Speech-based med reminders show promise with familiarity, repetition

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
 - iniate the call, phone, telephone

Physicians trying to improve older patients' adherence to medication may find a verbal reminder another tool in their arsenal, if it is delivered effectively. According to a study published in the January issue of Journal of the American Medical Informatics Association, patients remembered best when the medication name was already familiar and repeated more than once.

The research team led by Maria K. Wolters, PhD, of the University of Edinburgh in Scotland, analyzed recall ability among older patients. The 44 participants were tested with a series of messages about medication spoken by either a human voice or two synthetic voices, all speaking with a standard British accent. They asked patients to recall the medications listed in spite of levels of background noise or signal transmission quality.

They found that medications that were familiar were the easiest for patients to remember. When reminded of one medication at a time, patients recalled 89.3 percent of known medications correctly, but 59.3 percent of unknown medications. Voice did little to affect recall of known medications; however, when unknown medications were recited, human voice outweighed synthetic, 64.8 vs. 52.2 percent, respectively.

When using a four-medication list, patients recalled on average 1.86 medications while for synthetic voices they recalled between 1.5 and 1.6. Patients recalled more medications on average with repetition (two medications) compared to when explanations were provided about what the drug was for (1.8) or when a basic message was given (1.6). Repetition also helped listeners under difficult listening conditions understand synthetic voices better. Without repetition, listeners understood between 1.4 to 1.5 medications listed by synthetic voice. When repeated, recall increased to an average of 1.9 to 2.1 medications.

While recall was far from perfect, Wolters et al suggested that spoken medication reminders can work as long as “they build on what users know” and employed repetition to catch what was missed. “While repetition alone is not enough to help people recognize and recall information, in this case, it provides a second chance to catch an auditory glimpse of the speech signal, building a more robust percept that is in turn more likely to be recognized later,” they wrote.