Recurrent ketoacidosis is still associated with a high risk of death

Recurrent diabetic ketoacidosis (DKA) is still associated with an unacceptably high risk of death despite modern treatment. The rate of death associated with admission to hospital with DKA has historically been high. For clinicians managing this condition in hospital, such findings may perhaps be surprising as, intuitively, the mortality rate associated with DKA during the acute hospital stay feels low. This is not a reflection of particularly high standards of care during the hospital stay, however, as a survey by Dhatariya et al (2016) suggests. Seventy-two hospitals responded to the survey, and low potassium levels were reported in 55% of inpatients. Their study relates to the hour-by-hour management of the acute condition, and they conclude that this area of care could be significantly improved.

The article by Gibb and colleagues (summarised alongside) would perhaps suggest that this is missing the point. The authors conclude that recurrent DKA is still associated with a high mortality rate and that the deaths occur not in hospital but at home. We can predict who is at risk, as recurrent DKA tended to affect young, socially disadvantaged adults with very high HbA_1c levels. This in itself is not a novel finding; this is the group of individuals who would previously have been labelled as having brittle diabetes. Our understanding of this condition is that it does not represent an organic disease of insulin action but rather is the result of psychosocial problems (Gill et al, 1996).

A number of approaches have been described to try and address this issue, and many diabetes centres have programmes in place to try and help this at-risk cohort of individuals. The importance of this paper is that it shows us they are still not working. We can and should continue to improve the day-to-day management of DKA in hospital, but to reduce deaths we need better strategies to engage this troubled group of patients. Hospital-based diabetes services have traditionally focused on the technical aspects of insulin delivery. These data would suggest that this approach is not appropriate. The authors’ conclusion is that a combined medical, psychological and social approach is required.

To read the article summaries, please download the PDF