In a recent post, I gave a hypothetical example showing how voice could be an effective tool in a variety of healthcare scenarios a single patient may experience. In this blog post, I’m going to discuss a recent clinical trial in which a voice assistant was used to monitor people with diabetes to study the usefulness of voice in a specific population.
First, however, let's review some of the benefits of using voice as an interface.
Compared to other means of interface, we are the experts of voice
Voice, like typing on a keyboard or tapping on a tablet or smartphone, is a way to communicate. However, unlike the other two methods, using our voices—or speaking—comes naturally to us.
With voice, we can share what we think, wish to know, or want to accomplish without using an unnatural interface developed by someone else. Voice also gives us a sense of power over things we interface with. For example, how many times have you stared at a document on your computer and wished you could perform a specific task or apply a certain format. You scan the menu options, but the correct choice isn't clear. Admitting defeat and likely blaming yourself, you give up and start searching online for an answer, or you call the tech support team.
With voice, you have the opportunity to state your desires in the way that makes sense to you. If the person or device doesn’t understand, you are more likely to blame it, not yourself, for the gap in communication. Granted, the person speaking has to state what he or she wants clearly, and the voice assistant needs to be smart enough to understand. In this case, however, the user can feel more empowered.
A voice assistant has no ego
As human as Alexa, for example, may sound, it is a device. It has no emotion (sorry, it's not actually thrilled when you get all the Jeopardy answers right). It can’t get mad at you for needing to repeat something. It doesn’t get frustrated when it is asked the same thing over and over again.
Compare that to a very human doctor or nurse. Though they may try to keep an even-temper, the likelihood is great that he or she will become frustrated if a patient asks the same question visit after visit—or even during the same visit. It is likely that they’ve answered that same question a dozen times before seeing this patient.
Also, human-to-human contact also comes with inherent problems. A patient is likely to hold his or her doctors and nurses in high regard since they are professionals. If a physician tells a patient to take a medication three times a day every day and then asks about adherence in a follow-up visit, the patient may not be willing to admit that he or she has not followed the doctor’s orders. The patient’s emotions may get in the way as well, not wanting to feel shame for not doing what he or she was told.
A voice assistant does not judge. It has no ego to be bruised.
Other benefits of voice
In past blog posts, I’ve also noted:
- speaking is much faster than typing
- voice technologies provide a level of accessibility that keypads do not, and
- the ability to speak endures even when motor skills and eyesight decline
With those reminders out of the way, let’s look at how these possible benefits of voice were evaluated.
Studies show how voice changes people’s behavior around their health
To test the value of voice, Worrell, an Orbita partner, recently conducted a trial to examine the feasibility of using voice assistants to help gather patient insights and help manage chronic diseases.
Sixteen participants age 12 to 72 with type 1 or type 2 diabetes were recruited to determine if it is possible to use a voice assistant to capture meaningful data and engage patients. Each person in the study was given an Echo Dot with skills necessary to participate loaded onto it. The study lasted for two weeks and required three daily interactions:
- Taking a survey to check in on overall health. As part of the process, participants set goals. The next day, the skill asked how they did in achieving those goals
- Responding to daily prompts regarding their experience with diabetes. These cues were not delivered through the voice assistant, but instead came to the user via text messages or phone calls using native messaging function
- Telling their stories (or a heart-to-heart) to be shared with people newly diagnosed with diabetes. The participants interacted directly with the voice assistant.
Voice assistants aid in adherence and reporting
The study’s findings show that participants felt surprisingly committed to upholding their goals because they knew they had to report them. They were also willing to admit if they didn’t meet the milestone they had set for themselves (see no ego, above).
Participants also shared their very personal health stories with Alexa. Some even called the experience cathartic. They appeared to form a relationship with the voice assistant, and they started speaking in a freer, more open way.
It must be noted that there were some issues, such as Alexa misunderstanding things the participants said as well as some technical issues. These were reminders that Alexa is not a real person. Still, this study shows that voice assistants could be effective accountability partners, aiding in adherence and honest reporting of symptoms and side effects.
Review all the findings in the study
To learn more about this study, other trials Worrell has undertaken, and the benefits of voice, be sure to watch the Webinar “Making sense of voice experiences and conversational AI in healthcare"
Be sure to register for our next webinar
On June 27 at 1 PM EST, Orbita will present "How Voice Enabled Solutions Can Drive Innovation in Pharma: Insights from the Alexa Diabetes Challenge". Register to learn about insights gained from the Merck-sponsored Alexa Diabetes Challenge.