Hospitalization costs associated with warfarin-related bleeding events among older community-dwelling adults.

PURPOSE

A prior paper from this study demonstrated that patient report of receiving medication instructions from health care professionals is associated with reduced risk of warfarin-related bleeding hospitalizations. The objective of this analysis was to describe the hospitalization costs due to warfarin-related bleeding events in older community-dwelling adults and to estimate the hospitalization costs avoided due to the receipt of medication instruction from different sources.

METHODS

We estimated the expected hospitalization costs associated with four instruction sources based on the respective incidence rate of observed hospitalizations and mean hospitalization cost for warfarin-related bleeding episodes from a prospective cohort study of beneficiaries of the Pennsylvania Pharmaceutical Assistance Contract for the Elderly (PACE). We estimated hospitalization costs avoided due to each instruction source compared to no instructions using the payer's perspective. We conducted probabilistic sensitivity analysis to account for uncertainty in our parameters.

RESULTS

One hundred twenty-six warfarin-related bleeding hospitalizations occurred during the observation period with a mean cost of $10,819 (SD: $11,536). The mean expected hospitalization cost from a warfarin-related bleeding hospitalization without instruction was $835 per year per person. Hospitalization costs avoided with instruction from a health care professional ranged from $443 to $481 per year per person.

CONCLUSIONS

The costs per hospitalization associated with warfarin-related bleeding events are substantial. Instructions for warfarin management from a health care professional may reduce the number of warfarin-related bleeding hospitalizations and associated costs. Investments in interventions to improve communication regarding warfarin management may be justified economically based on the potential cost savings estimated in this study.