By Phil Marshall,
Co-founder & Chief Product Officer,
Conversa Health

Theodore Roosevelt once said, “Nothing in the world is worth having or worth doing unless it means effort, pain, difficulty.” And while the benefits of advanced technology in healthcare are certainly worth having, introducing it for the long haul also comes with its difficulties.

Seeing a new technology through to full adoption has many layers to it, particularly in patient experience management, as many different stakeholders are involved–patient, doctor, admins and board members–at times, siloed from each other.

There are many health systems today that feel they provide excellent care, but feel the need to provide more services and cater to larger populations; doing so will maintain loyalty and retention. But that bucks up against limitations, like the need for more staff.

To address this, we are seeing more and more health systems implement “digital front door” programs, essentially killing the neverending phone directory or navigation of complicated websites in exchange for quick and easy access. It isn’t outsourcing to a machine; at the end of the day, most patients just want to book an appointment or get more information on their conditions, on-demand. And all of them want to feel like they have continuous access to their care team.

The majority of providers are adopting automation in some way, whether for handling claims or texting appointment reminders. With the rise of pop-up clinics and retail care, a hospital’s success will increasingly depend on their use of this technology to improve patient outcomes and save billions.

As a physician, I was familiar with the areas that had room for improvement. But I wanted to ask the providers, not just the patients, how they were thinking about technology. For example, when i approached Northwell Health, they had been asking how they were going to innovate further with technology, how they’d be able to scale it, measure patient satisfaction, manage care coordination, and reliably collect patient-generated health data. Then, what about after that? How would they reign all of that into their care plans, and use it to improve the in-person conversations?

The Process

Shake on it. The first step sounds simpler than it is: there must be unanimous agreement to start the road to change. Business and clinical teams have to make the choice to use a technology to better their care and more easily communicate, together.

Collaborate. This is crucial for onboarding. That means having a tried-and-true methodology in place that is consistent, scalable, and comprehensive. One aspect is providing the tools care teams need to adapt, which is largely about education. Webinars, quick videos and in-person training will address different learning styles, increasing the rate of comprehension and acceptance.

Think about the end user. Walk care teams through the patient experience on the chat platform, review talking points and explain how data gets uploaded into the clinical workflow (and where it goes). This will let all users feel in control of what they need to access, when patients are having conversations or if there are any red flags.

Play the long game. Look at usage over time. Ideally, formulating a disciplined 30 / 60 / 90+ day plan with your technology provider will help pinpoint weaknesses and provide measurable outcomes to leadership.

Hold hands. Care providers should walk their patients through the initial use, with the goal of meeting the patient where he or she is. At Northwell, staff would ask patients during the discharge process about their technology use (do you Facebook your grandchildren?), and explain that this technology is a way to better care cand communicate during their recovery process. They should show the patient how it works in real time, and activate it on patients’ phones right there before they go. The hardest part about a cold shower is that first step into the water.

Make it visible. On the provider side, care teams should have all the information they need to provide to the patient on hand. This creates an experience between the care team and the patient that bridges over to the platform. Ultimately, introducing it this way enhances human involvement, not replaces it.


Clinicians have hundreds of patients to call for follow-up checks, so applying automation to their massive caseloads tends to be a priority. Making hundreds of calls daily, they struggle with voicemail, wrong numbers, no answer or someone who has trouble communicating.

When patients respond to a few simple questions via text using automated chat, that data is routed through the system to a clinical dashboard, making it immediately viewable. After AI analyzes the data, each patient is given a colored flag (red, yellow, green) that marks the severity of the conversation. So, at a glance you can see that Mrs. Jones is having a lot of pain and more swelling, but Mr. Smith says he is fine and taking his meds. Now staff can divert their time toward those who actually need it, with fewer patients falling through the cracks. And no voicemails for miles.

A longer term benefit of automated, smart technology is viewing a patient’s fit with their healthcare roadmap. Within one healthcare system, the provider was able to identify that a new plan was needed for 29% of patients, all based off the data from chats.

The Challenges

One of the most common questions I’ve heard about technology in patient care is that it takes two to tango–will seniors really embrace chatbots? Turns out, the answer is a resounding “yes”–the average age of patients using Conversa is 64. Younger generations already prefer text over email.

As with anything that seems to do a human-like job, there are fears to quell on both the provider and patient side. Care teams may fear that the technology will fail them in some way–a step might be overlooked, or a patient might be harangued by ten texts a day. To prevent that from happening, healthcare systems should make sure that the solution is customized to their processes. Leadership must help their care teams better coordinate the management of the resources they do have, and see that reflected in the chat-to-clinician process. Sharing a detailed plan and timing that the solutions group approves makes sure that the chatbot will always provide approved messaging. The healthcare system, not the technology platform, should decide how they want to offer their care. As a result, one provider has seen an 86% engagement rate with their care team because they felt confident that the technology was extending their touch and decreasing readmissions.

On the patient side, some users can find the intelligence of the conversation to be daunting. They also don’t want this to turn into a daily spam bot. Chats should be scheduled appropriately (this patient gets a check-in every three days, the other one needs daily contact). Providers determine the frequency of outreach, with the goal of being sensitive to people in the recovery process to keep the integrity and validity is one key point. Nowadays, we can program a bot to be empathetic.

What We Learned

Work closely with leadership and stakeholders. Hold clinical team sessions, collaborate with business teams. When it’s fully customers, the chat platform should be based inside their health system and branded as such (“Northwell Health Chats”). Patients don’t care which company is providing the hospital with the service, they want to see the hospital make it their own. Patient onboarding should be baked into the provider’s process. When Northwell Health did this, their chat programs saw an 80%+ patient engagement rate, with patients completing conversations 93% of the time. 95% of patients said they found the chats helpful.

Think about how we think. Integrating self-determination theory to create an empathetic exchange, but also motivate the patient, is a vital factor in successful outcomes. This kind of technology will lack an impact if the conversations are never finished, so psychological incentives must be woven into the user experience.

Put humans first. Engage patients and providers at the experience level, because a patient’s status can always escalate in a health chat. Patients will use the button the contact the care team if they know it’s there and how it works.

Using conversational AI and automated technology as a staple in patient care makes it easier for providers to meet the patients where they, integrates meaningful health data in a new, efficient ways, and helps reinforce the care team-patient relationship. But the real success will go to the providers who execute onboarding wisely.

Dr. Phil Marshall is co-founder of Conversa Health, a chatbot platform that uses automation and AI to improve patient outcomes.

Contact Us

Contact Us

Collaborative Innovation in Healthcare
7550 IH 10 West, Suite 400
San Antonio, TX 78229-5616

Phone: +1-877-463-7678

Connect With Us


You have Successfully Subscribed!

Share This