Texting for Health Revisited

Way back in the distant past (2011) I wrote a series of posts on the power of texting patients. It was actually three posts (one, two, and three.) The last one was focused on the first set of data published about the effectiveness of text4baby, the much hyped, very expensive texting campaign targeting underserved pregnant women. More on that in a bit.

At the time of writing those posts, I was pretty bullish on the power of SMS. I am still bullish. In 2011, I wrote:

In the short term, I think medication and appointment reminders, which both have a clear benefit if implemented correctly, will dominate. Though they aren’t very flashy, these will be the biggest short-term uses of SMS in healthcare. Applications of SMS in health, like text4baby, will eventually have results of their pilot projects published, and those will drive the widespread use of messaging services in the future.

We do have more data now, but we’re still not seeing a ton of SMS in healthcare. Medication adherence and appointment reminders are certainly more common, though I wouldn’t say anywhere close to the norm or scaled.

Also at the time of those posts, I wrote about the need for messaging services beyond SMS that address some of the security and privacy concerns of SMS:

Additionally, new, non-SMS messaging services like TigerText are using the familiar format of SMS, but securing it. These services still require apps, which limit the value of this for patient messaging.

TigerText has raised lots of money since that time and recently opened up its API so other apps can integrate secure messaging. More recently PingMD, Me-Visit, and DoctorBase have been marketing tools to do secure messaging between patients and providers. These apps and services offer two-way communication and some are even structuring it, something I haven’t seen any pure SMS campaigns doing (mostly because SMS is pretty hard to structure.)

I really thought SMS in healthcare would have grown by leaps and bounds, much more so than it has. But much of what is happening is still in the research stage. Maybe I’m wrong and SMS, because of its inherent lack of structure and security, is not made for healthcare prime time. It just seems like such as easier and more efficient way to communicate.

And now on to the new data. Text4baby is the model of SMS in healthcare in the US. It’s scaled pretty widely and bankrolled by third parties, so providers (HHS and I think mostly Johnson & Johnson) don’t have to worry about the business case. I was initially skeptical about text4baby, but I’ve come to see the value in it. It’s really hard to assess its ROI, it’s doing what it is meant to do in engaging an underserved population.

The original data about text4baby, which I wrote about in 2011, was most compelling because of the rate at which participants contacted services included in text4baby messages:

"I think the most important stat, is that 39 percent of all text4baby users — and 53 percent of text4baby users without insurance — contacted a service that was listed in a text4baby message. This is pretty amazing to me. About half of users were engaged and trusted text4baby enough to follow-up on a lead in a text message."

I still think that 39 percent is an amazing stat. That data was based on 150 users. The most recent data on text4baby is based on 631 users, but the finding is similar – 40 percent called a service or phone number listed in a text4baby message. Just like in the previous study, that proportion goes up to 50 percent for uninsured patients. Text4baby is effective as a lead generation tool for services for uninsured moms. There’s got to be a business case for that, right?

Other findings from the most recent study include:

  • 65 percent reported that text4baby helped them remember an appointment or immunization that they or their child needed;
  • 74 percent of participants reported that text4baby messages informed them of medical warning signs that they did not know;
  • 67 percent of participants reported talking to their doctor about a topic that they read on a text4baby message.

Those are all good things, but I think connecting uninsured moms to services and programs to which they might be eligible is the biggest benefit of text4baby.

We now have better data that confirms the data we had from 18 months ago. Does it matter? How do we use this knowledge, this evidence, as a mobile health community? I’m not really sure.

The main learning for me, which is again the same as it was 18 months ago, is that we should be using text messages to engage underserved populations. Can one-way text messages effectively engage patients that aren’t already engage? I consider pregnant women to be some of the more engaged patients in our health system. Voxiva, the tech platform behind text4baby, has programs for diabetes and smoking, though I’m betting the patients are less engaged and less likely to be paying attention to messages than expectant moms.

I feel like I’m writing the same post I wrote 18 months ago. New data yes, but nothing terribly groundbreaking or actionable. Denver Health, the research link I included above, is scaling its SMS program to underserved diabetics. That study will likely provide more answers, especially as it is targeting chronic disease and using two-way communication.

I should mention that the other text4baby study, conducted by George Washington University last year, found participants subjectively felt better prepared to be mothers. I’m not sure that data adds much to the discussion.

Travis Good is an MD/MBA involved with health IT startups. More about me.

↑ Back to top

Founding Sponsors

Platinum Sponsors