Patients receiving maintenance dialysis face elevated risk of bloodstream infections, with Staphylococcus aureus a frequent causative pathogen. Historically declining central venous catheter (CVC) prevalence shifted during the COVID-19 pandemic era; the resulting increase in catheter use prompted renewed concern about current infection rates and risk profiles. This study examined national trends across pre-, intra-, and post-pandemic periods to quantify the burden of S. aureus-related sepsis and identify associated risk factors in the US dialysis population.
This is a retrospective cohort analysis using administrative claims data from the Centers for Medicare & Medicaid Services together with United States Renal Data System records covering 2016 through 2023. The cohort included adults older than 18 years receiving maintenance dialysis. Cases of S. aureus-related sepsis were ascertained using ICD-10 diagnosis codes A41.01 and A41.02.
Event rates were calculated per 100 patient-years (PY) for both first events and multiple events. Rates were stratified by demographic and clinical characteristics including dialysis modality and vascular access type. Cox proportional hazards regression models were used to assess risk factors for S. aureus-related sepsis and to estimate adjusted hazard ratios. The abstract reports numerical results for prevalence, incidence rates, and a principal adjusted hazard ratio for catheter use compared with fistula use.
Over the 2016–2023 period, the prevalence of CVC use among the dialysis population increased from 16% in 2016 to 24% in 2023. The incidence of S. aureus-related sepsis for first events was 3.3 per 100 patient-years; when accounting for multiple events, the incidence was 4.1 per 100 patient-years.
First-event incidence varied markedly by access and modality. Hemodialysis recipients using CVCs experienced the highest first-event rate at 8.4 per 100 PY. By comparison, patients with arteriovenous fistulas had a first-event rate of 2.0 per 100 PY, and peritoneal dialysis recipients had a rate of 1.0 per 100 PY. These differences highlight the contribution of vascular access type and dialysis modality to sepsis risk.
Rates of first S. aureus-related sepsis events were not uniform across the population. Elevated rates were reported among younger patients and males. Clinical comorbidities associated with higher unadjusted rates included diabetes, heart failure, and conditions causing immunosuppression. The abstract provides these stratified patterns but does not report full subgroup numerical tables within the abstract text.
In multivariable Cox regression analyses, vascular access type emerged as a dominant risk factor. After adjustment, catheter use compared with fistula use was strongly associated with S. aureus-related sepsis, with a reported hazard ratio of 4.24 (95% confidence interval 4.15 to 4.33). The abstract indicates that other covariates (age, sex, diabetes, heart failure, immunosuppressive conditions) were considered in stratified rates and likely in adjusted models, but the abstract provides the catheter-versus-fistula hazard ratio as the principal adjusted estimate.
S. aureus-related sepsis continues to be a substantial burden among US dialysis patients, particularly among those using central venous catheters. The observed rise in CVC prevalence from 2016 to 2023 coincided with persistently high sepsis rates, underscoring a renewed infection control challenge for the nephrology community.
The authors emphasize the need for targeted prevention efforts, especially measures to optimize vascular access (favoring fistula use where appropriate) and focused management of high-risk subgroups such as patients with diabetes, heart failure, or immunosuppressive conditions. These conclusions derive from national claims-based data and reinforced the association between catheter use and markedly increased risk of S. aureus-related sepsis.
The abstract reports methods, key incidence rates, catheter prevalence trends, subgroup patterns, and the main adjusted hazard ratio for catheter use. Details not included in the abstract (and therefore not reported here) include full model covariates and coefficients beyond the catheter hazard ratio, granular subgroup numeric tables, microbiologic confirmation beyond ICD-10 coding definitions, and explicit discussion of study limitations or potential sources of bias. These items may be available in the full text but were not provided in the abstract.