Dear Editor,

We thank doctor Vargas and colleagues for their interest in our paper [1].

We read their letter, which arises some questions that deserve to be elucidated.

High-flow nasal cannula (HFNC) is a valid non-invasive respiratory support to treat respiratory failure in COVID-19 patients. The main reason why we did not consider its use as a first-line treatment of hypoxic failure was its low availability compared to H-CPAP support. For this reason, HFNC was applied, when possible, only in few patients who developed delirium, but only after the neuropsychiatric disturbance had emerged. Due to the high rates of hospitalization for COVID-19 in our non-ICU department, our treatment protocol only included H-CPAP supports and we have not had the opportunity to use other interfaces. All the components of the medical staff had received adequate training for COVID-19 management and respiratory care.

For what concerns treatment of delirium, patients who developed acute confusional state were mainly treated with haloperidol or quetiapine. In case of concomitant severe sleep or mood disorders, trazodone was taken into account as an alternative treatment.

In an attempt to establish an adequate doctor–patient relationship, we wrote our names on the full body protective suits, as a re-orienting measure. Patients were in contact with their families through video calls that were made three times a week. The calls were performed at the end of the work shifts. This part of the assistance was provided through formally organized work shifts. Patients’ mobilization was part of the standard care. Patients with H-CPAP were the most complicated to mobilize; however, we alternated periods of prone position with periods of supine position making an effort to prevent skin lesions of the neck caused by rigid collar pressure on the skin.

Finally, the concept of incidence and prevalence of delirium has recently become increasingly important in scientific literature, as shown in the meta-analysis cited in our prospective study [2]. We did not explore the presence of different predisposing or precipitating factors upon admission and during hospitalization. However, all the patients received the same protocol of care, including the same preventive measures. We found an incidence rate higher than the prevalence. Although this can be at least partially explained by the fact that hospitalization is itself a precipitating factor for delirium, we agree with the authors that possibly different pathophysiological mechanisms may have played a role during hospital stay.