Paddy Padmanabhan is founder and CEO of Damo Consulting, a growth strategy and digital transformation advisory firm that works with healthcare enterprises and global technology companies, and author of The Big Unlock: Harnessing Data and Growing Digital Health Businesses in a Value-Based Era.

My first instinct on being hit with the flu (after consulting Dr. Google, of course) was: I’m fine, I don’t need to go to an ER. I need to speak with my primary care physician (PCP), get him to prescribe some meds, have someone pick it up for me, stay hydrated and ride it out for the next few days.

I typed out a message via the myChart app on my iPhone (I had also lost my voice, so I was in no shape for a phone conversation) to my PCP, gave him a brief run-down of my symptoms, and waited. I had to wait two days before I received an answer — from an administrative staff member, who asked me to call their after-hours service. By then, the worst was behind me and when I got through to the on-call doc, she asked standard triaging questions, told me just to get rest and got off the phone as quickly as she could (I could hear a wailing child in the background, which may have had something to do with the briefness of my encounter).

Now it’s time to introduce a new character and a wonderful woman into the story: Grace.

Grace is a new chatbot that has been test-launched by Providence St Joseph in Washington State. Grace, whose job description would be “symptom triage,” appears as a pop-up chat window, and asks for symptoms before triaging the responses to decide on an action (e.g. whether to schedule a clinic visit).

The possibilities for Grace are tantalizing. Imagine if I could have had access to a Grace who could have determined after a series of mostly yes/no responses whether I needed to go to an urgent care center, or get some prescription meds, hydrate and rest. If required, Grace could have called me a rideshare, provided a heads-up to the clinic and sent over my medical records along with the chat transcript, notified my wife via text, and sent me on my way — after informing me what my out-of-pocket costs and copays were likely to be. Of course, if she found that my symptoms were not part of her library of scenarios, she could have transferred me to a live doc for the next order of diagnosis.

Grace and other such healthcare chatbots are here. What does that mean for healthcare providers and consumers?

Chatbots provide immediate access to care

For the vast majority of routine medical conditions, people don’t have to go to a doctor (and many don’t). Call this the password reset problem: you’ve lost your password, but you don’t need to talk to the CIO. A software tool will do the job. The healthcare sector has a staffing shortage problem that’s not going away anytime soon. At the same time, patients looking to access care struggle to get through to the physicians they want to talk to, for a host of reasons. The result is frustration for patients and lost revenue for healthcare providers. No wonder a recent poll listed access and convenience as the overarching goals for healthcare CEOs.

Chatbots will get you to the right caregiver and care setting more accurately and in a shorter time

Data from Providence’s deployment of Grace suggest 90% accuracy in its recommendations. Data from Advocate Aurora Health on 2000 patients who used a healthcare chatbot indicate that patients used it most for urgent or same-day care, and clicked on the symptom checker’s recommendation 35% of the time. The range of scenarios and the populations covered may be limited at present, but it’s a start. As with the use of chatbots in automating processes, the accuracy and convenience of the tool leads to greater adoption over time through machine learning and natural language processing (NLP) tools, potentially leading to tackling more complex care over time.

Chatbots can also cannibalize revenue (for your provider)

For the same reason that telehealth adoption is not growing faster, the use of chatbots will also struggle, especially among the vast majority of health systems still on fee-for-service (FFS) reimbursement models. The preponderance of FFS models may be one reason why big health plans like Anthem have launched symptom checker services to try and direct consumers towards the appropriate care options. Of course, this has implications for who gets to control the primary interface with a healthcare consumer at the time of need.

A related challenge is the possible proliferation of chatbots that can cause  “automation sprawl.” My firm’s work with a leading health system last year found dozens of 1-800 numbers and multiple web portals for patients trying to access care within the system. So much for digital front doors. The proliferation of chatbots, possibly from various technology vendors, can lead to the same situation, only now in addition to multiple phone numbers and portals, we also have numerous chatbots. All of these moving parts have to be carefully coordinated (maybe by another higher-order cadre of chatbots?) to minimize channel confusion.  At least one state,
California, has also introduced a Bot Law that requires chatbots to reveal their “artificial identity” to consumers.

Final thoughts

Chatbots, and in particular Robotic Process Automation (RPA), have made significant inroads into customer service and back-end organizational processes. However, delivering healthcare is different from selling credit cards or interest-free checking accounts because of the potential for harm arising from incorrect or confusing guidance. The anticipated benefits from healthcare chatbots are immense. If my own experience is anything to go by, chatbots can not only improve access but also enhance the quality of care advice over time. The technology is here, but is in very early stages.