SEQuential tUnneled Catheters and Infective Events in haemodiaLysis: A Five-Year Retrospective Cohort Study (SEQUEL Study)
Shing Shen Bay, Suek Xuan Pon, Xin Yu See, Soo Ying Yew, Elisya Liyana Abdullah, Khai Vern Poon, Chee Keong Thye, Shian Feng Cheng, Yee Wan Lee, Maisarah Jalalonmuhali, Wan Ahmad Hafiz Wan Md Adnan, Chye Chung GanBackground
Tunnelled-cuffed catheters (TCCs) are widely used for haemodialysis access when permanent vascular access is unavailable. Internal jugular vein (IJV) placement is guideline-preferred, but femoral vein (FV) catheters are frequently used when upper venous access is limited. Whether femoral access confers a higher infection risk than catheter sequence number remains unclear. This study evaluated these two competing determinants of catheter-related bloodstream infection (CRBSI).
Methods
This single-center retrospective cohort study included all TCC insertions at a tertiary nephrology unit between January 2016 and December 2021. The primary analysis examined CRBSI incidence per 1000 catheter-days stratified by catheter sequence number (first, second, third+ insertion per patient). Secondary analyses compared femoral versus internal jugular TCCs overall and within sequence strata. A registry-index cohort (patients with no documented prior TCC before January 2016, n = 135) was used to reduce, though not eliminate, sequence-related confounding in the site comparison; given only 24 CRBSI events, this analysis is exploratory and a clinically meaningful site difference cannot be excluded. Kaplan–Meier methods with log-rank testing were used for time-to-event outcomes. A cluster-robust marginal Cox model was used to account for repeated catheter insertions within patients, adjusting for sequence number, insertion site, age, sex, insertion setting, and exchange from a non-tunnelled catheter.
Results
Among 626 TCC insertions in 339 patients (389 FV, 237 IJV), contributing 98,727 catheter-days, CRBSI incidence escalated with catheter sequence: 1.17 per 1000 catheter-days for first insertions, rising to 1.98 for second and 2.49 for third-or-more insertions. Notably, by the second insertion, CRBSI rates were similar between FV and IJV access (31.7% each). In the index cohort (n = 135; FV = 95, IJV = 40), CRBSI-free survival did not differ significantly between sites (log-rank p = 0.340), nor did catheter survival (p = 0.950) or patient survival (p = 0.487). In the all-TCC cohort, after adjustment for catheter sequence number, age, sex, insertion setting, and exchange from a non-tunnelled catheter using a cluster-robust marginal Cox model, femoral access remained associated with a higher, though attenuated, CRBSI hazard after adjustment for the measured covariates (HR 1.43, 95% CI 1.02–2.01; p = 0.039), indicating a modest residual site hazard beyond the sequence effect.
Conclusions
Repeated catheterization was independently associated with increasing CRBSI risk, and catheter sequence number was the strongest predictor in the fully adjusted model. The femoral site effect was substantially attenuated after adjustment for sequence and was not statistically significant in the registry-index cohort analysis, though the wide confidence intervals preclude an equivalence interpretation. IJV access remains preferred; when IJV is unavailable, femoral TCC placement may be an acceptable alternative from an infection-risk perspective, but this must be weighed against thrombotic and limb complications, including deep vein thrombosis, which were not systematically ascertained in this study. Residual confounding cannot be excluded in this single-center retrospective study.