Risk factors for mortality in hospitalized moderate-to-severe COVID-19 patients: A single-center retrospective study
Leyla Bahadorizadeh1
, Soheil Rahmani Fard2, Arman Karimi Behnagh3, Seyed Ali Hashemi Kiapey3, Seyed Hassan Saneii4, Sara Minaeian2
1 Antimicrobial Resistance Research Center, Institute of Immunology and Infectious Diseases; Firoozabadi Clinical Research Development Unit (FACRDU), Iran University of Medical Sciences, Tehran, Iran
2 Antimicrobial Resistance Research Center, Institute of Immunology and Infectious Diseases, Iran University of Medical Sciences, Tehran, Iran
3 Firoozabadi Clinical Research Development Unit (FACRDU); Department of Internal Medicine, School of Medicine, Iran University of Medical sciences, Tehran, Iran
4 Department of Basic Sciences of Rehabilitation, Rehabilitation Research Center, School of Rehabilitation Sciences, Iran University of Medical Sciences, Tehran, Iran
Correspondence Address:
Dr. Sara Minaeian
Antimicrobial Resistance Research Center, Institute of Immunology and Infectious Diseases, Floor 3, Building no. 3, Hazrat-e Rasool General Hospital, Niyayesh St, Sattar Khan St, 1445613131, Tehran
Iran
Source of Support: None, Conflict of Interest: None
DOI: 10.4103/iahs.iahs_66_21
Aims: The aim of this study was to determine the mortality risk factors of hospitalized moderate-to-severe coronavirus disease 2019 (COVID-19) patients in Tehran. Materials and Methods: We retrospectively evaluated the baseline characteristics and clinical and paraclinical parameters of 223 deceased and discharged patients who were hospitalized in Firoozabadi General Hospital between March 1 and April 1, 2020. Results: According to our multiple logistic regression model, advanced age (odds ratio [OR], 1.06; 95% confidence interval [CI], 1.02–1.09; P < 0.01), reduced oxygen saturation (OR, 0.92; 95% CI, 0.87–0.96; P < 0.01), and hypertension (OR, 2.59; 95% CI, 1.04–6.46; P = 0.04) can be perceived as independent risk factors for mortality. Conclusion: Our results suggest that patients with older age, lower oxygen saturation, and hypertension are predisposed to an increased risk of mortality. Thus, to lower the COVID-19 mortality rates, patients with these characteristics should be the primary targets for early treatment, vaccination, or monitoring strategies.
Keywords: Coronavirus disease 2019, mortality, risk factor, SARS-CoV-2
Coronavirus disease 2019 (COVID-19), which first emerged from China in December 2019,[1] has since grown to be one of the deadliest pandemics in recent history and as of February 21, 2021, has infected more than 110 million and caused more than 2.4 million deaths.[2] Despite rigorous worldwide efforts to contain the pandemic, it is still spreading throughout the world and no definitive treatment is available.
SARS-CoV-2, the virus responsible for COVID-19, is a member of coronavirus family (the same virus family responsible for SARS and MERS outbreaks) and uses angiotensin-converting enzyme 2 (ACE2) for cell entry.[3] General symptoms are fever, cough, and fatigue,[4] but because of the wide distribution of ACE2 throughout the body and the multifactorial nature of COVID-19 pathogenesis encompassing inflammation, endothelial function, and coagulation,[5] far more diverse clinical symptoms have been reported showcasing cardiac, kidney, liver, digestive tract, and nervous system injuries.[6]
Unfortunately, the current high COVID-19 mortality rates in Iran show that a lot of areas regarding screening, monitoring, treatment, and preventive strategies are in dire need of improvement. Analyzing the epidemiological, clinical, and laboratory characteristics of COVID-19 patients is crucial for screening and treatment strategy development and could be very helpful in early identification of patients who are predisposed to a higher risk of mortality. This ensures that the medical and national resources are being used with utmost efficiency. Studies show that COVID-19 mortality risk factors include old age, hypertension, obesity, and decreased oxygen saturation.[7],[8] As similar studies are scarce regarding the local population,[9],[10] the current study was designed to evaluate the differences in epidemiological, comorbidities, clinical, and laboratory characteristics between deceased and discharged COVID-19 patients that were admitted to Firoozabadi General Hospital in Tehran between March and April of 2020.
Materials and MethodsThe present study was a cross-sectional study that has been performed on 223 hospitalized cases of COVID-19 who were admitted to Firoozabadi General Hospital (Tehran, Iran) from March 1 to April 1, 2020. Patients' relevant data were extracted from the hospital digital records. For the purposes of this study, based on the clinical outcome, subjects were divided into two groups: (1) discharged and (2) deceased.
Patient inclusion criteria were based on national guidelines and were consisted of having three or more of the following symptoms: (1) cough, (2) weakness, (3) fever of ≥38.5°C, (4) intense fatigue, (5) myalgia, (6) sore throat, (7) dyspnea, (8) low appetite/diarrhea/nausea, and (9) decreased awareness or having one or more of the following disease characteristics: (1) oxygen saturation value of <93%, (2) disease confirmation based on chest imaging results, and (3) a respiratory rate >24 breaths per min.
Disease severity classification was also based on national guidelines. Moderate disease severity was designated as having one or more of the following: (1) presence of respiratory symptoms (dyspnea and feeling of pain or pressure in chest), (2) oxygen saturation value between 90% and 93%, and (3) lung involvement of <50%. Severe disease severity was designated as having one or more of the following: (1) rapid progression of respiratory symptoms (dyspnea in particular), (2) respiratory rate of more than 30 breaths per min, (3) oxygen saturation value of <90%, and (4) lung involvement of more than 50%.
All selected patients had a positive SARS-CoV-2-specific reverse transcription polymerase chain reaction test result, and patients with negative test results were excluded from the study.
This study was reviewed and approved by the Scientific Advisory and Ethical Committees of Iran University of Medical Sciences (registration number: IR.IUMS.REC.1399.013) and conforms to the principles of Helsinki declaration.
Clinical, laboratory, and epidemiological data of the patients were collected including age, gender, hospitalization period, blood pressure, respiratory rate, body temperature, peripheral oxygen saturation, erythrocyte sedimentation rate, C-reactive protein (CRP), complete blood count, prothrombin time, partial thromboplastin time, creatinine, potassium, magnesium, alanine transaminase (ALT), aspartate aminotransferase, and troponin. Cell counts and biochemical analysis were conducted using a Sysmex KX21N (Sysmex, Kobe, Japan) and Hitachi 912 Autoanalyzer (Roche Diagnostics GmbH, Germany), respectively.
Patient chest imaging results and baseline comorbidities, including hypertension, diabetes mellitus, ischemic heart disease, cerebrovascular accident (CVA), lung diseases, and malignancies, were also collected. Chest imaging was conducted using a 16-row spiral computed tomography (CT) scanning device (Siemens AG). Any documented additional or necessary administered drugs and therapeutic procedures were also collected from the hospital digital database.
Descriptive statistics (e.g., mean and standard deviation or median and interquartile range [IQR] and simple proportions) were used to describe the data. The Kolmogorov–Smirnov test was used to assess the presence of normal distribution in data. According to the presence of normal distribution, the Student's t-test or Mann–Whitney U test were used to compare the mean scores between different subgroups. To compare proportions between different groups, the Chi-square or Fisher's exact test was used. Furthermore, the multiple logistic regression model was used to assess effective factors on the final status of COVID-19 cases after treatment (death, discharge). P < 0.05 was considered a statistical significance level. All analyses were performed using IBM SPSS Statistics for Windows, version 22 (IBM Corp., Armonk, N.Y., USA).
ResultsTwo hundred and twenty-three moderate-to-severe confirmed COVID-19 patients were enrolled in this study. The study population were predominantly male (132 [59.2%]), and the mean age was 58.52 ± 18.1 years. As this study was conducted in the early days of the pandemic, the mortality rates were quite high, and in our study population, 56 (25.1%) patients had perished. Analysis of basic characteristic data revealed that patients in the deceased group were significantly older than patients in the discharged group (68.96 ± 14.31 vs. 55.01 ± 17.91 years, P = 0.01). Furthermore, respiratory rate at admission date was also higher in the deceased group (19.11 ± 5.14 vs. 17.55 ± 3.70 breaths per min, P = 0.02) [Table 1].
Table 1: Baseline characteristics, comorbidities, and admission symptoms of moderate-to-severe hospitalized coronavirus disease 2019 patientsIn our study population, 57% of patients were suffering from one or more underlying condition including diabetes mellitus (32.7%), hypertension (31.8%), ischemic heart disease (20.2%), lung disease (8.5%), CVA (3.1%), and cancer (2.7%). Univariate analysis showed that hypertension (27 [48.2%] vs. 44 [26.3%], P = 0.01) and cancer (5 [8.9%] vs. 1 [0.6%], P = 0.01) were more frequent in the deceased group. In addition, a higher number of patients in the deceased group experienced dyspnea (49 [87.5%] vs. 119 [71.3%], P = 0.01) and neurological (15 [26.8%] vs. 23 [13.8%], P = 0.02) symptoms. Furthermore, weak myalgia (17 [30.4%] vs. 78 [46.7%], P = 0.03) and gastrointestinal symptoms (6 [10.7] vs. 38 [22.8], P = 0.05) were less frequent in the deceased group compared to the discharged group [Table 1].
Analysis of paraclinical test results revealed that in comparison to the discharged group, patients in the deceased group had higher neutrophil (8067.26 ± 4325.97 vs. 4983.43 ± 3090.20, P = 0.01), total white blood cell (9797.35 ± 4709.90 vs. 6882.38 ± 3230.75, P = 0.01), and neutrophil-to-lymphocyte ratio (9.86 ± 8.48 vs. 5.56 ± 6.82, P = 0.01). Despite that there was no significant difference in lymphocyte count between the study groups, lymphopenia (<1.1 × 109/L) was much more frequent in the deceased group (32 [65%] vs. 61 [43%], P = 0.01). These patients also had lower oxygen saturation (80.67 ± 14.23 vs. 90.69 ± 6.62, P = 0.01) on admission date and were more likely to have coagulation abnormalities (P = 0.04) [Table 2]. Furthermore, analysis of the existing laboratory test results revealed that patients in the deceased group were more likely to have higher CRP (65 median [IQR 86–31] vs. 35 median [IQR 63-14], P = 0.01), troponin (51.25 median [IQR 119.15–20.05] vs. 6.2 median [IQR 23-1.6], P = 0.01), and aspartate transaminase (AST) (53 median (IQR 99-37) vs. 32.5 median [IQR 64.75-25.25], P = 0.01) values. Although not reaching clinical significance, similar to AST, ALT was also generally higher in the deceased group (35 median [IQR 63-21] vs. 25 [IQR 43-15.25], P = 0.06), which is indicative of significant liver injury [Table 3]. In addition, no differences were observed in the frequency of chest imaging abnormalities between deceased and discharged groups [Table 2].
Table 2: Paraclinical test results and chest imaging abnormalities of moderate-to-severe hospitalized coronavirus disease 2019 patientsTable 3: Paraclinical test results of moderate-to-severe hospitalized coronavirus disease 2019 patientsThe results revealed that a higher number of deceased patients received corticosteroids (Cortone) (9 [16.1%] vs. 12 [7.1%], P = 0.05). Furthermore, the need for invasive oxygen support was much higher in the deceased group compared to the discharged group (49 [89.1%] vs. 15 [10.9%], P = 0.01) [Table 4].
Table 4: Medications and supportive procedures administered during hospitalization periodIn addition, the distribution of patients receiving no antibiotic treatment or receiving one, two, or three antibiotics was not significantly different between the study groups (three different types of antibiotics (56 patients [33.5%] vs. 27 patients [48.2%]) and no antibiotic treatment (26 patients [15.6%] vs. 5 patients [8.9%]) being the highest and lowest proportions for discharged and deceased groups respectively, P = 0.08). Similarly, no differences in the distribution of patients receiving no antiviral treatment and patients receiving single or combinational antiviral therapy were observed in the study groups (Kaletra/Tamiflu 91 [54.5%] vs. 31 [56.4%] and Kaletra/ribavirin [1 (0.6%) vs. 0] being the highest and lowest proportions for discharged and deceased groups respectively, P = 0.28) [Table 4].
Furthermore, the number and proportion of patients receiving intravenous immune globulin (0 vs. 2 [3.6%] for discharged and deceased groups, respectively, P = 0.06) and chloroquine (121 [72.5%] vs. 35 [62.5%], P = 0.16) were not statistically different between the study groups.
Based on initial symptoms and clinical characteristics of our study population, 86 patients matched the severe disease profile and 137 patients matched the moderate disease profile. Data analysis based on disease severity revealed that severe patients were older (62.65 ± 17.22 vs. 55.92 ± 18.21 years, P = 0.01), and a lower proportion of them suffered from CVA (0% vs. 5.1%, P = 0.05). Comparison of initial symptoms showed that more patients in the severe group suffered from dyspnea (86% vs. 68.6%, P = 0.01), and their respiratory rates were higher compared to the moderate group (19.36 ± 5.10 vs. 17.06 ± 3.14, P = 0.01) [Supplementary Table 1].
In addition, peripheral oxygen saturation was significantly lower in the severe group (79.84 ± 11.82 vs. 93.38 ± 2.51, P = 0.01), while white blood cell count, neutrophil count, and neutrophil-to-lymphocyte ratio (8952.44 ± 4700.93 vs. 6724.49 ± 2873.80, P = 0.01; 7209.49 ± 4556.96 vs. 4979.59 ± 2855.79, P = 0.01; and 8.77 ± 9.43 vs. 5.26 ± 5.50, P = 0.01, respectively) were found to be significantly higher compared to the moderate group. Moreover, abnormalities in coagulative processes were more frequent in the severe group (P = 0.04) [Supplementary Table 2]. Analysis of laboratory data revealed that patients in the severe group had higher troponin levels compared to the moderate group (27.65 median [IQR, 7.25–84] vs. 6.9 median [IQR, 1.75–27.45]) [Supplementary Table 3].
Furthermore, patients in the severe group were more likely to need invasive and noninvasive supplementary oxygen support (P = 0.01) and succumb to the disease (46.5% vs. 11.7%, P = 0.01) [Supplementary Table 4].
The results of the multiple logistic regression model suggest that advanced age (odds ratio [OR], 1.06; 95% CI, 1.02–1.09; P = 0.01) and hypertension (OR, 2.59; 95% CI, 1.04–6.46; P = 0.04) can increase the chance of mortality in COVID-19 patients. Furthermore, we found that oxygen saturation has a reverse relationship with mortality as a higher oxygen saturation reduced the chance of mortality in our study population (OR, 0.92; 95% CI, 0.87–0.96; P = 0.01) [Table 5].
Table 5: Multiple logistic regression model of risk factors for mortality in moderate-to-severe hospitalized coronavirus disease 2019 patients DiscussionCurrently, countries all around the world are struggling to deal with the devastating effects of COVID-19 pandemic.[11],[12] It has spread to more than 110 million people around the world with no sign of slowing down. There has been a great deal of concern regarding the high mortality rates of COVID-19 in Iran in the recent months, and identifying the risk factors in the local population can be very useful in early identification and treatment of patients whom are predisposed to a higher risk of mortality.
In the current study, out of the 223 moderate-to-severe COVID-19 patients that were included in the study, 56 have died. Analysis of epidemiological data revealed that patients in the deceased group were older and more frequently suffered from comorbidities such as hypertension and cancer. Furthermore, these patients more frequently experienced neurological symptoms and dyspnea and in contrast less likely to have gastrointestinal symptoms and myalgia upon admission. In addition, they had lower peripheral oxygen saturation and higher neutrophil-to-lymphocyte ratio and during the hospitalization period were more likely to need invasive supplementary oxygen support. Our Multiple Logistic Regression Model showed that age, peripheral oxygen saturation, and hypertension can act as independent risk factors for mortality.
Advanced age is probably the most well-known risk factor for COVID-19 mortality as it has been consistently reported as such in almost every risk assessment study.[7],[8] Although the exact underlying mechanisms responsible for higher mortality rate in older population are still not fully understood, but it has been suggested that this might be the result of cumulative effects of multiple different phenomena such as endothelium and clotting functions, immunosenescence, inflammaging, comorbidities, and mitochondrial dysfunction in older patients.[13],[14] A recent study conducted by Ho et al.[15] found that advanced age can increase the mortality risk by 13-fold. The authors further reported that even in older populations with no other underlying risk factors, the mortality rate is still 4 times higher than the general population. Our results also demonstrate that patients who have died due to COVID-19 complications were significantly older that patients who survived. These evidence corroborate the fact that older patients should be the priority in monitoring, therapeutic, or vaccination strategies.
Another well-documented characteristic for severe COVID-19 is reduced oxygen saturation. This is due to SARS-CoV-2 binding to ACE2 receptors in lungs and respiratory tract. Hypoxia progression can result in increased viral proliferation, cytokine storm, intravascular coagulation, and pulmonary hypoxic vasoconstriction, which are all characteristics of disease progression and are associated with higher rates of fatal outcome.[16] Comparably, patients in our study with reduced oxygen saturation were at a higher risk of mortality. In these patients, mitigating hypoxia through oxygen supplementation therapy may reduce the destructive consequences of hypoxia and consequently help reduce the mortality rates.
In the present study, mortality was much higher among patients suffering from hypertension. This is in line with several large-scale studies reporting a significantly higher mortality rate, intensive care unit admission, and severe disease course among patients with preexisting hypertension.[17],[18],[19] The exact mechanisms in which hypertension exacerbates the COVID-19 disease are not fully clear yet. Although renin–angiotensin–aldosterone system inhibitors were first thought to be the culprit by overexpressing the ACE2, it has since been proven by multiple studies that these drugs in fact reduce the mortality rate of COVID-19 patients with preexisting hypertension.[19],[20] As older patients are more susceptible to hypertension and other comorbidities, timely identification and intervention is vital in controlling the disease severity.
Several other studies have found that neurological symptoms,[21],[22],[23],[24] gastrointestinal symptoms,[25],[26] cancer,[27],[28] and changes in raw count and proportion of lymphocytes and neutrophils[29],[30] could also act as predictors for COVID-19 severity and mortality; thus, based on these findings and our own univariate analysis results, we strongly recommend that these symptoms and conditions not be overlooked and should be thoroughly investigated in patients so that appropriate treatments can be administered in time.
This study has several limitations and they should be considered before any interpretations of our results. First, although some of the risk factors that were found to be clinically significant in our univariate analysis were dismissed as independent risk factors in multivariate analysis; this might be duo to our relatively small study population and their importance should not be disregarded. Second, some laboratory tests were not performed for the entirety of the study population and thus were forcefully discarded as they would have damaged the integrity of our statistical analysis. Third, patients in the current study all suffered from moderate-to-severe COVID-19; this has limited the capacity of our statistical models to also assess the mortality risk factors in patients with mild disease. Fourth, patient readmissions were not recorded which may have produced bias in our mortality rate analysis.
ConclusionIn this single-center retrospective cohort study, we found out that older age, decreased peripheral oxygen saturation, and hypertension are independent risk factor for COVID-19 mortality in moderate-to-severe patients. Early identification and prioritizing treatments for these patients may help lower the COVID-19 mortality rates.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References
留言 (0)