Background and aim Catheter-related sepsis (CRS) is an important complication affecting neonates with indwelling percutaneously-inserted central venous catheters (PCVCs). Culture of a catheter tip sent at line removal is commonly used to assist diagnosis of CRS. We examined the relative utility of microbiological culture of three discrete segments of a PCVC for assisting diagnosis of neonatal CRS.

Design/methods Prospective study of infants with PCVCs indwelling >24 h in two tertiary neonatal intensive care units. PCVC skin exit site was disinfected before line removal and the PCVL cut into three 1 cm-long segments comprising tip, proximal portion (taken 1–2 cm beyond skin entry point) and mid-portion. Our case definition for definite CRS was signs of clinical sepsis associated with culture-positivity of both a peripheral blood sample plus any line segment.

Results 189 PCVCs were removed over a 14-month period. 47 neonates were clinically-septic at the time of line removal of whom 15 had definite CRS, an overall rate of 7.9% (15/189). In well babies, bacterial colonisation rates were significantly higher for proximal segments compared with tips (p=0.004). However comparative rates of culture-positivity did not differ significantly between line segments in clinically-septic babies overall (p=0.14) or in cases of definite CRS (p=0.14) and the positive predictive values for definite CRS were similar for all three line segments.

Conclusion While the proximal portion of a PCVC line is more often colonised than a line tip, preferential culture of a proximal line portion in clinically-septic babies may not permit better prediction of definite CRS.