News

At some point or other, you’ve no doubt uttered the phrase “Hey Alexa…” or “OK Google…” to a smart speaker device to do things like: check the weather, play a song, or ask a simple question… but what if you could use this technology to monitor your health? This thought, along with existing research ultimately led to the creation of Medly Voice. A study performed by Ware et al. investigated adherence rates to the Medly program (offered as a smartphone app) over a 12 month period (Ware et al.); results showcased an overall average decline as time went on. Other studies have shown that various patient demographics, specifically older adults and those with cognitive and physical impairments, struggle to use mobile technologies, such as touch screen devices and smartphones. So although mobile health (mHealth), defined as the use of mobile technologies for digital health, is one of the most popular platforms for chronic disease self-management, there are still some patient demographics who struggle when interacting with this technology. 

Cue voice apps! 

Not only are smart speaker devices such as Google Assistant and Amazon Alexa becoming more popular in household settings, but voice apps (deployable through these devices) are also an emerging technology in the healthcare field. This popularity stems from the defining features of a smart speaker device: simple setup, ease of use, and low cost. So far, voice apps have been used in a clinical setting to help with the patient registration process and physician note transcription; more notably voice apps have been designed to help people manage their chronic illness and live independently in their homes. Current research is limited since these voice apps are still in development and piloting phases, showcasing limited efficacy in testing to support final outcomes and conclusions. There are also a lack of voice apps being designed as a tool to provide more personalized, user-specific healthcare support.

Cue the Medly voice app! 

My MASc thesis aimed to expand the reach of the Medly program; the intent behind the project was to make Medly more accessible by offering it on another platform besides the smartphone. Given the recent emergence of voice apps in the digital health field and compelling attributes of the platform they are delivered on (i.e., smart speakers), we saw the potential it could deliver to a program like Medly. However - attributes such as ‘simple set up’ and ‘ease of use’ are hardly sufficient to confidently believe in the feasibility of deploying a tool like this in a clinic, to be used by physicians and patients. 

Cue my research! 

My first paper, Design of a Patient Voice App Experience for Heart Failure Management: Usability Study, employed a user centered design process to develop the voice app. 

A preliminary version of the voice app was created from existing Medly requirements and findings from a literature review and market scan. The voice app conversation captured the same data points as the Medly smartphone app and was designed to be succinct. A usability study with 8 Medly patients was then performed at the Peter Munk Cardiac Clinic in Toronto General Hospital to observe and understand how patients interacted with the voice app. While the quantitative data showed promise in patient acceptability of the voice app, the qualitative data gathered through observations and interviews identified deeper implications of this technology. Most notably was the lack of confidence levels showcased when interacting with the voice app; physical behavior always changed when compared to the interaction with the smartphone app. Participants typically sat up straighter, appeared more tense, and would constantly be looking over to the study coordinator for re-assurance as they engaged in conversation with the voice app. So, although it seems ‘natural’ to interact with a voice app since it’s similar to engaging in normal, human conversation, the findings from this study began to challenge this. Other barriers of entry were considered that move beyond the scope of this usability study, namely technological literacy … Does the patient have the required tools (for e.g., Internet connectivity) and knowledge (for e.g., ability to set-up a device and account) to successfully use this voice app in their home? These types of questions were the motivation for the second study of this project. 

The first study proved that a voice app version of Medly worked in a controlled environment while the second study, A Voice App Design for Heart Failure Self-management: Proof-of-Concept Implementation Study investigated how well the voice app worked in its intended setting, a patient's home. Twenty participants interacted with the app over a 4 week period. Similar to the first study, the quantitative data showed promise with acceptability and feasibility, while the qualitative data revealed more insightful findings. For example, the overall average accuracy rate (calculated by comparing the measurements inputted on the smartphone app with those recorded on the voice app) was 97.8%, but some participants used multiple attempts before the voice app correctly recorded their data. Sometimes the device would abruptly stop or would not understand what the participant was saying. When this occurred, participants often found themselves adapting their (natural) conversational style in an attempt to record the correct data. On the other hand, the device integrated well within patients’ households, with some using it to check the weather, play music, get the news, and even find new recipes to try. 

Perhaps the most interesting finding from this study was the patient profile we began to put together when thinking about who the voice app would be best suited for. The oldest demographic exhibited the highest and most consistent weekly engagement levels, and participants who were bedridden due to other medical complications greatly appreciated inputting their measurements through conversation rather than needing to use their smartphone device. Some participants, however, preferred using the smartphone specifically because it took less time to complete and was more reliable. Therefore, those who are older, have less busy schedules, are confident in using technology, or experience other symptoms may benefit more than others from using the voice app. 

This project begins to lay the foundation for the potential of other modalities, such as voice apps, to be used for accessing digital therapeutics. It’s time we begin to think about if and how these new and emerging technologies can be used to improve patient accessibility for chronic disease treatments. The future is bright!