So many tech products are about solving problems that aren’t really problems — like leveraging facial recognition technology to identify your friends’ faces in photos or location data to ensure you always have an Uber waiting for you at the same time every day, wherever you are. What if we were able to leverage machine learning and AI to solve real problems for real people that also have empathy for the user.
On my last visit to the emergency room, I had been feeling chest pains while trying to sleep for the fifth night in a row. On arrival and in the midst of a health crisis, a receptionist gave me a form and told me to fill it out. I went through the familiar administrative procedure of itemizing my medical history in the sterile-by-necessity waiting room. I knew the paperwork was necessary, but, because I was feeling vulnerable, it was even more of an annoyance than usual. Turns out I was just suffering from stress, and there wasn’t anything seriously wrong with me that quitting my stressful job, getting enough sleep, and eating well wouldn’t fix. But that experience got me thinking about how technology could make the introduction to a medical office or ER less stressful and traumatic.
Contrast that with how you are taken care of at home when you have a non-life-threatening, temporary health condition, such as a cold or the flu. If you are fortunate, someone will sympathize with you; bring you medicine, food, and hot beverages; and take care of things you are too sick to handle. Now imagine your caregiver asks you to fill out a form before handing you a tissue. Would you feel cared for?
After being greeted by a receptionist, the second point of entry into the typical medical-care experience is filling out a lengthy set of forms with one’s medical history and current medications. What if the experience could, at minimum, “pretend” to care? Imagine that, when you enter a hospital or doctor’s office, a human greets you, and then, rather than handing you a pen and clipboard, connects you to a patient network where a chatbot guides you through the medical questionnaire. Instead of a dry, frustrating form, you get a friendly, helpful personality asking for your basic information, answering questions if you are confused, and clarifying trouble spots with examples. The chatbot would be able to filter responses it recognizes as unhelpful to medical professionals, and it could prompt patients for more information when necessary. The output for health care professionals could be in a plain, easy-to-read format that matches their standard paper forms.
The idea, which is to give patients a more humane experience in the waiting room, came to me while listening to an episode of the podcast, Partially Derivative, where the hosts interviewed Joshua Browder about his chatbot, DoNotPay. The bot, which is designed to help people get out of parking tickets, asks for user information, then generates it in whatever format is required by the local government in question. When the City of Los Angeles was interviewed about the chatbot coming to their city, they feared losing revenue from individuals getting out of parking tickets, but they looked forward to seeing how it would streamline the appeals process, since the responses would be more consistent and understandable.
Consistent. Understandable. According to the study published in the Journal of the Royal Society of Medicine, poor handwriting is a “significant” problem in health care. The study’s findings were about doctors’ illegible handwriting, but imagine if we could reduce this problem on the patient side by ensuring the information they provide is legible, clear, and straightforward. Medical professionals are often overworked, tired, and are faced with sometimes hundreds of patients a day. Reducing error by reducing the number of instances that health care providers must struggle with or guess at long lists crammed into tiny boxes would save everyone time and frustration. And while the patient/doctor consultation eventually clears up many questions, there often isn’t time for extended conversations in scenarios like an overcrowded ER or a busy primary care practice with patient quotas. Confusion can be further avoided in cases in which a patient is seen by multiple health care professionals during the same visit.
Tools like MyChart by Epic Systems and other medical record management systems have made organizing patient information significantly easier for medical professionals and patients alike by incorporating user interfaces that the average patient can easily navigate. A patient-facing form chatbot could be its own product that integrates with an existing MyChart-type platform or act as a helpful new feature.
While there is potential to make the initial patient forms more consistent and user friendly for health care professionals, there is even greater opportunity to make the patient feel more welcome and more “cared for” when they first enter the health care experience. Chatbots in a health care environment should have a gentle and caring personality — they must have a “soul” — making them as empathetic as the human on the other side of the form. In most cases, no one is overjoyed to go to the doctor, and I can almost assuredly guarantee, no one is excited to go to the ER. Done well, a chatbot could soften some patients’ experiences and anxieties by shifting attitudes from “unhappy to be here” to a positive first impression. When I am a patient, I would take any humanizing improvements to my interactions at the doctor’s office, even if it came from a robot.
Chatbots have a lot to offer in the initial patient medical facility interaction by bringing warmth to an important aspect of our lives and efficiency to a system and industry that allows little room for error.