Antifungal stewardship in a tertiary-care institution: a bedside intervention.

Antifungal stewardship (AFS) programmes are needed in tertiary-care hospitals. Our aim is to describe a bedside non-restrictive AFS programme, and to evaluate its economic impact. During the first year of the AFS a bundle of non-interventional measures were implemented. During the second year an infectious diseases specialist visited 453 patients receiving candins, liposomal amphotericin B, voriconazole or posaconazole. Monthly costs were studied with an interrupted time series (ITS) analysis. The main prescribing departments were haematology (35%), medical departments (23%), and intensive care units (20%). Reasons to start antifungal therapy were: targeted therapy (36%), prophylaxis (32%), empirical therapy (20%) and pre-emptive therapy (12%). At the initial visit, diagnostic advice was provided in 40% of cases. The most common therapeutic recommendations were to de-escalate the antifungal drug (17%) or to suspend it (7%). Annual total antifungal expenditure was reduced from US$3.8 million to US$2.9 million over the first 2 years, generating net savings of US$407,663 and US$824,458 per year after considering the cost of additional staff required. The ITS analyses showed a significant economic impact after the first 12 months of the intervention (p 0.042 at month 13), which was enhanced in the following 24 months (p 0.006 at month 35). The number of defined daily doses decreased from 66.4 to 54.8 per 1000 patient-days. Incidence of candidaemia was reduced from 1.49 to 1.14 (p 0.08) and related mortality was reduced from 28% to 16% (p 0.1). A collaborative and non-compulsory AFS program based on bedside intervention is an efficacious and cost-effective approach that optimizes the use of AF drugs.

Prospective, randomised study of selective versus routine culture of vascular catheter tips: patient outcome, antibiotic use and laboratory workload.

In order to assess the value of vascular catheter tip culture in patients with negative blood cultures, all tip samples from hospitalised patients were prospectively randomised (1:1) to two different routines for processing catheters: culture of all tips (routine A) vs culture only of tips from patients with concomitant bacteraemia or fungaemia (routine B). Over a nine-month period, 426 catheters from 318 patients were randomly assigned to routine A and 429 catheters from 322 patients to routine B (n=40 [corrected] patients). We compared the outcome and costs from both groups. No statistically significant differences were found with respect to demographic data, mortality, hospital stay or antimicrobial use. In non-bacteraemic/fungaemic cases (N=517), days on antimicrobial therapy after catheter withdrawal were significantly higher in patients from group A [10.0 days (interquartile range, IQR): 6.0-14.0] vs 8.0 days (IQR: 4.7-12.2), P=0.03], as was the number of daily defined doses (DDDs) of antimicrobials [10.8 DDDs (IQR: 2.4-26.9) vs 7.5 DDDs (IQR: 1.5-20.0), P=0.03]. Median antimicrobial cost per treated patient was significantly higher in group A: €222.30 (IQR: €20.30-€1,030.60) vs €109.10 (IQR: €10.90-€653.20), P=0.05. If all vascular catheter tips were processed according to routine B, the microbiology laboratory workload would decrease by 77% for the total number of catheters processed. Microbiology laboratories should not routinely culture catheter tips in patients without bacteraemia or fungaemia.

Economic impact of HIV protease inhibitor therapy in the global use of health-care resources.

OBJECTIVE: To evaluate the clinical and economical impact of the introduction of HIV protease inhibitor (PI) therapy in the current clinical care of HIV-infected patients. METHODS: Cohort study with 155 HIV-infected patients with a full year of follow-up before and after the introduction of PI by June 1998. The setting was a large urban tertiary teaching hospital in Madrid, Spain. The main outcomes measures were clinical and immunological evolution, pharmacy, out-patient, emergency room and in-patient medical costs evaluated by diagnostic-related group classification, and the global economic costs of clinical care in HIV-infected patients (AIDS and non-AIDS). RESULTS: The cost of PI therapy was compensated fully by savings related to reduction of the number, length and severity of hospital admissions in AIDS cases. In contrast, more modest clinical effects with increased costs were observed in non-AIDS cases. Globally, there was an increase of about 20% in the total health-care costs of HIV-infected patients (P < 0.01). CONCLUSIONS: PI therapy is highly cost-effective in AIDS patients. Its value in less severely immunosuppressed patients requires further evaluation.