Clinical Risk Factors for Long-term Hospital Stay in Common Patients with coronavirus disease 2019(COVID-19)

This study aimed to investigate the potential risk factors associated with hospital stay in mild patients with COVID-19. Methods laboratory the and compared between and stay. Univariable and multivariable logistic regression methods were used to explore the risk factors associated with


Results
Of 109 COVID-19 patients, 61 patients were short-term stay (≤ 10 days) and 48 patients were long-term stay (> 10 days). The average age of patients in short-term stay were younger than those long-term stay(P = 0.01). Hypertension was the most common comorbidity (34%, 21/61), followed by diabetes (15%,9/61) and Cardiopathy (8%, 5/61). Fever and cough were the typical clinical manifestation in two group. Decreased WBC, Hemoglobin and increased Monocyte, MLR (Monocyte Lymphocyte ratio) and Hypersensitive CRP showed a long-term stay (all P < 0.05). The treatment of Resochin and Human immunoglobulin had a shorter hospital stay. Multivariable regression showed that MLR and CRP on admission were risk factors for predicting the hospital stay, with the HR (hazard ratio 2.03, 1.02-5.39; P = 0.022) and (1.32,1.05-3.24, P = 0.045) respectively.

Conclusions
The potential risk factors of MLR and CRP may help clinicians to predict the hospital stay of COVID-19 patients.

Background
The 2019 novel coronavirus (SARS-CoV-2) disease (COVID-19) [1][2][3] is a new and highly contagious respiratory disease, initially outbreaking in the end of 2019 from Wuhan China and quickly spread throughout the world [4][5]. Due to its extremely infectivity, as of 24:00 on April 14,2020, a total of 1932760 con rmed cases of COVID-19 and a total of 115676 deaths have been reported [6], causing a pandemic worldwide.
To better understand the disease, systematic and effective diagnosis and treatment, several diagnosis and treatment plans and guidelines have been reported [7][8], and the median hospital stay of forty-seven discharged patients was 10 days [9]. Some studies indicated that patients' maximum lung involved peaked at approximately 10 days from the onset of initial symptoms in Wuhan [10]. meanwhile, patients outside Wuhan with symptoms longer than 10 days were less severe than those in Wuhan [11]. Therefore, the hospital stay in COVID-19 patients is one of the prognostic indicators, the estimation of risk factors for hospital stay is important for preventing transmission and facilitating diagnosis to establish the guidelines and criteria.
Here, 109 cases of the hospital with complete diagnostic data that meet discharge standards from January 16 to March 15 of 2020, including clinical features and laboratory examination was included to explore the risk factors of hospital stay for COVID-19 patients.

Study cohort
This retrospective study was approved by the institutional review board of the hospital and the written informed consents were waived. All hospitalized 109 patients were laboratory con rmed COVID-19 according to WHO interim guidance. Laboratory con rmation of SARS-CoV-2 was performed in certi ed tertiary care hospitals. RT-PCR assays were performed in accordance with the protocol established by the WHO [12]. The diagnostic criteria and discharge criteria were in line with "The new coronavirus pneumonia treatment program (Trial seventh edition)" issued by the Ministry of Health on March 3, 2020 [13]. The inclusion criteria: a). Positive COVID-19 nucleic acid test (RT-PCR method); b). Admission in the hospital from the rst time since onset to discharge; c). The clinical classi cation was common; d). no lymphatic system disorders or other hematologic diseases to ensure a normal baseline values of blood parameters. Patients with acute coronary syndromes, renal or hepatic failure or systematic in ammatory diseases were excluded. In total, 48 COVID-19 patients suffered from long-term hospital stay and the other 61 short-term patients were collected. The mean age in long-term stay was 55.1 years (ranging from 22 years to 81 years).

Laboratory Procedures
PCR re-examination every other day after clinical remission of symptoms (fever, cough, et al) by throatswab or sputum specimens. The criteria for discharge were absence of fever for at least 3 days, Pulmonary imaging showed signi cant improvement in acute exudative lesions, Two consecutive negative respiratory tract nucleic acid tests (sampling time at least 24 hours apart) [13]. The hospital stay was the days from admission to discharge (average, 12 days, 6 ~ 28 days). In this study, the optimal cutoff days of hospital stay was 10 days according to previous studies [9][10][11], and then the patients were classi ed into short-term hospital stay (≤ 10 days) and long-term hospital stay (> 10 days).

Statistical Analysis
All statistics were analyzed with SPSS (version 25.0, IBM Crop. Armonk, NY, USA) and Medcalc (version 19.0, https://www.medcalc.org). Continuous variables were analyzed using student t test (mean ± standard deviation) and categorical variable with Chi-square (number(percentage)) test, as appropriate to compare differences between short-term and long-term hospital stay. Different parameters between groups(P < 0.05) were further assessed by univariable and multivariable logistic regression(stepwise). Only independent risk factors with the hospital stay at univariate analysis were included in the multivariable model. The results were shown with hazard ratio (HR) with 95% con dence intervals (CI). The predictive accuracy of the independent factors was quanti ed by the area under the receiver operator characteristic (ROC) curves (AUC). The statistical signi cance levels were all two-sided with statistical signi cance set at 0.05.
Five treatment strategies were used in the hospital, including Resochin, Lopinavir, Arbidol, Recombinant Human Interferon, Hormonotherapy and Human immunoglobulin ( Table 2). Lopinavir (100%) and recombinant Human Interferon (almost 100%) were most used in two groups. More patients were treated with Resochin and Human immunoglobulin in short-term hospital stay compared to long-term stay(P < 0.001). So, signi cant differences between the groups were found for patient's age, blood routine markers (WBC, Hemoglobin, Monocyte and MLR), treatment scheme (Resochin and Human immunoglobulin), the mean time from illness onset to COVID-19 negative, 7 days in short-term stay and 20 days in long-term stay.
The mean time from onset to discharge was 9.0 days in short-term stay and 21.5 days in long-term stay ( Table 2).
Identifying Independent Risk Factors for short-term and long-term hospital stay Clinical variables including age, blood routine markers (WBC, Hemoglobin, Monocyte, MLR and CRP), were included in the univariable analysis. As shown in Table 3, all above parameters except Hemoglobin were independent predictors associated with higher risk of long-term hospital stay. Multivariate logistic regression analysis showed that MLR and CRP were signi cantly correlated with long-term stay for COVID-19 patients (HR = 2.03, 95%CI = 1.02-5.39; P = 0.022 and HR = 1.32, 95%CI = 1.05-3.24, P = 0.045, respectively, Table 3).

Discussion
This retrospective cohort study identi ed several risk factors for long-term hospital stay with COVID-19 patients, which suggested the feasibility of this new methodology as a potential tool for COVID-19 evaluation in clinical outcome. In particular, higher MLR and CRP on admission were associated with the higher risk of longer hospital stay in COVID-19 patients with initial diagnosis of common type disease at admission. Additionally, older age, decreased WBC, more treatment scheme of Resochin and Human immunoglobulin were more commonly seen in long-term hospital stay.
Previously, older age has been reported as an important independent predictor for COVID-19 death [14], due to the stronger host innate responses to virus infection leading to de ciency control of viral replication and more prolonged proin ammatory responses, which may lead to poor outcome [15]. The white blood cell slightly decreased due to the lymphocyte abnormality after the novel coronavirus infection by invading the human body immune system, which may lead to a longer hospital stay. The antiviral treatment Resochin recommended by Chinese guidelines [13] and Human immunoglobulin by improving the immunity might improve clinical outcomes for shorter hospital stay in COVID-19. The time from illness onset to COVID-19 negative in long-term stay (21.5 days) was longer than short-term group (9.0 days), which consistent with the discharge time.
In this study, MLR was signi cantly higher in long-term hospital stay patients than short-term group. Previous evidence demonstrated that monocytes were susceptible to human coronavirus (HCoV)-229E infection by strongly restricted OC43 replication [16]. After infection by HCoV-OC43, monocytes still could alive for more than 6 days and no apoptosis [17]. These studies suggested that monocytes might be stable in function during HCoV infection . Meanwhile, SARS-CoV often attacks cytotoxic T lymphocytes [18][19]. Lymphopenia is one of hematological abnormalities, which could predict the severity and clinical outcomes of COVID-19 patients [20]. Previous study showed that lymphocytes were signi cantly decreased in SARS patients, with more remarkable CD4 + and CD8 + lymphopenia in severe clinical illness or died patients [20]. Similar with SARS-CoV, COVID-19 infection also was associated with lymphocytes loss, it was supported by Chinese guidelines [13]. So, the MLR increased especially in long-term stay patients with a predicting AUC value of 0.72. Further research needed to be continued.
C-reactive protein (CRP) is an indicator of systemic in ammatory response. Lobo SM [21] reported that higher CRP levels re ected a stronger in ammatory response, may be associated with severe acute respiratory illness (SARI) evolving to acute respiratory distress syndrome (ARDS) and the death in patients with in uenza A (H1N1) viral pneumonia. Therefore, it is not surprising that CRP was identi ed as an independent risk factor for severe condition in COVID-19 patients [22][23][24], which suggested that elevated CRP early in the course maybe a potential biomarker associate with risk of longer hospital stay. In this study, AUC of CRP was 0.67. Future research with multicenter and more sample size is needed to be researched.
The ndings of this study should be interpreted in the light of potential limitations. Firstly, the small sample size is relatively small. Second, only common type of COVID-19 was included, and severe type and mild types were excluded in this analysis. Third, the prognosis values for MLR and CRP didn't have powerful performance, a combination of clinical parameters and CT ndings are suggested in clinical practice. Four, the clinical and biochemical signs was retrospective collected from January 16 to March 15, treatment drugs and regimens were immature, which may in uence the result. Large-scale independent prospective multicenter study is needed for further analysis and validating the results.

Conclusion
We found several risk factors for long-term hospital stay with COVID-19 patients. higher MLR and CRP on admission were associated with the higher risk of longer hospital stay in COVID-19 patients with initial diagnosis of common type disease at admission. The information from the current study can be used to might assist clinicians in making appropriate decisions and optimizing the use of hospital resources, recognize those with high risk of longer hospital stay even at the rst outpatient visit, and facilitate early implementation of more appropriate interventions to decrease the risk of longer stay in COVID-19 patients initially diagnosed with common disease.

Declarations
Availability of data and material The datasets generated and/or analyzed during the current study are not publicly available due to an IRB decision which was made in the interest of ensuring patient con dentiality but are available from the corresponding author on reasonable request. Figure 1 The ROC curves of MLR for diagnosis short-term and long-term hospital stay in COVID-19 patients.