Is Voice the Right Technology for the Healthcare Problem You Are Trying To Solve?
By Devin Nadar, Partnerships Manager at Boston Children's Hospital
At Boston Children’s Hospital, we feel strongly that voice-assisted technology will play a key role in healthcare, both within and outside of the clinical setting. But as with any technology, it is important to make sure that you leverage voice appropriately based on the problem or pain point you are trying to solve. To that end, below we share some learnings and considerations in determining whether voice is the right platform for your project.
Speed
One of the first considerations is speed. Voice technology can often enable you to quickly surface a specific piece of information from an identified source. For example, if you need to know the dose of a particular medication based on a child’s weight and age, a smart speaker could give you that information almost immediately using the BCH KidsMD skill. Compared to opening up an app on your phone or logging into a workstation to search for that information, voice offers a much more efficient experience. However, if the smart speaker must ask you a long series of questions to determine an appropriate content response, it will likely be more frustrating than helpful. For example, when we presented clinicians with a voice skill that assessed asthma severity, they found that it took much too long to arrive at the classification.
Location
Another important consideration is the location of the smart speaker. Spending time in the location you want to deploy the smart speaker is the best way to determine if the location is suited to voice. Convenience is key- if the user has a question and the smart speaker is down the hall or in another room, they are likely not going to make that trip to ask a simple question. But ambient noise is also an important consideration. Nursing stations and other areas where multiple conversations are occurring simultaneously may not be a good spot for a smart speaker.
Reliability
One concern we have heard around the use of voice in healthcare is that it is hard to know if the information provided by a smart speaker is from a reputable, reliable source. It is imperative that users — whether consumers or clinicians — trust the information being provided. Think about what content you will have in your voice skill, the source of that content, how the content is curated and by who, and, perhaps most importantly, how you communicate the content source to your users. We work directly with clinicians on content to ensure that we are providing the best advice and information to our users.
Compliance and Security
Currently, there is no commercially available smart speaker that is HIPAA-compliant. For many healthcare organizations, this can present significant barriers to introducing voice technology. The sooner that an organization’s security and compliance teams can be involved in a voice project, the more success you will have with implementation. To date, we’ve chosen to pursue voice pilots that do not involve HIPAA-protected information and we believe there are a multitude of voice-based applications within these confines.
In the end, there is no technology that is best suited for every clinical pain point. When evaluating whether or not voice can be an effective tool, make sure you ask the right questions and take user feedback into account during the entirety of the project. Listen to your users, but do not be afraid to try and persuade them either. Finally, do not be disheartened by what seem to be limitations right now. Voice technology is rapidly evolving and some of the barriers you find today will likely be addressed in the near future.
Questions? E-mail devin.nadar@childrens.harvard.edu