Skip navigation

Intensive insulin treatment and heart attack

Diabetic patients are more likely to die after a heart attack than those without diabetes. A randomised trial of intensive insulin therapy (DIGAMI trial from Sweden) compared with usual treatment demonstrates that death rates can be reduced by 33% in the three and a half years after a heart attack.


Included patients were those with an acute myocardial infarction and a blood glucose of more than 11 mmol/L. In the treatment group they were given an insulin-glucose infusion for at least 24 hours (according to a defined protocol), followed by subcutaneous insulin for at least three months.


In the first year after their heart attack, 58/306 treated patients died, compared with 82/314 controls. Cause of death was cardiovascular in 97% of cases in this first year. Over a mean follow up of 3.4 years there were 138 deaths (44%) in the control group and 102 (33%) in the treated group. The 95% confidence interval around the mean reduction in death rate of 28% was 5% to 45%.

The main results (as NNTs) are in the Table. The one-year NNT was 14 and the 3.4 year NNT was 9. This means that nine patients with heart attack and blood glucose >11 mmol/L have to be treated with insulin therapy for one of them to be alive 3.4 years later, who would not have been if they had been given usual care.

Main results of DIGAMI study: effect of intensive insulin treatment on mortality
Patient group Mortality at year NNT (95% CI)
All patients 1 14 (7.3 to 164)
All patients 3.4 9.4 (5.5 to 33)
Low risk, no insulin 3.4 7.2 (4.1 to 27)
There was a pre-defined low-risk group (younger, no history of heart disease) with no previous insulin treatment, to which 44% of the enrolled patients belonged. For them, the NNT was 7.


Here is an excellent study from a good stable (Karolinska) and with a background of biological plausibility. The trial isn't enormous, and the confidence interval around the outcome is wider than we may like. But an NNT of 7 is a worthwhile figure, because it refers to mortality rather than some minor outcome, and because the intervention - getting control of hyperglycaemia - is one we might want to do anyway.

There is another nuance to the trial. Its title, DIGAMI, is, as Spanish speakers will recognise, close to "Diga me", how many people in Spain and South America answer the `phone. It means "talk to me" - what every clinical trial should be about.


  1. K Malmberg. Prospective randomised study of intensive insulin treatment on long term survival after acute myocardial infarction in patients with diabetes mellitus. British Medical Journal 1997 314: 1512-5.

previous or next story in this issue