**3.1 Patient demographics** **and ****p****athogens**

We analyzed 60 preterm and 33 full-term neonates, none of which tested positive for both pneumonia and extrapulmonary infections; thus analysis was conducted only on six groups. See Fig. 1 for numbers enrolled in, and excluded from, each group. The majority of preterm neonates (32/60) analyzed were male. The mean gestational age of preterm and full-term neonates was 28.9±2.2 weeks and 39.5±1.7 weeks, respectively.

The diseases identified in preterm neonates included neonatal respiratory distress syndrome, necrotizing enterocolitis, neonatal hyperbilirubinemia, neonatal cerebral ischemia, and neonatal septicemia. The diseases identified in full-term neonates included neonatal hypoxic ischemic encephalopathy, neonatal hyperbilirubinemia, congenital lactic acidosis, neonatal septicemia, and congenital megacolon.

**3.2 Indicators**

In terms of day 4 BAL-fluid sTREM-1 levels (Fig. 2a), there was a difference between the ptP+E- (508.25±180.70 pg/ml) and ptP-E- (115.50±49.01 pg/ml) groups (*t* = 9.0135, P <0.05), as well as between the ptP+E- and ptP-E+ (140.89±53.19 pg/ml) groups (*t* = 5.2855, P <0.05), but not between the ptP-E- and ptP-E+ groups (*t* = 1.1370, P >0.05). There was a difference between the ftP+E- (742.57±207.76 pg/ml) and ftP-E- (106.00±33.01 pg/ml) groups (*t* = 12.3704, P <0.05), as well as between the ftP+E- and ftP-E+ (138.64±40.86 pg/ml) groups (*t* = 6.9711, P <0.01), but not between the ftP-E- and ftP-E+ groups (*t* = 1.7552, P >0.05). There was also a difference between the ptP+E- and ftP+E- groups (*t* = 4.1855, P <0.05), but between neither the ptP-E- and ftP-E- groups (*t* = 0.5458, P >0.05), nor the ptP-E+ and ftP-E+ groups (*t* = 0.0843, P >0.05).

Regarding day 4 serum sTREM-1 levels (Fig. 2b), there were differences between the ptP+E- (209.85±102.64 pg/ml) and ptP-E- (75.50±27.93 pg/ml) groups (*t* = 5.4276, P <0.05), between the ptP+E- and ptP-E+ (487.24±305.65 pg/ml) groups (*t* = 4.4207, P <0.05), as well as between the ptP-E- and ptP-E+ groups (*t* = 5.8549, P <0.05). There were differences between the ftP+E- (480.58±180.83 pg/ml) and ftP-E- (67.75±20.47 pg/ml) groups (*t* = 5.8906, P <0.05), as well as between the ftP-E- and ftP-E+ (623.69±167.12 pg/ml) groups (*t* = 10.6367, P <0.05), but not between the ftP+E- and ftP-E+ groups (*t* = 1.6963, P >0.05). There was also a difference between the ptP+E- and ftP+E- groups (*t* = 6.8963, P <0.05), but between neither the ptP-E- and ftP-E- groups (*t* = 0.7749, P >0.05), nor the ptP-E+ and ftP-E+ groups (*t* = 0.9721, P >0.05).

Upon comparing day 4 BAL-fluid and serum sTREM-1 levels, there were differences within each group (group ptP+E-: *t* = 8.4948, P <0.05; group ptP-E-:* t* = 3.0084, P <0.05; group ptP-E+: *t* = 2.9537, P <0.05; group ftP+E-: t=3.9289, P <0.05; group ftP-E-: t=3.1141, P <0.05; and group ftP-E+: t=6.9060, P <0.05).

With respect to day 4 white blood cell count (Fig. 2c), there were differences between the ptP+E- (11,358.46±4,704.71/mm3) and ptP-E- (7,728.83±1,818.86 /mm3) groups (*t* = 4.7774, P <0.05), between the ptP+E- and ptP-E+ (17,867.51±1,340.04/mm3) groups (*t* = 6.9304, P <0.05), as well as between the ptP-E- and ptP-E+ groups (*t* = 9.8140, P <0.05). There were differences between the ftP+E- (17,374.20±1,361.47/mm3) and ftP-E- (6,493.33±733.33/mm3) groups (*t* = 29.4420, P <0.05), as well as between the ftP-E- and ftP-E+ (17,658.22±1,849.32/mm3) groups (*t* = 23.3148, P <0.05), but not between groups ftP+E- and ftP-E+ (*t* = 0.6450, P >0.05). There were also differences between groups ptP+E- and ftP+E- (*t* = 5.3034, P <0.05), as well as between groups ptP-E- and ftP-E- (*t* = 2.5348, P <0.05), but not between groups ptP-E+ and ftP-E+ (*t* = 0.2363, P >0.05).

With reference to day 4 serum C-reactive protein levels (Fig. 2d), there were differences between the ptP+E- (6.87±2.61 ng/ml) and ptP-E- (0.65±0.32 ng/ml) groups (*t* = 10.0254, P <0.05), between the ptP+E- and ptP-E+ (3.61±1.49 ng/ml) groups (*t* = 3.1819, P <0.05), as well as between groups ptP-E- and ptP-E+ (*t* = 6.1642, P <0.05). There were differences between the ftP+E- (8.92±2.14 ng/ml) and ftP-E- (1.52±1.47 ng/ml) groups (*t* = 7.1073, P <0.05), between the ftP+E- and ftP-E+ (5.28±3.98 ng/ml) groups (t=2.8448, P <0.05), as well as between groups ftP-E- and ftP-E+ (*t* = 2.7430, P <0.05). There was also a difference between groups ptP+E- and ftP+E- (*t* = 2.8028, P <0.05), as well as between groups ptP-E- and ftP-E- (*t* = 2.4435, P <0.05), but not between groups ptP-E+ and ftP-E+ (*t* = 1.0350, P >0.05).

Concerning day 4 serum procalcitonin levels (Fig. 2e), there was a difference between the ptP+E- (4.54±1.36 ng/ml) and ptP-E- (0.15±0.19 ng/ml) groups (*t* = 19.4827, P <0.05), as well as between the ptP-E- and ptP-E+ (5.38±0.94 ng/ml) groups (*t* = 15.1137, P <0.05), but not between groups ptP+E- and ptP-E+ (*t* = 2.6115, P >0.05). There were differences between the ftP+E- (4.86±1.29 ng/ml) and ftP-E- (0.39±0.45 ng/ml) groups (*t* = 9.5044, P <0.05), between the ftP+E- and ftP-E+ (9.23±3.32 ng/ml) groups (*t* = 7.7982, P <0.05), as well as between groups ftP-E- and ftP-E+ (*t* = 14.5059, P <0.05). There was also a difference between groups ptP+E- and ftP+E- (*t* = 1.9879, P <0.05), as well as between groups ptP-E- and ftP-E- (*t* = 0.8090, P <0.05), but not between groups ptP-E+ and ftP-E+ (*t* = 2.9529, P >0.05).

Bacterial culture results from serum and BAL fluid indicated the presence of various bacteria, as summarized in Table 1.

**3.3 Receiver operating characteristic analysis**** of indicators for pneumonia**

Fig. 3 illustrates the ROC analysis performed to determine the diagnostic value of indicators in groups ptP+E- and ftP+E-. Table 2 summarizes the results in terms of area under the curve (AUC), cutoff value, sensitivity, and specificity. For both groups, the AUCs of BAL-fluid sTREM-1, serum sTREM-1, and serum C-reactive protein were statistically significant (P <0.05). The AUCs of serum procalcitonin and BAL-fluid/serum sTREM-1 were only statistically significant for group ptP+E- (P <0.05). The AUC of white blood cell count was not statistically significant for either group (P >0.05).

We constructed multivariate logistic regression models incorporating different pairs of markers. Using prediction probability as diagnostic indicator, we determined diagnostic cutoff values from the ROC curve. The AUCs of the two-marker models were all larger than that of any single marker model. The optimal combination was BAL-fluid sTREM-1 and serum C-reactive protein concentration, for which AUC (95% confidence interval [CI]), sensitivity, and specificity was 0.968 (0.909-0.999), 98.9%, and 72.7%, respectively, in group ptP+E-, and 0.926 (0.918-1.000), 95.6%, and 92.4%, respectively, in group ftP+E-.

**3.4 Repeated measures ANOVA ****of indicators for ventilator-associated pneumonia**

In the ptP+E- and ftP+E- groups, 19/35 and 11/17 neonates, respectively, recovered. All indicators were more favorable at the third and fourth time points in neonates who recovered than in those who did not, in each group (all P <0.05; Table 3). We observed no differences at any time point for serum or BAL-fluid sTREM-1 concentrations between the ptP+E- and ftP+E- groups (all P >0.05), and none of the values changed over time in ptP+E- or ftP+E- neonates who did not recover (all P >0.05).