Analysis of Deaths and Favorable Developments of Patients with SARS-CoV-2 Hospitalized in the Largest Hospital for Infectious Diseases and Pneumo-Phthisiology in the West of the Country.

Purpose: Romania is one of the European countries that has been hit the hardest by the severe acute respiratory syndrome caused by the new coronavirus SARS-CoV-2, with over 1.91 million reported cases and over 59,257 deaths. The aim of this study was to identify the main predictors of death in hospitalized patients. Patients and
Methods: In the period from 1 March 2020 to 30 June 2021, an observational, retrospective, randomized, case-control study was conducted, which included a sample of 139 patients who died in hospital and another sample of 275 patients who had been discharged in an improved or healed condition. Confirmation of COVID-19 cases was performed by RT-PCR from nasopharyngeal and oropharyngeal exudates. Statistical data were analyzed by logistic regression, Cox regression and a comparison of survival curves by the log-rank (Mantel-Cox) test.
Results: The most powerful logistic regression model identified the following independent predictors of death: history of coagulopathy HR = 30.73 [1.94-487.09], p = 0.015; high percentage of neutrophils HR = 1.09 [1.01-1.19], p = 0.027; and decreased blood-oxygenation HR = 53881.97 [1762.24-1647489.44], p < 0.001. Cox regression identified the following factors that influenced the evolution of cases: history of coagulopathy HR = 2.44 [1.38-4.35], p = 0.000; O2 saturation HR = 0.98 [0.96-0.99], p = 0.043; serum creatinine HR = 1.23 [1.08-1.39], p = 0.001; dyspnea on admission HR = 2.99 [1.42-6.30], p = 0.004; hospitalization directly in the ICU HR = 3.803 [1.97-7.33], p < 0.001; heart damage HR = 16.76 [1.49-188.56], p = 0.022; and decreased blood-oxygenation HR = 35.12 [5.92-208.19], p < 0.001.
Conclusion: Knowledge of the predictors of death in hospitalized patients allows for the future optimization of triage and therapeutic case management for COVID-19.

Introduction

Since the onset of the Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic, more than 337 million cases and more than 5.5 million deaths have been confirmed worldwide, corresponding to an average fatality of 1.65%.1

In Romania, more than 1.91 million cases have been reported as of 16 January 2022, with 59,257 deaths attributable to coronavirus disease 19 (COVID-19), leading to an average fatality rate of 3.09%, which is higher than the international rate.2 The first Romanian studies on the risk factors for death with COVID-19 identified endogenous factors, including age, male sex and comorbidities such as cardiovascular, renal, and oncological diseases as well as immunosuppression.3

Later, diabetes and obesity were added to the list of risk factors, which is significant for a population in which the prevalence of diabetes is estimated at 11.6%, with 18.4% at the prediabetic stage, and where 31.9% and 34.7% of adults aged 20 to 79 years are obese or, overweight, respectively, according to the National Study on the Prevalence of Diabetes, Prediabetes, Overweight, Obesity, Dyslipidemia, Hyperuricemia and Chronic Kidney Disease (PREDATORR).4–6 In addition to these factors, there are those who are also undiagnosed, given that one in three diabetic European adults do not know their status.7 The literature also identifies other comorbidities that have an unfavorable influence on the outcome, such as: dyslipidemia, chronic neurologic diseases (Parkinson’s disease), chronic obstructive pulmonary disease (COPD), or pregnancy.8–11

In addition, 56.4% of the total Romanian population belongs to the urban environment, with a higher population density that favors the transmission of SARS-COV-2. The proportion of the population over 60 years of age reached 23.5% in 2020, and the number of doctors from 34 to 10,000 inhabitants is lower compared with other European countries, with a systematic migration in the last decades (according to international studies); these factors all contribute to the increase in case fatality rate.12–15

Hesitant behavior in the face of vaccination, which places Romania in the penultimate position among European Union countries (38.81% of the population completely vaccinated on 2 December 2021), the relaxation of the population, with non-adherence to preventive behaviors in 2021, and subsequent overcrowding has led the country to the highest COVID-19 mortality rate in the European Union (the second highest as of November 2021 with 586 deaths/day) and one of the highest per capita death rates in the world.16–19 There are several factors that have contributed to this situation:

- The establishment of severe restrictions, with mandatory hospitalization, even for asymptomatic cases, in the first months of the pandemic, thereby increasing the population’s anxiety, resulting in an avoidance of medical services, especially among the population with limited medical knowledge;

- the incomplete application of preventive measures and premature relaxation, without justification in vaccine coverage or pandemic trends;

- only performing a small number of PCR or antigen tests, with a subsequent underdiagnosing and underreporting of the cases;

- the excessive politicization of the COVID-19 pandemic measures, which led to a lack of trust in the established decisions;

- a pro-vaccine campaign not adjusted to the different population categories;

- the spread of false information, the presentation of conspiracy theories, not only in social media networks, but even in traditional media, with encouragement from the anti-vaccine movement;

- the behavior of a segment of healthcare workers refusing the vaccine and being very vocal on social media;

- the ambivalent role of the church.20–22

Under these conditions, an efficient management of material and human resources is vital, in a health system with limits accentuated by the pandemic context.

The main objective of the study is to identify predictive factors, both clinical and paraclinical, independent of the in-hospital mortality of COVID-19 patients hospitalized in the largest clinic for communicable diseases in western Romania.

Patients and Methods

Inclusion Criteria

In the period between 1 March 2020 and 30 June 2021, an observational, retrospective, randomized, case-control type 2:1 study was carried out that included 414 patients distributed in 2 samples:

a) EI (N1=139), in which there were included all the patients were included who met the criteria for the definition of a confirmed case of Acute Respiratory Syndrome with the new- coronavirus hospitalized in Clinic II of the Hospital for Infectious Diseases and Pneumophthisiology of Timisoara and who died during hospitalization;

b) EII (N2=275), which included control subjects, admitted to the same clinic for confirmed SARS-CoV-2 infection, with the condition at discharge discharged or improved, appropriate in terms of age and period of hospitalization.

For each deceased case, the first 2 patients from the list of those hospitalized with COVID-19 were included, in the same month as the case, who were included in the same age group (maximum range of 5 years).

Clinical criteria included the presence of at least one of the following signs and symptoms - cough, fever, shortness of breath, sudden onset of anosmia, ageusia or dysgeusia, gastrointestinal manifestations (eg, vomiting, accelerated transit), with radiological/radioimaging evidence of COVID-19 compatible lesions. The epidemiological criteria included at least one of the following: a) direct contact with a confirmed case of COVID-19, within 14 days prior to the date of onset; b) a resident or staff of institutions for the care of vulnerable persons in the period 14 days prior to the date of onset, and the institution was place in which the transmission of SARS-CoV-2 has been confirmed.

Exclusion Criteria

Patients under the age of 18 and those with psychiatric pathology with impaired discernment were excluded, as it was not possible to obtain informed consent.

Data were collected on demographic variables, symptoms, and comorbidities existing at the time of hospitalization, during period of illness and hospitalization, the need for support ventilation, paraclinical parameters at hospitalization, complications, and the main medicinal products administered. The case-fatality rate was calculated as the proportion of patients who died of COVID-19 out of the total number of patients hospitalized with COVID-19 in the monitored ward.

For a better understanding of the pathology included in the study, we will present the definition of cardiac, hepatic, renal, neurological and pancreatic lesions in COVID-19. Acute cardiac impairment associated with COVID-19 is a decrease in ejection fraction and an increase in troponin I; it is most commonly associated with high blood pressure, coronary heart disease and chronic heart failure.23–25 Liver damage occurs for a variety of reasons, some may be non-cancerous (benign) and some are cancerous.26 COVID-19 is often associated with abnormal liver enzymes (with acute liver failure being rare).27 The kidneys may be affected in COVID-19 and laboratory data will indicate: increased serum creatinine, albuminuria, proteinuria, hematuria and hematoproteinuria; the incidence of acute kidney injury was 5–23% (up to 35%).28 In patients who have already had acute or chronic kidney damage, the condition may worsen.29 COVID-19-associated neurological lesions, although often nonspecific, include COVID-19-associated necrotizing leukoencephalopathy, post-hypoxic leukoencephalopathy and clear encephalopathy, and the severity of neurological symptoms ranges from temporary anosmia to dizziness, confusion, convulsions and stroke.30 Pancreatic lesions caused by COVID-19 include the de novo onset of diabetes mellitus, through direct cell damage to β-pancreatic cells, implicitly numerous ketoacidosis and hyperosmolar coma in critically ill patients and disorders of carbohydrate metabolism.28

Molecular Method of Case Certification

The confirmation of acute respiratory syndrome with the new coronavirus was made by detecting SARS-CoV-2 nucleic acid by RT-PCR in the nasopharyngeal and oropharyngeal exudate. In some severe cases, collection was also performed from sputum or tracheobronchial exudate. Confirmation of the COVID-19 case was performed using the CFX-96 Real-Time IVD System (Bio-Rad).

Viral RNA was extracted using the NIMBUS automatic extractor, using the STARMag 96×4 Universal Cartridge Kit (Seegene) extraction kit and amplified with the Allplex 2019-nCoV kit (Seegene, Seoul, South Korea), with amplification of three viral targets and a gene control: a) RdRP gene (Cal Red 610); b) gena N (Quasar 670); c) gene E (FAM); d) internal control (HEX); 45 amplification cycles were used according to World Health Organization (WHO) recommendations. The samples were considered positive when specific SARS-CoV-2 genes were detected.

Statistical Analysis

Data analysis was performed using IBM SPSS Statistics 20 (SPSS Inc., Chicago, IL). Continuous variables were characterized by median and interval between quartiles (IQR), and category variables were characterized by value and percentage. Data distribution testing was performed with the Kolmogorov–Smirnov test. The numerical variables were compared using the Mann–Whitney U-test for independent samples, and the nominal ones were compared using the chi2 test (Fisher’s exact test). In conditions of statistical significance, an OR/RR>1 (95% CI > 1) was considered a risk factor. The variables that met the criteria of statistical significance in the univariate analysis were investigated by logistic regression, with the choice of the model according to the Nagelkerke R2 coefficient and the test for assessing the deviation from the Hosmer-Lemeshow theoretical model. Predictors of mortality during hospitalization were investigated by Cox proportional hazards regression models. The Kaplan-Meier method with the log-rank (Mantel-Cox) test was used to compare survival. All statistical tests were calculated with 2 extremities and the threshold of statistical significance was considered to be p ≤ 0.05.

Results

Over a period of one year and three months of pandemic, 139 deaths were recorded on the supervised ward, 84 (60.43%) in 2020 and 55 (39.56%) in the first 6 months of 2021, with a peak frequency in the month of December 2020 (Figure 1). Among the control subjects, 148 (53.82%) patients were discharged as cured and 127 (46.18%) were discharged as improved. The univariate analysis of all the variables studied revealed various variables (Table 1), symptoms (Table 2), comorbidities and complications of the disease (Table 3) and paraclinical investigations and treatment (Table 4).

Figure 1

Case-fatality rate and monthly percentage of total COVID-19 deaths between March 2020 and June 2021.

Table 1

Univariate Analysis of Study Variable

Items		EI N1=139	EII N2=275	p	OR [95% CI]
Age [median, IQR]		70.00(20)	70.00(18)	0.245	1.01 [0.99–1.02]
No. hospitalization days [median, IQR]		9.00(11)	11.00(6)	0.617	0.99 [0.96–1.02]
No. ICU hospitalization days [median, IQR]		2.00(9)	0.21(0)	<0.001	1.62 [1.40–1.87]
No. previous days of illness [median, IQR]		6.00(22)	7.00(22)	0.423	0.98 [0.93–1.03]
BMI		30.25(7.26)	29.41(7.12)	0.220	1.03 [0.98–1.07]
F [n, %]		65(46.76)	126(45.82)	0.917	1.04 [0.68–1.60]
M [n, %]		74(53.24)	149(54.18)	0.917	1.04 [0.68–1.60]
U environment [n, %]		97(69.78)	184(66.91)	0.579	1.14 [0.72–1.82]
R environment [n, %]		42(30.22)	91(33.09)
Occupation at risk - Yes [n, %]		5(3.60)	9(3.27)	1.00	1.10 [0.28–3.75]
Smoking status - Yes [n, %]		9(6.47)	23(8.36)	0.563	0.76 [0.32–1.78]
Direct admission to the ICU - Yes [n, %]		17(12.23)	0.00(0.00)	<0.001	40.39 [6.19–1688.70]
From closed communities –Yes [n, %]		7(5.04)	5(1.82)	0.116	2.86 [0.76–11.64]
Transmision	Community [n, %]	6(4.32)	8(2.91)	0.194	/
	Family [n, %]	30(21.58)	78(28.36)
	Unknown [n, %]	102(73.38)	181(65.82)
	Nosocomial [n, %]	1(0.72)	8(2.91)
Clinical form	Medium/severe [n, %]	130(93.52)	185(67.27)	<0.001	7.03 [3.29–15.52]
	Mild [n, %]	9(6.47)	90(32.72)

Abbreviations: EI, sample I (patients who dies); EII, sample II (patients discharged alive); N1/N2, number of patients from EI/EII; p, p-value; OR [95% CI], odds ratio and 95% confidence interval; No., number; [median IQR], [median, interquartile range]; [n, %], number and percentage of item; BMI, body mass index; F, female; M, male; U/R environment, urban/rural environment; ICU, intensive care unit.

Table 2

Univariate Analysis of Symptoms

Symptoms	EI	EII	p	OR [95% CI]
Fever – Yes [n, %]	79(56.83)	165(60.00)	0.597	0.88 [0.57–1.36]
Chills – Yes [n, %]	34(24.46)	84(30.54)	0.207	0.74 [0.45–1.20]
Fatigue – Yes [n, %]	126(90.65)	247(89.82)	0.863	1.10 [0.53–2.33]
Cough – Yes [n, %]	108(77.70)	206(74.91)	0.546	1.17 [0.70–1.95]
Odynophagy – Yes [n, %]	15(10.79)	28(10.18)	0.865	1.07 [0.52–2.17]
Rhinorhea – Yes [n, %]	8(5.75)	11(4.00)	0.459	1.47 [0.52–4.04]
Dyspnoea – Yes [n, %]	129(92.81)	190(69.09)	<0.001	5.77 [2.78–12.31]
Chest pain – Yes [n, %]	16(11.51)	31(11.27)	1.00	1.02 [0.51–2.03]
Myalgia – Yes [n, %]	39(28.06)	94(34.18)	0.222	0.75 [0.47–1.20]
Arthralgia – Yes [n, %]	21(15.11)	44(16.00)	0.887	0.93 [0.51–1.70]
Headache – Yes [n, %]	36(25.89)	120(43.64)	<0.001	0.45 [0.28–0.72]
Nausea – Yes [n, %]	20(14.39)	45(16.36)	0.669	0.86 [0.47–1.57]
Accelerated intestinal transit – Yes [n, %]	9(6.48)	35(12.73)	0.063	0.47 [0.21–1.07]
Vomiting – Yes [n, %]	11(7.91)	26(9.45)	0.716	0.82 [0.37–1.81]
Conjunctivitis – Yes [n, %]	0.00(0.00)	0.00(0.00)	/	/
Skin manifestations – Yes [n, %]	0.00(0.00)	4(1.45)	0.305	0.00 [0.00–2.99]
Irritability – Yes [n, %]	10(7.19)	14(5.09)	0.382	1.45 [0.58–3.58]
Coma – Yes [n, %]	6(4.32)	0.00(0.00)	0.001	14.42 [1.81–652.08]
Anosmia/ageusia – Yes [n, %]	6(4.32)	27(9.82)	0.051	0.41 [0.14–1.06]
Decreased/loss of appetite – Yes [n, %]	38(27.34)	78(28.36)	0.826	0.95 [0.59–1.54]
Vertigo – Yes [n, %]	4(2.88)	16(5.82)	0.187	0.48 [0.11–1.53]
Sweating – Yes [n, %]	6(4.32)	13(4.73)	0.850	0.91 [0.28–2.63]
% CT changes [median, IQR]	60.00(25)	35.00(20)	<0.001	1.08 [1.06–1.09]
% O2 saturation at admission [median, IQR]	86.00(10)	91.00(7)	<0.001	0.84 [0.81–0.88]
No. O2 days on the mask [median, IQR]	4.00(8)	6.00(10)	0.784	1.00 [0.97–1.04]
No. non-invasive ventilation days [median, IQR]	0.00(0)	0.00(0)	<0.001	4.12 [2.02–8.39]
No. Invasive ventilation days [median, IQR]	1.00(4)	0.00(0)	<0.001	/
No. ECMO days [median, IQR]	0.00(0)	0.00(0)	0.160	/

Abbreviations: %, percentage; CT, computed tomography; ECMO, extracorporeal membrane oxygenation.

Table 3

Univariate Analysis of Comorbidities and Complications

Comorbidities – Yes [n, %]	EI	EII	p	OR [95% CI]
Pregnancy	0.00(0.00)	1(0.36)	1.00	0.00 [0.00–77.16]
Confined under 6 weeks	0.00(0.00)	1(0.36)	1.00	0.00 [0.00–77.16]
Hypertension	99(71.22)	186(67.64)	0.501	1.18 [0.74–1.90]
Cardiovascular pathology	58(41.73)	73(26.54)	0.002	1.98 [1.26–3.12]
Previous diabetes mellitus	49(35.25)	82(29.82)	0.266	1.28 [0.81–2.02]
Overweight	29(20.86)	85(30.91)	0.036	0.59 [0.35–0.98]
Obesity	77(55.39)	129(46.91)	0.119	1.41 [0.91–2.16]
Chronic liver pathology	13(9.35)	13(4.73)	0.085	2.08 [0.88–4.94]
Neurological pathology	27(19.42)	34(12.36)	0.077	1.71 [0.95–3.08]
Chronic renal pathology	20(14.39)	16(5.82)	0.005	2.72 [1.29–5.74]
Chronic pulmonary pathology	15(19.79)	39(14.18)	0.359	0.73 [0.37–1.44]
Oncological pathology	12(8.63)	22(8.00)	0.851	1.09 [0.49–2.39]
Autoimmune pathology	5(3.60)	22(8.00)	0.095	0.43 [0.12–1.20]
Psychiatric pathology	10(7.19)	16(5.82)	0.669	1.25 [0.51–3.03]
Coagulopathy	15(19.79)	11(4.00)	0.010	2.90 [1.22–7.00]
Ulcerative Pathology	4(2.88)	4(1.45)	0.450	2.01 [0.37–10.93]
No. Comorbidities [median, IQR]	3.00(2)	3.00(1)	0.002	1.29 [1.11–1.51]
Complications - Yes [n, %]
Decreased blood-oxygenation	137(98.56)	1(0.36)	<0.001	18,769 [1472.91–22,026.47]
Hyperglycemia	110(79.14)	193(70.18)	0.060	1.61 [0.97–2.69]
Pancreatic damage	0.00(0.00)	5(1.82)	0.173	0.00 [0.00–2.15]
Secondary infections	20(14.39)	13(4.73)	0.001	3.39 [1.55–7.49]
Sepsis	79(56.83)	17(6.18)	<0.001	19.98 [10.64–37.95]
Thromboembolism	13(9.35)	10(3.64)	0.022	2.73 [1.09–6.93]
Heart damage	138(99.28)	9(3.27)	<0.001	4078.7 [543.88–22,026.47]
Kidney damage	93(66.91)	50(18.18)	<0.001	9.10 [5.56–14.95]
Liver damage	81(58.27)	126(45.82)	0.022	1.65 [1.07–2.55]
Neurological damage	20(14.39)	7(2.55)	<0.001	6.43 [2.52–18.42]

Table 4

Univariate Analysis of Paraclinical Investigations and Treatment

Paraclinic - [Median, IQR]	EI	EII	p	OR [95% CI]
No. leukocytes	9280.00(7040)	7090.00(4400)	<0.001	1.00 [1.00–1.01]
% neutrophils	86.70(10.10)	79.80(18.70)	<0.001	1.07 [1.04–1.09]
% lymphocytes	7.40(7.30)	13.20(13.60)	<0.001	0.92 [0.88–0.95]
C-reactive protein	112.80(123.08)	50.12(90.53)	<0.001	1.01 [1.00–1.01]
Procalcitonin	0.23(0.77)	0.06(0.11)	0.114	1.15 [0.97–1.37]
Fibrinogen	5.47(2.43)	5.19(2.40)	0.041	1.11 [1.00–1.22]
Eritrocyte sedimentation rate	55.00(62.00)	25.00(40.00)	<0.001	1.02 [1.02–1.03]
D-Dimers	0.96(1.73)	0.57(0.45)	<0.001	1.15 [1.05–1.25]
Serum creatinine	1.07(0.89)	0.88(0.34)	<0.001	2.44 [1.69–3.49]
Alanin aminotransferase	32.30(31.90)	30.00(29.50)	0.666	1.00 [0.99–1.01]
Aspartat aminotransferase	43.00(36.30)	31.10(25.80)	<0.001	1.01 [1.01–1.02]
Glycemia	162.00(112)	133.00(69)	<0.001	1.00 [1.00–1.01]
Interleukin 6	30.90(70.01)	7.10(18.22)	<0.001	1.01 [1.01–1.02]
Treatment - Yes [n, %]
Hydroxychloroquine	0.00(0.00)	0.00(0.00)	/	/
Tocilizumab	19(13.67)	18(6.54)	0.027	2.26 [1.09–4.70]
Corticosteroids	137(98.56)	225(81.82)	<0.001	15.22 [3.87–130.68]
Plasma administration	9(6.47)	2(0.73)	0.001	9.45 [1.91–90.61]
Paracetamol	108(77.70)	222(80.73)	0.518	0.83 [0.49–1.41]
Kineret	53(38.13)	72(26.18)	0.012	1.74 [1.10–2.75]
Clexane	22(15.83)	80(29.09)	0.003	0.46 [0.26–0.80]
Fraxiparine	112(80.57)	172(62.54)	<0.001	2.48 [1.49–4.16]
Without anticoagulant	2(1.44)	18(6.54)	0.022	0.21 [0.02–0.89]
Azithromycin	5(3.60)	31(11.27)	0.014	0.29 [0.09–0.79]
Cefort	95(68.35)	121(44.00)	<0.001	2.75 [1.75–4.32]
Meropenem	24(17.27)	4(1.45)	<0.001	14.14 [4.68–56.95]
Darunavir (Norvir)	47(33.81)	109(39.64)	0.294	0.78 [0.50–1.22]
Favipiravir	23(16.55)	80(29.09)	0.007	0.48 [0.28–0.83]/
Remdesivir	30(21.58)	31(11.27)	0.005	2.17 [1.21–3.89]
Resolsta	14(10.07)	22(8.00)	0.601	1.29 [0.60–2.74]

Logistic regression with all statistically significant variables (excluding those related to treatment) identified three independent risk factors of negative evolution: a) history of coagulopathy HR=30.73 [1.94–487.09], p=0.015; b) high percentage of neutrophils HR=1.09 [1.01–1.19], p=0.027; c) decreased blood-oxygenation HR=53881.97 [1762.24–1,647,489.44], p<0.001 in a model with a Nagelkerke R2 coefficient of 97.3% and the statistical significance of the test for assessing the deviation from the Hosmer-Lemeshow theoretical model p=0.868.

The logistic regression in which only the variables with the status of risk factors (OR/RR > 1 [CI 95% > 1]), statistically significant, were introduced, led to a model that explains 79.8% of the data variation (Hosmer-Lemeshow test p=0.934), in which the following were noted as independent risk factors for death: a) pulmonary changes identified by CT HR=1.05 [1.02–1.08], p<0.001; b) high percentage of neutrophils HR=1.05 [1.01–1.09], p=0.014; c) serum creatinine HR=2.51 [1.64–3.83], p<0.001; d) IL-6 HR=1.01 [1.00–1.02], p=0.004.

Factors influencing the probability of death during hospitalization, identified by Cox regression, were: a) O2 saturation - HR=0.98 [0.96–0.99], p=0.043; b) serum creatinine - HR=1.23 [1.08–1.39], p=0.001; c) history of coagulopathy - HR=2.44 [1.38–4.35], p=0.002; d) dyspnea on admission - HR=2.99 [1.42–6.30], p=0.004; e) hospitalization directly in the ICU - HR=3.803 [1.97–7.33], p<0.001; f) heart damage - HR=16.76 [1.49–188.56], p=0.022; g) decreased blood-oxygenation - HR=35.12 [5.92–208.19], p<0.001.

Kaplan-Meier curves show the factors that influence survival in Figure 2.

Figure 2

Kaplan Meier curves in the case of the main factors influencing survival. (A) Dyspnea on admission. (B) Coagulopathy. (C) Directly admitted to the ICU. (D) Decreased blood-oxygenation. (E). Heart damage.

The quantification of the differences in survival in the case of comorbidities, irrespective of the occurrence of complications during SARS-CoV-2 infection is presented in Table 5.

Table 5

The Impact of Comorbidities and Complications in Survival

Comorbidities	Log-Rank Mantel-Cox p	Onset of Complications	Log-Rank Mantel-Cox p
Cardiovascular pathology	0.004	Heart damage	<0.001
Previous diabetes mellitus	0.212	Hyperglycemia	0.820
Chronic liver pathology	0.045	Liver damage	0.485
Chronic renal pathology	<0.001	Kidney damage	<0.001
Neurological pathology	0.011	Neurological damage	<0.001
Coagulopathy	0.004	Thromboembolism	0.001
Gastro-duodenal ulcer	0.789	Pancreatic damage	0.224
Overweight	0.192	Secondary infections	0.448
Obesity	0.180	Sepsis	0.001
Hypertension	0.468	Decreased blood-oxygenation	0.001
Oncological pathology	0.782
Chronic pulmonary pathology	0.150
Autoimmune pathology	0.235
Psychiatric pathology	0.285

Discussion

The analysis of the first 15 months of the pandemic shows an upward trend in the case-fatality rate, but the R2 value is low due to the important variation in the monthly values, so the trend line cannot precisely predict the fatality value at a given moment. There was a decrease in the values in the spring and an increase in the autumn/winter months, with a peak reached in November 2020. This pattern of seasonality seemed to be repeated in 2021, when in the same section, in the context of wave 4 of the pandemic, the fatality rate reached 22.75% in September, 39.62% in October and 44.70% in November. The same seasonal pattern of COVID-19 severity in Europe, with the involvement of environmental factors (low temperature and/or solar radiation) in mortality variability has also been identified in other studies, along with immunological, social, behavioral factors (frequency of travel, compliance with preventive and control measures, quarantine, School functionality, etc.).31–33

The univariate analysis identified as risk factors for death the following variables: hospitalization directly on the ICU, number of days of hospitalization in the ICU, moderate and severe clinical form, dyspnea/coma at hospitalization, percentage of lung changes on CT, number of days of noninvasive/invasive ventilation, total number of comorbidities, history of cardiovascular/renal pathology, coagulopathy, or the onset of decreased in blood oxygen saturation, secondary infections, sepsis, thromboembolism or cardiac/renal/hepatic/neurological damage. A systematic review and meta-analysis of 42 studies and 423,117 patients showed the association that some chronic comorbidities, including acute kidney injury, COPD, diabetes, hypertension, CVD, cancer, increased D-dimer, complications, demographic variables, male gender, old age, being a current smoker and obesity are clinical risk factors for a fatal outcome associated with coronavirus.34 Furthermore, a series of paraclinical data associated to the clinical form, met the statistical criteria of risk factor: increased number of leukocytes at hospitalization, percentage of neutrophils, increased acute phase reactants, D-Dimers, serum creatinine, blood glucose, AST and IL-6. After analyzing 414 articles on COVID-19, 34 studies selected the following biomarkers for severe forms: CRP, amyloid A, IL-6, lactate dehydrogenase, D-dimer, high-sensitivity troponin and renal markers, leukocytosis, lymphopenia and thrombocytopenia.35

The high saturation of O2 at hospitalization and the high percentage of lymphocytes, but also the presence of headache at hospitalization and overweight, registered statistical values corresponding to some protection factors for the fatal evolution of the case. A retrospective study of 400 American patients confirmed with COVID-19 (31 patients died in hospital), for which machine learning logarithms were applied to predict hospital mortality using a prediction model based on demographic, clinical and paraclinical predictors, concluded that age > 65 years, SO2 < 88%, neurological/ cardiac comorbidities and diabetes are risk factors for increased in-hospital mortality.36

Since the beginning of the pandemic, obesity has emerged as a risk factor for the SARS-CoV-2 infection, for the severe evolution that requires hospitalization, intensive care, mechanical ventilation and increased mortality, due to impaired lung function but also chronic inflammation with impaired response immune and thrombogenic.36–38 While most studies have a linear relationship between BMI and the risk of severe COVID-19 development, in 2021 there were also studies showing a nonlinear curve, in which overweight patients (BMI 25–30 kg/ m2) appear to have the lowest mortality risks.39 In our study, too, overweight appears to be a protective factor against fatal evolution, although the result may be biased by smoking status, alcohol consumption or possibly other confounding factors.

Headache, one of the most common neurological symptoms of SARS-CoV-2 infection, commonly associated with fever, myalgia and arthralgia, possibly associated with an effective antiviral immune response, occurs according to the Centers for Disease Control and Prevention (CDC) in 14.8% of hospitalized patients.40,41 In these patients, the presence of headache is an independent predictor of reduced risk of mortality or hospitalization in the ICU.41–43 The univariate analysis of our study shows the same increased frequency of headache among patients with a favorable outcome.

Following the first logistic regression, three variables were detached as independent factors and predictors of death: history of coagulopathy, high percentage of neutrophils and decreased blood-oxygenation. The SARS-CoV-2 virus increases the risk of venous and arterial thrombosis by inflammation of the vascular endothelium, promoting blood stasis and systemic hypercoagulability. Comorbidities that affect coagulation (for example, atrial fibrillation) and initial coagulation abnormalities that predispose to prothrombotic status are identified as risk factors for adverse evolution of COVID-19 in other studies.44 The second logistic regression model, with only the analysis of risk factors, indicates four independent factors and predictors of death to be all of the paraclinical variables: pulmonary changes identified by CT; serum creatinine, IL-6 and a high percentage of neutrophils.

The Cox regression revealed the following: a) an increase by 1 unit of oxygen saturation decreased the probability of death by 2%; b) an increase by 1 unit of serum creatinine increased the probability of death by 23%; c) the presence of coagulopathy in the antecedents increased the probability of death by 144%; d) the presence of dyspnea increased the probability of death by 199%; f) direct hospitalization in the ICU increased the probability of death by 280.3%; g) heart damage during COVID-19 infection increased the probability by 16.76 times; h) decreased blood-oxygenation increased the probability by 35.12 times.

In the study conducted by Hosse, who conducted an analysis on the CT results of 265 German patients, chronic and extensive lung disease was shown to be associated with an almost twice as high instantaneous risk of death. The CT imaging characteristics of COVID-19 pneumonia showed 74.3% of patients exhibited peripherally accentuated bilateral consolidations -and this was more extensive in patients intubated or hospitalized at ICUs and opacities in “matte glass” at 89.1%, who rarely showed pleural effusions and lymphadenopathy.45 The analysis of 23 studies, with a total of 4631 Chinese patients, showed that elevated levels of CK, CK-MB, troponin T/L, LDH and IL-6 and emerging arrhythmia were associated with the development of severe forms of the disease, hospitalization at the ICU or death.46

Serum creatinine has been shown to be an independent predictor of death, consistent with the results of other studies. Thus, death rates in patients hospitalized for COVID-19 were significantly higher in those with renal impairment, including those with high serum creatinine/urea levels, with proteinuria, and multivariate analysis revealed that renal impairment from SARS-CoV-2 infection is associated with a statistically significant increased risk of death from changes in serum creatinine greater than 50%.47,48

Coagulopathy is an important prognostic factor, being associated in many studies with severe forms and higher mortality in COVID-19 infections.49–51 Bed rest, hypoxia and sepsis associated with the SARS-COV-2, all increase the risk of thrombogenesis. It should be taken into consideration the fact that prerequisite endogen factors that favor blood hypercoagulation may exist.

The differences between the survival curves were extremely statistically significant in the case of history of chronic renal pathology, but also cardiac comorbidity (excluding hypertension), chronic liver disease, neurological pathology and coagulopathy generated significant differences. In terms of complications, cardiac, renal, neurological damage, sepsis and thromboembolism have statistically significantly reduced survival. The results are in line with general scientific knowledge; - thus, according to the European Center for Disease Prevention and Control (ECDC), the risk factors for in-hospital mortality of COVID-19 patients, identified by age and sex-adjusted analysis, were renal pathology, diabetes mellitus, neurological pathology (dementia), heart ischemia/stroke, obesity and oncological pathology.52

In a 12-month longitudinal observational study, which included 244 Italian patients, analysis of the prevalence of comorbidities between survivors and the deceased showed that chronic kidney disease, coronary heart disease and heart failure, dementia and cerebrovascular disease were more common in the deceased patients.53 Additionally, a French cohort study showed that chronic liver diseases increase the risk of death of hospitalized patients for SARS-CoV-2 infection by 80%, while thromboembolic events, occur at a frequency of 20–30% of cases. According to other sources, this number has been shown to be 40–70%, and the risk of mortality was shown to increase by 74% (OR, 1.74; 95% CI, 1.01–2.98; p = 0.04) in a meta-analysis performed on 42 studies with 8271 patients.52,54

COVID-19 infection itself is a viral sepsis because it affects not only the lungs, but also causes cardiovascular, renal, hepatic or CNS dysfunction, because in many cases, the death is due to multiple organ failure, according to autopsies and the presence of RNA. The viral response has been identified in all organs, and the immune response is no different from that of sepsis from other causes. In addition, the need for hospitalization in the ICU of severe cases, namely the risk of mechanical ventilation and consequent bacterial pneumonia, the use of corticosteroids and sometimes antibiotics, generates additional risks of secondary bacterial or fungal infections, with the possibility of generalization. In fact, the prevalence of sepsis among COVID-19 cases was estimated at 39.9% (95% CI, 35.9–44.1) in a first meta-analysis published in December 2020.55

In the published studies, the high values of the evaluation indices of sepsis, septic shock, multiple organ failure, sequential organ failure assessment (SOFA), quick SOFA, acute physiology, chronic health evaluation II and systemic inflammatory response syndrome scores were associated with an increased risk of death from COVID-19, along with lymphocyte counts, PaO2/FiO2, IL-6 and CRP.56,57 In this study, the highest statistical values as a predictor of in-hospital death were obtained for decreased blood-oxygenation. This progression from viral pneumonia to ARDS has been associated with fatal cases, with a significantly higher risk when associated with acute renal failure or septic shock, and in other studies, in which the crude 30-day mortality rate reached 41% (95% CI 38–45%).58,59 In contrast, in our study, neither obesity nor diabetes were found to be risk factors, not even in the bivariate analysis.

The treatment of the viral infection is complex and according to the guidelines includes: antiviral drugs (remdesivir, chloroquine or hydroxychloroquine and/or azithromycin, lopinavir/ritonavir or other HIV protease inhibitors and, nitazozaxides), anti-SARS-CoV-2 monoclonal antibodies, plasma convalescent, SARS-CoV-2-specific and nonspecific immunoglobulin, immunomodulators (colchicine, corticosteroids, fluvoxamine, cell therapy, granulocyte-macrophage colony growth factor inhibitors, Alpha and Beta interferon and Il-1 inhibitors) 6-tocilizumab or sarilumab and kinase inhibitors), antibiotics, anticoagulants, anti-inflammatory drugs (and other symptomatic drugs), vitamin C, D and zinc supplements, as well as support for vital functions.60

The design of the study did not focus on establishing the effectiveness of the therapeutic preparations used. Regarding the therapeutic variables, the statistically significant differences are explained by the differences in the therapeutic protocol applied according to the clinical form of the disease, with the administration of Tocilizumab (humanized monoclonal antibody Ig G1 human antireceptor of Il-6, obtained by the recombinant DNA technology in ovarian cells of Chinese hamsters), plasma, corticosteroids, Remdesivir (when pneumonia requiring oxygen therapy has been combined), Kineret (Anakinra - r-metHULL-Ira, antagonist of human Il-1 receptors, produced in Escherichia cells sheets by recombinant DNA technology), Fraxiparine and antibiotics such as Cefort and Meropenem, significantly more common in more severe forms. At the same time, the administration of Azithromycin (especially at the beginning of the pandemic) and Favipiravir are characteristic of mild or moderate forms, and the lack of anticoagulants means a mild clinical form, which explains their preponderance among surviving patients.

In our study, Enoxaparin showed values of protective factors in bivariate analysis, in line with the results of recent research, while in others unfractionated heparin appeared to be more effective in.61,62 Some researchers claim that heparin, compared to other anticoagulants, improves the survival of the host in COVID-19 and through its direct anti-inflammatory and antiviral effects by controlling the cytokine storm and delaying the progression of the disease. A meta-analysis of 4421 potentially relevant studies (of which only 2107 were identified as relevant) including 11 studies on the impact of anticoagulant treatment on COVID-19 mortality and bleeding events, indicates a reduction in the risk of mortality through anticoagulant therapy, even in critically ill patients admitted to intensive care.63 The results of another prospective cohort study conducted on 6195 patients in 14 hospitals and 60 clinics in the United States show that SARS-CoV-2 is associated with a state of hypercoagulability and that the presence of cardiovascular, immunological and coagulatory comorbidities has been associated with a higher risk of death, especially for inpatients, ICUs and mechanically ventilated and non-ventilated patients.64–66

The Limitations of the Study

The extrapolation power of the results in this study has been reduced due to its unicentric nature. The design did not focus on determining the efficacy of therapeutic preparations, which is why these variables were not included in the multivariate logistic regression or Cox regression. The impact of outpatient therapy was not estimated before hospitalization - in most cases, the symptomatic preparations were similar (antipyretic, antitussive, vitamin C, D and Zn) and the differences (prescription of Azithromycin in the first half of 2020 and oral anticoagulant, sometimes replaced in mild forms with antiplatelet agent) were due to the updating of the therapeutic protocol, which was adopted at the national level. The samples did not include patients with severe immunosuppression (eg, HIV/AIDS infection, organ transplantation) and racial differences could not be investigated because all patients were Caucasian. The vaccine status of the patients in 2021 was not investigated because the percentage of the population vaccinated during that period was very small (23.45% on 30.06.2021) and most of the hospitalized patients were unvaccinated.18

Conclusion

The study identified several main factors influencing the evolution of the cases: history of coagulopathy, oxygen saturation, high percentage of neutrophils at hospital admission, serum creatinine and IL-6 and the increase of the lung injury with decreasing oxygen saturation during hospitalization. Survival was significantly reduced in patients with a history of cardiovascular, renal, liver and neurological diseases, coagulopathies. The outcome was unfavorable in patients who developed thromboembolic events, sepsis, neurologic damage, cardiac or renal failure. A future study investigating the impact of the anti-SARS-COV-2 vaccine on outcomes in COVID-19 patients regarding the vaccine type that is used and the infectious viral strain is needed. Knowledge of these predictors allows for the optimization of triage and therapeutic case management, especially in medical systems with limited material resources.