Skip to main content

Table 4 Cox proportional hazard regression analysis to identify the factors associated with port-infection-free survival

From: Port type is a possible risk factor for implantable venous access port-related bloodstream infections and no sign of local infection predicts the growth of gram-negative bacilli

Clinical features Univariate analysis Multivariate analysis—maximal model Multivariate analysis—reduced model
HR (95 % CI) p value HR (95 % CI) p value HR (95 % CI) p value
Port type (TYCO vs. Bard) 4.759 (1.137–19.927) 0.0327 7.643 (1.795–32.535) 0.0059 7.105 (1.688–29.904) 0.0075
Gender (male vs. female) 2.849 (1.408–5.766) 0.0036 1.974 (0.934–4.172) 0.0748   
Age (>65 vs. ≤65 years) 2.262 (1.154–4.436) 0.0175 2.083 (1.043–4.157) 0.0375 2.320 (1.179–4.564) 0.0148
Malignancy (lung cancer vs. others) 5.121 (2.610–10.050) <0.0001 4.885 (2.362–10.104) <0.0001 5.807 (2.946–11.447) <0.0001
Surgeon (surgeon A vs. others) 0.862 (0.273–2.718) 0.8000 0.872 (0.271–2.810) 0.8184   
Surgeon (surgeon B vs. others) 0.508 (0.171–1.511) 0.2234 0.86 (0.285–2.594) 0.7887