Pills and Patient Behavior:
Smarter Experience Tracking


Presented at American Geriatric Society annual meeting, Washington, D.C., April 2008. P.T. Ryder, University of Maryland School of Pharmacy, Baltimore, Maryland Vesta Brue, MedSignals®, San Antonio, Texas


BACKGROUND

Medication adherence is problematic in geriatric care. This trial evaluated an integrated system using MedSignals®, a 4-drug communicating pill case, to monitor medication intake in home settings, to collect responses to study questions, and to upload adherence data daily.

METHODS

Randomized crossover study:
Control Phase: Unprompted, naturalistic pill-taking patterns monitored and recorded for 3 weeks
Active Phase: Pillboxes set to remind participants when/how to take pills
Participants randomized to phase order
Participants electronically recorded answers to questions regarding medications twice daily in active phase

The device:
•  Recorded and time-stamped lid openings (presumptively pills taken)
•  Uploaded data on a daily basis to host servers via telephone line

Inclusions criteria:
•  Under medical supervision
•  Take at least 4 drugs
•  Age 65 or older
•  Have a landline telephone

Exclusion criteria:
•  Non-ambulatory
•  Require assistance with medications
•  ADL or cognitive impairment

Primary outcome measure adherence to:
•  Quantity of doses prescribed
•  Timing of doses

RESULTS

41 Participants: 31% male; 40% African American; 38% Caucasian; 22% Hispanic; average age 69.6 years (range 65-84).

Active phase yielded significantly greater adherence (p<.001) to:
•  Overall doses (ratio of lid openings to prescribed doses)
•  Number of dosing times per day per drug

Participants averaged one pill fewer per day than prescribed for TID and QID dosing in control phase

Participants answered only 56.7% of the prompted questions

Two-timing adherence measures showed significant treatment effects:
•  Number of day gaps (days without lid openings)
•  Proportion of doses taken within 15% of time interval indicated

Condition-by-order interaction (p<.001). Suggests that active phase had very strong carryover effects when it preceded the control phase.

CONCLUSIONS

The device appears to improve medication adherence, especially when medications are taken more frequently than once per day. With chronic conditions that rarely change, these elders reported that responding to questions did not seem meaningful.

ADDRESS FOR CORRESPONDENCE

Vesta Brue, via email vesta@medsignals.com

This project was funded by a grant from the National Institute on Aging, 5R44AG025612