Baseline characteristics of COVID-19 Italian patients admitted to Desio Hospital, Lombardy: a retrospective study

Abstract The correlation of clinical, radiological and laboratory findings of patients at admission in the Emergency Department (ED) with clinical severity and risk of mortality was investigated. Adult coronavirus disease 2019 (COVID-19) patients hospitalized in March 2020 in Desio Hospital, Lombardy, were retrospectively included in the study, and categorized in terms of disease severity and adverse outcome. Out of the 175 patients enrolled, 79% presented one or more comorbidities, with cardiovascular disease being the most frequent (62%). More than half of the patients showed lymphocytopenia and 20% thrombocytopenia. The patients in the severe group presented higher absolute neutrophil count (ANC), C-reactive protein (CRP), AST, LDH, procalcitonin (PCT) and BUN values compared to the non-severe group (p < .05). Increased odds of mortality associated with older age (OR = 22.43; 95% CI 5.22–96.27), partial pressure of oxygen (PaO2)/fraction of inspired oxygen (FIO2) ratio < 200 (OR = 4.97; 95% CI 1.55–15.84), clinical severity (OR = 21.32; 95% CI 2.27–200.13), creatinine > 106.08 µmol/L (OR = 2.87; 95% CI 1.04–7.92) and creatine kinase > 2.90 µkat/L (OR = 3.80; 95% CI 1.31–10.9) were observed on admission (p < .05). The above findings may contribute to identify early risk factors of poor prognosis, and to select the most appropriate management for patients.


Introduction
The novel coronavirus Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) was first identified in December 2019 in Wuhan, Hubei Province of China and is responsible for coronavirus disease 2019 (COVID-19) [1][2][3].The main routes of transmission are respiratory droplets and human-to-human contacts.A high proportion of infected subjects develops mild symptoms, such as fever, cough and asthenia.However, approximately 20% of the infected individuals presents serious illness, characterized by pneumonia, acute respiratory distress syndrome (ARDS), and in a few cases, multi-organ dysfunction syndrome (MODS) and death [1][2][3].The current estimated crude mortality rate (number of reported deaths divided by reported cases) ranges from 3 to 4% [4].
In January 2020, the COVID-19 outbreak rapidly spread worldwide.On 20 February 2020, the first patient with COVID-19 was diagnosed in Italy, followed by several cases that were confirmed afterwards, especially in the Lombardy region, in northern Italy [5].Up to 13 May 2020, a total of 221,133 people were positive for SARS-CoV-2, in particular 83,298 in Lombardy [6].Severe and critical conditions were found in 16.9 and 2.04%, respectively, calculated out of a total 45,896 cases considered [6].
Italy has been one of the most hard-hit countries by COVID-19 and much still remains to be evaluated and understood.With this in mind, this study aims to assess the correlation of baseline characteristics of a group of Italian COVID-19 adult subjects with illness severity and outcomes.

Study subjects and settings
A retrospective study was conducted including 175 COVID-19 adult patients admitted from 1 March 2020 to 31 March 2020 to Desio Hospital, a 344-bed facility located in Lombardy.Only SARS-CoV-2 cases confirmed through realtime reverse-transcriptase-polymerase-chain-reaction (RT-PCR) assays of nasopharyngeal swabs, in accordance with WHO guidance, were included in the analysis [7].RT-PCR was performed at San Gerardo Hospital, Monza, and subsequently at the Italian National Institute Of Health.Only hospitalized Covid-19 patients with confirmed diagnosis were enrolled in the study.Subjects who did not require hospitalization or were transferred to other hospitals were excluded from the study.
The patients were categorized into two groups in terms of disease severity: non-severe (37 patients) and severe (138 patients).Subjects were classified as severe if they presented one or more of the following symptoms: respiratory rate (RR) !30 breaths/min, finger oxygen saturation (SpO 2 ) 93% at rest and arterial partial pressure of oxygen (PaO 2 )/ fraction of inspired oxygen (FIO 2 ) 300 ratio [8,9].Additionally, the 175 patients were divided into non-survivors (62 patients) and survivors (113 patients), according to mortality outcome.

Data collection
Basic information such as age, gender, comorbidities, clinical and laboratory data, chest radiograph (when performed) and outcome (survival or non-survival) was obtained from the electronic medical records of each patient and stored by a password-protected database.Two researchers independently reviewed the database to verify the correctness of data collection.
The study was approved by the Ethical Committee of the Istituto Nazionale Malattie Infettive Lazaro Spallanzani, Roma 'Observational cohort study on the natural history of hospitalized SARS-COV-2 patients: the STORM trial (Studio OsseRvazionale sulla storia natural dei pazienti ospedalizzati per SARS-COV-2: studio STORM)' of the University of Milano Bicocca.

Clinical laboratory measurements
All clinical laboratory measurements were performed at admission.Complete blood count (CBCs) were performed on whole blood with K3-EDTA vacuum tubes and obtained by Sysmex XN-9000 platform (Sysmex, Norderstedt, Germany) .Blood chemistries were performed on lithium heparin plasma vacuum tubes using COBAS 8000 platform (Roche Diagnostics, Mannheim, Germany).Coagulation tests were determined using Sysmex CS-2500 analyzer (Siemens, Munich, Germany) on plasma samples collected in sodium citrate vacuum tubes.The glomerular filtration rate (GFR) was estimated by CKD-EPI equation [10].

Statistical analysis
Descriptive statistics were obtained for all study variables.Statistical analyses were performed using the STATA software version MP 16 (StataCorp LLC, College Station, TX) [11].Continuous variables were expressed as medians and interquartile ranges (IQRs) and compared with the Mann-Whitney U test; categorical variables were expressed as counts and percentages and compared by v 2 test or Fisher's exact test between non-severe and severe groups, and between survivor and non-survivor groups.
Univariate analysis of demographic and clinical data, as well as laboratory measurands available at admission for each patient was performed to determine the statistical significance of the correlation between these parameters and patient outcomes (survival or non-survival).A two-sided alpha of less than 0.05 was considered statistically significant.
Multivariate logistic regression models were constructed to identify independent demographic, clinical and laboratory factors associated with non-survivors in COVID-19.Based on univariate logistic regression and clinical constraints, the following measurands and parameters were assessed by multivariate analysis: age > 65 (yes 106.08 ¼ 0).Variables from the multivariate analysis were excluded if the number of events was too small to calculate the odds ratios or their between-group differences were not significant.

Demographic and clinical characteristics
A total of 175 COVID-19 patients were enrolled in this study.Demographics and baseline characteristics are shown in Table 1.The median age of the study population was 71 years (IQR 58-80), of whom 105 (60%) were aged over 65.Seventy percent of the patients were males and 30% females.Seventy-nine percent of the subjects presented one or more comorbidities, cardiovascular disease being the most common (62%).The most frequent symptoms at admission were dyspnea, fever and cough.The median time from onset of symptoms to hospital admission was 7.0 d.Of the 175 patients, 23% was admitted to the intensive care unit (ICU).

Laboratory findings
The hematological and biochemical characteristics of the patients are shown in Table 2. CBC showed leukocytosis in 18% and leukopenia in 14%.Furthermore, lymphocytopenia (55.4%) and thrombocytopenia (20%) were observed.
Compared to the patients in the non-severe group, those in the severe group had higher absolute neutrophil counts (ANCs), C-reactive protein (CRP), and procalcitonin (PCT) (p < .01).Additionally, the severe patients had significantly greater enzymatic (AST and LDH) and BUN values (p < .05).

Radiologic findings
The findings of individuals who had a chest radiography performed at admission in the ED (n ¼ 165) were evaluated.Of these patients, only seven did not present any alteration on the chest x-ray.The lack of alterations was more frequent in the non-severe group when compared to the severe one (11 vs. 2%; p ¼ .03).The most common finding was bilateral pneumonia (64%), with a higher prevalence in the severe than the non-severe group (69 vs. 46%; p ¼ .02).Less frequently, the radiologists described ground-glass opacity (26%), unilateral pneumonia (21%), interstitial pneumonia (12%) and pleural effusion (9%).No statistically significant difference was observed between the chest x-rays of survivors and non-survivors (p > .1).

Univariate and multivariate analysis
For all demographic, clinical and laboratory data (Tables 1  and 2), each statistically significant variable between the non-survivor and survivor group was evaluated using univariate analysis.Older age (> 65 years), presence of comorbidity, PaO 2 /FIO 2 ratio < 200 (moderate and severe ARDS), RR > 20/min and clinical severity were significantly associated with higher risk of mortality (p < .01)(See Supplementary Table 1).Moreover, the following laboratory measurands: neutrophilia, lymphopenia, neutrophil/lymphocyte (N/L) ratio higher than 3.53, anemia, elevated inflammation-related indices (CRP and PCT), elevated serum enzymes (AST, LDH and CK) and creatinine were associated with higher mortality rate in COVID-19 patients.
The above variables were assessed using a multivariate logistic regression model to identify independent prognosis indicators of COVID-19 patients.One hundred and fortythree patients (45 non-survivors and 98 survivors) with    .0001 APTT, ratio .0005  3).

Discussion
Comprehensive data analysis on the demographic, clinical, laboratory and radiological characteristics and outcomes of 175 hospitalized patients with non-severe and severe COVID-19 are presented.At the moment of the study, Italy was the second most affected country worldwide.The severe cases of COVID-19 in this study were by far the most prevalent and were characterized by older patients with more than one comorbidity, as previously reported [12,13].
Comorbidities, such as cardiovascular diseases (including hypertension) and diabetes, were associated with a 3.9-and 2.8-fold increase in poor prognosis, respectively.The most common symptoms found at admission were dyspnea (64%), fever (52%) and cough (46%).However, almost half of the patients were afebrile; indicating that this sign may not be a feature of the early stage of the disease, as recently described [14].Thus, it seems that body temperature below 37.5 C is not an efficient rule-out criterion.
Eleven percent of the non-severe patients and 2% of the severe did not show any apparent radiological alterations, underlining the importance of a deeper investigation of these cases with epidemiological history, clinical symptoms, laboratory and CT, as previously described by Guan and coauthors [14].In accordance with previous data, the most common chest x-ray finding was bilateral pneumonia [14,15].
Among severe patients and those with an adverse outcome, leukocytosis was more frequent than leukopenia, differently from other previous studies [14,16].Guan and coauthors reported that leukopenia was present in 33.7% of 1099 patients with laboratory-confirmed COVID-19 [14].Leukocytosis could reflect an excessive inflammation, and the finding of a higher CRP level in patients with severe COVID-19, as reported by Zhang and coauthors [17], may support this.Non-survivor patients presented a significantly higher neutrophil count and a higher N/L ratio, in accordance with other studies [14,16,18].The non-survivors of this study more frequently developed lymphopenia, as also reported by other authors [18,19].
Hospital mortality increased with PaO 2 /FIO 2 ratio 200, which corresponds to moderate or severe ARDS [20].Villar and coauthors reported that severe hypoxemia reduced lung compliance, and bilateral radiographic pulmonary infiltrates might explain the progression from ARDS to death of COVID-19 subjects [21].It was found that high creatinine and CK levels were associated with in-hospital mortality, making this data crucial for the timely identification of patients at higher risk of adverse outcome.In agreement with our data, Cheng and coauthors found that elevated baseline serum creatinine, baseline blood urea nitrogen, proteinuria and hematuria (acute kidney injury) were independent risk factors for in-hospital death [22].
Elevated creatinine levels in COVID-19 patients affected by acute kidney injury were observed, as other studies reported [22][23][24].The increased creatinine levels were significantly different in the non-survivor group vs. the survivor group.
The enrolled patients of this study usually presented normal levels of procalcitonin, while CRP results were always elevated, suggesting that these inflammatory indexes are important during the differential diagnosis process of a suspected COVID-19 case, as confirmed in others works [18,25,26].
Univariate analysis indicated that preexisting comorbidities increase the risk of death.This is in agreement with other results reported in the literature [16,27].The data from this study confirm that the odds of mortality are 20fold higher in patients aged over 65 years, as already described by other authors [13].The age-related immune dysfunctions, resulting from low-grade chronic inflammation [27][28][29] and multi-organ failures in older patients due to the scarce ability to compensate hypoxia could be a possible explanation [12].Multivariate regression analysis revealed that being classified, as a severe COVID-19 case is a significant risk factor associated with death.This confirms the importance of lung damage and consequent respiratory failure as one of the main determinants for not survival.In our multivariate analysis, age remained one of the significant predictors of in-hospital mortality, as confirmed by Bonetti and coauthors [18].
This study presents some limitations that should be considered.First, the study is retrospective and was performed in a single hospital setting.A large multicenter cohort study could improve our findings.Our laboratory data were not complete for some of the suggested tests making this another limit of the study.
In conclusion, this study suggests that older age, moderate or severe ARDS at presentation, laboratory abnormalities (creatinine, CK) and disease severity can be used to predict patients which are at higher risk of poor prognosis, in agreement with the studies conducted on COVID-19 Italian patients [18,29].

Table 1 .
Demographic and clinical characteristics of COVID-19 patients at admission in ED.

Table 2 .
Laboratory findings of COVID-19 patients at admission in ED.

Table 3 .
Multivariate regression: risk factors associated with in-hospital mortality.: confidence interval; z: regression coefficient divided by its standard error; PaO 2 /FIO 2 : arterial partial pressure of oxygen divided by fractional inspired oxygen concentration For the definition of severe patient, see Materials and Methods. CI