Skip to main content

Table 3 Univariate and logistic regression model for clinical response

From: An area under the concentration–time curve threshold as a predictor of efficacy and nephrotoxicity for individualizing polymyxin B dosing in patients with carbapenem-resistant gram-negative bacteria

Variable

Failure

(n = 144)

Response

(n = 249)

Pa

OR (95% CI)

Pb

Age, years

58.0 (51.0–67.0)

54.0 (47.0–63.0)

0.009

0.98 (0.96–1.0)

0.012

SOFA score

8.0 (7.0–12.0)

8.0 (5.0–10.0)

0.003

  

APACHE II score

19.0 (14.0–23.0)

17.0 (11.0–21.0)

0.001

  

Septic shock

69 (47.6%)

84 (33.9%)

0.007

0.55 (0.33–0.93)

0.025

Vasoactive drugs

87 (60.8%)

108 (44.4%)

0.002

  

Use of nephrotoxic drugs

117 (80.7%)

172 (69.4%)

0.014

0.44 (0.24–0.83)

0.011

Laboratory data

GFR, mL/min·1.73m2

96.0 (51.8–114.4)

102.7 (73.7–118.0)

0.049

  

Albumin, g/L

29.2 (26.2–33.2)

30.5 (27.3–35.0)

0.080

  

Platelets, 109/L

139.0(73.0–244.0)

172.5 (85.5–271.0)

0.008

  

C-reactive protein, μg/L

89.4 (45.9–150.0)

67.6 (30.8–130.7)

0.072

  

Procalcitonin, ng/mL

1.2 (0.5–4.0)

0.9 (0.3–3.5)

0.080

  

Polymyxin B dose and exposure

Dose/weight, mg/kg/day

2.0 (1.6–2.3)

2.1 (1.8–2.5)

0.011

  

Daily dose ≥ 150 mg/day

82 (56.6%)

185 (74.6%)

 < 0.001

  

AUCss,24 h ≥ 50 mg h/L

63 (43.4%)

175 (70.6%)

 < 0.001

4.39 (2.56–7.47)

 < 0.001

cC0h > 1.01 mg/L

79 (54.9%)

169 (67.9%)

 < 0.001

  
  1. OR odds ratio; CI confidence interval; GFR glomerular filtration rate; and AUCss,24 h the area under the curve across 24 h at steady state
  2. aderived from univariate analysis
  3. bderived from logistic regression analysis
  4. cC0h was not included in the logistic regression model due to collinearity with AUCss,24 h