The Problem: Improving AI Chatbot Interactions

AI chatbots can be helpful or discouraging to patients. Some people are used to chatbot interactions from web portals and have no problem interacting with these services. Other people would like to have a real person interacting with them during a health portal session and are annoyed when chatbots pop up.

Chatbots can also be annoying if the person moderating the chat is trying to accommodate too many conversations at one time or the response performance of the chatbot is poor. No one likes to wait for an interaction, especially if they are not feeling well.

The four biggest challenges of chatbots are:

1. The integration with context and meaningful interactions between the chatbot and the person.
2. Holding the user’s attention throughout the interaction.
3. Continually testing and updating the chatbot natural language models.
4. Creating an effective and well-planned strategy relative to the business goals of the organization.

Other challenges of chatbots are related to susceptibility to data security breaches, misunderstanding user sentiment, the language or dialect spoken by the user, and interrupted user experiences. If the chatbot misconstrues a user input, the outcome will likely not deliver the results expected by either party.

The other potential challenge with chatbot is that people must type. Speaking to a real person for more complex issues is more satisfying.

The Solution: Focused Social Audio Solutions Will Improve Patient Satisfaction

What is social audio? It consists of four components: social media, digital audio, virtual events, and conversations. Social audio allows listeners to participate in real time with the speakers. Social audio capabilities that are likely to drive the highest return on investment are influencer marketing (e.g., a cancer patient interacting with other cancer patients), branded chats (e.g., chat sessions with expectant mothers), social listening (e.g., discussing a health service with targeted patients), and virtual events (e.g., introducing new medical staff).

Social audio is like the video conferencing solutions that exist with Zoom, GoToMeeting, and Teams but without the video. In some cases, people prefer to not participate in the video component of virtual interactions. People leading the social audio interaction in healthcare must be well trained in the subject matter they are supporting, must be good conversationalists, must like people, and must be pleasant and polite.

As social audio evolves to incorporate artificial intelligence to extract and codify discrete data from these conversations, it will provide a database of consumer and patient data for improving care delivery and patient satisfaction. Eventually, voice-based biometrics could be integrated into these environments to provide real-time wellness screening.

While social audio provides a high-touch tool for patient interactions, it should be thought of as a supplemental solution with existing chatbot and video conferencing solutions being used to support patient care.

The Justification: Using High-Touch Patient Interactions for Targeted Populations

As healthcare providers continue to extend their patient engagement solutions to support key services and/or populations of patients, the social audio technology provides another approach that will resonate with some consumers and patients. Elderly patient populations are likely to have a better experience with social audio services hosted by a real person. Social audio programs could be targeted to people who are more proficient in communicating in other languages.

Social audio programs will incur higher resource costs relative to using people versus AI chatbots, but their potential impact for driving improved clinical outcomes and patient satisfaction for specific healthcare services should be evaluated in context to current patient engagement programs.

The Players: Emerging Companies and Social Media Players Are Diving In

While new companies, such as Clubhouse, are driving early market awareness, other well-known social media companies are jumping into the fray. Examples include the following:

• Clubhouse – Generating lots of marketing buzz
• Discord – Being acquired by Microsoft
• Spaces on Twitter – Extending services to voice
• LinkedIn – Creating real-time virtual discussions via user profile matches
• Fireside – Being invested in by Mark Cuban

Remember, social audio is an immature and emerging technology!

Success Factors

1. Review patient engagement touchpoints to determine where a real-time audio interaction may have the highest impact.
2. Determine which current staff members have the verbal skill sets to effectively support consumers and patients with social audio.
3. Create a prototype service for patient groups with outcomes and engagement benchmarks to determine whether social audio is a fit.

Summary

Social audio is an emerging technology that supports real-time, interactive, virtual voice conversations between a discussion leader and the audience or between a group of people discussing a specific topic or event. It will likely be used as a component of patient engagement by healthcare providers to support specific healthcare services programs for targeted populations of patients. The technology could also be used by healthcare professionals for addressing challenges or issues that impact patient care.

Could this technology have been useful for healthcare professionals who were treating COVID-19 patients across disparate facilities? Could successful treatment protocols have been discussed and shared earlier to prevent patient deaths? Could this technology have been useful for engaging patients who were trying to discern whether they had COVID-19?

With the COVID-19 examples, the cost of supporting social-audio programs versus the benefits would have to be evaluated relative to expected outcomes.

It is important to note that social audio technologies will not be a replacement for current AI chatbots or virtual conferencing and telehealth programs, but they could supplement those services.

Photo credit: taniasv, Adobe Stock