Iflexion
Chatbots to replace mHealth apps in patient engagement

Chatbots to Replace mHealth Apps in Patient Engagement

Lead: Some providers still doubt whether patients themselves needed patient portals to get data about their health status. Others tend to disregard the demand in mHealth applications to ensure convenient care delivery. Now, chatbots are new in the healthcare town. Would they have to fight with rejection as well?

One of the reasons why healthcare struggles to balance costs and patient health outcomes is rooted in the lack of patient engagement. As soon as patients step out of the facility doors, they have to face new routine, medications, nutrition limitations and exercise therapy all by themselves.

For the past few years, providers invested in healthcare software development to bridge the gap in communication with patients between appointments. Earlier, caregivers adopted patient portals, later they advanced with mHealth apps that allowed a more interactive way to improve patient engagement and involve patients in everything related to their wellbeing. Patient could track their vitals, journalize nutrition and symptoms, and keep up with the treatment plan and physical exercises.

With all the benefits that mobile patient apps bring in, they are still not enough to ensure patient engagement. For some patients, any reminder about their disease or limitations related to it is painful. It is frequent when patients stop taking prescribed medications only because either they don’t want to do it or they feel well. Thinking about illness is hard.

Why chatbots for healthcare

Such visionary technology companies as Microsoft, Amazon, Facebook and Apple put major stakes on chatbots, which are automated agents designed to simulate conversation in mobile messaging apps. Since messengers are among the most used applications today, different industries harness the power of the ‘conversation UI’ (another name for chatbot) and apply it anywhere from e-commerce and enterprise to news, politics and entertainment. The question is if this technology can be equally effective in healthcare and beat mHealth apps in patient engagement.

How ‘conversation UI’ can assist patients

Chatbots can open up a door into the AI-driven healthcare, where patients will communicate their symptoms within a messenger and get a consultation. A chatbot may ask a series of casual questions about the user’s age, gender and health complaints. The backend will seamlessly fill in a patient profile, which can later be used to automatically complete medical forms during admission or appointment.

By asking patients personalized questions and composing a profile, the UI then pinpoints the patient’s symptoms and determines the probability that these symptoms signal this or that disease. Then, a chatbot can compose a diagnosis suggestion with treatment options or advise a patient to schedule an appointment with their physician to take additional tests.

Accordingly, allowing patients to interact with their caregivers via one of those messengers is much more effective than asking individuals to download and open another standalone app. Instead, patients will have their own pocket healthcare assistant already set up and waiting for their requests. No additional actions needed.

This idea can also help elderly patients, who may be confused with a complex mHealth app UI, but certainly know how to write a message to their children or provider. Therefore, speaking with a chatbot about their health concerns feels more natural to them.

What makes chatbots stand out

Overall, chatbots seem to allow cutting down on unnecessary consultations and creating an advanced medical data model. However, chatbots can feel‘“robotic’, which hinders patient engagement in the same way as it happens with mobile patient applications.

To make chatting applications feel realistic, the healthcare software development industry taps into the recent advances in AI:

  • Repositories of public chat scripts, which allow chatbots to learn, understand and mimic human conversations with random words, exclamation marks and other flaws.
  • Tools to embed NLP, sentiment analysis and concept extraction technologies into chat scripts.
  • Algorithms on detecting emotions in the natural language to enable chatbots engaging patients empathetically.
  • Improvements in image recognition allowing chatbots to process pictures, hand-written notes and even QR barcodes.

Healthcare chatbots around the globe

In recent years, Project RED from the Boston University Medical Center has developed  nurse Louise, who can explain a condition to patients, teach how to take their medications and go over their home care needs at the hospital discharge time.

Your.MD is an UK robotic health assistant powered with AI. It allows a patient to enter the symptoms and get additional questions about gender, age and other symptoms to pinpoint the diagnosis. Upon receiving information about the suggested condition, a patient can be referred to a local caregiver. This chatbot also allows providing patients with therapeutic education on sexual health, mental health, wellbeing, family health, nutrition and exercises. Your.MD is available on major messaging platforms, such as Facebook Messenger, Kik, Skype, Telegram and Slack. Patients can also download the Android or iOS standalone mobile app, allowing to synchronize data from health and fitness apps to see the progress.

In California, Sense.ly has created Molly the virtual nurse engaging discharged patients to keep up with their personalized treatment plan, record their vitals and adhere to the physical activity routine. The application analyses the submitted data and notifies the health specialist in case any health risks emerge. While Molly is technically more than a chatbot, it enables the same conversation elements as messenger bots.

Babylon Health is an UK application that uses AI chat functions to navigate patients through their symptoms and then connect them with hospitals or health specialists. Moreover, it also provides patients with the ability to schedule, check medical records and lab tests.

Coming from the Stanford University, MedWhat is an artificial intelligence medical assistant. It can answer medical questions, help patients with their daily health and wellness routine and manage reminders. With the help of MedWhat, patients can have a Health Record profile with the history of answered questions, diagnoses and health state information, including steps count, weight and sleeping hours. This AI can also help physicians and nurses with personalized medical questions tied to patient EMRs and medical literature. It is currently available as Android, iOS or Windows Phone standalone apps.

Engaging patients via chatbots

As we know, patient engagement starts with interaction between a health specialist and their patient. The level of patients’ dedication to their own health gradually lowers as they step out of the facility and start fighting with the condition all by themselves. While mHealth patient apps are more effective in maintaining the distant patient-caregiver bond than patient portals, they still aren’t easy and natural for many patients.

Chatbots seem to be the next distant care delivery tool, and by offering a conversation-like experience, they may have what it takes to make patients even more engaged. However, we still need to remember that a personal relationship between a physician and their patient may never disappear.

Adriana Blum

  • Micheline Logan

    The problem with the personal relationship between a patient and their physician is that it is generally conducted in an 8-10 minute consultation. I think you will agree that this is not a very satisfactory relationship. Maybe a chatbot is just what the patient with a life-threatening illness requires, when their insomnia at 2am caused by chemotherapy has them in despair. The chatbot can lend a sympathetic ear and record the problem, and suggest mitigations based on the patient’s drug regime. Managing a patient’s health is a 24/7 occupation, and the combination of chatbots, smart drugs and wearables can monitor the patient’s condition and alert the physician to an impending medical crisis.