Management of Chills During Hemodialysis
Immediately obtain blood cultures and start empirical antibiotics with vancomycin PLUS gram-negative coverage, as chills during hemodialysis represent catheter-related bloodstream infection (CRBSI) until proven otherwise, and delays in treatment increase mortality by 10% per hour. 1, 2
Immediate Actions (Within Minutes)
- Draw at least two sets of blood cultures before antibiotics: one from the catheter AND one from a peripheral site if feasible, or from the bloodlines connected to the dialysis circuit if peripheral access is unavailable 1, 2
- Start empirical antibiotics immediately after cultures are drawn: vancomycin 20 mg/kg (actual body weight) loading dose during the last hour of dialysis PLUS gram-negative coverage (third-generation cephalosporin, carbapenem, or β-lactam/β-lactamase combination based on local antibiogram) 1, 2
- Avoid aminoglycosides due to substantial risk of irreversible ototoxicity in dialysis patients 2
The rationale for immediate empirical antibiotics is compelling: 60% of hemodialysis patients with chills have a documented infection, and 33.5% have bacteremia 3. In patients with catheters, the risk is even higher, with catheter access conferring a 6-fold increased risk of bacteremia (OR 6.2; 95% CI, 3.2-12.0) 3.
Clinical Assessment for Complications
- Inspect the catheter exit site for erythema, warmth, purulent drainage, or tunnel infection 1, 2
- Assess for metastatic infections: endocarditis, suppurative thrombophlebitis, or osteomyelitis 1, 2
- Check for hemodynamic instability or septic shock 1, 2
- Document fever presence: fever is an independent risk factor for bacteremia (OR 1.6; 95% CI, 1.1-2.3) 3
Staphylococcus aureus is the leading cause of catheter-related bacteremia in hemodialysis patients, with chronic hemodialysis patients having a 50-60% nasal carriage rate 1.
Catheter Management Algorithm
For S. aureus, Pseudomonas species, or Candida species:
- Always remove the infected catheter immediately and insert a temporary (non-tunneled) catheter at a different anatomical site 1, 2
- If absolutely no alternative sites are available, exchange the infected catheter over a guidewire 1
For other pathogens (gram-negative bacilli other than Pseudomonas, coagulase-negative staphylococci):
- Continue empirical antibiotics without immediate catheter removal 1, 2
- Remove the catheter if:
- If symptoms resolve within 2-3 days and no metastatic infection:
Antibiotic De-escalation Based on Culture Results
- For methicillin-susceptible S. aureus: switch from vancomycin to cefazolin 20 mg/kg (actual body weight, rounded to nearest 500-mg increment) after dialysis 1, 2
- For vancomycin-resistant enterococci: use daptomycin 6 mg/kg after each dialysis session OR oral linezolid 600 mg every 12 hours 1, 2
- If cultures are negative at 48-72 hours AND symptoms resolved: discontinue antibiotics if no other infection source is identified 4
Treatment Duration
- 10-14 days: for uncomplicated infection (symptoms resolve within 2-3 days, no metastatic infection) 1, 2
- 4-6 weeks: if persistent bacteremia/fungemia >72 hours after catheter removal, endocarditis, or suppurative thrombophlebitis 1, 2
- 6-8 weeks: for osteomyelitis 1, 2
Follow-up Monitoring
- Obtain surveillance blood cultures 1 week after completing antibiotics if the catheter was retained 1, 2
- If surveillance cultures are positive: remove the catheter and place a new one only after obtaining negative blood cultures 1, 2
- Place a new long-term hemodialysis catheter once blood cultures with negative results are obtained 1
Special Considerations for Low-Risk Patients
Only in patients with fistula or graft access (not catheter), presenting without fever, leukocytosis, or hypoalbuminemia, and with no obvious source of infection, the risk of bacteremia is low (6%) and investigation without prompt antibiotic treatment may be considered 3. However, all other patients should receive antibiotic coverage immediately following a chills episode 3.
Common Pitfalls to Avoid
- Do not delay antibiotics waiting for culture results in symptomatic patients, as mortality increases by 10% for every hour's delay 1, 2
- Do not use aminoglycosides for empirical coverage due to ototoxicity risk 2
- Do not assume low risk in catheter patients: catheter access confers a 6-fold increased risk of bacteremia compared to fistula/graft 3
- Do not forget to obtain cultures from both catheter and peripheral sites when feasible for optimal diagnostic yield 1, 2