Limited prognostic accuracy of the CRB-65 and qSOFA in patients presenting with pneumonia and immunosuppression

Eur J Intern Med. 2020 Nov:81:71-77. doi: 10.1016/j.ejim.2020.08.006. Epub 2020 Aug 7.

Abstract

Background: Scores for risk prediction used in immunocompetent patients with sepsis or pneumonia are poorly evaluated in immunocompromised patients. Therefore, we evaluated the prognostic value of the qSOFA- and CRB-65-criteria in immunocompromised patients presenting with pneumonia.

Methods: Retrospective cohort study including consecutive patients hospitalized with pneumonia and immunosuppression without treatment restrictions. The qSOFA and CRB-65 criteria were documented in the emergency department. Outcome was defined as need of mechanical ventilation (MV) or vasopressor support (VS) and/or hospital-mortality.

Results: 41 of 198 (21%) patients reached the outcome and 10% died. Both, the CRB-65 and qSOFA- were independently associated with the outcome (all p<0.01), but age was not predictive. ROC curve analysis showed moderate predictive potential for both scores (CRB-65: AUC 0.63 and qSOFA: 0.69). With scores of 0, the negative predictive values were below 90% (CRB-65: 9/60 and qSOFA: 12/105 missed patients). With scores > 1, the positive predictive values were 36% (CRB-65) and 58% (qSOFA), respectively.

Conclusions: Both, the qSOFA and the CRB-65 only showed moderate prognostic value, and negative predictive values were inadequate to exclude organ failure or death in patients with immunosuppression. In this population, age was not a predictive parameter. Patients with > 1 positive vital sign criterion measured by both scores should be assessed for organ failure.

MeSH terms

  • Hospital Mortality
  • Humans
  • Immunosuppression Therapy
  • Organ Dysfunction Scores
  • Pneumonia*
  • Prognosis
  • ROC Curve
  • Retrospective Studies
  • Sepsis*