Clinical course of severe and critically ill patients with coronavirus disease 2019 (COVID-19): A comparative study.

To the editor:

We read with great interest the article by Dr. Zheng Z and colleagues in the Journal of Infection titled “Risk factors of critical & mortal COVID-19 cases: A systematic literature review and meta-analysis”, published online in April 2020. The authors analyzed the risk factors associated with critical illness or death in all COVID-19 patients to help assessing patient status and identify critical patients early. They found male, age >65, smoking patients were at higher risk of developing into the critical or mortal condition. Some comorbidities, clinical manifestation and laboratory examination might suggest the progression of COVID-19. In this study, we compared the clinical course of severe and critically ill patients with COVID-19, which is of paramount importance to identify indicators for progression from severe to critical illness and thus reduce mortality .

From January 01 to March 17, 2020, we included 31 severe and 20 critically ill patients with COVID-19 from three designated hospitals. All the patients were positive for COVID-19 via real-time fluorescence polymerase chain reaction tests. The patients were diagnosed as severe or critical ill cases according to the trial version 7 of guidelines in China. We collected epidemiological, clinical, laboratory findings, and treatment from medical records. Two-sample t-test, Mann-Whitney U test, χ² test, or Fisher's exact test were used to compare differences between severe and critically ill cases where appropriate.

Among the 51 patients, 44 (86%) reported COVID-19 exposure history. The mean age of the patients was 59.5 ± 13.6 years (range, 29–89 years) (Table 1 ). Most patients (67%) were male. 51% patients had chronic diseases, with hypertension (33%) being the most common one, followed by diabetes (14%) (Table 1). At onset, fever and dry cough were the most frequent symptoms (Table 1). Two severe patients’ infection were asymptomatic (Table 1). Critically ill cases were concomitant with more chronic illnesses than severe cases (70% vs 39%, p = 0.029) (Table 1).

Table 1

Demographics and clinical characteristics of patients with COVID-19.

Characteristics	All patients(n = 51)	Severe patients(n = 31)	Critically ill patients(n = 20)	P value
Age, years	59.5 ± 13.6	57.6 ± 13.7	62.5 ± 13.3	0.210
Age range, years
20–29	1 (2)	0	1 (5)	0.392
30–39	5 (10)	5 (16)	0	0.143
40–49	5 (10)	4 (13)	1 (5)	0.636
50–59	10 (20)	6 (19)	4 (20)	1.000
60–69	18 (35)	10 (32)	8 (40)	0.572
70–79	8 (15)	3 (10)	5 (25)	0.237
80–89	4 (8)	3 (10)	1 (5)	1.000
Sex				0.680
Female	17 (33)	11 (35)	6 (30)
Male	34 (67)	20 (65)	14 (70)
Chronic diseases	26 (51)	12 (39)	14 (70)	0.029
Hypertension	17 (33)	7 (23)	10 (50)	0.043
Diabetes	7 (14)	3 (10)	4 (20)	0.410
Chronic pulmonary disease	2 (4)	0	2 (10)	0.149
Chronic kidney disease	3 (6)	1 (3)	2 (10)	0.553
Cancer	4 (8)	1 (3)	3 (15)	0.287
Chronic cardiac disease	3 (6)	3 (10)	0	0.271
Cerebrovascular disease	1 (2)	1 (3)	0	1.000
Others	4 (8)	4 (13)	0	0.145
Symptoms
Fever	42 (82)	23 (74)	19 (95)	0.072
Dry cough	31 (61)	17 (55)	14 (70)	0.279
Fatigue	7 (14)	5 (16)	2 (10)	0.690
Diarrhoea	3 (6)	2 (6)	1 (5)	1.000
Headache	2 (4)	2 (6)	0	0.514
Rhinorrhoea	1 (2)	0	1 (5)	0.392
Myalgia	1 (2)	1 (3)	0	1.000
Asymptomatic infection	2 (4)	2 (6)	0	0.514
Complications
ARDS	13 (25)	4 (13)	9 (45)	0.010
Septic shock	2 (4)	0	2 (10)	0.158
Myocardial injury	4 (8)	0	4 (20)	0.019
Arrhythmia	2 (4)	1 (3)	1 (5)	1.000
Acute kidney injury	3 (6)	1 (3)	2 (10)	0.553
Acute liver injury	3 (6)	2 (6)	1 (5)	1.000
Hyperglycaemia	12 (24)	4 (13)	8 (40)	0.042
Gastrointestinal hemorrhage	1 (2)	0	1 (5)	0.392
Bacterial co-infection	11 (22)	4 (13)	7 (35)	0.085
Fungal co-infection	6 (12)	3 (10)	3 (15)	0.668
Treatment
Antiviral drugs	49 (96)	30 (97)	19 (95)	1.000
Antibiotic drugs	51 (100)	31 (100)	20 (100)	1.000
Antifungal drugs	8 (16)	3 (10)	5 (25)	0.237
Corticosteroids	39 (76)	21 (68)	18 (90)	0.095
Immunoglobulin	39 (76)	24 (77)	15 (75)	1.000
Albumin	31 (61)	17 (55)	14 (70)	0.279
Convalescent plasma therapy	2 (4)	0	2 (10)	0.158
Traditional Chinese medicine	36 (71)	20 (65)	16 (80)	0.236
Invasive mechanical ventilation	9 (18)	0	9 (45)	<0.001
Non-invasive mechanical ventilation	36 (71)	21 (68)	15 (75)	0.579
High-flow nasal cannula oxygen	30 (59)	15 (48)	15 (75)	0.059
ECMO	1 (2)	0	1 (5)	0.392
CRRT	3 (6)	0	3 (15)	0.055
ICU admission	28 (55)	10 (32)	18 (90)	<0.001
Clinical outcome
Discharge	43 (71)	30 (97)	13 (65)	0.028
Death	8 (16)	1 (3)	7 (35)	0.004

Note: Data are n (%) or mean ± standard deviation. p values were calculated by t-test, χ² test or Fisher's exact test, as appropriate. Abbreviations: ARDS, acute respiratory distress syndrome; ECMO, extracorporeal membrane oxygenation; CRRT, continuous renal replacement therapy; ICU, intensive care unit.

Laboratory findings at admission showed partial pressure of carbon dioxide (p = 0.020), lymphocyte count (p = 0.007) and albumin (p = 0.033) in critically ill cases were significantly lower than that in severe cases (Table 2 ). However, neutrophil count (p = 0.009), aspartate aminotransferase (AST) (p = 0.016), fasting blood glucose (FBG) (p = 0.001) in critically ill cases were remarkably higher than that in severe cases (Table 2).

Table 2

Laboratory findings of patients with COVID-19.

Characteristics	All patients(n = 51)	Severe patients(n = 31)	Critically ill patients(n = 20)	P value
Partial pressure of oxygen, mmHg	76.1 (61.5–89.5)	78.5 (69.6–89.6)	63.5 (54.5–76.1)	0.126
Oxygen saturation,%	95.9 (90.8–98.4)	96.1 (94.5–98.8)	94 (88–97.7)	0.084
Partial pressure of carbon dioxide, mmHg	35.7 (32–39.8)	38 (34.9–40)	32.5 (30.7–39.3)	0.020
White blood cell count, × 10⁹ per L	5.40 (3.77–8.09)	5.00 (3.42–6.47)	6.57 (4.10–8.96)	0.195
Neutrophil count, × 10⁹ per L	3.30 (2.22–5.59)	2.83 (2.12–4.58)	5.46 (3.08–7.73)	0.009
Lymphocyte count, × 10⁹ per L	0.86 (0.57–1.21)	1.03 (0.71–1.63)	0.65 (0.52–0.91)	0.007
Lactate dehydrogenase, U/L	164 (132–213)	164 (132–209)	161 (131–276)	0.602
Hemoglobin, g/dL	144 (120–166)	148 (123–171)	129 (113–161)	0.243
Platelet count, × 10⁹ per L	132 (65–198)	121 (46–172)	146 (80–219)	0.255
Albumin, g/L	36.2 (34.8–39.1)	36.9 (35.9–39.3)	34.9 (30.8–38.2)	0.033
AST, U/L	32.6 (23.9–44.5)	29.7 (21–38.7)	37.0 (30.7–58.8)	0.016
ALT, U/L	25.5 (14–43.3)	23.3 (14.2–36.9)	35.7 (13.3–49.7)	0.372
APTT, s	31.3 (29.3–35.6)	31.1 (29.2–35.5)	32.1 (29.4–41.7)	0.539
Prothrombin time, s	12.4 (11.9–13.2)	12.7 (11.9–13.2)	12.4 (11.9–13.3)	0.789
Creatine, μmol/L	84 (65–379)	80 (63–309)	202.7 (67.1–509)	0.300
Creatine kinase, U/L	81 (50.8–185.5)	70 (52.5–156)	88 (45–231)	0.446
Creatine Kinase Isoenzyme MB, U/L	12.1 (7.6–14.9)	13.2 (7.6–15.5)	11.8 (7.4–14.6)	0.505
High-sensitivity C-reactive protein >5 mg/L	20 (39)	11 (35)	9 (45)	0.497
Lactic acid, mmol/L	2.6 (1.4–3.85)	2.4 (1.48–3.89)	3.06 (1.25–3.8)	0.744
Fasting blood glucose, mmol/L	7.16 (5.9–8.68)	6.15 (5.39–7.35)	8.72 (7.53–10.21)	0.001

Note: Data are median (interquartile range) or n (%). p values were calculated by Mann-Whitney U test, χ² test, or Fisher's exact test, as appropriate. Abbreviations: AST, aspartate aminotransferase; ALT, alanine aminotransferase; APTT, activated partial thromboplastin time.

As compared with severe patients, critically ill patients were more likely to develop comorbidities, including acute respiratory distress syndrome (ARDS) (45% vs 13%, p = 0.010), hyperglycaemia (40% vs 13%, p = 0.042), and myocardial injury (20% vs 0%, p = 0.019) (Table 1). However, there were no significant differences in septic shock (p = 0.158), arrhythmia (p = 1.000), acute kidney injury (p = 0.553), acute liver injury (p = 1.000), gastrointestinal hemorrhage (p = 0.392), bacterial co-infection (p = 0.085), and fungal co-infection (p = 0.668) (Table 1). Critically ill patients required more intensive care unit (ICU) care (p <0.001) and invasive mechanical ventilation (p <0.001) than severe patients (Table 1). Specifically, two (10%) critically ill patients were transfused with convalescent plasma (CP), one (5%) was given extracorporeal membrane oxygenation (ECMO), and three (15%) were treated with continuous renal replacement therapy (CRRT) (Table 1). Critically ill patients had significantly higher mortality than severe patients (35% vs 3%, p = 0.004) (Table 1).

Based on previous studies, evidence suggests that older, male patients are the most susceptible to COVID-19. 48% of COVID-19 patients had comorbid conditions, commonly cardiovascular diseases and diabetes. This rate was significantly higher for critically ill patients, in this study, 70% critically ill cases had more than one chronic disease, such as hypertension and diabetes. Elderly people with underlying diseases are at increased risk of becoming critically ill or dying if they have COVID-19.

Laboratory tests might provide some key clues to indicate critical illness of COVID-19. Lymphocytopenia was a prominent feature of patients with COVID-19 because targeted invasion by viral particles damages the cytoplasmic component of the lymphocyte and causes its destruction. Lymphocytopenia may reflect the severity of COVID-19 [3]. The elevation of AST level was more frequent and significant than the increase of ALT in severe and critically ill patients on hospital admission. Admission AST might be a good indicator of disease severity because AST elevation was positively correlated with the increase of neutrophil counts and the decrease of lymphocyte counts at baseline. Critically ill patients had significantly higher FBG level, which may attribute to pre-existing diabetes and stress-related hyperglycemia. Diabetes is characterized by chronic hyperglycemia affecting the immune response to the coronavirus. Patients having diabetes were more likely to develop ARDS and require ICU and mechanical ventilation as compared with non-diabetic patients, indicating patients with diabetes had higher risk of progressing to critically ill cases. However, the impact of pre-existing diabetes may be smaller than stress-related hyperglycemia because only 14% patients reported a known history of diabetes. Stress hyperglycemia is a well-described body's response and maladaptive mechanism, which may lead to an abnormal inflammatory and immune response contributing to the progression of the COVID-19. A well-controlled hyperglycemia during COVID-19 may result in a decrease of inflammatory cytokines release and an improvement of prognosis.

A recent large study showed that 5% of the cases were critically illness characterized by respiratory failure, septic shock, and/or multiple organ dysfunction or failure. To date, no therapeutics have yet been proven effective for the treatment of critically illness except for supportive care, including treatment with antiviral drugs, antibiotic drugs, corticosteroids, immunoglobulins, and mechanical ventilation. The principal feature of patients with critical illness is the development of ARDS. ECMO is recommended by WHO interim guidelines to support eligible patients with ARDS, while the use of which is restricted to specialised centres globally and technology challenges. CP had been used as a last resort to improve survival rate of critically ill patients with COVID-19. It can significant reduce the ICU stay and risk of mortality of patients, which might because that antibodies from convalescent plasma might suppress viraemia.

This study suggested that critically ill patients with COVID-19 had high proportion of underlying diseases and high risk for developing multiple organ failure, which made the treatment more challenging. A well-controlled hyperglycemia is crucial for critically ill patients. Intensive supporting and careful monitoring are necessary to reduce mortality in critically ill patients before effective drugs and vaccines to be developed against COVID-19.

Declaration of Competing Interest

The authors declare no competing interests.