Aim: To identify patients aged 10-30 years with probable hyperglycemic hyperosmolar syndrome (HHS), to describe demographic and clinical profiles, and to attempt to assess risk factors for poor outcomes.
Study design: Retrospective cohort study (medical records review).
Setting: A 944-bed tertiary care teaching and research hospital and a 425-bed affiliated facility.
Patients: 10-30 year-old patients with a primary or secondary discharge diagnosis of HHS or diabetic ketoacidosis (DKA). Patients with a serum glucose >600 mg/dl in the absence of significant ketoacidosis (possible HHS) were profiled. Further stratification based on measured or calculated serum osmolarity >320 mOsm/kg (probable HHS) was undertaken.
Interventions: Patients received treatment for hyperglycemic crises, consisting primarily of fluids, electrolyte replacement and insulin.
Measurements and main results: Of the 629 admissions, 10 with a diagnosis of HHS and 33 with a diagnosis of DKA met the initial study criteria for HHS. 60% were African Americans and 89% were new-onset diabetics. From this group, 20 admissions had serum osmolarity > or =320 mOsm/kg. Fisher's exact test and Pearson coefficients were used to examine associations between risk factor and poor outcomes and correlations between admission data and length of hospital stay, respectively. Serious complications occurred in four patients (including two deaths, 10% mortality) and were limited to those with unreversed shock over the first 24 hours of admission and who received <40 ml/kg of intravenous fluids over the first 6 hours of treatment.
Conclusions: HHS was underdiagnosed in this population and occurred disproportionately in African Americans. Serious complications occurred exclusively in those with unreversed shock and inadequate fluid resuscitation.