mHealth Intervention to Support Diabetes Medication Adherence

HbA1c was extracted as percentage which is calculated following the standard formulae [HbA1c(%)=(HbA1c(mmol/mol)+23.5)/ 10.93] based on its concentration in blood samples. It identifies average plasma glucose concentration (1).

The most recent HbA1c data registered between 6th of April and 12th of November in 2021 were extracted for all participants at baseline.

At post-intervention, we extracted the most recent data which was available between 21st September 2022 and 21st February 2023.

We calculated adherence in terms of medication possession ratio (MPR), defined as the number of days with treatment as medication being dispensed from the pharmacy to the patient (numerator), out of the total days of treatment prescribed by the doctor (denominator) (2):

MPR = [Days with treatment (prescription dispensed) / Days with treatment as prescribed by the doctor] x 100

At baseline it was calculated as the mean adherence for all the glucose lowering drugs prescribed during the 6 months previous to recruitment excluding insulin.

At post-intervention it was calculated as the mean adherence for all the glucose lowering drugs prescribed during the 12 months follow-up excluding insulin.

Number of participants classified as adherent based on Self-reported adherence to glucose medications was measured with a 7-items ad hoc questionnaire adapted from Chaves-Torres et al. for people with type 2 diabetes. Participants who obtained 7 points were considered adherent while the ones with < 7 points were non-adherent.

The 5-level EuroQol 5-dimensional questionnaire (EQ-5D-5L) questionnaire was completed at baseline and post-intervention interviews. The index score was calculated using STATA syntax code and values with the Spanish value set according to Ramos-Goñi JM et al. Information about the EQ-5D index, derived using a value set, can be presented in much the same way as the EQ VAS data, i.e. using measures of central tendency and dispersion, such as mean values and the SD (or standard error). If the data are skewed, the median values and the 25th and 75th percentiles could be presented. Note, when reporting index values, a maximum of three decimal places is usually sufficient. Index values can range between 0 to 1. Higher EQ-5D index scores mean better quality of life related to health status.

The validated scale known as diabetes management self-efficacy scale in Spanish (DSES-S) was completed at baseline and post-intervention interviews. The scales consist of 8 Likert-type 10-point items. The score for each item was the number circled. If two consecutive numbers were circled, the lower number (less self-efficacy) was coded. If the numbers were not consecutive, the item was not scored. The final score for the scale is the mean of the eight items (points). Therefore, DSES-S scores range between 1 to 10. If more than two items were missing, we did not score the scale following instructions. Higher number indicates higher self-efficacy.

The 14-point Mediterranean Diet Adherence Screener (MEDAS-14) questionnaire was registered at baseline and post-intervention. Participants were classified as low adherents (≤5), moderate adherents (6 to 9 points) or high adherents (≥10 points) according to the results obtained that can range between 1 to 14 points. After that, we joined moderate adherents with high adherents and tagged them as adherents to the Mediterranean Diet, while low adherents were considered non-adherents to the Mediterranean diet. Therefore, we presented the number of participants adherent to the Mediterranean Diet.

A 6-items adapted from the short version of the International Physical Activity Questionnaire (IPAQ) (11) was registered at baseline and post-intervention. Participants were classified as having a low, moderate or high level of physical activity based on metabolic equivalent of task (METs) calculation. After that, we joined the participants that presented moderate or high level of physical activity and tagged them as adherents to Physical Activity recommendations, while low levels of Physical Activity was considered as being non-adherent to the Physical Activity Recommendations. Therefore, we presented the number of participants adherent to the Physical Activity Recommendations.