Mortality of Patients With Severe COVID-19 in the Intensive Care Unit: An Observational Study From a Major COVID-19 Receiving Hospital

Objective To determine the mortality of patients with severe COVID-19 in the intensive care unit (ICU) in relation to age, gender, co-morbidities, ventilatory status, and length of stay (LOS). Methods This was a cross-sectional study based on data retrieved for 204 patients admitted to the ICU of Hayatabad Medical Complex, Peshawar, Pakistan, from April to August 2020. Study variables were age, gender, co-morbid conditions, ventilatory status, and length of stay (LOS). The data were analyzed using SPSS version 21 (IBM Corp., Armonk, NY). The independent t-test and the chi-square test were used to compare the means and frequencies of variables. Multivariate regression analysis was used to predict the likelihood of mortality. Results The overall mortality was 77%. Non-invasive ventilation (NIV) was administered to 61.8% of patients. Mortality was higher for invasive mechanical ventilation (IMV) (93.6% vs 66.7%, p<0.001) and for over 60 years (87.3% vs 72.3%, p=0.019). Mortality without co-morbidities was 75.2%. Comparative mortality rates for at least one co-morbidity (79.7%), diabetes mellitus (80.0%), hypertension (100%), diabetes mellitus and hypertension both (87.1%), and chronic obstructive pulmonary disease (75%) were insignificant. The LOS for survivors was longer (8.9±8.9 versus 5.4±5.2 days, p=0.017). The LOS < 24h was associated with higher mortality (85.9% vs 72.9%, p=0.040). On multivariable regression, the likelihood of mortality was high for IMV (7.330, 95% CI 2.667 - 20.143, p<0.001) and elderly (>60 years) patients (2.607, 95%CI 1.063 - 6.394, p=0.036). Mortality decreased with LOS longer than 24h (0.412, 95%CI 0.173 - 0.982, p=0.045). Co-morbidities did not have any effect on mortality. Conclusions Age more than 60 years and IMV were independent risk factors for higher mortality. Longer ICU stay, specifically more than 24 hours, was associated with lower mortality but LOS less than 24 hours might not have a causal relationship with mortality. The odds of survival were not affected by co-morbidities.


Introduction
Coronavirus disease 2019 (COVID-19) started as atypical pneumonia in Wuhan, China, in December 2019 and soon became a global pandemic [1]. Globally, there are 28,285,700 confirmed cases as of September 12, 2020, and 911,255 have succumbed to the disease [2]. Pakistan reported its first case on February 26, 2020, and as of September 12, 2020, there are 300,955 confirmed cases with a death tally of 6,373 [3]. The clinical manifestations of COVID-19 vary from asymptomatic seroconversion to mild upper respiratory tract infection, severe pneumonia, and multiorgan failure [1].
The mortality of COVID-19 pneumonia is higher than other viral pneumonia [4]. After the outbreak, a large number of patients requiring respiratory support placed an unprecedented demand for intensive care services [5]. This necessitated a rapid expansion of intensive care infrastructure, capacity building, and staffing in many countries [6]. The data in terms of mortality, ventilatory support, comorbid conditions, and length of hospital stay is conflicting because different authors have reported the outcomes of a fraction of admitted patients and at variable durations since admission [7][8]. Multiple factors influence ICU mortality in COVID-19. These include the severity of the illness itself, gender, ethnicity, age, comorbid conditions, blood type, mode of ventilatory support, in addition to the availability of trained staff in the ICU [9]. Studies have reported nearly 100% mortality amongst patients on invasive mechanical ventilation (IMV) during the peak of the pandemic [10]. Mortality among patients on IMV was 88.6%, 97%, 43%, 31%, 88.8%, 22%, and 40%-60% in New York [7], China [11], UK [12], Spain [13], Australia [14], and India [15], respectively. With time, the mortality in ICUs declined to around 40% [4]. In the early days of the outbreak, there were no international guidelines for managing patients admitted to the ICU, and the rationing of resources based on local policies in overwhelmed ICUs might have added to divergent data. It is likely that due to the pressure of the pandemic on ICU services, there has been widespread use of other respiratory support (NIV or high-flow nasal oxygen) outside ICUs, and, therefore, patients admitted to the ICU were disproportionately sicker.
Fatality rates vary significantly by country, as the magnitude and velocity of surge fluctuated in different regions of the world on different time frames [7][8]10]. Moreover, the preparedness of the health services, the response of the community, and cultural and religious beliefs might have added to the outcome [16].
Pakistan is a developing country where one-third of the population is living below the poverty line (less than one dollar a day) [17]. The health system is weak and already overstretched. COVID-19 further added to this. There was a social stigma attached to a COVID-19 diagnosis [18]. Patients would often hide epidemiological risk factors & even clinical features of the disease to avoid being diagnosed. Many patients would not come early even with severe disease to the hospital and relatives would keep them at home till very late. Moreover, the shortage of ICU beds, ventilators, and trained staff was another worry, adding to the nightmare of the worst outcome for the patients of COVID-19 in Pakistan [19].
There was no harmony in the vision and actions of different stakeholders and policymakers of the government to prevent the spread of the virus [20]. Amidst the pandemic and prevailing ground realities, the general perception of the medical community was that Pakistan's overall mortality, in general, and ICU mortality, in particular, will be strikingly different from the rest of the world [21]. This necessitated determining the survival of COVID-19 patients in relation to comorbid conditions and ventilatory support in our population.

Materials And Methods
This cross-sectional study was carried out in a major COVID-19 receiving tertiary care hospital in Peshawar, Pakistan. It is a 1200-bedded hospital, caring for patients coming from all over the province. Since there were no national or international guidelines for treating severely ill patients with COVID-19 in the early days of the pandemic, a local criterion for admission to the ICU was devised. While devising the criteria, the severity of the disease itself, the degree of hypoxia, hemodynamic stability, co-morbid conditions, and multiorgan involvement were considered. The study was approved by the institutional review board of Khyber Girls Medical College/Hayatabad Medical Complex, Peshawar, Pakistan. Permission was obtained from the medical director. Data were retrieved for 204 patients admitted from April 1, 2020, till August 31, 2020. Study variables were age, gender, comorbidities, ventilatory status, length of stay (LOS), and outcomes in terms of survival and death. Ventilatory status was taken as the highest level of respiratory support received, either non-invasive ventilation (NIV) (continuous positive airway pressure therapy (CPAP) or bilevel positive airway pressure (BiPAP)) or IMV. Survival was defined as shifting from the ICU to a general isolation ward when the patient improved clinically and did not need further ICU care. The data were analyzed using SPSS version 21. The independent t-test and chi-square test were used to determine the statistical significance of differences in age, gender, comorbidities, ventilatory status, and LOS between survivors and non-survivors. The multivariable regression model was used to predict the likelihood of mortality for the above-mentioned variables.

Results
The mean age of the study population was 55.7 ± 11.6 years. Out of the total of 204 patients, 151 (74%) were male. Around two-thirds of patients (69.1%) were under the age of 60 years. One-hundred twenty-five (61.3%) patients did not have any co-morbidity. The most common co-morbidities were diabetes mellitus (DM) and hypertension (HTN) (15.2% of patients had both DM and HTN in addition to 12.3% who were diabetic only). The mean LOS in ICU was 6.2 ± 6.5 days, and 64 (31.4%) patients spent less than 24 hours in the ICU. Demographic parameters are summarized in Table 1.     [8].
The reasons for the difference in survival compared to the developed countries could be due to the better availability of trained and dedicated staff for ICUs, the timely preparedness of the health system for the pandemic, sufficient resources, and the late phase of the pandemic in their regions. Moreover, in most of the above-mentioned studies, a significant number of patients were still receiving ICU care when the studies were published while in this study the outcome of all the admitted patients has been reported.
The mean age of patients admitted to the ICU was 55.7 ± 11.6 years. This contrasts with studies reported from the USA [22], Italy [8], and Spain [13], where the mean age of patients ranged from 63 to 73 years. Pakistan has only a 3.50 % population over the age of 65 years [25]. This might be the reason for the lower mean age of the patients in the present study.
Co-morbid conditions have been reported to contribute to the severity and worse outcomes in COVID-19. However, the distribution of comorbidities varies by region [4]. In the present study, DM was the leading comorbid condition. In addition to 15.2% of patients who had both DM and HTN, 12.3 % had DM alone followed by COPD (4%) and HTN (2.9%). In other studies, HTN has been the most common co-morbid condition [8,11,13]. The high proportion of DM in this study may be due to the high prevalence of type 2 DM and pre-DM in Pakistan (16.98 % and 10.91%, respectively) [26]. This varied co-morbid profile also warrants further studies to better understand the impact of individual patient risk factors.
In the present study, 61% of patients did not have any co-morbid condition in contrast to that reported from Italy [8] and China [27], where 68% and 68.6% of patients, respectively had at least one co-morbid condition. The reason for the difference might be the younger age of our study population.
NIV has been the mainstay of treatment (61.8%) while 38.2% were treated by IMV. Yu  . This might be explained by the fact that these studies were conducted during the early days of the pandemic when recommendations about the most appropriate modality of ventilation were not clear, and IMV was considered a more effective option. Moreover, there was a fear that NIV could generate an aerosol. It was quite late into the pandemic that NIV was advocated as the preferred alternative, especially for patients without significant co-morbidities [28].
A higher proportion of patients with DM only, HTN only, DM and HTN both, and COPD did not survive. But, when compared to those without any co-morbidity, the difference in mortality for any of the co-morbid conditions is not significant. This contrasts with published literature, where the presence of these comorbidities increased the odds of mortality from COVID-19 [29]. However, a metanalysis (articles = 27, n=22,753) by Bajgain et al. concluded that the presence of one or more co-morbidity is not associated with a higher fatality rate [30].
Overall, the mortality was higher (93.6%) for those on IMV than those on NIV (66.7%). Treatment with NIV significantly increased the odds of survival (11.581, CI 3.307 -40.553, p=<0.001). This is consistent with the mortality rate observed by Wang et al. and Yang et al. in China [10][11]. The high mortality with IMV patients could be due to inadequate intensive care services and lack of enough experience with treating COVID-19 related acute respiratory distress syndrome. Moreover, late shifting to the ICU at a stage where patients had developed multiorgan failure also added to the high mortality in this study.
When compared to younger patients (age less than 60 years), elderly patients had significantly higher mortality (OR 2.607, CI 1.063 -6.394, p=0.036). Higher mortality among the elderly has been consistently reported [8,10]. it is a punishment from God. The response of the government was also far from ideal. Tactics of using the police, rangers, and army for a selective lockdown of infected neighborhoods, escorting infected patients to isolation centers, the presence of guards at the cases' residences, and repeated check on their family members at their homes regarding symptoms and test results, were tantamount to treating patients and their families like criminals. All these culminated in the social stigma attached to a COVID-19 diagnosis [18]. This led to the development of a reluctant attitude in the public to seek timely medical attention.

Conclusions
The overall mortality in the present study was 77%. Age more than 60 years was an independent risk factor for higher mortality. Mortality was high for patients on IMV. The odds of mortality were not affected by DM, HTN, DM and HTN both, and COPD. Longer ICU stay, specifically, more than 24 hours, was associated with lower mortality but LOS less than 24 hours might not be associated with higher mortality.  2020 to conduct the following study: "Mortality of severe COVID-19 patients in intensive care unit; an observational study from a major COVID-19 receiving hospital" The institutional review board needs to be informed and approached for revision of approval shall any significant change(s) are made in the study design/protocol. Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue. Conflicts of interest: In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.