TABLE 4

Bivariate Analysis Illustrating the Proportion of Children Who Had Blood Culture Obtained, Had Local Guideline-Appropriate Decision, and Had Bacteremia, N = 330

Characteristic% Getting CultureP% Appropriate DecisionP% True Positive Culture ResultP
Age, y.12.07NRa
 <251712.9
 3–539811.3
 6–1142830
 12–1757670
Disposition<.0001<.0001.02b
 Discharge from PED11930
 Admitted from PED75601.1
 Admitted from OSH69663.7
Chest radiograph findings<.0001<.0001<.0001b
 Normal45820
 Viral versus RAD47660
 Focal infiltrate44730
 Effusion/empyema969617.9
WBC count (×103/μL)<.0001.09<.01
 >15 00077685.6
 <15 00039780.4
  • OSH, outside hospital; PED, pediatric emergency department; RAD, reactive airway disease.

  • a P value not reported (NR) because χ2 is not a valid test with frequency of true positive blood culture result of zero in 6- to 11-year-olds and 12- to 17-year-olds. The Fisher exact test was not appropriate for 4 × 2 table.

  • b Fisher exact test was run for these variables due to cell frequencies of zero. For disposition, patients admitted from PED and OSH were collapsed into 1 group. For chest radiograph findings, normal, viral versus RAD, and focal infiltrate were collapsed into 1 group.