Predictors of hyperglycaemic crises and their associated mortality in Jamaica

Abstract

The objective of the study was to determine the clinical characteristics and mortality of patients with hyperglycaemic hyperosmolar syndrome (HHS) and diabetic ketoacidosis (DKA) at a Jamaican tertiary care hospital. In a retrospective study of 1560 admissions for diabetes during the period 1998–2002, 980 dockets were reviewed and 164 individuals met the ADA diagnostic criteria for DKA or HHS.

Patients with HHS were older than DKA patients (64.5 years [95% CI: 60.7–68.4] versus 35.9 years [95% CI: 30.2–41.6]), but were not more likely to be non-compliant with medications, infected, or male. Overall, 24% had a mixed DKA/HHS syndrome. Most DKA patients had type 2 diabetes (62%). Only 2% of HHS and 6% of DKA/HHS patients had type 1 diabetes.

Syndrome specific mortality was: DKA 6.7%, HHS 20.3%, and DKA/HHS 25% (p for trend = 0.013). Mortality increased significantly with age, especially in patients ≥50 years. Significant univariate predictors of mortality were altered mental status on admission, co-existing medical disease, increasing age, older age at onset of diabetes, acute stressors, and DKA/HHS. In multivariate models, only altered mental status was significant (OR = 3.59; 95% CI: 1.24–10.41). Hence, hyperglycaemic crises in a Jamaican tertiary care hospital are associated with significant mortality especially in patients who are older or with altered mental status.

Introduction

Chronic diseases are the most common cause of death in Western countries and approximately 1 million deaths in 2002 were attributable to diabetes mellitus [1]. Diabetes affects 171 million people worldwide and this number is expected to rise to 366 million by the year 2030; 298 million of whom will live in developing countries [2]. Diabetes mellitus afflicts about one in six urban Jamaicans, occurring in up to 16% of men and 24% of women [3], [4] and is the fifth leading cause of death.

Hyperglycaemic crises (diabetic ketoacidosis and hyperglycaemic hyperosmolar syndrome) are among the deadliest complications of diabetes and are a common cause of hospital admissions. Diabetic ketoacidosis (DKA) occurred in 3–9% of diabetic hospitalizations in the United States [5], [6] with a prevalence of 4.6–8 episodes per 1000 diabetic individuals [6], [7]. The rates of hospital admissions for hyperglycaemic hyperosmolar syndrome (HHS) are lower and are estimated to be <1% of all diabetes hospitalizations [5], [8] with a rate of 17.5 episodes per 100,000 person-years [9].

Previously, DKA and HHS were thought to define separate clinical entities; however, these syndromes could be considered as the extremes of the same clinical spectrum [10]. Up to one-third of patients with hyperglycaemic crises have a mixed state of acidosis and hyperosmolarity [11], [12]. Hyperglycaemic crises are not only a significant cause of morbidity, but also have high mortality: 1–9% for DKA and 5–45% for HHS [5], [13], [14], [15], [16] especially in the older hyperosmolar patient [10], [17]. The mortality for the mixed entity can be intermediate of DKA and HHS [10], but can be higher [13]. Infection is the most common precipitant for hyperglycaemic crises [2], [18], [19], [20] although non-compliance with insulin is a major factor in urban African-Americans [8].

To our knowledge, there is little data on the epidemiology of hyperglycaemic crises in the Caribbean or Central America. The objective of this study was to describe the clinical characteristics and mortality of patients with hyperglycaemic crises who presented to a Jamaican tertiary care hospital. The study also investigated the factors that predicted the development of either type of hyperglycaemic crisis in teenagers and adults.