[Diagnostic value of serum procalcitonin in identifying the etiology of non-responding community-acquired pneumonia after initial antibiotic therapy].

OBJECTIVE: This study was to investigate the diagnostic value of serum procalcitonin(PCT) in identifying the etiology of non-responding community-acquired pneumonia (CAP) after initial antibiotic therapy.

METHODS: A retrospective analysis was performed for 232 hospitalized CAP patients admitted to the People's Hospital of Zhengzhou University during June 2013 and January 2014. Early treatment failure was defined as the presence of persistent fever (>38 °C) and/or clinical symptoms (malaise, cough, expectoration, dyspnea) or deterioration after at least 72 h of initial antimicrobial treatment, or development of respiratory failure requiring mechanical ventilation, or septic shock. Bronchoscopy or transthoracic lung biopsy was performed in case of early treatment failure when indicated. Serum level of PCT was detected by double antibody sandwich method. The differences between 2 or more groups were compared using 2-independent student t test, one-way ANOVA; Mann-Whitney U test, Kruskal-Wallis rank sum test, or χ(2) test. Risk factors and odds ratios for nonresponsiveness were analyzed by setting up a Logistic regression model. The diagnostic values of PCT were determined by receiver operating characteristic curves (ROC curves).

RESULTS: Of the 232 CAP patients enrolled, 124 were male and 108 were female, with an average age of (46 ± 20) years. Thirty-six patients failed to respond to the initial antibiotic therapy. As shown by Logistic regression analysis, the risk factors for treatment failure included hypoalbuminemia, type 2 diabetes, previous history of splenectomy , PSI 4-5 grade, and lung infiltration ≥ 3 lobes. The most common causes of non-responsiveness were antimicrobial insufficiency (n = 23), and misdiagnosis of noninfectious mimics of pneumonia (n = 11), with 2 cases of unidentified etiology. The serum PCT level in admission was 0.19 (0.07-0.66) µg/L in the antimicrobial insufficiency subgroup, which was significantly higher than that in the misdiagnosis subgroup [0.06(0.05-0.08)µg/L; P < 0.01]. The antimicrobial insufficiency subgroup included 11 cases of bacterial infection (5 of G(+) cocci and 6 of G(-) bacilli) and 12 cases of nonbacterial infection; their PCT levels were 0.66(0.19-5.80) µg/L and 0.08(0.05-0.20) µg/L, respectively (P < 0.01). There was no statistically significant difference among PCT levels of the 4 subgroups of nonbacterial infections (4 tuberculosis, 3 fungi, 3 atypical pathogens, 2 viruses) (F = 3.025, P = 0.094). The cut-off values of PCT were >0.13 µg/L and >0.115 µg/L for differentiating non-responsiveness originated from bacterial infection or other causes, and infection vs non-infection, which yielded a sensitivity of 100% (11/11) and 65% (14/23) , specificity of 83% (19/23) and 91% (10/11) , and AUC of 0.955 and 0.802, respectively.

CONCLUSIONS: Antibiotic failure to cover the microbial pathogens, infectious complications and misdiagnosis are the most common causes of early treatment failure in patients with CAP. Serum PCT level fails to predict non-responsiveness, but is suggestive of bacterial infections in hospitalized CAP patients with early treatment failure.