Objective. In patients with central venous catheters and inflammatory response symptoms, catheter-related sepsis (CRS) is often suspected but is often not the etiology. Unnecessary interventions and undertreatment of CRS both may have associated morbidity and cost. This study examines the cost-effectiveness of 10 strategies for managing patients with suspected CRS.
Method. Decision-tree analysis and cost-effectiveness analysis. Two strategies required catheter replacement and catheter tip and peripheral blood cultures with: (NSR) new site replacement and (GWX) guidewire exchange. Three strategies required a through-catheter culture (TC), leaving it in place and treating through infected catheters with: quantitative (TCquant), qualitative (TCqual), and paired quantitative with peripheral blood quantitative (TCPB) cultures. Strategies were (a) with or (w) without empiric antibiotics pending culture results. All positive cultures were repeated and treated with antibiotics through the replaced or infected catheter. Catheters were replaced at a new site if CRS persisted despite antibiotics. Cost were estimated from 1997 charges and the literature. Probabilities were estimated from systematic reviews and expert consultation. Estimates were varied within their range of uncertainty. Effectiveness was a score with 1 = effective treatment or uninfected patients and 0.05 = untreated CRS. We subtracted 0.05 for intervention-related complications and for unnecessary interventions(antibiotics or catheter replacement in uninfected patients).
