The purpose of this study is to determine whether alerting primary care providers by email about low values of BMI, HbA1c% or cholesterol will affect treatment and improve overall survival and other health indexes of people older than 75 years.

Scientific background

Interventions aimed to ameliorate malnutrition are important for elderly health and include dietary counseling and discontinuing unnecessary medicines.

Emailing an alert regarding low BMI was found to improve dietary counseling numbers.

Correlation between death and HbA1c% is U-shaped, with increased mortality under a 6.5% level in patients taking two anti-diabetic medicines. Sending an email alert regarding an over-tight control of diabetes was followed by a reduction in mortality.

Death and cholesterol correlation is also U-shaped, with increased mortality and morbidity under 160 mg%. The investigator found no interventional study for this situation.

Objectives

To check whether alerting the primary care providers by email, about low values of BMI, HbA1c% or cholesterol will affect treatment and improve health indexes of people older than 75 years.

Working hypotheses

During a year, and relative to the control group, intervention emails may result in the following:

• A decrease in mortality.
• An increase in dietary counseling percentage and a decrease in prescribing anti-diabetic and cholesterol-lowering medicines.
• A decrease in medical expenses and in other morbidity indexes.

Type of research and methods of data collection

This randomized controlled trial will be conducted entirely through the existing computer system. The participants (patients) will be assigned to the two Arms/Groups "Intervention Email" and "Control". It has three separate interventions: a. Alerting about a significant drop in BMI. b. Alerting about a low HbA1c% level in patients taking anti-diabetics. c. Alerting about a low cholesterol level in patients taking cholesterol-lowering medicines. The alerts will be sent to the primary clinicians.

Method(s) of data analysis

Differences between intervention groups and control groups will be analyzed using Chi-square test (or Fishers' exact test) for categorical variables and using T-test (or Two-sample Wilcoxon test) for continuous variables.

Uniqueness and relevance

Health service policy regarding signs of malnutrition and excessive medicinal treatment needs a relevant scientific knowledge base. Nutritional counseling and revision of medicinal treatment may dramatically affect health. This research deals with questions that have no commercial interest, but are important to the public.

Inclusion Criteria:

• All found by computerized search in the data base of Clalit Health Services North and South districts:
• 1. A drop in BMI of 2 Kg/m^2 or more during previous two years AND
• BMI less than 23 Kg/m^2 AND
• No dietitian counseling during previous year
• OR
• 2. Last HbA1c% level of 6.5% or less AND
• dispensing anti-diabetic medicines during previous 2 months
• OR
• 3. Last total cholesterol less than 160 mg/dL AND
• dispensing cholesterol-lowering medicines during previous 2 months

Exclusion Criteria:

• Patients whose their primary doctor and nurse email address is unobtainable
• For criterion 3: Patients diagnosed to have had a myocardial infarction, an ischemic heart disease, a transient ischemic attack or an ischemic stroke.

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request to email nir.tsabar@clalit.org.il. Shared data will not include exact dates, exact lab results or any data that might expose the identity of patients. Also, commercial data such as drug brands or costs may not be shared.

After publishing the results in a peer reviewed journal.

Any clinical researcher; Any ethically approved clinical research

Adverse event data were collected for 1 year

Primary physicians and nurses who were recipients of the trial e-mail alert, were asked to report any adverse event to the principal investigator. No such report was received.