Assessment of dietary compliances among patients with type diabetes mellitus receiving text message (SMS) reminder : a randomized control trial

Date of Award


Document Type


Degree Name

Master of Science in Epidemiology & Biostatistics (MSc Epidemiology & Biostats)


Community Health Sciences


Background It has been estimated that 285 million people are suffering from Diabetes Mellitus (DM) worldwide. Furthermore, 80% of deaths attributable to DM take place in developing countries. Compliance to a healthy diet is very important for effective management of type 2 DM. Use of electronic devices in the field of health care has been successful in motivating patients to improve their health behaviors. We can send text messages to improve compliance to healthy diet for management of type 2 DM. Studies have proved that by motivating patients with reminders through SMS, compliance to healthy diet can be improved which can lead to better diabetes control. However such studies have not been conducted in Pakistan. Objective To assess the difference in dietary compliance in patients with type-II diabetes, who were reminded through text message (SMS) vs. those provided standard Care. Methods A block randomized control trial is being conducted at Aga Khan University and Hospital in Karachi, Pakistan. Diabetic patients age 30 to 70 years, who have been diagnosed and conformed as a known cases of type 2 DM in the last 2 to 15 years, have a cell phone are being invited to participate in the study. In the intervention arm three text message reminders are being sent per week using Frontline SMS software for 3 months. Compliance to diet is being assessed in 2 ways. Firstly, overall compliance to dietary guideline is being measured by assigning score to the dietary intake assessed by Food Frequency Questionnaire (FFQ). Secondly, compliance is being measured by assessing responses to fortnightly two-item (fruits and vegetables) questionnaire that was sent to all participants via cell phone. Interim analysis was conducted after follow-up of 1.5 months instead of per protocol follow up of 3 months on 30 participants from each arm, because of time line issues for thesis submission. Results of interim analysis are presented in this thesis. Results There were no significant differences in distribution of variables in both the study arms at the baseline. The mean dietary compliance score (assessed by FFQ) in the intervention arm at the baseline was 7.90 and after follow-up was 7.80 while in control arm, the mean compliance score at the baseline was 7.77 while after follow-up it was 7.87. There was no significant difference between the two arms (F (group) = 0.065; p-value= 0.79) as well as within the arms (F (time) = 0; p-value>0.99). Marginal interaction was observed between group variable (intervention and control arm) and factor variable (mean dietary compliance score by time) (F (interaction) = 3.702; p-value= 0.059). No significant difference was found in dietary compliance between the two arms assessed by response to fortnightly two-item questionnaire (p-value=0.22). There was no significant effect of the intervention status as well as response to fortnightly two-item questionnaire on the mean dietary compliance score assessed with FFQ ( p-value (Intervention status)= 0.65; p-value (Response status)= 0.29). Interaction between the intervention status and response status was also insignificant (p-value =0.85). Conclusion We are unable to make conclusion about the relationship of dietary text message reminders and dietary compliance in type 2 diabetic patients. The reasons might be reduced follow up time as compared to per protocol and low post-hoc power of interim analysis.

This document is available in the relevant AKU library