Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Internet diabetes monitoring


Considerable effort goes into outpatient clinics, where patients with chronic diseases return at intervals for monitoring or advice. Very often samples are taken for tests, and more visits are needed or letters are written. It really is a great deal of work for professionals and healthcare systems, and effort and fuss for the patients. What if someone invented a way of communicating over the telephone system using television, when patients could monitor their own condition at home? It's called the Internet, and randomised trials of Internet interventions are beginning to emerge [1].


The setting for this trial was Korean patients with type 2 diabetes and Internet access. Severe concomitant disease was an exclusion criterion, or previous participation in any similar programme. Participants underwent examination and laboratory tests before and after 12 weeks in the study.

Patients consenting to participate were randomised to usual or Internet care. Usual care involved monthly visits with two or three visits with senior staff during a 12 week period. The Internet intervention consisted of a portal in which patients could enter pre- and postprandial blood glucose results, with information on type and dose of glucose lowering drugs, weight, exercise, and any other important changes. There was also an opportunity to ask questions.

Participants in the intervention group were encouraged to use the Internet three times a week, and participants in both groups were given a supply of monitoring sticks. At the hospital results were examined and a multidisciplinary team provided electronic responses with recommendations regarding medicines according to extant Korean guidelines.


One hundred and ten patients were randomised, with 101 completing the final examination. Most losses to follow up were because patients stopped visiting the centre. At randomisation groups were identical, were two-thirds men, and had an average age of 54 years. In the intervention groups the average number of log-on times per patient was 42, roughly once every two days. These patients asked an average of 14 questions over the 12 weeks.

The average frequency of blood glucose monitoring was twice as high in the intervention group (average 72 sticks used) as in the usual care group (38 sticks), though a few high users probably skewed these results. HbA1c levels were significantly lower in intervention than in control groups, for all completed patients, and those with baseline HbA1c levels below or above 7% (Figure 1). There were no significant changes to lipids or fasting blood glucose between baseline and the end of the study in either group.

Figure 1: HbA1c levels in Internet users and usual care controls


This is the first randomised trial of the use of the Internet for outpatients Bandolier has seen. The implication is that better management is achieved when patients use electronic access to information, and they seem to be very keen. How this would work over longer periods remains to be seen.

Patients using the Internet monitoring system were certainly motivated. They used twice as many blood glucose strips. We know that increased use of strips in type 2 diabetes leads to lower HbA1c levels (Bandolier 93), and that lower HbA1c levels lead to lower overall costs (Bandolier 84).

There is no economic analysis with the paper. As best one can see, using the Internet meant that outpatient visits were not needed, while patients remained interested in their condition and treatment. What would be the result on hospital staff utilisation and overall costs remains unknown, but it would be brave to bet against it being cost saving to the hospitals and to society.


  1. HS Kwon et al. Establishment of blood glucose monitoring system using the Internet. Diabetes Care 2004 27: 478-483.

previous or next story