At Södersjukhuset, Stockholm, a project to start HFNO treatment was initiated in the department of infectious diseases, with 15 single rooms with modern ventilation and negative suction pressure, built to be used for airborne diseases. This makes it possible to change the whole air in the room 8 times/hour, so the HFNO that is aerosol-generating would be safer for health workers. A negative pressure room is regarded as preferable for patients receiving HFNO according to the ANZICS guidelines (5). Initially, four rooms were supplied with HFNO and after four weeks the number increased to eight rooms. The number of rooms varied depending on the necessity, which was determined on a weekly basis, as well as the hospital’s resources. Health care givers used maximized protection when patients had HFNO.
An anesthesiologist or intensive care physician holds the decision-making about which patients are going to the infection ward to receive HFNO taking into consideration the ARDS severity that they have. This is evaluated with parameters including PFI (PaO2 to FiO2 Index) which uses the partial pressure of the oxygen in blood (PaO2) and the fraction of inspired oxygen (FiO2), and this comparison helps to determine the degree of any problems, such as ARDS, with how the lungs transfer oxygen to the blood (10). We used arterial blood gas analysis to measure PaO2 (Point-of-care testing). PFI was used to determine the need of mechanical respiration inclusive of HFNO. Severe ARDS is defined as a PFI of <13, while moderate PFI is defined as a PFI of <26. Normal PFI is > 50.
The target for admission to infection ward was to offload the ICU. Patients who were stabilized with HFNO (irrespective of the previous need of invasive ventilation) in ICU; or needed HFNO under assessment of the necessity of intensive care were candidates for this treatment. Even patients who were unsuitable for tracheal intubation but were experiencing respiratory deterioration despite standard nasal oxygen or facemask oxygen were able to get HFNO treatment in the ward.
HFNO was initiated at 50 L/min (AIRVO2 or Optiflow, Fisher & Paykel Health Care Ltd., Auckland, New Zealand) with temperature set at 34C°. Nasal cannula size was determined by the patient's nostril size (50%). FIO2 was adjusted to maintain SpO2 at 94%. Flow and temperature were adjusted based on patient comfort and clinical response. These settings were accepted as standard settings except if the patient had other settings from the ICU when discharged to the ward. The settings ordered from the ICU were then retained.
In this retrospective, observational cohort study, we followed all patients with Covid-19 (defined with positive nasopharyngeal polymerase chain reaction for Severe Acute Respiratory Syndrome Corona Virus 2, SARS-CoV2-RNA PCR test) who were admitted to the infection ward to get HFNO between April and June 2020. We counted how many days the patients had HFNO treatment and how that saved valuable time in ICU. This was achieved by prolonging the time before initiating ICU care (which is called ‘’step-up’’ and means primary usage of HFNO with frequent controls to evaluate the need of ICU and invasive ventilation), or by skipping it if the patient is stabilized with HFNO in the infection ward. Another possibility was to use the infection ward as a ‘’step-down’’, which means helping patients who initially needed ICU care and then needed supportive respiratory care. Another patient category was patients with do-not-resuscitate and do-not-intubate orders where HFNO was considered as a viable treatment when other possibilities were limited as there was no intention to transfer to ICU despite deterioration.
The usage of HFNO as a step-up or step-down was assessed by looking at which ward the patient came from and was transferred to after the time in the infection ward. We even looked at the other diseases the patients had and studied other risk factors. The outcomes were the number of days of treatment with HFNO in the infection ward, the result of treatment (cure, death, or transfer to other hospital or care form), complications; and if patients had a do-not-resuscitate order (DNR) and how that affected the result. We compared the group as all with a subgroup of Covid-19 deaths who also had been treated with HFNO in the same ward.
The project has been approved by the Swedish Ethical Review Authority (DNR 2020-03760).
Statistical analysis
Values are means (interquartile range) unless otherwise stated. P-values for differences obtained using Fisher’s exact test and Kruskal Wallis test. Statistical significance was set at p <0.05. All statistical analyses were performed with the use of SPSS.