Health systems around the world have been severely compromised in recent months due to the health emergency caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Hospitals have had to manage large numbers of very sick patients. Spain was one of the European countries hardest hit by the pandemic, with almost 5 million cases and more than 85,000 deaths due to coronavirus disease 2019 (COVID-19). In our region, the Principality of Asturias, we have accumulated a total of 70,892 cases and 2068 deaths to date for a population of 1,011,560 people. We are currently leading our country, with more than 90% of the target population (people aged 16+ years) fully immunized. The current incidence in our region is about 10–15 cases per day with a test positivity rate of 0.55% on 30 September 2021. Our centre, the Hospital Valle del Nalón, is a public hospital located in the region of Langreo, which serves a population of 71,300 people.

One of the most concerning aspects of the pandemic has been severe respiratory failure as the final and most feared outcome of SARS-CoV-2 infection, due to the scarcity of resources in intensive care beds and mechanical ventilation equipment. Several management strategies have been tried, including non-invasive mechanical ventilation techniques or the use of high-flow nasal cannulae (HFNC). These strategies, initially controversial, have been accumulating evidence in recent months, consolidating their presence in clinical practice guidelines.1 Proof of this is that in the spring of 2020 several Spanish scientific societies drew up a consensus guideline,2 which established the standards for the management of acute respiratory failure by non-invasive ventilation techniques, including HFNCs, without delaying the use of invasive measures when necessary.

Despite an increase of up to 300% in the usual number of intensive care beds, these were insufficient, with units being installed in dressing rooms or in hospital physical therapy areas. Hence, our hospital drew on previous experience3 to develop a progressive care strategy, depending on the occupation of intensive care unit (ICU) beds, both in our own centre and in other ICUs in the surrounding area. All this was based on the transversality of care and the multidisciplinary assessment of patients, which resulted in a consensus protocol for the management of severe respiratory failure due to SARS-CoV-2. This protocol established that patients with SARS-CoV-2 pneumonia admitted to a conventional internal medicine hospital ward and receiving standard medical treatment who, despite the use of a 40%–50% venturi mask, did not obtain saturations of at least 92%, should be evaluated by their internist or pulmonologist and an intensivist. In a consensual and regulated manner, it was decided to initiate high-flow oxygen therapy in the conventional internal medicine ward or in the ICU, always in collaboration with the nursing staff. In the event of clinical worsening despite HFNC, patients should be transferred to our ICU or to other nearby hospitals. This management strategy on the ward was also applied to those who, due to their baseline situation and comorbidities, did not require intensive clinical management.

High-flow therapy was performed with HF-2990D devices (Linde Medica). In all cases, the medical and nursing staff performed standard precautions for the management of patients with COVID and patients were protected using surgical masks over goggles.4 Oxygen saturation and respiratory rate were monitored hourly during the first 4 h of HFNC and whenever dyspnoea was reported by the patient or deemed necessary by the nursing staff. Vital signs were recorded on an electronic medical record form, including calculation of the ROX index. The on-call internal medicine physician was alerted whenever the ROX index was less than 4 or saturation was less than 92%. Failure of HFNC therapy was considered to entail the need for intubation and connection to mechanical ventilation and in no case was this measure delayed when necessary.

Between September 2020 and April 2021, this protocol was applied to a total of 87 patient, whose outcomes have been analysed. Of these patients, 69 (79.3%) were potential candidates for ICU, while the other 18 were not because of comorbidities, so HFNC was considered as a therapeutic ceiling. Of the patients who were potential ICU candidates (N = 69), overall survival was 69.56%. Patients whose HFNC was initiated in the ICU (N = 15) were analysed versus those whose HFNC was initiated in the conventional ward (N = 54), with no significant differences in terms of survival or failure of high-flow therapy. Sex, age, days of evolution of COVID-19 or respiratory frequency did not influence the success or failure of high-flow therapy or survival. Of these 54 patients who received high-flow therapy on the ward, 40.7% progressed satisfactorily, avoiding intubation and admission to the ICU and were discharged home in 100% of cases. We found no differences in terms of survival between patients who received high-flow therapy on the conventional ward or in the ICU. There were no deaths or emergent orotracheal intubation on the inpatient ward (Figure 1).

Algorithm—summary of patients included in the study and their clinical outcomes. HFNC, high-flow nasal cannulae; ICU, intensive care unit; OTI, orotracheal intubation

During this evaluation period, eight cases of SARS-CoV-2 infection were reported among healthcare personnel on the internal medicine ward, representing 1.4% of the staff, compared with 14.1% of cases among other hospital staff on other wards where COVID patients were not admitted. There were no deaths or emergent orotracheal intubations on the inpatient ward.

After this experience in the use of high-flow therapy, our proposal for the future is to use high-flow devices in conventional hospitalization wards and not exclusively in ICUs, regardless of the situation of care context. We believe that it is only necessary to establish minimum monitoring and alert systems and to work collaboratively between different medical specialties and nursing staff. We believe that collaboration has been fundamental to the successful planning, execution and positive evaluation of this ventilatory management protocol in patients with COVID-19, which has avoided stays in more complex units with high morbidity without detriment to the safety of patients or healthcare staff.

ACKNOWLEDGEMENT

The authors thank the Biostatistics and Epidemiology Platform of the Institute of Health Research of the Principality of Asturias (ISPA).

HUMAN ETHICS APPROVAL DECLARATION

The study was approved by the Ethical Committee in Research of the Principality of Asturias.