Experience of the proned patients

We selected 10 patients (8 males and 2 females, mean age 73) as candidates for prone positioning. Demographic and laboratory findings are reported in table 1. The attempt failed due to lack of compliance in one patient (sliding and moving in lateral position), inadequate pain control even with ongoing treatment in one patient and refusal by two patients to prone positioning.

Six patients (five males and one female) completed a cycle of prone ventilation. Their complaints included fever plus other clinical manifestations, including alterations of taste and smell, syncope, diarrhoea, vomiting or dry cough, for a median of 8±5 days. Every patient had a positive nasopharyngeal swab for COVID-19.

All the patients had peripheral capillary oxygen saturation (SpO₂) less than 90% on room air, with tachypnoea (RR >22 breaths per minute). Only two had fever (body temperature >37.5°C). Four patients had cardiovascular comorbidities. All the patients had laboratory findings as reported in COVID-19 infection,1 such as lymphocytopenia, increased aspartate transaminase, elevated lactate dehydrogenase and C-reactive protein (table 1). Lung US showed B lines, pleural line irregularities in most fields and spared areas bilaterally. In all the cases high-resolution CT confirmed interstitial pneumonia ranging from 25% to 70% of lung parenchyma involved.

Initially, we provided oxygen to the patients via high-flux reservoir mask, but after a temporary improvement in SpO₂ values, we observed a further worsening of the respiratory failure with persistent dyspnoea. An ICU bed was not available. For this reason, we started NIV CPAP via StarMed helmet with an FiO₂ of 100% and a positive end-expiratory pressure ranging from 10 to 20 mm Hg. After a median time of 19 hours in NIV CPAP, all the patients developed severe hypoxaemia with a median PaO₂/FiO₂ ratio of 68±5 mm Hg.

We put the patients in a prone position with the help of specialised nurses (video 1). The patient was initially put supine on the border of a two-point movement bed with a water antidecubital mattress. A pillow was slid laterally under the chest and abdomen and the patient was prepared for turning with one arm on his abdomen and a leg crossed over the other. The nurses then slowly rolled the patient into a prone position and adjusted the helmet and the position of the head and arms (figures 1 and 2). The physician checked the helmet and the NIV pressures to verify the compliance of the patient with the procedure. In order to increase the compliance to NIV and reduce the anxiety linked to the situation, a continuous morphine infusion of 20–30 mg/day was started, which improved the patient’s compliance. After 1 hour of proning at the above settings, ABG showed an improvement in PaO₂/FiO₂ ratio for all the patients (median 97±8 mm Hg). Lung US after 1 hour of prone position showed no differences in B line quantity and distribution.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Proning in non-invasive ventilation (NIV).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Prone position.

At the time of writing, all the six patients eventually were admitted to ICU and underwent cycles of proning during invasive endotracheal ventilation. Patient 1 has been recovering in a rehabilitation ward with low-flow nasal prongs; patients 2, 3, 4 and 5 were moved to other ICUs outside Piacenza province to make new beds available; patient 6 died. Among the patients who did not receive prone NIV, patients 7, 9 and 10 died. Patient 8 recovered after a period of time in ICU.

In comparison to the study of Caputo et al,3 our cohort of patients are older and with a longer stay in the ED. Also, our patients had already been assessed by an intensivist and considered not to be admitted to ICU straight away. We decided to attempt proning in order to try everything possible to help these awake and compliant patients with severe respiratory failure in a crowded ED who were waiting for an ICU bed to become available for them.

We are aware that our results are preliminary and further investigation will be required to confirm our observation. We may have failed to spot differences in lung US before and after proning, because we performed lung US too early to detect significant changes.

Based on our experience, we believe that prone NIV CPAP could be an option for patients non-responsive to traditional NIV CPAP with helmet from whom ICU care is not quickly available.