Curbside Consult with Dr. Jayne 9/11/23

I was recently in a meeting of healthcare executives where the topic in question was how to best leverage technology when working with “older patients.” Compared to my physician experiences with the senior population, I found some of their attitudes to be not only outdated, but their opinions to be inaccurate. I heard statements such as “retirees are technologically naïve” and “most of them don’t have smartphones.”

I’d like to see those executives join me in an exam room when a Medicare-age patient pulls out their phone and logs into their patient portal to provide data that I can’t otherwise access. They might also find themselves educated when a patient pulls out their phone to show me pictures that their grandchildren texted to them.

According to data from The Senior List, 98% of people older than age 60 have mobile phones, which splits to 81% smartphone users and 17% basic mobile phone users. Approximately 75% of seniors text at least once daily, with most indicating that texting is a preferred communication method.

A whopping 65% of adults in that age group spend three hours or more on their phones daily, and seniors have an average of 25 apps installed on their phones. The most popular apps are those addressing weather, shopping, social media, games, and news. Health tracking apps are present in around 49% of seniors’ app installations. As far as app use, social media leads the way as far as percentage of app usage, with Facebook being the favorite followed by Amazon, Google, and The Weather Channel. The data in question was gathered in 2021, and based on what I see in the community, I woud guess that smartphone use and app utilization have both grown since then.

One of the health system executives kept citing “the data” indicating that smartphones and seniors don’t mix, but didn’t have an actual citation. He said he had seen “multiple research studies in the news this week,” so I decided to try to figure out what he was talking about. I found a couple of articles in the lay press from mid-August that talked about COVID, seniors, and text messages not mixing, but they all linked back to the same scholarly article, so I hardly consider that an overwhelming amount of new research.

The study in question looked at text messages sent to more than 17,000 Medicare-age seniors during the initial distribution phase of the COVID vaccine. Patients were sent a text invitation to receive the vaccine, but 28% reported technical problems. The authors dug deeper into 4,200 responses, coding them based on issues experienced by the recipients. Approximately 7% of seniors had issues responding to the text invitation, including responding in the wrong format, delays in responding, or inability to get help when they requested it.

Based on the results, researchers recommended interventions including the use of natural language processing, flexible response formats, and improved feedback processes when conducting such text-based campaigns in the future. Based on my experiences as a consulting CMIO during the pandemic, I would also recommend that organizations look at the scripting they are sending out and whether the text of the messages is contributing to the confusion.

I was hired by a couple of organizations during the pandemic to support their efforts reaching patients, and some of the draft language that I was sent to proof-read or polish had definite room for improvement. Common themes I encountered during that time included reading levels that were too high, complex sentence structure, and questions with responses that didn’t have an appropriate response choice listed.

It should also be noted that some of these patients might have been receiving competing or conflicting messages from different organizations that added to the confusion. For example, they could have received different messages from their primary care physician, a subspecialty physician, a health system where they’ve had diagnostic or laboratory testing, a senior center, a retail pharmacy, and an insurance carrier. Duplicative messages might just have been ignored, confounding the data.

The article lists the specific language that was used as part of the campaign studied, and even I found it confusing. Recipients were prompted to select from three options, which were presented to them with a request to indicate their choice by keying in either “A” or “G” or “D” which were presented in that order. The offering of such choices ordered in that manner is counter to some key usability and patient engagement principles, especially in a population where recipients might be more conditioned to choose from an ordinal list – “ABC” or “123,” for example, rather than “AGD.”

The article noted that “the response options letters in the message were chosen from a set of letter groups and assigned at the time of delivery based on the next message service’s automated batch-processing system,” which is a statement I don’t really understand.

Had this script come across my consulting desk, I might have made it even simpler with yes/no questions and a bit more branching. Although it may have created complexity for those configuring the system, it would have been more straightforward for the potential patients receiving it.

Those patients who wanted to schedule online were told to enter their email address and that they would receive a link within two days. That requires additional steps for the recipient versus just providing a link right in the text exchange.

The study authors looked at some of the “non-conforming” messages that patients sent that the system was unable to process. Some of them included statements recognizing typographical errors (such as “Sorry, I sent wrong response, D is the response”) and requests to just go to a website. Others included statements around the patient having already received a vaccine or auto-generated replies that were sent because the patient was driving when the text message was received.

The authors also noted that the campaign wasn’t optimized to address the needs of patients whose primary language wasn’t English. Other limitations included medical records anomalies that may have led to the generation of messages to the wrong recipients.

It should also be noted that many of these campaigns were being done in the heat of the moment due to unpredictable vaccine availability. On some of the campaigns where I assisted, clients offered a premium for rush turnaround of content. Several of the outreach programs were built in a couple of days, and one was actually configured in less than 12 hours. Health systems were under immense pressure at the time and might not have had the luxury of constructing a campaign compliant with usability best practices.

Those all add to the idea of not using this type of data (which is now two years old in addition to the other flaws noted) to throw the proverbial baby out with the bath water as far as understanding how seniors will interact with technology. It will be interesting to see follow-up studies on the topic.

Does your organization do anything specific when tailoring campaigns to different demographic groups? Have you considered specific testing on your target audience? Leave a comment or email me.

Email Dr. Jayne.