I. Introduction

Corona Virus Disease 2019 (COVID-19) is an infectious disease caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Since the discovery of the first group of cases in Wuhan City, Hubei Province, in China on December 31^st, 2019, the initial COVID-19 outbreak has become one of the most significant global health threats.

It was declared a public health emergency of international concern on January 30^th, 2020, by the World Health Organization (WHO) and a pandemic on March 11th, 2020 [1, 2]. By 2021, COVID-19 had affected almost all countries around the world.

According to the WHO situation reports, as of October 4^th, 2022, the number of confirmed COVID-19 cases reached more than 615,777,700 cases and 6,527,192 deaths (1.06% case- fatality rate) [3]. In the African region, the first case of COVID-19 was reported on February 14^th, 2020, in Egypt and on February 17^th, 2020 in sub-Saharan Africa (Nigeria). Since then, all 54 African states have been affected. As of September 21^st, 2021, the WHO African Region has recorded 8,166,634 cases and 206,740 deaths (2.53% case-facility rate) [4–7]. The Democratic Republic of Congo (DRC) recorded its first case on March 10^th, 2020. Since then, the cumulative number of cases up to September 3^rd, 2022 was 92,942 cases, of which 92,940 were confirmed, with 1,357 deaths, representing a case-fatality rate of 1.5% [8]. According to the epidemiological report in North Kivu, as of may 22^nd, 2022, the province recorded 10,049 cases, of which 9,144 recovered (91.0%) and about 600 died (5.9%). These figures place the province in second place after the provincial city of Kinshasa in terms of disease burden [9]. On the same date, the report indicates that the city of Butembo, with its two health zones, had 1280 cases, of which 1157 (90.4%) recovered and 116 (9.1%) died. These two health zones concentrated 20% of the province’s mortality [9].

The signs and symptoms of COVID-19 vary considerably, with clinical features ranging from asymptomatic presentation to fatal respiratory distress and multiorgan failure [10, 11]. COVID-19 does not exert an equal impact on the different affected regions in the world, with a wide variation in the reported proportion of patients with severe disease and death [12, 13]. Morbidity accounts for 6% of cases in the Eastern Mediterranean region, 30% in Europe, 39% in the American region, 19% in South Asia, 3% in the Pacific and 4% in Africa [4, 14]. The consequences are palpable worldwide at all levels with significant economic, social, health and financial costs [15].

Strategies to prevent COVID-19 transmission have included border closures, the creation of quarantine centers, self-isolation, enforcement of containment measures, surveillance, routine screening, and, recently, vaccination, with the first vaccine deliveries to Africa in March 2021 [15].

Studies were done worldwide to determine COVID-19 risk factors and patient characteristics and recently the impact of vaccination on hospitalization [1, 2, 6, 15]. Overall, risk factors for severe disease and death included advanced age, male sex, history of comorbidity, poor nutritional status, a severe form of the disease on admission, vaccination status, delay in admission to intensive care, etc. To our knowledge, by the time we conducted our research in 2021, such a study had not yet taken place in the region. Therefore, our study proposed to identify the predictors of mortality of COVID-19 patients who were hospitalized in the city of Butembo from January 1^st to 31 December 31^st, 2021.

II. Methods

This study was conducted in six COVID-19 treatment centers (CTC) in the Katwa and Butembo health zones of the city of Butembo, North Kivu province, Democratic Republic of Congo. The CTCs were located at the Hospital of Women Committed to the Promotion of Integral Health, the Kitatumba Referral General Hospital, the Graben University Clinics, the Katwa Referral General Hospital, the Matanda Hospital and the Ngote Hospital.

A retrospective cohort study was conducted from 1 January to 31 December 2021.The study population included all patients hospitalized in one of the CTCs in the city of Butembo in 2021 with a positive reverse transcriptase-polymerase chain reaction result (RT-PCR +). The statistical units were the COVID-19 patients registered in each health center. We excluded patients without investigation forms or whose clinical records could not be found. In total, 303 Covid-19 patient records were found. Delay between symptom onset and hospital admission was the exposure variable. Patients were divided into two groups: 188 patients hospitalized for ≤ 5 days (Short delay) and 115 patients hospitalized for> 5 days (long delay). Data was extracted by the authors who are all medical doctors.

Variables

The time to event (death) was the outcome variable. Predictor variables were age, sex, occupation, clinical stage of disease based on signs and symptoms, ambient air oxygen saturation (SpO2), respiratory rate, NEWS-modified score (National Early Warning Score), history of comorbidities, number of comorbidities, types of comorbidities, body mass index (BMI), duration of disease before hospitalization, date of discharge or last news about the patient. NEWS comprises seven parameters: respiration rate, SpO2, any supplemental oxygen, temperature, systolic pressure, heart rate, level of consciousness (Table 1). It allows the classification of COVID-19 patients into three severity categories: low (aggregate 1–4), medium (aggregate 5–6), and high (aggregate 7 or more). Because some parameters (blood pressure, supplemental oxygen) in the patient’s records were missing, we modified the scoring of some parameters to obtain the NEWS-modified score with five parameters. The time variable was calculated from the date of enrolment (admission at the hospital) to the end point (death, discharge or end of the study).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Score NEWS–adapted from the NEWS2 scoring matrix.

https://doi.org/10.1371/journal.pgph.0002020.t001

III. Results

There were 303 participants who had a mean age of 53 ± 22 years. Those who consulted within 5 days of symptom onset were younger than those who consulted ≥ 5 days (51 vs. 57 years; p = 0.014). Overall, 43% were at least 60 years old, 46% were male, and 6.4% were healthcare professionals. The distribution of patients in the population from which the two groups were drawn was identical according to sociodemographic characteristics, except for age and site (Matanda and Katwa particularly) (Table 2).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Socio-demographic characteristics and clinical signs of Covid-19 patients on admission to the Covid-19 treatment centers in the city of Butembo from January 1^st, to December 31^st, 2021.

https://doi.org/10.1371/journal.pgph.0002020.t002

Of 303 patients included, 45.9% had at least one comorbidity of whom 25.2% had at least two comorbidities. The most frequent comorbidities were hypertension 18.8% (57/303) and diabetes 17.8% (54/303). The distribution of patients in the ≤ 5 days (short) and > 5 days (long) hospitalization groups was similar in terms of comorbidities (Table 2).

Cough and asthenia were present in more than 80% of cases, anorexia, headache and dyspnea in 50–79% of cases, and fever, digestive, and taste disorders and other signs in less than 50% of cases. The distribution of patients between the short and long-hospitalization groups differed in terms of symptoms, except for cough, headache, loss of taste, and digestive and cardiac disorders (Table 2).

The mean oxygen saturation on admission was 85.60% ± 12.72% (81.91 ± 14.29 in the exposed (> 5 days (long) hospitalization group) and 87.85 ± 11.11 in the unexposed (in the ≤ 5 days (short) hospitalization group)). Patients admitted early had higher oxygen saturation than those who visited late. Patients with a severe form accounted for 61% of all participants based on the m-NEWS score, 49% based on their SpO2 level and 55% based on symptoms at admission. Regardless of the classification used, the distribution of patients between the short and long-hospitalization groups differed according to disease severity upon admission (Table 3).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Clinical stage of inpatients on admission to Covid-19 treatment centers in the city of Butembo from January 1 to December 31, 2021.

https://doi.org/10.1371/journal.pgph.0002020.t003

The fatality rate was 33.8 deaths per 1000 patient-days. The rate was 46 deaths per 1000 patient-days in the exposed group (> 5 days (long) hospitalization group) versus 25.6 deaths per 1000 patient days in the unexposed group (in the ≤ 5 days (short) hospitalization group) (Table 4).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Predictors of mortality of Covid-19 inpatients in the city of Butembo from January 1 to December 31, 2021.

https://doi.org/10.1371/journal.pgph.0002020.t004

After adjustment, four predictors were independently associated with inpatient death: older patients (≥ 60 years) had a higher risk of death than the reference group(<40 years) (adjusted HR: 9.90; 95% CI: 2.68–36.27) (see Fig 1 and Table 4), having at least one comorbidity had a higher risk of death than patients without comorbidity (aHR: 11.39; 95% CI: 3.19–40.71) (see Fig 2 and Table 4); patient admitted after “long delay” between symptom onset and hospital admission duration of illness of > 5 days before hospitalization had a higher risk of death than patients admitted after “short delay” (aHR:1.70, 95% CI: 1.04–2.79) and Patients with SpO2 < 90% had a higher risk of death than patients with SpO2 ≥ 95% (HRa = 14.02, 95% CI: 2.23–88.32) (Table 4). The interaction age and comorbidity was significant and was included in the model.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Survival of hospitalized COVID-19 patients by age and time from disease onset to admission.

https://doi.org/10.1371/journal.pgph.0002020.g001

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Survival of hospitalized COVID-19 patients by presence of comorbidity and disease stage (based on saturation) at admission.

https://doi.org/10.1371/journal.pgph.0002020.g002

IV. Discussion

This study focused on the predictors of survival of Covid-19 patients hospitalized from 1 January to 31 December 2021 in the city of Butembo. Advanced age above 60 years, low free air SpO2 on admission, history of comorbidity and late admission (> 5 days after symptom onset) were predictors of mortality. Overall, these predictors have been previously reported in studies conducted in China, Europe, the USA and other African countries [2, 4, 10, 17–27]. Another new predictor introduced in the present study was the duration of illness > 5 days before hospitalization. Knowledge of these factors is an important advocacy tool for covid-19 patients in humanitarian and resource-limited settings. The use of NEWS-modified or original NEWS score, the measurement of oxygen saturation and search for co-morbidities must be systematic. These parameters may help to select the most at-risk patients with a poor prognosis and to intensify surveillance. In almost all hospitals in Africa in general, and in the DRC in particular, the measurement of oxygen saturation was not systematically used of times due to the absence of an oximeter.

Our study found that the risk of death increases with age. Compared to patients under 40 years of age, the risk of death is four times higher in patients > 60 years. It is well established that older age groups are more likely to have comorbidities and are more susceptible to acquire SARS-CoV-2, severe forms of the disease, and higher mortality risk than younger patients [28]. The difference between the age groups lies in the ability to fight the infection [29]. Several studies have proposed multiple pathogenic mechanisms for severe disease in the elderly population, including low levels of angiotensin-converting enzyme 2 (ACE2) in the elderly [30], age-related difficulty in clearing particulate matter from small airways [31], excessive release of inflammatory mediators in the elderly ("inflammaging") [32]. This finding is similar to recent retrospective studies from Egypt [33], Iran [27] and Pakistan [34] and Kinshasa [35–38]. This is also consistent with other reports from outside the Eurogio Meuse-Rhine (EMR) countries [39, 40].

Our study also revealed that the risk of death for patients with a history of at least one comorbidity (the most common being hypertension and diabetes) is more than double. Indeed, the high expression of ACE2 receptors in highly differentiated airway epithelial cells underlies susceptibility to SARS-CoV-2. When hypertensive patients are infected with the SARS-CoV2, blood pressure regulation becomes more complicated and difficult to control, and the cardiovascular risk is amplified [27, 41]. Diabetic patients are more likely to have severe Covid-19 complications because their high blood glucose levels promote viral growth and impair their immune function and ability to resist infection, setting the stage for secondary bacterial and viral co-infections. In addition, if diabetes complications occur, the risk of multiorgan failure and death is significantly increased. One study reported that mortality and multiorgan injury were significantly higher in Covid-19 patients with type-2 diabetes than in patients without diabetes (HR, 1.5). Such an association has been demonstrated by numerous studies worldwide, reminding us of the need to control chronic non-communicable diseases and pay special attention to people with comorbidities during epidemics [23, 26, 27, 41–50].

Regarding SpO2 at admission, our study demonstrated the risk of death for patients admitted with SpO2 < 90%, which is more than 13 times compared to patients with normal saturation (≥95%). Low ambient air SpO2 is strongly associated with poor outcomes on admission to the hospital and, therefore, is part of the classification of disease severity [47, 51]. Indeed, the degree of ambient air O2 saturation reflects the level of the respiratory system function [38]. Such observation has been made by other authors and reiterates the need to take into account SpO2 levels in the prioritization of patients for admission to intensive care units [35, 38, 39, 44]. The wide confidence interval for SpO2 is due to the small number of events in the reference group, we have only 2 cases of death in mild severe patients.

In our study, delay in consulting was significantly and independently associated with mortality, with a risk of more than doubling. Indeed, in all infectious diseases, the duration of symptoms before hospitalization has an impact on the outcome. In fact, delayed onset of supportive care affects clinical outcomes by enabling immuno-inflammatory and thrombotic responses [44]. This result mirrors the findings of other studies, such as that of Wen-Hua Liang et al. in China [52].

On the other hand, our study did not establish a significant association between gender and death as in the study of Cummings MJ et al. [44]. Similarly, Nasiri et al. showed that there was no significant difference between men and women in terms of admissions to intensive care units [53]. Another Egyptian study of 260 patients with COVID-19 showed no significant association between male gender and risk of severe disease [54]. In Italy, Ciceri et al. reported no significant association between the female gender and the risk of severe disease [55]. Other reports from the Eurogio Meuse-Rhine countries demonstrated similar findings [27, 33, 37]. Although recent surveillance has shown that male patients were more likely to acquire severe infection and die with COVID-19 [56], This sex-specific difference was observed in previous SARS outbreaks [57].

The interpretation of the results of the present study should consider some of its limitations. First, although we tried to be as complete as possible, the retrospective nature of this study limited the availability of data due to issues of archiving records, quality of the data found in the patient records, and possible errors in assessing patient parameters at admission. Second, record incompleteness, and the limited capacity in diagnosing other comorbidities could have biased the outcome as we could not input more predictors that could have explained mortality. The fact that we did not have the opportunity to analyze the impact of COVID vaccination in our study is another limitation. Regarding COVID-19 vaccines, the roll-out started on April 19th, 2021. However, the uptake was slow initially, by the time of data collection, only about 7% of the target population (entire population >18) had received at least one dose in the entire country. This low coverage may affect case fatality. This may have also substantial impact on study results as not all cases may have ended up to the hospital thereby creating reporting bias of the outcome.

Finally, because the date of symptom onset is based on self-report, a bias related to the patient’s recollection may have misclassified patients at admission. However, to reduce the risk of error, we linked the different sources of data (patient and investigation forms, registers, and databases). We ensured the streamlining of clinical assessment and equipment used to collect patient parameters upon admission in each CTC.

Despite these limitations, this study has nevertheless the merit of having shown a new factor in the arsenal of predictors of survival of patients hospitalized for COVID-19 in the city of Butembo from January to December 2021. These results can also be used to prioritize actions according to the characteristics of the population and the triage of patients upon admission to the health facilities in the city of Butembo and in similar resource-limited contexts.

V. Conclusions

Advanced age, history of comorbidity, low level of ambient air SpO2 at admission and long duration of the disease before hospitalization were predictors of survival of patients hospitalized for COVID-19 in Butembo. Therefore, the following actions are recommended: (a) health policymakers and authorities must strengthen barrier measures and make them mandatory to protect the elderly and those with a history of comorbidity; (b) health authorities, along with technical and financial partners, should incentivize symptomatic patients to seek care at dedicated health structures as early as possible; and (c) health care providers should measure ambient air SpO2 as part of criteria for admission to intensive care units.

Supporting information