Skip to main content

Table 3 Bivariate and Multivariable logistic regression analysis for predicting the independent contribution of potential predictors of antimicrobial resistance among patients with UTIS

From: Trends and predictors of antimicrobial resistance among patients with urinary tract infections at a tertiary hospital facility in Alexandria, Egypt: a retrospective record-based classification and regression tree analysis

Variables

Drug Resistance

Yes

n(%)

No

n(%)

Unadjusted OR (95%CI)

Sig.

Adjusted OR

(95%CI)

Sig.

Hospital acquired UTI(N = 220)

197(89.5%)

23(10.5%)

5.3

(3.2-9)

P < 0.001*

5.7

(3.058–10.76)

P < 0.001*

Female(N = 218)

173(79.4%)

45(20.6%)

1.3

(0.83–2.14)

0.236

  

Age:≥60(N = 101)

89(88.1%)

12(11.9%)

2.7

(1.4–5.22)

0.003*

2.9

(1.266–6.720)

0.012*

Tumor:(N = 108)

92(85.2%)

16(14.8%)

2

(1.11–3.64)

0.021*

  

Stones:(N = 164)

117(71.3%)

47(28.7%)

1.8

(1.09–2.82)

0.021*

  

Diabetes mellitus:(N = 88)

84(95.5%)

4(4.5%)

8.3

(2.9–23.4)

P < 0.001*

3.8

(1.243–11.837)

0.019*

Abnormal structure of urinary tract:(N = 81)

61(75.3%)

20(24.7%)

0.9

(0.5–1.56)

0.663

  

Use of urinary catheter:(N = 190)

154(81.1%)

36(18.9%)

1.6

(0.96–2.5)

0.072

  

Recurrent UTI: (N = 246)

220(89.4%)

26(10.6%)

6.7

(3.9-11.25)

P < 0.001*

2.6

(1.083–6.202)

0.033*

Continuous stay in hospital:≥15 days(N = 73)

64(87.7%)

9(12.3%)

2.4

(1.15–5.1)

0.020*

  

Previous use of antibiotics within last three months:(N = 284)

252(88.7%)

32(11.3%)

9.3

(5.5-15.87)

P < 0.001*

5.2

(2.186–12.307)

P < 0.001*

PCNL:(N = 121)

83(68.6%)

38(31.4%)

1.9

(1.19–3.18)

0.008*

  
  1. *Statistically significant variables by bivariate analysis were initially included in multivariate logistic regression using backward likelihood ratio method were stones, tumor, diabetes mellitus, recurrent UTI, duration of stay in hospital, previous use of antibiotics, old age, patients with hospital- acquired UTI, PCNL.
  2. *44.8% of the variance of the drug resistance outcome is explained by significant independent variables included in the model (McFadden R2 = 0.448), Model Summary:X2 = 136.022, P < 0.001
  3. Hosmer-Lemeshow X2 = 8.743, P = 0.189, PCNL: Percutaneous nephrolithotomy