Cost comparison of antibacterial therapies for serious infections. A New Zealand 3-hospital study.

OBJECTIVE

The first aim was to identify and determine the economic costs of the regimens currently used in 3 New Zealand hospitals in the treatment of bacterial infections in haematology patients with febrile neutropenia and in intensive care patients with severe infections. The second was to develop a spreadsheet-based decision analytic model for use by hospital decision-makers as an aid in evaluating the comparative cost of drug regimens.

DESIGN AND SETTING

The research utilised time and motion and microcosting techniques. The analytical perspective adopted for the study was that of a hospital administrator or clinical manager.

PATIENTS AND INTERVENTIONS

Patients were eligible for inclusion in the study if either they were treated with the imipenem/cilastatin monotherapy, or could have been treated with this regimen. The final analysis considered 360 patient-treatment days and 8 antibacterials.

MAIN OUTCOME MEASURES AND RESULTS

Drug acquisition cost ranged from 4.52 New Zealand dollars ($NZ; 1997 values) per patient-treatment day for gentamicin to $NZ104.81 for imipenem. The cost per patient-treatment day (when other cost components such as fluid additives, giving sets and needles were added) ranged from $NZ8.75 for gentamicin to $NZ129.12 for tazobactam. Drug acquisition cost, as a percentage of total drug preparation and administration cost, ranged from 52% for gentamicin to 93% for piperacillin. Giving sets and intravenous (i.v.) fluids were found to be important cost items when they were required specifically for the treatment regimen. There was a mean monitoring rate of 0.40 at a cost of $NZ6.41 per patient-treatment day for gentamicin. It was estimated that nephrotoxicity could add between $NZ23 and $NZ43 per day to the cost of aminoglycoside treatment.

CONCLUSIONS

Although the small sample sizes of the study mean that results should be regarded as indicative rather than conclusive, there were sufficient information to construct a working model and show how the total cost of an antibacterial regimen could be evaluated in practical terms. The important cost drivers were found to be drug cost, the use of fluids and giving sets, and monitoring.