Medication Reminders by Text: What 16 Trials Say, in Doses per Year

Half of all medication is not taken the way it was prescribed. That is not a guess. It is the figure the CDC uses, and it has a large and well-measured cost. The question for anyone managing a daily prescription, their own or a parent's, is what actually moves that number. Text messages are one of the few interventions with a pooled randomized answer.

The scale of the problem

The CDC's 2017 Public Health Grand Rounds on medication adherence, published in MMWR, lays out the national picture:

• 3.8 billion prescriptions are written in the United States each year.

• Approximately one in five new prescriptions is never filled.

• Of those that are filled, approximately 50 percent are taken incorrectly, with errors in timing, dosage, frequency, or duration.

• Nonadherence accounts for an estimated $100 billion to $300 billion in U.S. health care spending each year.

Working those first three numbers through: one in five of 3.8 billion is about 760 million prescriptions that are never picked up. Of the remaining 3.04 billion, half taken incorrectly is another 1.52 billion. Combined, about 2.28 billion prescriptions a year, roughly 60 percent of everything written, are either not filled or not taken as directed.

The CDC's own data also shows who carries the most doses. Among adults 65 and older, the share taking five or more prescription drugs in the past 30 days rose from 33.3 percent in 2001 to 2004 to 43.0 percent in 2017 to March 2020. More than two thirds, 67.7 percent, take three or more.

What the trials found

In 2016, Jay Thakkar and colleagues at the George Institute pooled every randomized trial they could find of text message reminders for medication adherence in chronic disease. The meta-analysis in JAMA Internal Medicine covered:

• 16 randomized controlled trials

• 2,742 patients, median age 39

• A median trial length of 12 weeks

• Text reminders approximately doubled the odds of adherence (the paper reports an odds ratio of 2.11)

• Adherence rose from 50 percent in the comparison groups to 67.8 percent with texts

The effect held regardless of whether messages were personalized, whether they were sent daily, whether patients could reply, or what disease they had. The authors caution that most trials measured adherence by self-report and ran only a few months, so the true long-run effect could be smaller. But across 16 independent trials, the direction was consistent.

Converting that into doses

An odds ratio is hard to feel. Doses are not. I took the pooled before-and-after adherence rates (50 percent and 67.8 percent), applied them to a simple daily schedule, and counted the doses a person would take or miss in a year. The method assumes one dose per medication per day and treats the adherence rate as the fraction of doses taken, which is a simplification of how the trials defined adherence, but it keeps the arithmetic honest.

For a person on five once-daily medications, the CDC's 43 percent group:

• 1,825 scheduled doses per year

• At 50 percent adherence: about 913 taken, 912 missed

• At 67.8 percent: about 1,237 taken, 588 missed

• Difference: 325 more doses taken per year, or about 27 a month, a little over 6 a week

For a person on two medications:

• 730 scheduled doses per year

• At 50 percent: 365 missed

• At 67.8 percent: 235 missed

• Difference: 130 more doses taken per year

And in the simplest terms, per 100 patients, the trials move 18 people from the nonadherent column to the adherent one.

Why text, specifically

Nothing in the trials depended on a clever app. The intervention was a plain message at the time a dose was due. That fits what we know about why people miss doses in the first place: not refusal, but the pill time arriving while attention is somewhere else. A text lands on the lock screen, on any phone, without needing to be opened.

It also fits the population that takes the most medication. Adults over 65 are the heaviest users of prescriptions and the least likely to run a dedicated app, but almost all of them have a phone that receives texts.

With Reminder Textapp, the setup is the sentence you would say out loud: "blood pressure pill every day at 8am" or "antibiotic every 8 hours". You get a text at those times. Reply "done" to clear it or "remind me later" to get it again in an hour, which covers the case where the dose is due but breakfast is not ready.

What a reminder cannot do

A reminder does not change the prescription, and it does not help if the reason for skipping is cost, side effects, or doubt about whether the drug is needed. Those are conversations for a prescriber or pharmacist. The CDC piece is explicit that adherence is a team problem, and it reports that team-based care after hospital discharge raised adherence to 89 percent versus 74 percent without it. A text reminder is one piece of that, cheap and immediate, for the share of missed doses that are simply forgotten.

If you take medication on a schedule, or you help someone who does, the trials suggest the forgotten share is large enough to be worth a message.

Try it by text

Text Hi to (720) 315-7147 to start Reminder Textapp. Then text your schedule in plain English: "take metformin every day at 8am and 6pm" or "eye drops every 4 hours". You will get a confirmation, then a text at each time. Works on any phone that can text, no app, no account. The first 20 messages are free. More at textapp.co/apps/reminder.