Determinants of 90-Day Infection Risk After Tunneled Dialysis Catheter Placement: Patient, Device, and Procedural...

Tunneled hemodialysis catheters remain widely used but carry substantial infection risk. While patient and procedural factors are established drivers of infection, the contribution of catheter design features is less characterized. We evaluated associations between device features and 90-day systemic infection-related events after tunneled hemodialysis catheter insertion.

We conducted a retrospective observational study using deidentified electronic health record data from the Truveta database. We analyzed 47,750 tunneled hemodialysis catheter insertions among 37,074 patients between 2008 and 2024; 19,774 insertions among 16,517 patients met complete case criteria for primary adjusted modeling. We fit multivariable logistic regression models to estimate adjusted odds ratios (aORs), evaluating grouped device features and individual brands while adjusting for demographics, comorbidity burden, insertion site, catheter number, and calendar period.

In complete case brand-agnostic models, higher adjusted odds were observed for femoral insertion (aOR 1.71), fourth or later catheter placement (aOR 1.41), and pre-curved configuration (aOR 1.29). Lower adjusted odds were observed for age ≥ 80 years (aOR 0.70), Black patients (aOR 0.86), Hispanic patients (aOR 0.66), COVID (aOR 0.72) and post-COVID periods (aOR 0.66), catheter length ≤ 24 cm (aOR 0.84), and French size < 14.5F (aOR 0.34). In full-dataset sensitivity analyses, unknown race, unknown sex, unknown insertion site, second and third catheter placement, higher comorbidity burden, and split-tip design were associated with higher odds. Brand-specific models showed heterogeneity relative to GlidePath beyond broad design categories.

Infection risk after tunneled hemodialysis catheter insertion was most consistently associated with insertion context, catheter history, calendar period, and demographics. Device-related findings were smaller and varied by model specification, with brand-level differences not fully explained by broad design categories. Future studies should incorporate granular device descriptors and facility-level context to clarify which design choices translate into meaningful differences in infection risk.