Tunneled catheters (TCs) for hemodialysis (HD) have been positioned as alternatives to native arteriovenous fistulas. Although the KDOQI1 guidelines limit their use to very specific situations and do not consider them as a first treatment option because of their high morbidity and mortality, TCs represent an indispensable form of vascular access in most nephrology departments.

According to the Information System of the Autonomous Coordination of Transplants of Andalusia (SICATA),2 43% of patients who started HD in 2024 in the Andalusian community did so through the use of a TC. In the FORTHCOMING study, 45% of patients started HD in 2008 through a central venous catheter, with an annual increase in the percentage of catheter use in incident patients from 24.7% to 29.5%.3 According to data from the Registry of Renal Diseases of Catalonia, since 2009, fewer than 50% of patients have started HD in Catalonia through a native arteriovenous fistula.4 Recent data are available at a global level; specifically, according to the 2023 report of the US Renal Data System, between 2018 and 2022, the percentage of people who started HD with a catheter (with or without permanent access in maturation) increased by 3.9% and reached a value of 84.7%.5

The main limitation for the use of TC versus arteriovenous fistulas has involved their increased risk of infection. Infection is among the main causes of mortality in patients receiving HD, with bacteremia related to the tunneled catheter (BRC) being among the most serious complications observed in patients with catheter dependence. In 2020, 21.7% of patients receiving HD in Andalusia died from infectious causes.2 The 2023 report of the US Renal Data System5 reported that 14.7% of deaths were due to infectious causes. More than half (55.9%) of deaths with known causes have been reported to be related to cardiovascular diseases (including stroke).5 These records do not identify the focus of infection; therefore, the mortality rate secondary to BRC is unknown.

The widespread use of TCs will likely continue in the future. The challenge for nephrologists involves the development of safe and effective interventions that minimize the risk of complications associated with TCs. The few available studies that attempted to analyze the optimal combination of prophylactic measures to avoid infection in patients using TCs had limited sample sizes and follow-up periods.

The objectives of our study were to describe the clinical and demographic characteristics of patients who received a TC for HD in the healthcare area of a tertiary hospital; to determine the incidence and etiology of BRC; and to analyze the impact of prophylaxis measures prior to implantation of TCs and the survival of this group of patients over a long follow-up period.

The study complied with the guidelines established by the Declaration of Helsinki and the Declaration of Istanbul. The collected data were anonymized. This research was approved by the local Ethics Committee of the Virgen Macarena University Hospital (Appendix B; Supplementary Material 1). Given the retrospective nature of the study, informed consent was waived.

BRC is among the most serious complications of patients receiving HD and is associated with high morbidity and mortality. Our study investigated the clinical characteristics of a population of patients receiving HD with TCs, as well as the incidence and etiology of BRC and their mortality, with the premise that catheter implantation was performed by nephrologists who strictly followed a protocol that was agreed upon by the infectious diseases department.

The overall incidence rate of BRC in our study was very low (0.36 per 1,000 days of TC use) compared with the literature, which describes rates between 0.2 and 5.5 episodes per 1,000 days of catheterization according to previous studies.6,7

More than 80% of BRC cases were caused by gram-positive microorganisms, which is consistent with the traditionally reported etiology.8,9 However, unlike that in the available literature, the main isolated microorganism was S. epidermidis (41.1%), followed by S. aureus (27.4%). The detection of methicillin-resistant strains was anecdotal (observed in only 3 of the 95 cases during a follow-up period of 20 years) and much lower than that reported in the literature, wherein the rate of detection of methicillin-resistant S. aureus is approximately 10%.9 These results could be related to prevention strategies for methicillin-resistant S. aureus, among which the detection and treatment of patients who are nasal carriers are noteworthy. The clinical guide of the Spanish Multidisciplinary Group of Vascular Access (Grupo Español Multidisciplinar del Acceso Vascular, or GEMAV)10 does not recommend the systematic detection and decolonization of patients who are nasal carriers of S. aureus because of the possible development of antimicrobial resistance and the limited evidence of its benefit in patients receiving HD.

This supposed increase in antibiotic resistance was not detected in our study. Our protocol (in accordance with the infectious diseases department and following hospital policy to prevent resistance) includes only one decolonization regimen using topical mupirocin. Patients in whom the nasal smear continues to demonstrate positive results despite this first round of topical antibiotic therapy are referred to the infectious diseases department, where the need for oral treatment will be assessed after considering the clinical characteristics of the patient, the risk of infection related to catheter use and the risk of developing antibiotic resistance.

Moreover, in 2021, Vanegas et al.11 published a prospective study involving 210 patients receiving HD with a central venous catheter who were screened for S. aureus at the beginning of the treatment and at 2 and 6 months of follow-up. Among the 141 patients who completed the 3 S. aureus screenings, 44.7% (n = 63) were intermittent carriers, and 12.8% (n = 18) were persistent carriers. Fifty patients (23.8%) developed bacteremia, most of which was caused by S. aureus (n = 28; 39.4%). Four patients developed methicillin-resistant S. aureus bacteremia (1.9%), all of whom were colonized and developed recurrent infections due to this microorganism. Survival analysis revealed no association between BRC and initial colonization by S. aureus; however, it did reveal a significant association when the state of colonization was included at 2 and 6 months (hazard ratio: 4.93; 95% CI [1.89–12.88]). The authors concluded that colonization by S. aureus increases the risk of BRC and the recurrence of S. aureus infection. It is necessary to continue generating scientific evidence to review current protocols and propose the addition of decolonization as a necessary strategy to reduce infections, serious complications and increased costs in these patients.

The risk of infection was proportional to the duration of catheter use. The majority of the BRC cases observed in our study occurred after 6 months of TC implantation, with only 4 (4.2%) being detected in the first 30 days after implantation. In previous studies, an infection peak was commonly detected in the first 30 days of catheter placement, which is possibly due to insufficient asepsis during the procedure and the presence of S. aureus on the patient’s bodily surface. TC manipulation in the first 24 h also favors infection; hence, another measure used in our study involved waiting 24 h until the first use of the catheter.

On the basis of these data, we believe that the following 3 key points of our protocol—screening and treatment of nasal carriers of S. aureus, bathing with chlorhexidine and prophylactic antibiotic therapy in situations with increased risk of infection—have allowed us to (1) decrease the incidence of S. aureus and shift the germ most frequently related to BRC to S. epidermidis, which has less potential to be considered as an opportunistic pathogen and demonstrates less virulence and a lower rate of serious complications than S. aureus does; (2) avoid an increase in the incidence of antibiotic-resistant strains; and (3) delay the onset of BRC.

Our low overall incidence rate, the minimal percentage of BRC detected in the first 90 days and the lower isolation rate of S. aureus (with a minimal percentage of methicillin-resistant strains, during the years of follow-up could indicate that the prevention strategies implemented in our UGC have contributed to these results. Notably, as previously mentioned, the screening and decolonization of patients who are nasal carriers of S. aureus is a measure that has been discouraged by expert consensus in clinical guidelines 10 owing to the probability of increasing antibiotic resistance, although scientific evidence of this is limited.

During the follow-up period, 75.6% of the patients died from any cause. Infection was determined to be the leading cause of death in our cohort. However, only 9 of the 325 patients included in the study died from BRC, which represents only 2.7% of the study population. The data reported on mortality in patients with TCs are very limited. Xue et al.12 evaluated the mortality of a cohort of 37,826 incident patients receiving HD via a central venous catheter in the United States between 1995 and 1997. The mortality rates were 15.1% at 90 days, 26.9% at 6 months, and 41.5% at one year. More recently, Shimamura et al.13 analyzed a cohort of 64 patients (representing a smaller number of patients) receiving HD with TCs between 2012 and 2019. At 2 years, 27 patients had died (overall mortality rate of 42%). Although the mortality rate of approximately 41% detected in both studies was slightly lower than that described in our analysis, the short follow-up period results in difficulty in making a direct comparison with our results, as the average survival exceeded 5 years.

To obtain more long-term data, we must access medical records. According to the Spanish Registry of Renal Patients (REER),14 the survival of patients receiving HD regardless of vascular access is 52.2% at 5 years. The European Registry,15 with a survival rate that is somewhat lower than 5 years (46.7%), allows patients to be stratified according to age. If we focus on the age range that more closely matches that of the patients in our study (65–74 years), the 5-year survival rate decreases to 42.2%. As in the case of the REER data, this percentage includes patients receiving HD regardless of vascular access, specifically including those with an arteriovenous fistula, in which a higher survival would be expected. The data stratified by age are very similar to those reported in our study, with a survival of 37.9% at 5 years being reported. Based on these data and compared with the survival observed in our study within the corresponding age range, it can be concluded that the use of TCs in our population did not negatively impact 5-year survival. Therefore, in the context of patients with limited life expectancies and based on our results, the option of TCs could be seriously considered, carefully individualizing the decision between the creation of an arteriovenous fistula and the use of a TC and assessing the balance between risks, benefits and quality of life.

This study has several limitations that should be considered when the results are interpreted. First, due to the retrospective nature of the study, residual confounding factors may exist, despite the adjustments for demographic, clinical and catheter-related variables. Second, this study was developed in a single center in which TCs are exclusively implanted by nephrologists, which limits the possibility of generalizing the results to other hospitals or care centers where the TC can be implanted by other specialists. The differences in clinical practice, the organization of health services and the specific characteristics of each center can influence the applicability of the results to other populations. Third, we did not include a control group of patients for whom the TC implantation protocol was not performed or who received another type of vascular access. Additional prospective studies are needed to determine other possible risk factors for BRC, as well as the impacts of additional prevention and management measures.

The present study has 3 main strengths, including a long and careful follow-up period, a large sample size and a detailed evaluation of possible confounding factors, including comorbid conditions, variables related to TCs and data from a specific preventive protocol.

The use of a specific protocol for the implementation of a TC based on infection prevention strategies was associated with a low incidence of BRC. These infections were characterized by a late onset (beginning at the sixth month after implantation) and were predominantly caused by S. epidermidis, which is a microorganism that is less virulent than S. aureus, thus likely causing less severe episodes of BRC.

Among the adopted measures, the systematic screening and decolonization of nasal carriers of S. aureus contributed to a significant reduction in the number of BRC cases caused by this pathogen, and an increase in the proportion of methicillin-resistant strains over a long follow-up period was not observed.

In our cohort, the use of TCs was associated with low mortality attributable to BRC and did not demonstrate a relevant negative effect on overall survival at 5 years.

All of these results demonstrate that, in the context of a rigorous application of prevention and management protocols, TCs can be a valid and safe option in certain patient populations in whom the creation of an arteriovenous fistula is difficult. We believe that the obtained evidence requires a critical reconsideration of the role of TCs in the strategy of vascular access, beyond the traditional consideration of these catheters as an alternative of last resort.

This research has not received specific support from public sector agencies, the commercial sector or nonprofit entities.