With the help of the Behavioral Insights Team, New Orleans experiments with different text messaging campaigns to improve turnouts in primary healthcare visits.
This post is part of a series about our work with the 2nd cohort of the City Accelerator, an initiative from Living Cities and the Citi Foundation focused on municipal government and public engagement. In the second cohort led by the Engagement Lab, 5 cities rethink and reinvent public engagement, especially as it pertains to lower-income residents. In this series we’ll be sharing progress of the city’s projects as well as best practices and lessons for public engagement.

How can text messages encourage people to see a doctor?

Lessons from the New Orleans City Accelerator Project

By: Joseph Boskovski and Owen Phillips (Associate Advisors with the Behavioral Insights Team North America)

If you haven’t seen a doctor in a long time, the thought of picking up the phone to schedule an appointment can be terrifying. Sometimes it’s easier to adopt an “out of sight out of mind” mentality. But the people who haven’t seen a doctor in years are the ones most likely to benefit from a check-up.

The City of New Orleans and 504HealthNet (a non-profit member association of New Orleans health clinics serving the poor and uninsured) is currently experimenting with innovative ways to encourage take-up rates of primary health care. Their focus is on patients qualifying for the Greater New Orleans Community Health Connection (GNOCHC), a Medicaid waiver program with similar eligibility requirements of the expanded Medicaid program available in many states (but until recently, not Louisiana) thanks to the Affordable Care Act. Community health leaders found that while the GNOCHC program provides critical primary and behavioral healthcare coverage, only about half of covered patients had utilized the free care available to them in the past 24 months.

In a project supported by Living Cities’ City Accelerator program, the City of New Orleans and 504HealthNet engaged patients in the GNOCHC program through phone interviews, cohort convenings, and a day-long “design day” to understand their challenges in accessing care and to generate ideas for improving health care utilization.

One of the recurring themes that came out of that six-month long engagement was the need for better and more direct communication regarding the free health care available in a network of clinics established after Hurricane Katrina when the iconic Charity Hospital shut its doors for good.

With this feedback from patients, the New Orleans team asked the Behavioral Insights Team to design and conduct a randomized control trial to test different SMS text messages that encourage recipients to schedule a free check up. This trial was supported by Bloomberg Philanthropies’ What Works Cities initiative and is part of the City’s NOLAlytics initiative, aimed at leveraging data science to improve City services in New Orleans.

Text messages were sent to over 21,000 low income adults (aged 19–64) in the GNOCHC program. Importantly, none of the recipients had seen a primary care physician in the past two years.

Those eligible were randomly allocated into three groups; the Simplicity group, the Ego group, and the Social Motivation group.

Every message included three parts. First, a warm greeting followed by instructions to text “YES” to be contacted by a healthcare representative to set up a free appointment, and an option to text “STOP” to unsubscribe.

But each message differed slightly in what was emphasized.

For example, the Simplicity group just received straightforward instructions to “Txt YES to be contacted to set up a FREE doctor’s appt.”

Meanwhile, the messages sent to the Ego group included an additional line that said, “You have been selected for a FREE doctor’s appt.”

And in the Social Motivation group recipients received, “Take care of yourself so you can care for the ones you love.”

Five days after the messages were sent we tallied up the number of “Yes” responses. The proportion of recipients who replied “Yes” in each group are displayed in the graph below:

Many of the people we polled in advance (ourselves included) were certain that the Social Motivation message would be most effective. But we were surprised to learn that it actually performed significantly worse than the Simplicity text.

It turns out, recipients in the Ego group were the most likely to respond “Yes”. Indeed, twice as many people responded “Yes” in the Ego group than in the Social Motivation group.

All differences are statistically significant at the 95% confidence level, meaning we can be very confident the results are not due to chance.

In the coming months we will follow up with the recipients to find out if these texts convert into actual appointments. But the results are clear: simple, low cost interventions, like the use of text messages, continue to prove to be a promising method for improving communication practices with underserved populations to improve public health outcomes.