Conversational AI—fad or the future of patient access?

2023-10-02
Conversational AI—fad or the future of patient access?
Preview
来源: FierceHealthcare
None
Despite large financial investments, digital front door solutions still fall short, leading to patient leakage and additional work for already overburdened staff.
Yet, the idea of the digital front door is still the right one. The problem is the execution.
Patients overwhelmingly prefer to use digital tools for tasks like finding a doctor or scheduling an appointment. Solutions like search bars, chatbots, and portals just aren’t delivering.
Conversational AI is seen by many as the missing key to finally realizing the promise of the digital front door. But can it live up to this promise, or is it just hype?
An overview of conversational AI and how it works
Conversational AI is a way of harnessing LLMs (large language models) to understand spoken language, respond to questions, follow commands, and engage in interactive conversations—think Alexa and Siri, but with voice and text.
Thanks to advancements in the last few years, the technology has become (and continues to be) exponentially more capable and configurable.
In 2019, OpenAI introduced GPT-2, a first-of-its-kind, large language model (LLM) trained on 8 million web pages that could take text input, analyze all the information it was trained on, and generate unique, human-like responses.
That core technology (LLMs) can now be trained on proprietary data and customized for individual organizations and specific use cases.
These advancements—LLMs and the ability to train them on proprietary data—make it possible for healthcare to implement their own, and much smarter, Siri and Alexa-type assistants for patients and staff (conversational AI).
Why conversational AI has so much potential
Portals and websites contain vast amounts of information but are often cumbersome to navigate. Patients end up making unnecessary calls to contact centers about information that was accessible but hard to find—even with a search bar.
Chatbots were meant to relieve the frustration of making calls and the burden on the call center; however, because chatbots cannot understand the nuances and context necessary to provide helpful responses, patients find themselves on hold again.
Conversational AI solves this by being able to understand the nuance and context of an inquiry based on all the data it has at its disposal—including website data, call center documentation, patient EHR/CRM data, and more.
And because conversational AI can understand context as well as access information across systems, it can do more than answer questions; it can also help complete tasks like scheduling appointments, making payments, and refilling prescriptions.
Finally, the technology improves over time through machine learning, which enables the system to learn from interactions, adapt, and improve its responses.
Lessons learned from early conversational AI implementations
Pioneering health systems are already leveraging conversational AI in a variety of ways.
On the patient access side, conversational AI has been deployed as a virtual assistant that helps patients find information, schedule appointments, pay bills, and refill prescriptions just by typing a question or command.
Data from Notable Assistant, a leading conversational AI solution, reveals several trends in how patients are interacting with the technology. The most common interactions are related to:
Scheduling appointments. “I’d like to schedule an appointment.” Patients will often go to the website, and, when unable to self-schedule, call in to make an appointment. Conversational AI addresses this by enabling existing patients to authenticate and take action on their care journey 24/7 without phone calls.
Urgent care. “Where can I get care on a weekend?” Patients who do not feel well want quick and straightforward answers on where to go. Conversational AI can instantly surface the nearest location, contact information, and directions.
Finding providers. “I need an orthopedic specialist.” Rather than scroll through a long list of providers in a directory, patients are using conversational AI to find specialists and primary care providers with open availability with simple prompts.
Clinical questions. “My knee hurts,” “sore throat,” “depressed.” Patients frequently seek quick answers related to symptoms they are experiencing. This represents a risk and an opportunity. Rather than provide clinical diagnoses and recommendations with AI, most healthcare providers that have implemented conversational AI surface the most appropriate office or resource to contact.
Notable Assistant has an aggregate user satisfaction of 97% for the responses generated, and providers are seeing fewer unnecessary calls hit their contact centers.
Living up to the hype
It’s still early for conversational AI, but the feedback is overwhelmingly positive. Patients are giving conversational AI rave reviews, and health systems are doubling down by expanding the scope of the technology to do more than answer basic questions.
To see conversational AI in action for scheduling, bill pay, patient navigation, prescription refills, and more, check out Notable Assistant and its live demo of a fictional health system.
更多内容,请访问原始网站
文中所述内容并不反映新药情报库及其所属公司任何意见及观点,如有版权侵扰或错误之处,请及时联系我们,我们会在24小时内配合处理。
适应症
靶点
-
药物
-
来和芽仔聊天吧
立即开始免费试用!
智慧芽新药情报库是智慧芽专为生命科学人士构建的基于AI的创新药情报平台,助您全方位提升您的研发与决策效率。
立即开始数据试用!
智慧芽新药库数据也通过智慧芽数据服务平台,以API或者数据包形式对外开放,助您更加充分利用智慧芽新药情报信息。