Background
Severe COVID-19 infection causing respiratory failure requiring high-level care is unfortunately an ongoing global problem, posing considerable strain on hospital resources. The optimum management of such cases with different modalities of respiratory support and their effectiveness in certain groups of COVID-19 patients has not been extensively reported.[1] This is mainly due to the uncharted waters we are all navigating in with this new virus. In England, expert groups and guideline committees have rolled-out clear concise guidance on patient selection regarding escalation of care and eligibility for invasive ventilation, however, the outcomes of patients who are not fit for escalation or invasive ventilation but require treatment with CPAP have not been yet reported. 
Objectives
In our study, we investigated the mortality (in hospital) outcomes of CPAP use in patients with respiratory failure secondary to severe COVID-19 infection who were deemed not fit for invasive ventilation, and compared it to patients who were managed on oxygen alone.
Methods
No formal ethical approval for this study was sought as this was considered as a service evaluation. All interventions were carried out at Kettering General Hospital (United Kingdom) which is 600 bedded secondary care hospital. We retrospectively investigated all -19 (probable or confirmed) patients with severe respiratory failure who required FiO2> 0.6) and admitted to our hospital between 15th March and 4th May 2020. All confirmed cases had a positive rRT-PCR swab for COVID-19. Patients meeting the case definition of ‘probable COVID-19’ as per WHO case definition[2] were also included, if they were managed  clinically as COVID-19 infection by the treating physician, as false negative results were common with rRT-PCR testing [3][4]. Patients requiring NIV for acute or chronic type 2 respiratory failure due to pre-existing conditions were excluded from the study. A decision on fitness for invasive ventilation including DNAR was recorded in the medical notes at the time of admission after senior clinician review and discussion with the patient  as per national guidelines and did not interfere with the eligibility for CPAP. CPAP was considered for patients who met the BTS criteria for its initiation at the discretion of the attending physician in conjunction with the respiratory team and / or critical care outreach team [5]. Those who were deemed to be too ill to benefit from CPAP were managed with oxygen alone. We retrospectively analysed the data of all the patients admitted with suspected COVID-19 and also analysed their Vital Signs recorded online. Those meeting the criteria for CPAP {requiring FiO2> 0.6 to maintain SpO2 >92% (88-92% in COPD)} were analysed in greater depth. Patients who met the criteria for CPAP but were managed on oxygen therapy alone were included as the control group. CFS (clinical frailty score) and data on comorbidities well known to affect mortality in COVID-19 infection like hypertension, diabetes, cardiovascular disease (CVD), cerebro-vascular accident (CVA), Neutrophil / Lymphocyte ratio (NLR) on admission and chronic obstructive pulmonary disease (COPD) was collected to compare the two groups (Table 1).
To improve the validity of the study, whilst analysing the data for above two groups, we also prospectively collected data (at arm's length) for any patient who met the same criteria from 6th May to 8th June 2020. Combined data (retrospective and prospective) was also analysed to detect any statistically important differences between the groups (Table 2). 
Inclusion criteria
Exclusion criteria
Statistical analysis
The data was summarized using descriptive statistics and results are reported as means and standard deviations, and any differences between the two groups were analysed using a two tailed T test. Categorical variables are summarized numerically and percentages with any differences analysed using Chi squared test.
Results
Between 12th March and 04th May 2020, 71 patients fulfilled the inclusion criteria and were included in the study. 16 of them were treated with CPAP and the rest were treated with oxygen therapy alone. Their baseline characteristics are listed in (Table 1). A total of 55 patients were included in the control group and received standard oxygen administration methods ranging from Venturi, Humidified oxygen or non-rebreathe masks. Patients in the CPAP group were treated with either NIPPY 3® ventilator in the CPAP mode or StarMed Ventukit®  Up CPAP hoods (Intersurgical SpA, Italy). Mortality in the CPAP group was 93.7% (n=16) compared to 92.7% in the control group (n=55). There was a statistically significant difference between the 2 groups in terms of age and clinical frailty score in favour of the CPAP group. Despite this, there wasn’t any statistical or clinically significant difference in mortality between the two groups. 
The prospective arm of the study included a total of 33 patients of which 8 patients received CPAP with the rest receiving high flow oxygen only. The mortality in this group was also high with 91% dying (7 CPAP and 23 in the oxygen group). Mortality remained above 90% when both the retrospective and prospective groups were combined. Similarly there was no difference in mortality in patients with proven COVID-19 infection and those ‘highly suspected’ cases who were treated clinically as COVID-19 infection.