. Scand J Trauma Resusc Emerg Med . 2020;28:66. doi: 10.1186/s13049-020-00764-3 [DOI] [PMC free article] [PubMed]
- 12. Jo S, Yoon J, Lee JB, Jin Y, Jeong T, Park B. Predictive value of the National Early Warning Score–Lactate for mortality and the need for critical care among general emergency department patients. J Crit Care . 2016;36:60-68. doi: 10.1016/j.jcrc.2016.06.016 [DOI] [PubMed]
- 13. Nickel CH, Kellett J, Cooksley T, Bingisser R, Henriksen DP, Brabrand M. Combined use of the National Early Warning Score and D-dimer levels to predict 30-day and 365-day mortality in medical patients. Resuscitation . 2016;106:49-52. doi: 10.1016/j.resuscitation.2016.06.012 [DOI] [PubMed]
- 14. None None
- 15. Lippi G, Favaloro EJ. D-dimer is associated with severity of coronavirus disease 2019: a pooled analysis. Thromb Haemost . 2020;120:876-878. doi: 10.1055/s-0040-1709650 [DOI] [PMC free article] [PubMed]
- 16. Yang X, Yu Y, Xu J, et al. Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study. Lancet Respir Med . 2020;8:475-481. doi: 10.1016/S2213-2600(20)30079-5 [DOI] [PMC free article] [PubMed]
- 17. Zhang B, Zhou X, Qiu Y, et al. Clinical characteristics of 82 cases of death from COVID-19. PLoS One . 2020;15:e0235458. doi: 10.1371/journal.pone.0235458 [DOI] [PMC free article] [PubMed]
- 18. Cai YQ, Zhang XB, Zeng HQ, et al. Prognostic value of neutrophil-to-lymphocyte ratio, lactate dehydrogenase, D-dimer and CT score in patients with COVID-19. Res Sq 2020. . doi:10.21203/rs.3.rs-30959/v1 doi: 10.21203/rs.3.rs-30959/v1 [DOI]
- 19. Zhang L, Yan X, Fan Q, et al. D-dimer levels on admission to predict in-hospital mortality in patients with COVID-19. J Thromb Haemost . 2020;18:1324-1329. doi: 10.1111/jth.14859 [DOI] [PMC free article] [PubMed]
- 20. Gupta N, Zhao YY, Evans CE. The stimulation of thrombosis by hypoxia. Thromb Res . 2019;181:77-83. doi: 10.1016/j.thromres.2019.07.013 [DOI] [PubMed]
- 21. Dundar ZD, Kocak S, Girisgin AS. Lactate and NEWS-L are fair predictors of mortality in critically ill geriatric emergency department patients. Am J Emerg Med . 2020;38:217-221. doi: 10.1016/j.ajem.2019.02.006 [DOI] [PubMed]
- 22. Carr E, Bendayan R, Bean D, et al. Evaluation and improvement of the National Early Warning Score (NEWS2) for COVID-19: a multi-hospital study. BMC Med . 2021;19:23. doi: 10.1186/s12916-020-01893-3 [DOI] [PMC free article] [PubMed]
- 23. Thakur B, Dubey P, Benitez J, et al. A systematic review and meta-analysis of geographic differences in comorbidities and associated severity and mortality among individuals with COVID-19. Sci Rep . 2021;11:8562. doi: 10.1038/s41598-021-88130-w [DOI] [PMC free article] [PubMed]
- 24. Lyman GH, Culakova E, Poniewierski MS, Kuderer NM. Morbidity, mortality and costs associated with venous thromboembolism in hospitalized patients with cancer. Thromb Res . 2018;164(Suppl 1):S112-S118. doi: 10.1016/j.thromres.2018.01.028 [DOI] [PubMed]
- 25. Miljković AR. D-dimer rise in outpatients with COVID 19 and comorbidities. Opšta Medicina . 2021;27:47-52. doi: 10.5937/opmed2102047M [DOI]
- 26. Chen FJ, Li FR, Zheng JZ, et al. Factors associated with duration of hospital stay and complications in patients with COVID-19. J Public Health Emerg . 2021;5:6. doi: 10.21037/jphe-20-74 [DOI]
- 27. Richardson S, Hirsch JS, Narasimhan M, et al. Presenting characteristics, comorbidities, and outcomes among 5700 patients hospitalized with COVID-19 in the New York City area. JAMA . 2020;323:2052-2059. doi: 10.1001/jama.2020.6775 [DOI] [PMC free article] [PubMed]
- 28. Palaiodimos L, Kokkinidis DG, Li W, et al. Severe obesity, increasing age and male sex are independently associated with worse in-hospital outcomes, and higher in-hospital mortality, in a cohort of patients with COVID-19 in the Bronx, New York. Metabolism . 2020;108:154262. doi: 10.1016/j.metabol.2020.154262 [DOI] [PMC free article] [PubMed]