Web-based insulin titration: Improving diabetes care in the Netherlands. A pilot study of the PANDIT® system.

Diabetes mellitus is a disease characterized by an absolute or relative deficit
of insulin, a hormone necessary for the uptake of glucose into the cells. As a
consequence, the blood glucose levels of patients with diabetes are high,
leading not only to acute complaints such as thirst and fatigue, but also to
serious long term complications such as blindness, kidney failure and
cardiovascular disease. The occurrence of these long term complications can be
significantly reduced by adequate treatment, aiming for a reduction of blood
glucose levels to non diabetic values by diet, oral anti-diabetic treatment or
insulin therapy.
The effect of this treatment is evaluated by the widely accepted marker HbA1c
(glycosylated haemoglobin), which reflects the average blood glucose level over
a period of six to eight weeks. Current guidelines recommend an HbA1c goal of
7% or lower, which has been shown to result in a significant reduction of
diabetes related morbidity and mortality. This, however, has proven to be an
elusive goal, the average HbA1c level in the Netherlands being 7.6% for type 2
diabetes patients and 7.8% in type 1 diabetes patients.
A new strategy to achieve the appropriate HbA1c value in these patients is
urgently needed, as none of the novel pharmacotherapeutic options can be
expected to deliver such a result. Insulin therapy is still the most potent
agent in the therapeutic arsenal, being cost-effective and without an upper
dose limit. However, treating patients with insulin requires active dose
titration, increasing or decreasing the insulin dose according to blood glucose
response, diet, physical activity and the occurrence of possible side effects
such as hypoglycaemia (low blood glucose). The average 3-4 clinical visits per
year fail to provide sufficient titration opportunities and may cause the
treating physician or nurse to be cautious with dose increments, thus not
reaching the treatment*s full potential and HbA1c levels below 7%.
Therefore, we propose to develop an online, computerized titration algorithm to
guide patients in self-titration, the PANDIT system (Patient Assisting
Net-based Diabetes Insulin Titration). The use of an online, insulin titration
algorithm will not only allow for intensive treatment of patients with diabetes
but also shifts the focus of their treatment to self-management.
The Diabetes Research Group and Clinical Informatics Division of the Academic
Medical Center will develop the program that gives advice on insulin doses. An
algorithm for adjusting insuline doses that has already been proven to be
effective will be used. Patients are given a unique account and password to
the website. The Pandit be developed for patients with diabetes mellitus type
2, using once daily insulin injections.

The main purpose of this pilot study is to evaluate the safety of the PANDIT
system by testing the system in a group of patients, under strict supervision
of a diabetes nurse. The main objective is to check whether PANDIT provides
correct insulin dosing advices. Furthermore, after the safety checks, patients
will be asked whether the system is user-friendly, easy to use and if they have
suggestions for improvement. Based on the evaluations of these 20 patients we
will improve the PANDIT system.

This will a pilot study divided in two phases of respectively 4 and 8 weeks. In
total 20 patients will participate in this study, divided in two study phases,
resulting in 10 patients for each study phase.

During the first phase of the pilot 10 patients will make use of the PANDIT
system at their home for 4 weeks. Every dosing advice will be given by both a
diabetes nurse (DN) and the PANDIT system. Only the dosing advice given by the
DN will be displayed to the patient. When all patients have finished the first
phase, both advices will be compared and all discrepancies bewteen the DN and
the PANDIT system will be evaluated by a panel.

During the second phase of the pilot, another 10 patients will use the PANDIT
system at their home for 8 weeks.
During the 8 weeks they will receive the dosing advice directly from the PANDIT
system. This is checked by the DN on a daily basis. The DN will intervene only
if necessary (foreseeing any danger). An explanation will be given when
rejecting the PANDIT advice. Similarly, when all patients have finished this
second phase, all rejected advices will be evaluated by a panel.

The patient has to measure his/her fasting glucose value with the glucose meter
on a daily basis. After three days patients have to enter their fasting plasma
glucose values in the PANDIT system, confirm if they have used the previously
advised insulin dose and whether they experienced symptoms of a hypoglycaemia.
In addition the patient has to indicate whether they have started using
corticosteroids. Entering the information can also be done each day, but the
insulin dosing advice will always be provided after three days if an adequate
number of fasting glucose values has been collected.
After receiving the insulin dosing advice on the third day (if enough values
have been entered) the patient has to use the newly recommended dose on the
following three days.
The patient will be reminded by sms or e-mail if three days have passed and no
fasting glucose value has been entered yet.

.

Participation mainly requires daily blood glucose measurements and daily use of
a computer program on the Internet. PANDIT® titration will be performed at home
with a fixed interval of three days.
A site visit will be performed at the beginning and at the end of the study. In
total two venapunctures and a usability evaluation will be performed by the
patient.
Affirmation of the safety of patients using a web-based insulin titration
system will indicate that an improvement in HbA1c can be possibly achieved
through this system. This knowledge can justify the performance of a randomized
controlled trial in which the clinical efficacy of the algorithm will be
testes.