Interactions between healthcare professionals and patients although important, are often standard and repetitive. Analysis and results of these interactions help healthcare professionals to diagnose ailments and provide appropriate patient care. This process generally occurs by printing patient entry forms (containing a basic questionnaire) which a patient or kin fills in. Automating these interactions and analysis can not only improve overall patient experience in terms of better care, patient engagement and satisfaction but also result in significant cost savings. Besides healthcare professional-patient interaction, other interactions such as searching for nearest clinic or physician, benefits or claim status inquires, etc. could also be automated with the help of chatbots.
Healthcare aware and AI powered bots offer an alternative to human beings, where chatbots can ask a set of related questions in a guided conversation with patients. Healthcare professionals can take over once the basic analysis is complete thereby saving time and effort on rudimentary tasks. Additionally, increasing popularity of messenger apps such as WhatsApp, Facebook Messenger, etc. for personal messaging on smart phones, users/patients are getting comfortable in conversing using chat apps. Powering chatbots with AI and ML in addition to integrating them with existing healthcare systems can result in improved patient engagement, better patient care, better decision making and faster information exchange to name a few.
Typical Healthcare Use Cases
Popular use cases where chatbots are being effectively used in healthcare are:
- Primary Diagnosis – Conversational chatbots use techniques such as NLP, internal/external symptom checker DB, to ask relevant questions (e.g., patient age, gender, etc.) to assess patient symptoms and in primary diagnosis. It can suggest most likely conditions and propose next course of action based on the severity of the symptoms
- Appointment Scheduling – Intelligent Chatbots can be used to assist patients by asking for relevant information and scheduling appointments with specialist physicians. Based on the responses, it can also support them with pre-appointment and post-appointment actions. This would ensure that patients stay engaged with their providers post-appointment resulting in better care management
- Coverage and Claims Management – Payers can integrate their existing systems with conversational chatbots to help their members with coverage check, claim filing, claim status tracking, etc. Payers can also support providers in other coverage and claims related activities such as pre-authorization, billing, etc. Chatbots can efficiently handle claims related queries and management in a cost-efficient manner
- Clinical Trials – Chatbots can be designed to match patients (with help of AI) with appropriate trials. Chatbot starts with basic questions and then based on patient responses move towards relevant trial specific questions. Chatbots can be beneficial for both trial participants and investigators to a large extent. It can help them schedule site visits, report adverse events in real time, remind participants, monitor health status and perform other time-consuming tasks virtually online
- Healthcare IT Support – IT Support chatbots can solve common support cases such as password reset, resource access request, etc. Chatbots can help reduce the number of support calls / incidents by summarizing most common support requests within chatbot commands and relevant workflows thereby reducing human intervention, cost and increasing efficiency. Chatbots can also maintain and integrate with existing knowledge base to enrich, search and share relevant solutions to end users
As the market for chatbots and virtual assistants expands, major cloud providers too have released their version of chatbot services. These services not only provide readymade integration with existing chat applications (Skype for Business, Facebook Messenger, Telegram, Kik, etc.) but also provide SDKs to extend the service.
- Azure Bot Service – Provides healthcare organizations with an integrated environment purpose-built for bot development, enabling organizations to build, connect, test, deploy, and manage intelligent bots, all from one place. Also supports integration with Cortana, Skype for Business, Facebook, Kik, Slack, Telegram, etc. Using LUIS and Microsoft Translator, API developers can translate patient utterances into supported language. Azure Bot service works with multiple cognitive services such as face recognition, text analysis, speech support, spell check APIs, etc.
- AWS Lex – It’s suitable for building conversational interfaces in applications using voice and text. AWS Lex supports integration with Alexa, Facebook, Kik, Slack and Twilio channels. It provides support for speech and text support, NLP, and automatic speech recognition (ASR)
- Google Dialogflow – It builds engaging voice and text-based conversational interfaces powered by AI. It incorporates Google's machine learning expertise and products such as Google Cloud Speech-to-Text. Dialogflow supports integration with Google Assistant, Slack, Skype, Twilio, Facebook, Kik, Telegram, etc. It works with ‘Chatbase’ (analytics for chatbots), NLP, text to speech, speech recognition, automatic spell correction
Case Study: Interactive Symptom Checker
A leading healthcare ISV wanted to upgrade their current NLP, clinical, vocabulary and symptom-checker platform by developing a chatbot. CitiusTech developed an interactive and intelligent web based chatbot application which integrated with the ISV’s existing APIs. Additionally, Microsoft LUIS (Language Understanding Intelligent Service) for NLP, sentiment analysis and 3rd party services like Gender API were implemented to identify user’s gender based on their names.
Chatbots when powered with AI and integrated with healthcare systems can perform repetitive and routine tasks in an intelligent and cost-effective manner. Intelligent and conversational chatbots either through text or voice enabled AI interface can augment complete patient experience and help healthcare organizations provide better care and services.