USE OF INNOVATIVE, REAL-TIME VOICE ANALYTICS TO GUIDE CLINICIANS' TELEPHONIC INTERACTIONS WITH INDIVIDUALS IN CLINICAL PROGRAMS

Purpose: Managed care organizations implement clinical programs to provide health education, promote healthy behaviors, and improve health outcomes. Like many other health-promoting activities, it can be difficult to maintain engagement in these programs. This pilot study assessed whether a real-time voice analytics tool used for telephone consultations within a clinical program could effectively improve clinical program enrollment and retention; ultimately leading to an individual’s better engagement in their health.

Method: Clinicians conducted program invitation and health coaching telephone calls between October and December 2013. Calling clinicians were divided into two groups; voice analytics tool (VAT) or reference. The VAT clinician group viewed a real-time conversation analysis tool on their computer desktops as they conducted calls. The VAT uses real-time, non-traditional analytics to represent clinician and participant speech patterns visually on computer desktops to aid clinicians’ in their interactions with participants.  The reference group conducted calls without the tool. The primary outcomes were program enrollment and program retention.  Program enrollment was measured as the percentage of people who accepted the invitation and then completed a clinical (i.e. non-introductory) call among those offered.  Program retention was measured as the percentage of participants whose were continuously available for each coaching call until they reached their program goals (time in program varied based upon goals set at program start).

Result: The VAT group made 1,841 program invitation calls and the reference group made 456 invitation calls over the 3-month period. Program enrollment was higher in the VAT group than the reference group, 34% vs. 30% respectively, but the difference was not statistically significant. The clinical program attempted several strategies to reduce the rate at which they retained participants, one of which included the conversation analysis tool. In the VAT group, the retention rate improved more than in the reference group (4% improvement vs. 2% improvement pre-to-post). The difference was not statistically significant.

Conclusion: Program enrollment and retention trended higher in the VAT group although not statistically significant. The assessment of this technology is ongoing to determine if the initial benefits observed in this pilot study reach statistical significance with longer follow-up and larger sample size. The benefits of real-time conversation analysis technology have the potential to extend to overall clinical program engagement and improve health outcomes.