EVIDENCE SUMMARY
A meta-analysis of 6 RCTs studied the effect of smartphone self-care applications on A1C, weight, blood pressure, and lipids in adult patients with T2DM. All the interventions comprised 4 components: electronic self-management prompts and reminders, personal measuring devices, patient-driven data upload, and remote analysis of the data with feedback. The review excluded studies that used phone calls or lasted fewer than 3 months.
Some improvement in A1C found, but no effect on weight
Telehealth interventions improved A1C more than usual care (6 trials, 884 patients; mean difference = –0.40%; 95% CI, –0.69% to –0.11%).1 A subset of 4 studies with 560 patients evaluated changes in weight. Patients had a mean age of 61 years and average weight of 84 kg (in 3 of 4 studies reporting baseline weight). Aggregate weight loss was insignificant after 3 to 12 months (mean difference = –0.84 kg; 95% CI, –2.04 kg to 0.36 kg, P = .17). Investigators reported no harms. Limitations of the analysis included high heterogeneity in the main outcome of A1C (I2 = 70%) but low heterogeneity within the 4 studies assessing weight (I2 = 30%).
Other, small studies found no change in A1C
Two subsequent small RCTs came to different conclusions than the meta-analysis. One compared the impact of individualized physical activity–based text messages in response to pedometer readings with pedometer use alone.2 It included 126 adult patients (mean age, 50.5 years) with T2DM who had an A1C > 7% and access to an Internet-connected computer. Researchers excluded patients who were unable to perform moderate physical activity or who had cognitive deficits.
At enrollment, researchers supplied all patients with a pedometer and an appointment with a counselor to set goals for physical activity. They sent 2 text messages daily to the intervention group (and none to the control group) based on uploaded pedometer data. One message detailed physical activity progress and the second encouraged increased physical activity. The primary outcome was mean step counts per month; secondary outcomes included A1C and weight measured at 6 months.
The groups showed no significant difference in A1C (mean difference = 0.07%; 95% CI, –0.47% to 0.34%, P = .75) or weight loss (mean difference = 3.1 lb; 95% CI, –24.5 lb to 18.3 lb, P = .77). Many patients (43%) reported difficulty uploading step counts, receiving texts, and responding to texts. The dropout rate was 24%.
A second RCT with 150 patients, using a less elaborate protocol, assessed the effectiveness of tailored text-message reminders compared with nontailored text messages to improve A1C and body mass index (BMI).3 Patients were adult Iranians (mean age, 52.5 years) with T2DM who owned a cell phone and could receive and read text messages.
Patients filled out a diabetic self-care assessment to identify barriers to improving care and were randomized into 3 groups. The first group received tailored text messages (75% addressing the patient’s top 2 barriers to self-care and 25% general messages). The second group received nontailored text messages of encouragement. The control group received no text messages.
Continue to: After 3 months...