OBJECTIVES:

To avoid the potential harm of drug-drug interactions (DDIs), clinical decision support (CDS) systems using electronic health records have been installed in multiple healthcare systems. We aim to estimate whether DDI-CDS systems are cost-effective compared to prescribers’ decisions alone considering DDIs.

METHODS:

A cost effectiveness model was developed from a U.S. healthcare system payer perspective comparing a DDI CDS alert software to prescriber’s alone decision for patients with a prescription order that includes a potential DDI. The population of interest included individuals of 72+/- 10 years or older. The model incorporates the risk of override of the DDI alert and of the physician prescribing the potential DDI. The primary health states are severe adverse event (SAE), including life threatening event, hospitalization, and medically significant event; minor AE (mAE); and no AE. As death due to DDI has been estimated in patients admitted to hospitals, we included this probability in the hospitalization branch. Disutility values were used to calculate quality adjusted-life-years (QALYs). Endpoints included costs, QALYs, and incremental cost-effectiveness ratios (ICERs) with a willingness-to-pay (WTP) threshold of $150,000 per QALY.

RESULTS:

Prescriber’s alone decision had an incremental cost of $23 compared CDS-DDI ($511 vs. $488, respectively), providing similar QALYS (0.85 and 0.87 respectively). CDS-DDI alert was dominant compared to prescriber’s alone decision. The utility of a mAE, followed by the probability of DDI alerts being overridden are the most relevant variables. The cost-effectiveness acceptability curve indicates there are no WTP values where the DDI-CDS alert has a lower probability of being more cost-effective than the physician’s prescribing decision alone.

CONCLUSIONS:

This analysis suggests that CDS-DDI alerts are cost-effective compared to prescriber's decision alone. Regardless that most DDIs are overridden, DDI CDS is a cost-effective tool to avoid harm.