Glucose concentration in blood, better known as glycemia, is a very underestimated physiological parameter in neonatal intensive care units. Its value is currently monitored by performing few measurements per day, acquired through heel pricks or arterial samplings. Unfortunately, the frequency of those samplings is quite low in preterm newborns, who are very low birth weight infants, due to their fragility. As a result, hyper-or hypo-glycemia events occurring between two separate measurements remain often unnoticed. A medical research group working at the University Hospital of Padova already proved the clinical benefits of continuous glucose monitoring for preterm newborns hospitalized in the neonatal intensive care unit. However, glucose data are still sampled through individual devices that are not integrated into a central monitoring system. Thus the health care staff has to visit regularly the cradles, watch the glycemia trends and pay attention to alarms. Our contribution deals with the implementation of a system based on mobile technologies that sends the data acquired through commercial glucometers to a central database, so that they become available in real time for the staff in the ward room. Moreover, we also integrated into that system an algorithm described in the literature for recommending the most appropriate glucose infusion rates based on the last readings of the patient. The system runs on an Android device and has been successfully tested in two different clinical trials involving twelve and five patients respectively.