Conversational AI in Healthcare: The Chatbot Will See You Now


Telehealth, chatbots, voicebots, and other personification technologies are center stage these days, particularly in healthcare. Across the entire business and social landscape, many applications are looking to take full advantage of conversational AI capabilities, such as virtual healthcare assistance, e-commerce AI-powered bots, and virtual tutors in the educational sector. With the digital changes stemming from the current COVID-19 pandemic, touchless applications in the travel and hospitality industry can also expect to see a significant increase in popularity.

The downside to chatbots and voicebots? In healthcare in particular, it is common for patients to prefer human-to-human interaction, instead of human-to-machine interaction. A quick look at the technology helps explain the limitation.

Conversational AI is essentially a set of automated technologies used to present understandable interactions between humans and computers. This is done by identifying the Intent and Entities of each sentence to guide the machines in understanding the human language, also known as Natural Language Processing (NLP). The gathered information is then processed with other contextual information to finalize the exact meaning, also known as Natural Language Understanding (NLU).

Combining NLP and NLU allows the computing system to create the most appropriate response – such as automated text or speech. If the response is generated with human-like language, instead of a hard-coded one, then it is referred to as Natural Language Generator (NLG). The combination of all three makes up the workings of conversational AI, even though in today’s world, most conversational AI systems only utilize the help of NLP and NLU.

And therein lies the issue. The absence of a Natural Language Generator means there could be a strong lack of humanity in this interaction between patient and caregiver. With the technology requiring high levels of computation power, there is the possibility of failure when it comes to assessing complex sentences. This can leave the system seeming apathetic to human feelings, possibly causing a hurdle in the adoption and taking up more processing time than originally assumed.

Does the upside outweigh the “human touch” issue? Is there so much promise, and is automation so pervasive already, that it won’t be a daunting issue? I believe there is cause for hope.

Healthcare has never shied away from technology, albeit that some recent improvements can be hit-and-miss – electronic health records are only as useful as the data that has been input, for example. However, there is a consensus among providers and medical professionals that the industry is willing to accept some disruption if it has a positive impact on the overall patient experience.

In terms of technology with the power to both disrupt and improve the way that people engage with digital health, an argument can be made that conversational AI can be the missing link. After all, if your “smart speaker” (Alexa, Siri, can answer questions about the weather, start your car in the morning and tell jokes, then there is no reason why virtual assistants can’t provide you with the medical advice that you need, when you need it.

The Chatbot will See You Now

According to the World Economic Forum, virtual assistants – and, by association, chatbots – are not only being used in a diverse set of industries (healthcare, education, retail, tourism, and more), but also offer opportunities for companies to integrate NLP into routine or mundane activities. For instance, Amazon’s Alexa is the chatbot that the average person associates with the tech, but the key element in every interaction is the ability of NLP to understand what it is being asked and respond with the appropriate information.

Admittedly, building the chatbot is the easy part. Often, it is the conversational aspect that throws a digital wrench into the works. When you are talking to a finance bot or trying to return a pair of shoes, the stilted nature of these exchanges is rarely an issue. In a sector such as healthcare, it is the conversation and the information that is being dispensed that is a key part of how comfortable an end user is with knowing that they are not engaged with a human being.

In rural parts of the world, the use of so-called healthbots has already helped alleviate some of the pressure on localized health providers. It is ironic that the rest of the world may in fact embrace a trend that started –of necessity—in an underserved area.

Part of the challenge may also be one of perception: Conversational AI and chatbots are not new per se, but their acceptance can be impacted by falling under a too-broad definition of what they actually are. Simply put, a “chatbot” is AI software that simulates human conversation with end users – this can be text or voice – with the aim being to leverage machine learning algorithms and NLP to deliver required outcomes.

The healthcare sector has been experimenting with these solutions for some time, with Mobile Health News reporting in May 2020 that a leading healthcare provider was integrating bots into its Emergency Department EHR system. More recently, the World Health Organization launched a women’s health chatbot dedicated to breast cancer messaging – for context, this new digital experience follows hot on the heels of bots that delivered information on COVID-19, mental health and smoking cessation.

So, the bots are already here, but the question that we need to address is how we can use conversational AI and/or chatbots in addressing not only patient concerns but also alleviate the worry or concerns that a person has. Additionally, we should consider both the challenges to overcome and the advantages of using virtual assistants to deliver care and wellness.

An unbiased chatbot or conversational AI may be able to address awkward or pertinent questions in real-time, offering medically appropriate opinions and advice. With so many people having access to the ubiquitous computer in our pockets, the opportunities for engagement are essentially unlimited.

We all interact with digital experiences every day of our lives, so it makes perfect sense that the technologies we often use without thinking should be mirrored in the healthcare sector. And while the industry still relies far too much on legacy technology (disparate IT systems, pagers instead of wearables), the integration of digital health solutions is picking up pace.

These tools are here to stay and there is a defined need for them to become not only part of the digital health landscape but also a familiar resource for people to turn to. The effective use of chatbots to provide the right medical or clinical information at the right time is not a futuristic concept, these solutions are here now. What matters is how healthcare and life sciences companies choose to use them.

Over the last decade, the continued evolution of digital healthcare and the impact that next-generation technology will have on patient wellness and diagnostic outcomes has been a recurring conversation. Providers have been looking to make the most of the tech-centric tools that are increasingly available, while those in need of care or the right advice at the right time now expect the medical sector to be investing in digital solutions to physical problems.

The integration of tech into public health and wellness is nothing new, but the demands of the connected society has only raised awareness of what is possible, as well as flagged up the opportunities for engagement in the digital space. For more information click here:

Author Image
Apurv Doshi
Infostretch Corporation

Apurv Doshi is Senior Solutions Architect at Infostretch Corporation.

Infostretch Corporation
Company Logo