PLoS ONE (Jan 2021)

COVID-19 pneumonia in Galicia (Spain): Impact of prognostic factors and therapies on mortality and need for mechanical ventilation.

  • Luis Pérez-de-Llano,
  • Eva María Romay-Lema,
  • Adolfo Baloira-Villar,
  • Christian Anchorena,
  • María Luisa Torres-Durán,
  • Adrián Sousa,
  • Dolores Corbacho-Abelaira,
  • José Paz-Ferrin,
  • Carmen Diego-Roza,
  • Laura Vilariño-Maneiro,
  • Pedro J Marcos,
  • Carmen Montero-Martínez,
  • Fernando de la Iglesia-Martínez,
  • Vanessa Riveiro-Blanco,
  • Nuria Rodríguez-Núñez,
  • José Abal-Arca,
  • María Bustillo-Casado,
  • Rafael Golpe

DOI
https://doi.org/10.1371/journal.pone.0253465
Journal volume & issue
Vol. 16, no. 6
p. e0253465

Abstract

Read online

IntroductionThis study was aimed to identify risk factors associated with unfavorable outcomes (composite outcome variable: mortality and need for mechanical ventilation) in patients hospitalized in Galicia with COVID-19 pneumonia.MethodsRetrospective, multicenter, observational study carried out in the 8 Galician tertiary hospitals. All Patients admitted with confirmed COVID-19 pneumonia from 1st of March to April 24th, 2020 were included. A multivariable logistic regression analysis was performed in order to identify the relationship between risk factors, therapeutic interventions and the composite outcome variable.ResultsA total of 1292 patients (56.1% male) were included. Two hundred and twenty-five (17.4%) died and 327 (25.3%) reached the main outcome variable. Age [odds ratio (OR) = 1.03 (95% confidence interval (CI): 1.01-1.04)], CRP quartiles 3 and 4 [OR = 2.24 (95% CI: 1.39-3.63)] and [OR = 3.04 (95% CI: 1.88-4.92)], respectively, Charlson index [OR = 1.16 (95%CI: 1.06-1.26)], SaO2 upon admission [OR = 0.93 (95% CI: 0.91-0.95)], hydroxychloroquine prescription [OR = 0.22 (95%CI: 0.12-0.37)], systemic corticosteroids prescription [OR = 1.99 (95%CI: 1.45-2.75)], and tocilizumab prescription [OR = 3.39 (95%CI: 2.15-5.36)], significantly impacted the outcome. Sensitivity analysis using different alternative logistic regression models identified consistently the ratio admissions/hospital beds as a predictor of the outcome [OR = 1.06 (95% CI: 1.02-1.11)].ConclusionThese findings may help to identify patients at hospital admission with a higher risk of death and may urge healthcare authorities to implement policies aimed at reducing deaths by increasing the availability of hospital beds.