Severe COVID-19 pneumonitis is associated with poor outcomes in older frail patients as determined by the Clinical Frailty Scale [1]. Advanced age, male gender and significant medical co-morbidities are also associated with worse outcomes [2]. Subsequently, many older frail patients with multiple co-morbidities admitted to the hospital with COVID-19 pneumonitis are deemed inappropriate for invasive mechanical ventilation due to the likelihood of poor outcomes and distress caused by these interventions.

The optimal respiratory support strategy for these patients is unknown. Previous work done in our hospital suggests that this cohort is unlikely to benefit from non-invasive continuous positive airway pressure (CPAP) ventilation [3], although other authors appear to have observed better outcomes in a small retrospective analysis [4]. However, the later study did not include a contemporaneous control group for comparison, and the study included patients with likely, less severe diseases as determined by oxygen requirement at the time of starting CPAP. Additionally, patients often find tightly fitted CPAP masks difficult to tolerate, and alternative strategies are subsequently required. High flow nasal oxygen therapy (HFNO) is considered a potential alternative to CPAP.

HFNO has previously been shown to potentially reduce the need for invasive ventilation in patients with severe hypoxemic respiratory failure [5]. Several recently published observational studies suggest that HFNO may also prevent the need for invasive ventilation in COVID-19 pneumonitis [6-15]. These studies are summarised in a focussed literature review presented in Table 1.

It is, however, important to note that these studies primarily evaluated HFNO in younger patients deemed appropriate for invasive ventilation. They did not address the question of whether HFNO is an appropriate ceiling of treatment for older frail patients. As a result, it is not clear whether a benefit relative to standard oxygen therapy (SOT) would be observed in this patient group. A single pre-print was found examining the potential of HFNO as a ceiling of treatment in frail patients [16]. They observed a 75% mortality in an elderly frail group that as failing to maintain oxygen saturation with 15 L/min O2 via a standard mask with a non-rebreathe reservoir. The authors did not include a comparison group that utilised an alternative treatment strategy. Although the study included older frail individuals, the median age of the eight survivors was 69.5 years.

In our institution, patients with severe COVID-19 pneumonitis deemed inappropriate for invasive ventilation and requiring an inspired concentration of O₂ greater than 60% were considered for HFNO. We retrospectively identified all patients meeting these criteria within a four-month period in our institution. The patients were reviewed in order to compare the outcomes of those who received HFNO to those who received SOT.

After consultation with the chair of the local research ethics committee, no formal ethical approval for this study was sought as this was considered a service evaluation. All interventions were carried out at Kettering General Hospital (United Kingdom, a 600 bedded secondary care hospital serving a population of 330,000).

We prospectively identified all consecutive COVID-19 cases with severe type 1 respiratory failure requiring FiO₂ at 0.6 admitted to our hospital between 23 September 2020 and 24 January 2021 that were deemed inappropriate for invasive ventilation or CPAP. Clinicians made decisions regarding the suitability of patients for the trial of HFNO based on CFS. Their outcomes were recorded and compared with a similar group of patients who received standard oxygen as a ceiling of treatment. Patients were followed up until death or discharge.

Eighty-one patients were included following compliance with the aforementioned inclusion criteria and exclusion criteria. Twenty-four and fifty-seven patients were allocated to the HFNO and SOT groups, respectively. The groups did not exhibit a statistically significant difference in terms of age, gender, BMI, or any of several co-morbidities identified as being strongly associated with mortality in the OpenSafely study [2] (Table 2). The HFNO group had a significantly lower Clinical Frailty Scale score (Table 2).

Fig. (1). Survival time compared between HFNO and SOT (Oxygen) groups.

The rate of deterioration of both groups from admission to needing 60% oxygen therapy was similar (Table 3). The SOT group was started on FiO₂ > 60% at around the same time as when the HFNO group commenced HFNO therapy. All 24 patients who received HFNO therapy died. Forty-six patients (80.7%) in the SOT group died and the remaining patients survived until hospital discharge. The primary outcome of mortality was significantly worse in the HFNO group (Table 3). All survivors were in the SOT group and a comparison of survivors to non-survivors within this group showed that they were younger with different co-morbidity profiles (Table 4). Survival time analysis of up to 50 days similarly revealed shorter survival in the HFNO group (Fig. 1).

The data obtained in this study suggest that HFNO offered no benefit to older frail patients with severe COVID-19 requiring FiO₂ > 60% and deemed inappropriate for invasive ventilation. All patients that received-HFNO died, while 11 out of a total of 57 patients survived to discharge in the SOT group. This result suggests the possibility of potential harm from HFNO in this group.

This study has some important limitations that need to be considered. The total HFNO sample size was less than initially desired, with 24 patients included in this single centre study. This was the case because once poor outcomes became clear to the clinical decision-makers, they were less inclined to consider HFNO therapy. This was an observational study, and therefore our findings do not exclude other possible interpretations of the data. The groups were very similar ages (mean age 79.9 years [HFNO] vs 79.7 years [SOT]), and although the HFNO group had a higher proportion of males (62.5% vs 54.4%), this was not a significant difference. We, however, acknowledge the discrepancies in baseline demographics between the two groups. The standard oxygen therapy (SOT) group demonstrated a higher clinical frailty score (CFS) compared to the high flow nasal oxygen (HFNO) group. This was because the SOT patient cohort was initially reviewed by a respiratory physician and subsequently considered inappropriate and unlikely to benefit from HFNO therapy due to reasons such as severe frailty. As a result, they were instead treated with SOT.

Relative to other-data available in published literature, this study has the strength of including an imperfect but meaningful comparison group.

However, it should be noted that the Clinical Frailty Scale score of those who received HFNO was, on average, two points lower in comparison to the SOT group. As stated above, a higher score on this scale has previously been shown to be associated with worse outcomes in a large observational study [1]. The rate of deterioration from hospital admission until needing FiO₂ 60% to maintain O₂ saturations was very similar in both groups, suggesting that the severity of respiratory failure was similar in both groups. However, this does not exclude the possibility that the HFNO group may have been deteriorating more rapidly after this point.

Despite these limitations, the observations in this study are stark. It suggests that future studies for the use of HFNO in severe COVID-19 pneumonitis should be closely monitored for safety signals, particularly in the elderly frail cohort. The lack of benefit is consistent with the results of the RECOVERY-RS study preprint [18]. The RECOVERY-RS is a multi-centre, three-arm randomised control trial comparing CPAP, HFNO or conventional oxygen therapy with the primary outcome presented as a composite of invasive mechanical ventilation or mortality within 30 days. No significant difference was observed when comparing HFNO with conventional oxygen. Our study findings, however, contrast with the findings presented by van Steenkiste et al. in pre-print [16]. Van Steenkiste et al. observed 25% survival when HFNO was used as a last resort in elderly frail patients. However, the median age of their survivors was 69.5 years and a meaningful comparison group was not included/presented. Therefore, conclusions regarding the benefit of HFNO over SOT cannot be made from their study. The mean age of survivors in our study (all in the SOT group) was similar (70.8 +/- 6.07 years). Further, it may be pertinent to note that in the van Steenkiste study HFNO was utilised when patients had developed a higher O₂ requirement. While the findings of the RECOVERY-RS trial have guided current clinical practise, it is important to consider the difference in demographics between our study and the population recruited for the trial. The patients recruited for RECOVERY-RS were not at their ceiling of care and considered appropriate for escalation to invasive mechanical ventilation. As a result, we believe further RCTs are necessary to assess suitable treatments options for patients currently at ward-based ceiling of care and unsuitable for critical care intervention.

There are several potential mechanisms by which we may hypothesise that HFNO may be associated with harm. The nasal mucosa appears to be the site with the highest viral burden, thus, aspiration-mediated seeding virus to the lung is a plausible explanation for the distribution of viral infection and pneumonitis observed [19]. It may be possible to suggest that HFNO has a role in the facilitation of aspiration and seeding. Furthermore, relatively higher concentrations of O₂ may exacerbate organ injury via alternative mechanisms that have been observed in other conditions [20].

Additionally, the need to wear high-level personal protective equipment when caring for elderly frail patients receiving HFNO (deemed an aerosol-generating high risk procedure) may compromise or affect the quality of their nursing care (in comparison to those receiving SOT that were cared for by staff using lower-level personal protective equipment). However, patients requiring HFNO had a lower ratio of patients to nursing staff to compensate for these issues, i.e., nurses cared for a maximum of 4 patients receiving HFNO vs. 7 patients receiving SOT.

In summary, our study does not show any evidence of survival advantage from HFNO treatment in COVID-19 pneumonitis when compared to conventional oxygen therapy amongst patients who were unsuitable for invasive mechanical ventilation or CPAP. In fact, findings from the study even suggest potential harm from HFNO in this particular population group.

While HFNO facilitation in the ward setting has been demonstrated to be feasible, this study has conveyed some important points for consideration with respect to treatment efficacy and mortality. As a result, current treatment strategies for this patient population may require further reflection and evaluation in order to develop robust treatment strategies.