Characteristics, outcomes, and risk factors for in-hospital mortality of COVID-19 patients: A retrospective study in Thailand

Introduction Data on the characteristics and outcomes of patients hospitalized for Coronavirus Disease 2019 (COVID-19) in Thailand are limited. Objective To determine characteristics and outcomes and identify risk factors for hospital mortality for hospitalized patients with COVID-19. Methods We retrospectively reviewed the medical records of patients who had COVID-19 infection and were admitted to the cohort ward or ICUs at Siriraj Hospital between January 2020 and December 2021. Results Of the 2,430 patients included in this study, 229 (9.4%) died; the mean age was 54 years, 40% were men, 81% had at least one comorbidity, and 13% required intensive care unit (ICU). Favipiravir (86%) was the main antiviral treatment. Corticosteroids and rescue anti-inflammatory therapy were used in 74 and 6%, respectively. Admission to the ICU was the only factor associated with reduced mortality [odds ratio (OR) 0.01, 95% confidence interval (CI) 0.01–0.05, P < 0.001], whereas older age (OR 14.3, 95%CI 5.76–35.54, P < 0.001), high flow nasal cannula (HFNC; OR 9.2, 95% CI 3.9–21.6, P < 0.001), mechanical ventilation (OR 269.39, 95%CI 3.6–2173.63, P < 0.001), septic shock (OR 7.79, 95%CI, 2.01–30.18, P = 0.003), and hydrocortisone treatment (OR 27.01, 95%CI 5.29–138.31, P < 0.001) were factors associated with in-hospital mortality. Conclusion The overall mortality of hospitalized patients with COVID-19 was 9%. The only factor associated with reduced mortality was admission to the ICU. Therefore, appropriate selection of patients for admission to the ICU, strategies to limit disease progression and prevent intubation, and early detection and prompt treatment of nosocomial infection can improve survival in these patients.


Introduction
COVID-19 (Corona virus disease 2019) is an emerging disease declared by the WHO (World Health Organization) as a Public Health Emergency of International Concern in January 2020. Since the first case reported from China in December 2019, more than 500 million people have been infected worldwide, with an overall mortality rate of 1.14% (1), a mortality rate 17-28% for hospitalized patients, and 49-60% for mechanically ventilated patients (2)(3)(4)(5)(6)(7)(8)(9)(10). Risk factors for increased mortality from previous studies included older age, pre-existing medical illnesses, high Sequential Organ Failure Assessment (SOFA) score, receipt of IMV, acute respiratory distress syndrome (ARDS), and elevated D-dimer (2,6,9,11,12). However, most data were reported from China and developed countries, while the epidemiological data for COVID-19 diseases from lowand middle-income countries are lacking. The limitation of healthcare resources and treatment capacity may impact patient outcomes. This study aimed to determine the characteristics and outcomes and identify risk factors for in-hospital mortality for hospitalized COVID-19 patients in Thailand.
Protection Unit of Siriraj Hospital Faculty of Medicine, Mahidol University. The requirement for written informed consent was waived.
In Thailand, the COVID-19 pandemic has occurred since January 2020. All COVID-19 patients who require admission to our hospital are treated in cohort wards and intensive care units (ICU) with modified airborne infection isolation rooms (AIIR). The patients requiring vasopressors, high flow nasal cannula (HFNC), non-invasive ventilation (NIV), or invasive mechanical ventilation (IMV) were admitted to the ICUs when beds were available. Guideline for critical care management of severe COVID-19 patients in our hospital has been previously published (13)(14)(15). Moreover, the recommendation for antiviral and anti-inflammatory therapies for COVID-19 patients in Thailand have been changed regularly according to the updated published data (16,17). Since the publication of the preliminary report of the Randomized Evaluation of COVID-19 Therapy (RECOVERY) trial (18), dexamethasone has been recommended for all patients with a resting oxygen saturation of < 96% or a reduction in oxygen saturation of ≥ 3% after exercise-induced desaturation or those requiring respiratory support. However, before the publication of the RECOVERY study, some critically ill patients who had rapidly worsening respiratory symptoms with evidence of hyperinflammation and were less likely to have superimposed bacterial infections received corticosteroid therapy as previously described (13).
In this study, we investigated all adult patients aged ≥ 18 years diagnosed with COVID-19 and admitted to the cohort wards and ICUs at Siriraj Hospital from January 2020 to December 2021. The diagnosis of COVID-19 disease was confirmed by detecting severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) from any respiratory specimens by reverse transcription-polymerase chain reaction (RT-PCR). Patients who were readmitted due to diagnoses other than reinfection of COVID-19 were excluded.

Data collection
We initially obtained data from electronic medical records (EMRs) in collaboration with Siriraj Informatics and Data Innovation Center (SiData +) and then manually reviewed the EMRs of all patients to verify the diagnosis of COVID-19 infections and recorded the data that could not be collected electronically.
The data collected included demographic data of the patient, comorbidities, vital signs, and laboratory results at admission, antiviral and anti-inflammatory medications used for the treatment of COVID-19 disease, admission to the ICU, the requirement for renal replacement therapy (RRT) and respiratory support, including oxygen therapy, HFNC, NIV, IMV. The respiratory rate oxygenation index (ROX), the ratio of oxygen saturation measured by pulse oximetry (SpO 2 )/FiO 2 to respiratory rate, was also calculated in all patients (19).
Patient outcomes included survival at hospital discharge, ICU and hospital length of stay, complications during admission, including ARDS according to the Berlin Definition (20), nosocomial infections, septic shock, venous thromboembolism, and bleeding complications.

Statistical analysis
Categorical variables were presented as numbers and column percentages, while continuous variables were presented as mean with standard deviation (SD), or median with 25th and 75th quartiles, as appropriate. Categorical variables were compared using Chi-square or Fisher exact tests, while continuous variables were compared using the t-test or Mann-Whitney U-test, as appropriate. A multivariate logistic regression model with a stepwise forward method was used to identify independent clinical risk factors associated with inhospital mortality. We report risk factors with odds ratios (OR) and 95% confidence intervals (CI). All p-values < 0.05 were considered statistical significance. All statistical analyses were performed using IBM SPSS Statistics version 18 (SPSS, Inc., Chicago, IL, USA) (21).

Study patients
From January 2020 to December 2021, 2,900 patients with SARS-CoV2 viruses detected by RT-PCR were admitted to the cohort wards and ICUs of Siriraj hospital. Of these patients, 470 were excluded due to readmission for diagnoses other than COVID-19 reinfections; the remaining 2,430 patients were included in the final analysis ( Figure 1); 2,201 (91%) survived and 229 (9%) died. In Thailand, the first case of the delta variant was reported in December 2020, and the first case of the omicron variant was found in December 2021. In our cohort, most patients (n = 2,329) were admitted from December 2020 to December 2021. The mortality rates changed over time; the highest mortality (14%) occurred at the same time as the Study population. SARS-CoV-2, severe acute respiratory syndrome coronavirus-2.
highest admission rate (1,204 hospitalized patients) in July-September 2021 (Figure 2). Only 101 patients were admitted between January and April 2020; no mortality was observed in that period, and no COVID-19 patient was hospitalized from May to November 2020.
Overall in-hospital mortality was 9%; the mortality rate increased to 31% (187 of 605) in patients requiring HFNC or NIV, and 56% (99 of 178) in patients requiring IMV. The mortality rate for patients admitted to the ICU was 24% (74 of 307). Deceased patients were admitted more frequently to the ICU and received antiviral and anti-inflammatory medications, respiratory support, and RRT than survived patients ( Table 3). Nosocomial infections (6.4%) were the most common hospital complications, followed by ARDS (4.7%), septic shock (4.1%), pulmonary embolism (1.4%), and bleeding complications (0.6%). All complications occurred more frequently in patients who did not survive than in those who survived.

Discussion
In this large cohort of patients hospitalized for COVID-19 infections in Thailand, the overall mortality rate was 9%; the rate was higher in patients requiring admission to the ICU (24%) and IMV (56%). In multivariate analysis, admission to the ICU was the only factor associated with a reduction in mortality. Patients who were elderly, required high-level respiratory support, especially those with a lower ROX index, had renal dysfunction and thrombocytopenia at admission, were complicated by septic shock, and received hydrocortisone were at high risk of mortality in hospital.
A better understanding of the natural history of the disease and the availability of antiviral and anti-inflammatory therapies can lead to a reduction in mortality over time. Prior studies were carried out during the early stage of the pandemic, January-April 2020, in which information on the natural course and effective treatment strategies for emerging diseases was limited (3,4,6), while our study period was longer (January 2020-December 2021) and included the later stage of the pandemic when more scientific knowledge about COVID-19 was available (18, [22][23][24][25][26][27]. The difference in mortality of hypoxemic COVID-19 who did not require mechanical ventilation at admission was also found in two randomized controlled trials conducted in different periods. The RECOVERY trial revealed the effectiveness of dexamethasone in COVID-19 and was carried out between March and June 2020. That trial reported that the mortality at 28 days was 23% in the dexamethasone group and 26% in the usual care group (18). The Adaptive COVID-19 Treatment Trial 4 (ACTT-4) compared baricitinib and dexamethasone in combination with remdesivir for the treatment of hospitalized patients with COVID-19 was carried out from December 2020 to April 2021. ACTT-4 reported that mortality at 60 days was only 7 and 8% in the baricitinib and dexamethasone groups, respectively (28). The different variants of the virus and the availability of COVID-19 vaccines may be the other important reasons for the decreased mortality in our cohort (29). A prior study also reported a lower fatality of COVID-19 patients during the outbreak of the omicron variant compared to that of the delta and beta variants (30). At the end of 2021, 63% of the Thai population had already received at least two doses of we also demonstrated the shortage of cohort ICU beds in our hospital, since almost one-third of our patients received HFNC or IMV, but only 13% of all patients were treated in the ICU.
Therefore, increasing the capacity of the ICU with AIIR is an essential plan to prepare for the next pandemic. Previous studies have reported an association between the degree of hypoxemia and increased mortality; our results supported this finding (4,9,11,37). The lower ROX index and use of respiratory supports represented more severe hypoxemia in our patients. Since the risk of death was much higher in patients requiring IMV than in those requiring HFNC and lowflow O 2 therapy (Table 4), strategies to limit disease progression and prevent intubation, including early antiviral treatment for patients at high risk of disease progression, and rescue antiinflammatory therapy with tocilizumab or baricitinib for those who had rapid worsening despite corticosteroid treatment, appear reasonable (25-27, 38).
Hydrocortisone is the only type of corticosteroid found to be associated with mortality in this study (   (39,40). We found that septic shock was associated with high mortality in the multivariate model. This association between hydrocortisone and mortality could be explained by the fact that in our hospital hydrocortisone was used primarily in patients with septic shock who received high-dose vasopressor treatment. An observational study in Argentina also found that septic shock and refractory hypoxemia were the major causes of death in patients with COVID-19 requiring IMV, and the use of vasopressors on admission was an independent predictor of hospital mortality (10), the rate of nosocomial infection in our cohort was 6% (155 of 2,430) and 65% (100 of 155) of these patients developed septic shock ( Table 3). Therefore, early detection and treatment of superimposed bacterial infections are essential to improve patient outcomes. The associations between increased mortality and older age, thrombocytopenia, and renal dysfunction on admission in patients with COVID-19 have been reported in previous studies (2,4,8,11,(41)(42)(43), but older age is not modifiable, and thrombocytopenia and renal dysfunction appear to be attributable to the severity of the disease (44-47). Thrombocytopenia, defined as platelet counts < 150 × 10 3 /mm 3 , is common in COVID-19 disease. The proposed pathophysiological mechanisms include reducing platelet synthesis from the bone marrow due to direct viral infection or cytokine storm, increasing platelet destruction by the immune system, and increased platelet consumption resulting from aggregation and formation of microthrombi in damaged lung parenchyma and pulmonary endothelial cells (48). The possible pathophysiology of acute kidney injury (AKI) in COVID-19 consists of direct effects of the virus that causes endothelial and tubular epithelial damage and indirect impact from volume depletion, nephrotoxic drugs, sepsis-associated AKI from superimposed bacterial infection, increased renal venous pressure complicated from mechanical ventilation with high intrathoracic pressure, organ crosstalk from lung injury or cardiorenal syndrome, and rhabdomyolysis (46,47). Patients with any of these factors need close observation for severe disease.
The strengths of our study are that we reported the characteristics and outcomes of a relatively large cohort of hospitalized patients with COVID-19 and performed a multivariate analysis to determine the independent factors associated with in-hospital mortality. We also performed manual searches on the EMR to verify the diagnosis of COVID-19 infection in all patients and to obtain the data that could not be collected electronically. We also acknowledge several limitations of this study. First, this is a single-center study, which may limit the generalizability of our findings. Second, we did not have data on underlying chronic lung disease, SOFA score, and D-dimer level, which increased mortality risks in previous studies. Third, concurrent life-threatening medical or surgical diseases necessitating urgent treatment were not recorded; these conditions may impact patient outcomes. Fourth, we were unable to obtain clinical outcomes after hospital discharge. Finally, similar to other observational and retrospective studies, unmeasured confounders may affect the results.

Conclusion
The overall mortality of hospitalized COVID-19 patients in Thailand was 9%. Admission to the ICU was the only protective factor. Patients who received invasive or noninvasive respiratory support and were complicated by septic shock had the highest risk of death. Therefore, appropriate patient selection for admission to the ICU, strategies to limit disease progression and prevent intubation, and early detection and prompt treatment of nosocomial infection may improve survival. Although unmodifiable, elderly patients and those with thrombocytopenia and renal dysfunction need close observation for the development of severe disease.

Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement
The studies involving human participants were reviewed and approved by the Human Research Protection Unit of Siriraj Hospital Faculty of Medicine, Mahidol University (Certificate of Approval no. Si 335/2020). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

Author contributions
TN had full access to all data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis, designed the study, performed manual searches for additional data collection and interpretation, drafted and revised the manuscript. TV designed the study, reviewed, interpreted the data, drafted, and revised the manuscript. ST designed the study, performed the statistical analysis, drafted, and revised the manuscript. AD, TP, RW, RR, PP, PT, AP, and CP assisted in data collection, data interpretation, and critically reviewed the manuscript. All authors have read and approved the final manuscript and agreed to be responsible for all aspects of the work.