Multiplex PCR Pneumonia Panel in Critically Ill Patients Did Not Modify Mortality: A Cohort Study

Authors:
Open Access
Publisher website
Google Scholar
Abstract
In critically ill patients, identification of the pathogen may allow for the timely adjustment of antibiotics and improved outcomes. Background/Objectives: The aim of the study was to assess whether performing a multiplex PCR pneumonia panel (PN-panel) in patients with pneumonia in the intensive care unit (ICU) had any effect on mortality or other important clinical outcomes. Methods: A retrospective cohort study was conducted on adult patients with pneumonia who required ICU admission in four institutions in Bogotá between November 2019 and June 2023. Mortality at 30 days, the length of the hospital and ICU stay, the duration of antibiotics, and their association with the PN-panel performance were evaluated using an inverse probability of the treatment weighting to adjust for covariates and potential confounders. Results: A total of 304 patients were included, including 150 with PN-panel, with a mean age of 65.0 years (SD 14.6). SARS-CoV-2 was the primary etiologic agent in 186 (61.2%) patients, and 256 (84.2%) patients had community-acquired pneumonia. No association was found between 30-day mortality and the PN-panel, with a HR of 1.14 (CI 95% 0.76–1.70), although the assessment by an infectious disease specialist was associated with a lower mortality HR of 0.29 (CI 95% 0.19–0.45). There was no association between the PN-panel and antimicrobial therapy duration or other clinical outcomes. Conclusions: The use of the PN-panel was not associated with changes in mortality, the duration of antibiotics, or hospital and ICU stays. To acquire greater rational decision-making, microbiological data produced by this test should be interpreted with aid of an antimicrobial stewardship program oriented by an infectious disease team that could take the clinical data and integrate the information provided.
Funding Information
  • Universidad Nacional de Colombia (Hermes code 57576, QUIPÚ 202010040463)

This publication has 45 references indexed in Scilit: