Skip to page content

By Dr H Iftikhar, Dr S Alaee, Dr J Bennett, Dr A Creamer, Dr R Kaminski, Dr D Windsor, Dr C Sharp

Background and problem

COVID-19 is associated with significant mortality and morbidity in high risk groups requiring ventilatory support as per the Intensive Care National Audit & Research Centre (ICNARC).

Mechanical (IMV) and non-invasive ventilation modalities [Continuous positive airway pressure (CPAP) / High Flow Nasal Oxygen (HFNO)] support acute respiratory failure in COVID-19 but the mortality data comparing these modalities is limited.

Gloucestershire NHS Foundation Trust admitted a total of 860 COVID-19 patients, 130 requiring ventilatory support between February–July 2020; the highest number in the South-West.

Respiratory High dependency (HDU) and Intensive care units (ITU) were reconfigured in anticipation of clinical demand with HDU expanded to 31 beds compared to a normal capacity of 10 and ITU expanded to 36 beds, compared to a usual capacity of Patients requiring CPAP only were managed on HDU unless deemed at high risk of deterioration to require IMV.

Methods

We conducted a prospective observational study to assess comparative mortality in all COVID-19 patients admitted to HDU/ITU with acute respiratory failure and treated with invasive mechanical versus non-invasive ventilatory modality (CPAP/HFNO).

Parameters assessed included age, gender, clinical frailty score (CFS), co-morbidities, smoking and resuscitation status. Comparative mortality was assessed statistically by calculating relative risk ratio and p-value using Welch’s t-test.

Results

130 patients were treated with CPAP/HFNO, IMV or both. Only 1 patient was treated with HFNO with no mortality in this group. Overall mortality was 33% (n=43).

Resuscitation status and treatment escalation plans were reviewed for all patients on admission.

1.5% patients (n=2) had a pre-existing DNAR and CPR was not deemed appropriate for 23% patients (n=30) 62% patients (n=58) required IMV out of 72% patients (n=93) deemed suitable for it.

Printable version of this page