Paperonium

Diabetes care · JAMA Network Open · 2021

Summary: an app beat usual care on the bottom blood-pressure number. Its table: no clear difference.

The paper’s summaryapp did better (gives no number)Its own table, 6 months2.4, but could be no differenceIts own table, 3 months2.6, but could be no difference−1 to 5.7−0.96.2−2 points0246no differenceworseapp better
The paper’ssummaryapp did better(gives no number)Its own table,6 months2.4, but could beno differenceIts own table,3 months2.6, but could beno difference−1 to 5.7−0.96.2−20246no differencepointsworseapp better

How many more points the app group’s bottom number fell than usual care’s

The paper says

A trial gave 204 adults with type 2 diabetes a lifestyle app or usual care (their normal treatment, no app). The paper’s summary says the app group did better on weight, blood sugar, diet and the bottom blood-pressure number (the 80 in “120 over 80”).

Its own data show

Its own results section says no blood-pressure difference was seen at 6 months. Its table agrees: the app group’s bottom number fell 2.4 points more than usual care’s, but allowing for chance, the true gap could be anything from 1.0 point worse to 5.7 better with the app. At 3 months it is similar.

Even with no real effect, chance makes two groups differ, so a trial reports a range: the true effects its data can’t rule out. This one leans toward the app but includes zero, no difference. So the trial didn’t show the app beat usual care here; the summary shouldn’t say it did. Weight, blood sugar and diet aren’t in question.

“The summary” means the paper’s abstract: the short summary printed at its top, which is the part most readers see.

Read the paper →Full technical write-up →

Effect of a Smartphone App on Weight Change and Metabolic Outcomes in Asian Adults With Type 2 Diabetes · JAMA Network Open (2021) · doi:10.1001/jamanetworkopen.2021.12417