Aims: to explore the value of TREM-1 (triggering receptor expressed on myeloid cells-1) ,PCT(Procalcitonin), as well as CPIS (clinical pulmonary infection score) in the diagnostic and prognostic assessment of VAP (ventilator associated pneumonia); and to make a comparison with WBC (white blood cells) and CRP (C-reactive protein) level as well as SOFA (Sequential Organ Failure Assessment) Score Methods: There were 92 subjects of sepsis, who were either receiving endotracheal intubation or had undergone tracheotomy and were exposed to mechanical ventilation. The subjects were divided into the VAP group (32) and the Non-VAP group (60), the criterion being the contraction of VAP 48 hours after ICU admission. Etiological culture was conducted in BALF (bronchoalveolar lavage fluid). And sTREM-1 density was determined by examining serum sTREM-1, PCT, WBC, CRP and EVC (exhaled ventilator condensate). Meanwhile, the CPIS and SOFA score were worked out. With a 28-day survival as the demarcation line, the VAP group was further divided into the survivors group, who stayed alive for 28 days or more , and the non-survivors group, who died within 28 days. The sTREM-1 and PCT level were denoted as meridians (range interquartile), while the WBC and CRP level as well as the CPIS and SOFA score, means±standard deviations (SD).
Results: Averagely, the patients would contract clinically-confirmed VAP 6.9 days after admission, which was mainly traced to Gram-negative bacilli infection. On the very day of diagnosis, compared with the Non-VAP group, the VAP group showed a higher level of serum sTREM-1, PCT, WBC and CRP as well as CPIS and SOFA score（295.6pg/ml vs.143.5pg/ml, P<0.001；4.5ng/ml vs. 1.4ng/ml，P=0.008；16.7×10∧9/L vs.10.9×10∧9/L, P<0.001；11.5mg/dl vs. 7.7mg/dl，P=0.012； 6.0vs. 1.9, P<0.001；10.0vs. 7.5, P=0.017）, AUC (area under the receiver operating characteristic curve）turned out as follows :sTREM-1: 0.73（95% CI 0.61-0.85）;PCT : 0.70（95% CI 0.57-0.83）;WBC: 0.73（95% CI 0.60-0.85）.The CPIS score, which was proved by logistic regression analysis as the sole risky factor to VAP, amounted to 0.96（95% CI 0.91-1.00）. Combined prediction probability containing all the data was calculated in accordance on the relative regression equation. sTREM-1+WBC+CPIS score proved to be most reliable for diagnosis. AUC turned out as 0.98. With 0.277 as the cut-off point, sensitivity measured 0.97, specificity, 0.9 and YDI, 0.87. There were only 5 VAP subjects whose sTREM-1 density could be detected in EVC. The VAP patients were divided into a survivors group (n=15) and a non-survivors group (n=17) with a 28-day survival as the demarcation line. The non-survivors group demonstrated a higher PCT level and higher CPIS & SOFA score than the survivors. （3.0ng/ml vs. 15.3ng/ml，P=0.032；5.4vs. 6.6, P=0.03；8.1vs.11.7 P=0.049). AUC worked out PCT 0.752（95% CI 0.547-0.956）and CPIS 0.764（95% CI 0.575-0.953）. Calculations on the regression equation showed the PCT+CPIS score was most reliable for prognostic assessment. AUC turned out as 0.848. With 0.516 as the demarcation line, sensitivity measured 0.867, specificity, 0.818 and YDI, 0.685.
conclusion: WBC + CPIS helps improve VAP diagnosis; PCT+CPIS may be used for VAP prognostic assessment. Taking two items into consideration will be of guiding value in VAP treatment as well as mortality rate reduction. The sTREM-1 level in EVC，however，may be devoid of value for VAP diagnosis.
Inclusion Criteria:
Exclusion Criteria: