Optimizing Medication History Value in Clinical Encounters with Elderly Patients

Some electronic prescribing (e-prescribing) systems connect to networks that allow for real-time information on medication history from commercial pharmacies to be available to prescribers. Electronic medication histories that include prescription fill information enhance a clinician’s ability to provide quality medication management. Knowledge about a patient’s adherence to previously prescribed regimens can improve communication between patients and clinicians. This may allow for the engagement of patients in making decisions about their medications, particularly around regimens that a patient is more likely to adhere to. However, additional training on the use of these electronic medication histories is needed to optimize improvements in the quality of medication management during a patient encounter.

This project developed geriatric-specific algorithms for use in e-prescribing systems to alert clinicians to potential problems during medication management. The project team developed continuing medical education (CME) modules to educate clinicians about the use of medication information when prescribing to improve safety and patient-provider communication. Finally, they evaluated the impact of the algorithms and the CME modules on medication management. 

The main objectives of this project were to:

• Develop geriatric-specific algorithms to identify potential issues with medication management (e.g., polypharmacy, potentially inappropriate medication use, duplicative therapy, and nonadherence) using community pharmacy-generated medication history.
• Develop structured, problem-oriented frameworks for organizing medication history information during visits (triggering) for common issues identified by the algorithms developed in the first aim. 
• Develop and pretest modules to teach clinicians how to improve geriatric patient-provider communication relating to medication management with the use of technology (training). 
• Test the impact of these interventions on clinician behavior using a randomized controlled trial with two arms: 1) delivery of triggers; and 2) delivery of triggering and training interventions. 
• Develop "tool-kit" resources and intervention products for use by non-physician providers in other ambulatory settings (e.g., pharmacists in community pharmacy settings).

Thirty-three physicians were recruited and randomized to two interventions: 14 to the algorithms alone, and 19 to the algorithms combined with the CME modules. A multi-modal evaluation was completed using surveys of patients, focus groups of physicians, analysis of taped clinical encounters, and the tracking of e-prescribing data.

The project team found that physicians were open to evidence-based treatment algorithms, as long as alerts were designed to promote efficiency and reduce redundancy. It was found that alternatives to potentially inappropriate medications should be provided to clinicians during prescribing to make it easier for the clinicians to choose those alternatives. The study did not show a change in clinician perception of e-prescribing, nor of patient perception of patient-physician communication. Physicians overrode alerts frequently across both arms of the study. Physicians indicated frustration with alerts that provided commonly known information, and repetitive, previously overridden alerts. The project team noted a need to suppress alerts when medications were renewed for a patient currently taking and tolerating a medication well.