Background:

Clinical decision support systems are useful tools to discourage the use of potentially inappropriate medications (PIMs) in the geriatric inpatient population. These interventions have had variable success, in large part because geriatric medication decisions are difficult and not always amenable to resolution via a computerized prompt.

Purpose:

To develop an electronic PIMs dashboard which facilitates surveillance and clinical pharmacist intervention among geriatric inpatients who are prescribed potentially inappropriate medications (PIMs).

Description:

We created an electronic PIMs dashboard that identifies, in real–time, patients in a 658–bed university hospital who are age 65 or older and who have been prescribed at least one PIM. PIMs are determined according to Beer’s Criteria as well as other evidence–based lists of high–risk medications. The dashboard synthesizes and displays information about the prescription of up to 240 PIMs, as well as provides a score to quantify the anticholinergic burden, and the 48–hour cumulative narcotic and benzodiazepine doses. The dashboard also shows each patient’s age, gender, estimated kidney function, complete list of diagnoses, names of the inpatient and outpatient physicians, and a link to the full electronic health record. Thus, the dashboard supports efficient review by a clinical pharmacist, who contacts the treating physician(s) when a change in therapy is recommended. In the initial three–week pilot, approximately 1000 hospitalized patients were screened by the dashboard’s automated functions, which identified 73 patients with 203 PIMs that warranted pharmacist review. The pharmacist determined that 80 medications (39%) were actually not inappropriate, based on the patient’s clinical circumstances, leaving 123 medications (prescribed to 22 patients) for intervention. When contacted by the pharmacist, providers indicated over 90% acceptability for personal contact, and completed 75% of the recommended medication changes. Forty percent of the PIMs were stopped by discharge.

Conclusions:

We successfully developed an electronic dashboard integrated into the electronic health record that allowed focused pharmacy review of medication appropriateness among vulnerable hospitalized geriatric patients. The pharmacist’s time was greatly leveraged by this tool, the majority of recommended changes were made, and the intervention had a high degree of acceptability by providers.