Background
Long-term central venous catheters (CVC) are essential for management of chronic medical conditions, including childhood cancer. Serious CVC complications, including bloodstream infection, venous thrombosis and fracture, are associated with preceding occlusion. Prediction and pre-emptive treatment of occlusion might help prevent these complications, but no validated predictive technique is available.
Methods
We conducted a prospective trial to determine feasibility, acceptability and efficacy of a novel technique to predict CVC occlusion in paediatric patients by catheter resistance monitoring (CRM). Participants with long-term tunnelled CVCs underwent weekly CRM for up to 12 weeks. At each visit, resistance was assessed by measuring inline pressure at 4 different flow rates. Resistance was estimated from the gradient of a pressure-flow curve. Resistance was not estimated if turbulent CVC flow was suspected (R2 <85%), and results of these visits were categorized as ‘indeterminate’.
Findings
Ten patients were evaluated for a total of 113 CRM visits. Measurements were obtained at all visits. A change in CVC resistance of >8.8% between measurements was strongly associated with developing acute CVC occlusion over the subsequent 10 days (OR=6.2, p<0.01; sensitivity 75%, specificity 67%), and a combined prediction model comprising either change in resistance >8.8% or ‘indeterminate’ result also predicted subsequent occlusion (OR=6.8, p=0.002; Sensitivity 80%, Specificity 63%). The intervention was rated as highly acceptable by participants.
Interpretation
In this paediatric oncology population, CRM is feasible, acceptable, and predicts subsequent CVC occlusion. Larger studies are required to assess predictive value for other clinical outcomes and the impact of pre-emptive therapy.