Texting MRI appointment reminders cuts no-shows, captures revenues

Texting on top of phoning outpatient MRI patients to remind them of their scan appointments gets them to show up, all right, although it doesn’t help spur them to arrive on time.

That’s according to Harvard researchers at Massachusetts General Hospital who prospectively tested the dual-reminder strategy against phoning alone on nearly 7,000 patients at two outpatient sites.

The Journal of the American College of Radiology published the team’s findings online June 2.

Chang Liu, MA, H. Benjamin Harvey, MD, JD, and colleagues randomly assigned 3,086 MRI patients to texting days (text plus phone) and 3,903 patients to nontexting days (phone only) over four months in 2016.

The team analyzed not only appointment attendance versus no-show rates but also whether or not patients checked in a half-hour ahead of the scan time, as instructed.

They found the texting add-on yielded a 25 percent decrease in no-shows, from 5.1 percent to 3.8 percent, compared with phoning alone.

The researchers calculated this translates into around $325,000 in revenue per year at their institution.

Showing their math, they underscore that the gain would owe largely to their implementing the texting system using email via SMS gateways. This approach meant text messages were free to send and patients could be contacted directly, without the use of a texting provider.

“What’s more, the potential for savings is even greater with increasing text message adoption and application of the system to other imaging modalities,” the authors write in their discussion, explaining that they started with MRI because of its lags between ordering and scheduling, long exam times and high costs.

Interestingly to the researchers—and no doubt frustratingly to imaging staff—the texting had no significant effect on patient punctuality. Around 40 percent of patients in both groups arrived later than 30 minutes prior to scan time.

Meanwhile, not surprisingly, the patients who were unreachable by texting were quite a bit older than those who were texting-accessible.

The authors acknowledge their single-site study design as a limitation, as it leaves open the question of whether their results are generalizable to facilities not operating in urban academic settings. They also note that, because they randomized their experimental intervention by day rather than by patient, they could not rule out that day-specific factors such as bad weather skewed some of their data.

“However, we demonstrate the economic viability of the text message reminder strategy, and thereby empower other institutions to test this strategy without a significant financial investment,” they write.