Mortality after discharge from hospital following an episode of diabetic ketoacidosis.

Abstract

<h4>Aims</h4>The rate of inpatient mortality associated with diabetic ketoacidosis (DKA) has steadily decreased in recent decades. However, there remains a significantly increased outpatient death rate following an episode of survived DKA. We undertook this study to investigate the observed increase in mortality following an episode of DKA.<h4>Methods</h4>We completed a retrospective cohort study to investigate rates and causes of death in people admitted to our hospital with DKA between 2013 and 2018. DKA was confirmed by pre-defined biochemical parameters and cause of death data was extracted from multiple sources. Follow-up was for two years after discharge for all participants with one-year mortality being the main time point for analysis.<h4>Results</h4>We identified 818 admissions to hospital with DKA, affecting 284 people. Twenty people died as inpatients and a further 40 people died during the two-year follow-up. Of these 60 participants, cause of death was able to be determined for 41 (68%), with most deaths occurring due to infection or macrovascular disease. Risk factors for death within a year of hospital discharge included older age, vascular complications of diabetes, intellectual impairment and residential care living. Those who survived an episode of DKA had a one-year age-corrected mortality rate 13 times higher than the general population. This was more marked in the younger cohort with those aged 15-39 years being 49 times more likely to die in the year after surviving a DKA admission compared to their general population counterparts.<h4>Conclusion</h4>An episode of diabetic ketoacidosis is associated with a significant outpatient mortality risk with most deaths due to infectious or macrovascular causes. This study should prompt investigation of predictive scoring tools to identify those at increased mortality risk after DKA and encourage the development of targeted interventions to reduce mortality.