Performance of prognostic index in severe Clostridium difficile-associated infection: retrospective analysis in a university hospital

By consensus severe, Clostridium difficile-associated infection (CDAI) is one that results in hospitalization in ICU, colectomy or death within 30 days. Multiple prognostic indices (IP) attempt to predict these adverse events. To evaluate the performance of 4 PI in predicting severe CDI. Hospitalize...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Revista chilena de infectología 2014-12, Vol.31 (6), p.659-665
Hauptverfasser: Hernández-Rocha, Cristian, Tejos Sufan, Rodrigo, Plaza-Garrido, Ángela, Barra-Carrasco, Jonathan, Agüero Luengo, Carlos, Inostroza Levy, Gonzalo, Ibáñez Lazo, Patricio, Guzmán-Durán, Ana M, Paredes-Sabja, Daniel, Molina Pezoa, M Elena, Álvarez-Lobos, Manuel
Format: Artikel
Sprache:spa
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:By consensus severe, Clostridium difficile-associated infection (CDAI) is one that results in hospitalization in ICU, colectomy or death within 30 days. Multiple prognostic indices (IP) attempt to predict these adverse events. To evaluate the performance of 4 PI in predicting severe CDI. Hospitalized patients ≥ 18 years old with ICD were retrospectively evaluated. Patients with recurrent infection or hematological cancer were excluded. Four PI were evaluated: UPMC version 1, Calgary version 1, Hines VA and Calgary version 2. Seven of 81 patients (8.1%) met the definition of severe CDI. Positive predicted value (PPV) and negative predicted value (NPV) of PI ranged from 20-75% and 91.3-95.7%, respectively. Only Hines VA index had a satisfactory Kappa index (0.74; 95% CI 0.41-1) with a PPV of 75% and NPV of 95,7%. However, because of the variables included, this PI could be calculated only in 32.6% of patients. Hines VA index has the best predicted value and agreement to rule out a severe CDI. Like others PI it has the limitation of including difficult variables to assess in all patients and tends to overestimate an unfavorable course.
ISSN:0716-1018
DOI:10.4067/S0716-10182014000600003