[Non-conditional logistic regression analysis on risk factors of central venous catheter-related infection].

Objective: To study the correlation of indwelling central venous catheter and systemic infection, to study the causative microorganisms and factors contributing to infections, and the diagnosis and treatment of such infection.

Methods: All the central venous catheters were produced by American or German companies. Subclavian vein, internal jugular vein, or femoral vein was chosen. Skin swab, blood in the catheter and peripheral blood were cultured. Clinical data including sex, age, disease, times of puncture, site of puncture, indwelling duration, type of catheter, caliber of catheter (single or double chamber, hemodialysis), nursing care, ways of use of the catheter were recorded, and they were subjected to non-conditional logistic regression analysis, then independent risk factors were determined with multivariate stepwise non-conditional logistic regression analysis.

Results: Unvaried non-conditional logistic regression analysis showed that sex, primary disease, site of puncture, nursing care, and ways of use of the catheter showed no correlation with CRS. Multivariate stepwise non-conditional logistic regression analysis showed that indwelling duration [odds ratio (OR)=2.682, 95% confidence interval (CI): 1.156-6.226, chi2=5.275, P=0.022], caliber of catheter (OR=2.702, 95% CI: 1.183-6.175, chi2=5.559, P=0.018) and white blood cell count (WBC)<4x10(9)/L (OR=1.371, 95% CI:1.044-1.800, chi2=5.142, P=0.023) were independent factors. The clinical study of 1,100 patients showed that the total correlated infection rate was 2.5% (28/1,100).

Conclusion: Central venous catheterization is the major cause of bacteremia in hospital. The risk factors of infection are maintenance duration, location, and individual difference, and they are obviously correlated.