C-reactive protein (CRP): An important diagnostic and prognostic tool in nursing-home-associated pneumonia

Abstract 

Pneumonia is the second most common infection in long term care (LTC) residents and is a leading cause of death from infection in those groups of patients. Atypical presentations and fewer presenting signs and symptoms in older patients complicate diagnosis and delay initiation of adequate treatment. The aim of this study was to compare laboratory CRP levels to pneumonia severity scores, in prediction of short-term death from pneumonia. Diagnosis of pneumonia was performed according to the criteria of McGeer for the identification of pneumonia at an LTC facility. The severities of pneumonia and mortality prediction were assessed by three indices: PSI (pneumonia severity index), Missouri study index and the nursing home associated pneumonia (NHAP) severity index. A strong positive correlation was found between CRP levels and PSI (r=0.445, p<0.001), Missouri study index (r=0.315, p<0.001) and NHAP severity index (r=0.246, p=0.002). The initial values of CRP were significantly higher in patients with short term mortality and positively correlated with rate of death (r=0.493, p<0.001). By multivariate regression analysis, the variables that were independently and significantly associated with the rate of death included presence and duration of fever, respiratory rate, serum CRP and albumin levels, lymphocyte count, number of comorbid diseases, CHF, and DM (the R² was 0.711 and 0.685 when adjusted). Because presentation of nursing home acquired pneumonia is not specific, it is suggested that CRP should be performed in every patient with a suspicion of pneumonia.

Keywords: Pneumonia, Nursing home acquired pneumonia, Long-tem care of elderly, C-reactive protein