CAT

Raj Subramanian, 1/17/02

Tanriover, et al. Bacteremia associated with tunneled dialysis catheters: Comparison of two treatment strategies.

Kidney International, Vol. 57(2000), pp. 2151-2155.

Clinical Question:

Population: Pt with ESRD on HD with cuffed HD catheter with line sepsis/infection

Intervention/Exposure: Catheter exchange

Comparison/Control: Catheter replacement

Outcome: Recurrent Sepsis (microbiologic / clinical severity)

Death/Catheter malfunction/thrombosis/endocarditis & other embolic phenomenon

Background:

Tunneled cuffed catheters play an increasingly important role in the delivery of HD. However, bacteremia is a particularly important problem. The gold standard (thus far) in treating line sepsis is catheter removal and delayed replacement. In the HD patient, this issue is complicated by the need to continue dialysis treatments. Catheter removal requires use of temporary catheters and associated complications. It mandates multiple procedures, period of hospitalization, and increased costs. Importantly, catheter removal may be associated with loss of central venous access sites because of stenosis or thrombosis. Hence the evaluation of treating line sepsis in select patient populations by changing the catheter over a guidewire under antibiotic coverage.

Study Design:

Retrospective cohort. Setting: University of Alabama hospital. They analyzed the outcomes of all cases of catheter related bacteremia (CRB) during 2 year period (Jan 1, 97 – Dec 31, 98). All patient immediately treated with emperic broad spectrum antibiotics after drawing blood cultures.

Exclusions: 1) cases with exit site infection 2) cases with “severe sepsis’ (persistent shaking chills or hypotension 3) cases with persistent fever 48 hours after initiation of antibiotics 4) if a patient had multiple episodes, only the first episode in the two year period was included. 5) cases in which the replacement catheter not inserted within 10days of removal of infected catheter (perm access, persistent fever, patient death) 6) cases assumed treated like CRB but remained culture negative. ?? Patients whose catheter malfunctioned was electively removed (permanent access ready to use) or who died with functioning catheter were considered censored.

In the remaining cases of CRB, two treatment strategies were followed at the “discretion” of the nephrologist (sepsis severity not scored). Group A: Catheter exchange over guidewire within few days after bacteremia clinically resolved (no fevers/chills). Group B: Catheter removal within 1-2 days and replacement with new catheter 3-10 days later with interim HD with temp femoral catheter.

Critical review (based on EBM guidelines – JAMA May 25, 1994 Vol 271(20), pp. 1615-1619)

Are the results of the study valid ?

- There was a clearly identified comparison group (retrospective cohort)

- Some of the basic prognostic factors pertinent to the outcome were similar as in Table 1. However, prior episodes of line sepsis or Severity of sepsis (APACHE scores) in the two cohorts not provided.

- No recall or interview bias (retrospective information from charts over fixed 2 year period)

- There is an assignment bias (between excluding the patient and placing patient in catheter replacement group => soft inclusion criteria) especially given incomplete/unavailable medical records in general (not mentioned in article)

- Unclear regarding intention to treat (could flip back and forth between two groups)

- Unclear if opportunity to catheter exchange was equal to all patients (“discretion” of nephrologist)

- Only the first case within the 2 year observation period (multiple cases of CRB’s in same patient not included) – which is arbitrary

- All cases accounted for (x ?? censored cases/> 10d exclusion) however, arbitrary regarding duration of follow-up since infection free days is important outcome but observation was abruptly cut off Dec 31, 98.

- Catheter exchange/replacement did preceed outcome.

- There may be a dose-response gradient but this not accounted for since multiple cases of CRB in same patient excluded.

What are the results ?

- Results are Group A: B (Infection 52%:42% (p=0.86); Elective removal 16%:18%; Malfunction 26%:29%; Death 6%:10%) AND Kaplan-Meir survival curves (proportion infection free vs days with new catheter)

- Language of article confusing regarding censured patients. It appears that the Kaplan-Meir curves exclude death, malfunctioning catheters.

- No statistical difference noted between the two cohorts but it tends towards indicating that catheter exchange may cause harm.

- The article is definitely not overestimating harm and given Kaplan-Meir curves, appears to be underestimating harm done by catheter exchange.

- Calculation of relative risk or confidence interval not possible based on results.

Will the results help me caring for my patients ?

- patient demographics, morbidity, and other prognostic factors similar to our patients.

- treatments and exposures similar

- Cannot evaluate absolute risk, however harm may be underestimated

- Cannot tell if we should attempt to stop the practice of catheter exchange.