Theresa Pinn-Kirkland, Houston Methodist Hospital: When we were looking at this, we wanted to make it seem very personable…not that patients look at their discharge packets, but in their additional information, it does say you’ll get a text or a phone call.
There is a live nurse behind it. And so we’ve really tweaked verbiage probably about a dozen times at this point and are trying to make it very obvious that there are people responding to you on the other side because it’s a painful process to be like, “If you have questions, please tell me more. Yes, great. How can I help you? Medications. Please tell me more about your med questions. What can I help you with?” And then they’re like, “Oh, you’re not AI.” No, no, I am not AI. I just answered five things that were very specific to you. I promise I’m not AI.
We found that your octogenarians love emojis. Who knew? We get patients that text pictures in. They have all sorts of emojis. We send back emojis, so it’s very, very casual human touch.
Ashley Trambley, Vanderbilt University Medical Center: Well, we definitely took a different approach. And we’ve talked a lot about conversational messaging, which Vanderbilt’s general motto is making healthcare personal. So we did want to be personable, and for people to know there were humans behind it.
But this was really Vanderbilt’s first foray into using texting for a clinical use case. We have it for appointment reminders and billing and surveys like others. But our questions are things like, “Do you have your medications?”
And so we weren’t really sure, one, how our staff would respond to that, but then how patients would respond. And so we were trying to think through what’s the Pandora’s box if we open this up to our own teams to just converse free form?
And we were more conservative. We were a little hesitant because the technology was new and Artera has been lovely to let us take their appointment system and retrofit it into this clinical use case. So we were really learning the system and the features. And so we were like, let’s focus on technology first because some of the issues we anticipated potentially were branding. In communication, you lose body language, tone of voice, inflection, there’s language barriers. So we didn’t really want to put our staff in that situation to have to make those decisions in the moment while they’re learning this new technology interface that sits with them beside Epic and Teams and all the other things that they have open.
We also had the privacy concerns because as you said we we put out a “do you have med questions, yes or no?” And we predominantly rely on phone communication with some exceptions, but we leverage the quick responses and really try to lock down that communication.
Because patients will freeform. And even though we say one for yes, two for no, they’ll write a whole book about their headache and their blood pressure and photos of areas of their body that I can’t believe they texted into a healthcare provider. And so that was a privacy concern, too. It’s not so much on our part, but the patients do not understand that they’re coming in potentially through an unsecured channel. So we had that concern.
We also had some safety concerns and quality about we don’t want some nurses to think a cough can be really benign, but it can also be really emergent. We don’t want somebody making the decision of back and forth about talking in text and potentially having a chat open with somebody else and missing something clinically important.
So we went with the more conservative approach of really locking down what our staff can send. We put that through patient education and translations to make sure that it all comes across professional. We definitely want to move toward conversational, but we’ve gotten a lot of learnings on how patients use the system as well as our own team.
