Memorias de investigación
Communications at congresses:
Successful replacement of weekly face-to-face visits by unsupervised smart home telecare in diet-treated gestational diabetes (GD)

Research Areas
  • Artificial intelligence (neuronal nets, expert systems, etc),
  • Cronic diseases (diabetes, asthma, others),
  • Telemedicine

We have developed a computer-based smart telemedicine system with the aim of giving automated support to GD patients while insulin is not required. The smart telemedicine system combines a platform for remote monitoring of diabetes-related parameters with a decision-support system (DSS) based on expert knowledge that generates automatic feedback to patients and/or clinicians. The DSS generates initial and follow-up diet treatments and detects the need to start insulin. Blood glucose (BG) data downloaded to the system from the patient?s glucose meter is automatically classified into mealtime intervals and moments of measurement (preprandial,postprandial) by a classifier based on a decision tree. After downloading BG data and informing on ketonuria fasting status, the patient receives an evaluation of the data and, if needed, a proposal of diet adjustment. In case insulin therapy is advised, the system also informs the responsible doctor who schedules a face-to-face appointment. Sixty-nine patients diagnosed of GD following the NDDG criteria were randomized (2:1) to use the system (active group) or to attend the usual face-to-face visits (control group). At baseline, groups were comparable regarding all the clinical variables tested. During the follow-up period (36 days (1-141)), no correction of the automated-proposed treatment was done by doctors. Mean number of BG downloads by patients was 10.2±8 (1-29) and the mean number of changes in diet automatically proposed was 0.46. Mean number of BG values/day, mean BG and the % of BG values above 140 mg/dl, pre-partum HbA1c, and all the perinatal outcomes tested were similar between the groups. Mean number of face-to-face visits performed including first visit and training was 4.8±2.8 for the control group and 1.4±0.6 for the active group (p<0.001). In conclusion, this computer-based smart telemedicine system successfully replaced face-to-face follow-up visits in women diagnosed of GD while insulin therapy was not required.
American Diabetes Association's 75th Scientific Sessions
Boston, United States of America
Start Date
End Date
From page
To page
American Diabetes Association's 75th Scientific Sessions. 64 - Suppl. 1
  • Autor: Mercedes Rigla Cros Hospital de Sabadell
  • Autor: Gema Garcia Saez UPM
  • Autor: María Villaplana Hospital de Sabadell
  • Autor: Estefanía Caballero Ruíz UPM
  • Autor: Belén Pons Hospital de Sabadell
  • Autor: Anna Méndez Hospital de Sabadell
  • Autor: Montserrat Aguilar Hospital de Sabadell
  • Autor: Enrique Javier Gomez Aguilera UPM
  • Autor: Maria Elena Hernando Perez UPM

Research Group, Departaments and Institutes related
  • Creador: Departamento: Tecnología Fotónica y Bioingeniería