Biomarkers improve mortality prediction by prognostic scales in community-acquired pneumonia
- R Menéndez1,
- R Martínez1,
- S Reyes1,
- J Mensa2,
- X Filella3,
- M A Marcos4,
- A Martínez1,
- C Esquinas5,
- P Ramirez6,
- A Torres5
- 1Servicio de Neumología. Universitary Hospital La Fe, Ciber de enfermedades respiratorias (CIBERES), Valencia, Spain
- 2Servicio de Infecciosas, Hospital Clinic, IDIBAPSBarcelona, Spain
- 3Servicio de Bioquímica, Hospital Clinic, IDIBAPS Barcelona, Spain
- 4Servicio de Microbiología, Hospital Clinic, IDIBAPS, Barcelona, Spain
- 5Servicio de Neumología. Hospital Clinic, IDIBAPS, Ciber de enfermedades respiratorias (CIBERES), Barcelona, Spain
- 6Unidad Cuidados Intensivos. Hospital La Fe, Valencia, Spain
- Dr R Menéndez, Servicio de Neumología, Hospital Universitario La Fe, Avda de Campanar 21, 46009 Valencia, Spain;
- Received 25 July 2008
- Accepted 10 December 2008
- Published Online First 8 January 2009
Background: Prognostic scales provide a useful tool to predict mortality in community-acquired pneumonia (CAP). However, the inflammatory response of the host, crucial in resolution and outcome, is not included in the prognostic scales.
Methods: The aim of this study was to investigate whether information about the initial inflammatory cytokine profile and markers increases the accuracy of prognostic scales to predict 30-day mortality. To this aim, a prospective cohort study in two tertiary care hospitals was designed. Procalcitonin (PCT), C-reactive protein (CRP) and the systemic cytokines tumour necrosis factor α (TNFα) and interleukins IL6, IL8 and IL10 were measured at admission. Initial severity was assessed by PSI (Pneumonia Severity Index), CURB65 (Confusion, Urea nitrogen, Respiratory rate, Blood pressure, ≥65 years of age) and CRB65 (Confusion, Respiratory rate, Blood pressure, ≥65 years of age) scales. A total of 453 hospitalised CAP patients were included.
Results: The 36 patients who died (7.8%) had significantly increased levels of IL6, IL8, PCT and CRP. In regression logistic analyses, high levels of CRP and IL6 showed an independent predictive value for predicting 30-day mortality, after adjustment for prognostic scales. Adding CRP to PSI significantly increased the area under the receiver operating characteristic curve (AUC) from 0.80 to 0.85, that of CURB65 from 0.82 to 0.85 and that of CRB65 from 0.79 to 0.85. Adding IL6 or PCT values to CRP did not significantly increase the AUC of any scale. When using two scales (PSI and CURB65/CRB65) and CRP simultaneously the AUC was 0.88.
Conclusions: Adding CRP levels to PSI, CURB65 and CRB65 scales improves the 30-day mortality prediction. The highest predictive value is reached with a combination of two scales and CRP. Further validation of that improvement is needed.
See Editorial, p. 556
Funding: The study was funded by: CIBER de enfermedades respiratorias (CIBERES), CIBERES is an iniative of ISCIII; FIS 041136; FIS 020616 and Fundació Marató TV3 TV040530. Funding sources provided grants for research not related to any industry and they did not have any role in the design, analysis or interpretation of the results or manuscript approval. An independent statistician, Ivan Arribas, participated in statistical analyses.
Competing interests: AT has been a speaker at a symposium and has received a research grant from Brahms. All other authors have no conflict.
Ethics approval: The study was approved by the Ethics Committees of the two hospitals involved.