Introduction

Hemodialysis (HD) catheters are crucial for initiating treatment in most patients with end-stage renal disease (ESRD). Central venous catheters (CVCs) used in HD are linked to a significant proportion of infections, with catheter-related bloodstream infections (CRBSIs) being a leading cause of mortality among these patients [1, 2, 3, 4]. The incidence of CRBSI ranges from 0.6 to 6.5 episodes per 1000 catheter days, influenced by factors such as the definition used, local catheter placement and care policies, and the duration of catheter use [1, 5, 6, 7], highlighting the importance of local epidemiological data for guiding prevention strategies. A recent study has highlighted a notable escalation in the incidence of catheter-related bloodstream infections (CRBSIs) caused by multidrug-resistant organisms (MDROs), with a prevalence of 44.5% [8]. This study addresses the limited regional data on antimicrobial resistance patterns in catheter-related bloodstream infections (CRBSI) among hemodialysis patients, focusing on a tertiary care center in Northern India.

This prospective observational study was conducted at the Department of Nephrology, a tertiary care center in Northern India, over a one-year period from September 2023 to September 2024. All patients initiated on hemodialysis through central venous double-lumen catheters (both cuffed and non-cuffed) were enrolled. Demographic data, clinical variables including age, gender, type of catheter placed (cuffed or non-cuffed), site and date of catheter insertion, and cause of ESRD were collected at the initial visit. Patients who developed symptoms and signs of CRBSI like fever, chills, hypotension, and /or catheter exit site discharge on follow-up, were hospitalized and evaluated using peripheral vein and central venous catheter blood cultures, alongside baseline blood tests. The 2006 National Kidney Foundation Kidney Disease Outcomes Quality Initiative (NKF KDOQI) Clinical Practice Guidelines for Vascular Access classification were followed for classifying culture-positive probable CRBSI and culture-negative possible CRBSI. All patients received empirical intravenous antibiotics targeting both Gram-positive (including Methicillin resistant Staphylococcus Aureus MRSA) and Gram-negative organisms, reflecting global guidelines for CRBSI management. Empirical Gram-positive cover was given by vancomycin or teicoplanin depending on physician discretion and empirical Gram-negative cover was given by piperacillin tazobactam or meropenem A 2D Echocardiography was done to rule out infective endocarditis when clinically indicated.

Statistical methods

Continuous variables were reported as mean ± standard deviation (SD), while categorical variables were expressed as percentages (proportions). Chi-square tests assessed associations between categorical variables. Mann-Whitney U tests compared non-normally distributed continuous variables. All analyses were performed using IBM SPSS Statistics for Windows (version 25, Armonk, NY, USA).

Results

A total of 240 catheters (in 240 patients) were inserted during the study period, cuffed and non-cuffed dialysis catheters accounted for 195/240 (81.3%) and 45/240 (18.8%) respectively. Men and women accounted for 141 (58.8%) and 99 (41.3%) respectively. Hypertension and diabetes mellitus were the most common comorbidities, present in 212 (88.3%) and 189 (78.8% respectively). The mean age was 52.5 ± 15.7 years. Most patients were in the 30 to 60-year age group, followed by 60 to 90 years, accounting for 142/240 (59.2%) and 75/240 (31.3%). A total of 48/240 (20%) developed CRBSI with 41/48 (85.4%) having culture-positive probable CRBSI and 07/48 (14.6%) having culture-negative possible CRBSI, as defined by the 2006 NKF-KDOQI criteria. Intradialytic chills followed by fever were the most common presenting complaints present in 91.7% (44/48) and, 89.5% (43/48) patients respectively Mean Age of CRBSI vs. Non-CRBSI Patients was 51.6 ± 15.6 and 56.1 ± 15.8 years respectively; p-value 0.15. Out of 48 cases of CRBSI, men and women accounted for 23 (48%) and 25 (52%) cases respectively; p-value 0.09. A total of 41/48 (85.4%) episodes occurred in patients with cuffed dialysis catheters and 07/48 (14.6%) in non-cuffed dialysis catheters; p-value 0.536. The mean catheter days were 177.5 ± 94.9 days. The median catheter days for un-cuffed and cuffed dialysis lines were 28 days and 210 days respectively p-value < 0.001. The mean and median time to CRBSI were 204.6 ± 87.1 days and 231.5 days respectively. The incidence rate for un-cuffed and cuffed dialysis lines was 4.31 and 1.32 episodes per thousand catheter days. Concomitant exit site infection was present in 09/48 (18.8%). The overall CRBSI incidence was 1.46 episodes per 1000 catheter days (48 cases/32,782 catheter days × 1000). When stratified by classification, culture-positive probable CRBSIs (n = 41) occurred at a rate of 1.25 per 1000 catheter days, while culture-negative possible CRBSIs (n = 7) had a lower incidence of 0.213 per 1000 catheter days. Hypertension and diabetes were the most common comorbidity present in 45/48 (93.8%) and 40/48 (83.3%) of CRBSI patients respectively. Gram-positive organisms were isolated in 15/48 cases (31.3%), while Gram-negative bacteria accounted for 26/48 cases (54.2%), and 7/48 cases (14.6%) had negative cultures. Coagulase-negative Staphylococcus aureus (CoNS) was the most common Gram-positive pathogen isolated (07/15; 46.7%), all were methicillin-resistant. Among the remaining Gram-positive isolates, methicillin-sensitive Staphylococcus-aureus (MSSA), methicillin resistant Staphylococcus aureus (MRSA), and Enterococcus species were identified. Klebsiella pneumonia, Pseudomonas, and Acinetobacter were the most common among Gram-negative organisms grown (n = 6 each; 23.1%), followed by Escherichia coli, Pantoea agglomerans, Stenotrophomonas Maltophillia and Serratia Marcescens. Among 8 cultured Gram-negative bacterial species, resistance patterns observed was Ampicillin: 4/4 tested (100.0%), Quinolones: 4/6 tested (66.7%), Cotrimoxazole: 3/5 tested (60.0%), Carbapenems: 3/6 tested (50.0%), Gentamicin: 2/5 tested (40.0%), Amikacin: 1/5 tested (20%), and Piperacillin-Tazobactam: 1/5 tested (20.0%). Most Gram-positive isolates demonstrated susceptibility to the empirically administered antibiotics, while a predominant proportion of Gram-negative organisms were identified as extended-spectrum beta-lactamase (ESBL)-producing strains. The CRBSI group had a mean age of 56.1 ± 15.8 years, while the non-CRBSI group averaged 51.6 ± 15.6 years; p-value 0.150. In 60 to 90-year age group, 17/75 (22.7%) developed CRBSI followed by 29/142 (20.4%) in 30 to 60-year age group. Men and women accounted for 23/48 (47.9%) and 25/48 (52.1%) of CRBSI cases respectively; p-value 0.08. Leukocytosis (>11,000 cells/mm³) was present in 44/48 (91.7%) patients and 3/48 (6.3%) had normal total leucocyte count.

Catheter salvage was achieved in 39 of 48 cases (81.3%). Concurrent exit-site infections requiring catheter removal occurred in 9 patients (18.8%) with CRBSI. No mortality was reported during the study period. The results are summarized in Tables 1,2,3,4,25,6.

Discussion

In resource-limited settings, low socioeconomic status, limited health literacy, and inadequate reimbursement systems contribute to delayed medical care and late referrals for arteriovenous fistula (AVF) creation, often leading to emergent dialysis initiation via central venous catheters (CVCs). This is reflected by the historical data where non-cuffed dialysis catheters accounted for about 80 to 100% of all central venous hemodialysis catheters in developing nations [9, 10, 11]. In our study cuffed tunneled double-lumen catheters accounted for 81.3% of all central venous hemodialysis catheters. These results are favorable reflecting a trend towards more use of tunneled cuffed double-lumen HD catheters. Non-tunneled catheters exhibit a 2–3 times higher CRBSI incidence compared to tunneled catheters, as demonstrated by multiple studies [12, 13] and Cuffed double-lumen hemodialysis catheters (e.g., PermCath) serve as effective intermediate-term vascular access while awaiting arteriovenous fistula (AVF) creation and maturation. A total of 48/240 (20%) patients developed CRBSI during the study period. Our results are better than Shingarev R et al. [14] where 54% had a CRBSI by six months of a newly placed tunneled hemodialysis catheter. This reflects improvement in aseptic technique and better catheter care over the years. However, the study by Shingarev R et al. [14] included more than 450 participants and all the patients had cuffed dialysis catheters as against our study which included both cuffed and non cuffed dialysis catheters. In our study 41/48 (85.4%) had culture-positive probable CRBSI and 07/48 (14.6%) had culture-negative possible CRBSI. Our culture positivity rate is better than the study done by Sethi et al. [15] where the culture positivity rate was only 62.2%. The improvement in culture positivity rates reflects early recognition of symptoms and appropriate collection of septic screens before empirical antibiotic exposure. All patients with culture-negative CRBSI in our cohort had received IV antibiotics in the community before the collection of septic screens which was the most likely reason for negative cultures. Intradialytic chills followed by fever were the most common presenting complaints present in 91.7% (44/48) and, 89.5% (43/48) of our patients respectively. The most common age group involved in our study was 60–90 years which is similar to a study reported by Tao et al. [16] and Murea et al. [17]. This phenomenon may stem from factors such as immune-senescence, atypical clinical presentations in older adults, and delays in diagnosis. Cuffed catheters showed numerically higher CRBSI rates but the difference was not statistically significant (p-value 0.536). This looks contrary to the expected. The likely reason for such a discrepancy in our study is the under-representation of patients with non-cuffed dialysis catheters and significantly fewer at-risk catheter days in non-cuffed dialysis catheters. This underscores the need for larger studies and risk-stratified protocols to clarify this association. The median catheter days for un-cuffed and cuffed dialysis lines were 28 days and 210 days respectively (p-value < 0.001). The historical incidence of CRBSIs associated with hemodialysis catheters falls within the range of 3–6 per 1000 catheter-days [18]. Studies report widely differing CRBSI rates, from 0.4 to 1.27 cases per 1000 catheter-days [19, 20, 21]. The CRBSI incidence rate in our cohort is 1.46 episodes/1000 catheter days (48/32,782 × 1000) compared to an incidence of 1.30 episodes/1000 catheter days by Mark A. Little et al. [22]. The high incidence rate can be decreased by implementation of standard catheter insertion and catheter care protocol. The median time to CRBSI in our cohort was 136.5 days, higher as compared to 24.5 days reported by Agrawal et al. [23]. The main reason for this discrepancy could be partly due to better infection control practices in our center and the fact that study of Agrawal et al. [23] is done in patients on non-tunneled catheters which are at higher risk of infection and have less longevity. While most of the centers in the world encounter predominantly Gram-positive CRBSIs [14, 24, 25], Gram-negative organisms accounted for the majority of catheter-associated bloodstream infections (CRBSIs) in our study. Similar results have been reported from many other Indian studies [9, 26]. Given the fecal origin of most Gram-negative pathogens, we hypothesize that suboptimal hygiene practices contribute to the high incidence of Gram-negative CRBSIs. The other reason for increased incidence of Gram negative CRBSI could be due to indiscriminate use of broad-spectrum antibiotics. Staphylococcus aureus emerged as the most frequently isolated pathogen in the study cohort. The most prevalent Gram-negative pathogens isolated included Klebsiella pneumoniae, Acinetobacter spp. and Pseudomonas spp. While Western literature frequently associates percutaneously inserted catheters with Staphylococcus-aureus and Enterobacteriaceae-related CRBSIs, this study observed a Gram-negative predominance alongside notable Gram-positive contributions.

Among Gram-positive isolates, coagulase-negative staphylococci (CoNS) were predominant, with all exhibiting methicillin resistance. Additional Gram-positive pathogens included methicillin-sensitive Staphylococcus-aureus (MSSA), methicillin resistant Staphylococcus-aureus (MRSA), and Enterococcus species.

A majority of Gram-negative isolates demonstrated extended-spectrum beta-lactamase (ESBL) production, aligning with global trends of rising ESBL prevalence in Gram-negative infections.

The variations in microbiological profiles and antibiotic susceptibility patterns underscore the necessity for individual centers to analyze their local pathogen distribution and integrate these findings into empirical antibiotic selection protocols.

Our study has certain limitations. It is a single-center study with a limited sample size, a larger multicenter study needs to be done to reproduce the results.

Conclusions

CRBSI remains a major clinical challenge in patients receiving hemodialysis via central venous catheters, compounded by the growing prevalence of multi-drug resistant Gram-negative infections. This evolving resistance landscape necessitates enhanced preventive strategies, including rigorous sterile protocols, improved hygiene, ongoing surveillance, and accelerated transition to long-term vascular access (e.g., AV fistulas). Appropriate cultures should be taken before starting empirical antimicrobial therapy and every center should frame its antibiotic policy based on regional antibiotic susceptibility patterns.