[Prognostic factors in an individualised approach to non-pharmacological therapy of COVID-19: from oxygen and mechanical ventilation to extracorporeal membrane oxygenation].

Abstract

Background: Our centre followed a stepwise approach in the nonpharmacological treatment of respiratory failure in COVID-19 in accordance with German national guidelines, escalating non-invasive measures before invasive mechanical ventilation (IMV) or extracorporeal membrane oxygenation (ECMO). The aim of this study was to analyse this individualized approach to non-pharmacologic therapy in terms of patient characteristics and clinical features that may help predict more severe disease, particularly the need for intensive care.

Method: This retrospective single-centre study of COVID-19 inpatients between March 2020 and December 2021 analysed anthropometric data, non-pharmacological maximum therapy and survival status via a manual medical file review.

Results: Of 1052 COVID-19-related admissions, 835 patients were included in the analysis cohort (54% male, median 58 years); 34% (n=284) received no therapy, 40% (n=337) conventional oxygen therapy (COT), 3% (n=22) high flow nasal cannula (NHFC), 9% (n=73) continuous positive airway pressure (CPAP), 7% (n=56) non-invasive ventilation (NIV), 4% (n=34) intermittent mandatory ventilation (IMV), and 3% (n=29) extracorporeal membrane oxygenation (ECMO). Of 551 patients treated with at least COT, 12.3% required intubation. A total of 183 patients required ICU treatment, and 106 (13%) died. 25 (74%) IMV patients and 23 (79%) ECMO patients died. Arterial hypertension, diabetes and dyslipidemia was more prevalent in non-survivors. Binary logistic analysis revealed the following risk factors for increased mortality: an oxygen supplementation of ≥2 L/min at baseline (OR 6.96 [4.01-12.08]), age (OR 1.09 [1.05-1.14]), and male sex (OR 2.23 [0.79-6.31]).

Conclusion: The physician's immediate clinical decision to provide oxygen therapy, along with other recognized risk factors, plays an important role in predicting the severity of the disease course and thus aiding in the management of COVID-19.