Concurrent diabetic ketoacidosis with hyperosmolality and/or severe hyperglycemia in youth with type 2 diabetes

Abstract Introduction Prevalence of diabetic ketoacidosis (DKA) complicated by severe hyperglycaemia and hyperosmolality and its outcomes in youth with type 2 diabetes (T2DM) are not well‐described. Our aim is to determine the frequency and clinical outcomes of isolated DKA, DKA with severe hyperglycaemia (DKA + SHG) and DKA with hyperglycaemia and hyperosmolality (DKA + HH) in youth with T2DM admitted for acute hyperglycaemic crisis. Methods Through retrospective medical record review, patients with T2DM were identified and categorized into isolated DKA, DKA + SHG (DKA + glucose ≥33.3 mmol/L) and DKA + HH (DKA + glucose ≥33.3 mmol/L + osmolality ≥320 mmol/kg). Results Forty‐eight admissions in 43 patients ages 9‐18 were included: 28 (58%) had isolated DKA, six (13%) had DKA + SHG and 14 (29%) had DKA + HH. Subgroups’ demographics and medical history were similar. Seventeen patients (35%) had acute kidney injury (AKI). Odds of AKI were higher in DKA + SHG and DKA + HH relative to isolated DKA (P = .015 and .002 respectively). Frequency of altered mental status (AMS) was similar among groups. Three patients (6%) had concurrent soft‐tissue infections at presentation with no differences among subgroup. Three patients (6%) had other medical complications. These occurred only in patients with AKI and DKA + SHG or AKI and DKA + HH. Conclusions In youth with T2DM, severe hyperglycaemia ± hyperosmolality frequently complicates DKA. Youth with DKA and features of hyperglycaemic hyperosmolar syndrome, including isolated severe hyperglycaemia, have increased odds of AKI.


| INTRODUC TI ON
As prevalence of type 2 diabetes (T2DM) rises among America's youth, 1 paediatric providers will be increasingly responsible for treating acute hyperglycaemic complications of T2DM, including diabetic ketoacidosis (DKA) and hyperglycaemic hyperosmolar syndrome (HHS). In DKA, insulin deficiency is severe enough that in addition to hyperglycaemia, lipolysis begins and ketosis develops. 2 It is defined by the presence of hyperglycaemia with a serum blood glucose of > 11 mmol/L (approximately 200 mg/dL), acidosis with a pH less than 7.3 and/or a bicarbonate of < 15 mmol/L and ketosis. 3,4 In HHS there is enough insulin activity to prevent ketosis, but insufficient insulin activity to prevent hyperglycaemia. 2 HHS is characterized by minimal acidosis with a venous pH > 7.25 and serum bicarbonate > 15 mmol/L, minimal to no ketonuria, severe hyperglycaemia with a serum blood glucose > 33.3 mmol/L (>600 mg/ dL) and hyperosmolality with serum osmolality of > 320 mmol/kg. 4 Previously considered two distinct entities, in clinical practice the conditions overlap significantly. [5][6][7] A mixed presentation of DKA complicated by severe hyperglycaemia and hyperosmolality is common in patients hospitalized with acute hyperglycaemic crisis, occurring in up to 14% of children 8 and 27% of adults. 6 Retrospective cohort studies and case reports report high rates of mortality, altered mental status (AMS), acute kidney injury (AKI), thrombosis, rhabdomyolysis and other medical complications in those with mixed DKA/HHS and isolated HHS when compared to isolated DKA. 5,6,8,9 Persistent hyperglycaemia and osmotic diuresis lead to progressive volume depletion. Pre-renal injury from dehydration causes a decline in glomerular filtration rate, reducing renal glucose clearance, exacerbating hyperglycaemia and contributing to hyperosmolality. 10 A recent large retrospective case series in adults showed a higher rate of mortality in patients with mixed DKA/HHS compared to isolated DKA and isolated HHS. 6 Children's of Alabama, a tertiary care university children's hospital in the southeastern United States where T2DM prevalence is high, 11 is uniquely positioned to review complications and frequency of isolated DKA and DKA complicated by severe hyperglycaemia and/or hyperosmolality in youth with T2DM. We therefore aimed to determine how frequently DKA is complicated by severe hyperglycaemia and/or hyperosmolality in paediatric patients admitted for acute hyperglycaemic complications of T2DM. Our secondary aim was to evaluate the frequency of acute medical complications, including AKI and AMS, in this patient cohort. Finally, we aimed to identify any risk factors associated with hyperosmolality.

| Data source
A retrospective electronic medical record review was conducted. Initial query identified admissions to the intensive care unit (ICU) or stepdown ICU with International Classification of Disease 10 diagnosis code of diabetes (E8.xx, E10.xx, E11.xx or E13.xx) or hyperglycaemia (R73.9) and who were admitted from January 2013 through December 2018. Inclusion criteria were age 0-20 years, admission to the ICU or step-down ICU, and diagnosis of T2DM. Exclusion criteria included steroid-induced diabetes, type 1 diabetes mellitus, and insufficient data to confirm diagnosis of T2DM. Diagnosis of T2DM was confirmed based on documented clinical assessment, negative autoantibodies when measured, and elevated c-peptide if autoantibodies were not assessed. Investigator JS reviewed individual medical records at diagnosis and follow-up to ensure accuracy of T2DM diagnosis. For patients with multiple admissions, each admission was evaluated as a unique event.

| Variables
Cases were identified as belonging to one of three possible diagnostic subgroups: isolated DKA, DKA with severe hyperglycaemia (DKA + SHG) or DKA with severe hyperglycaemia and hyperosmolality (DKA + HH). Isolated DKA was defined by presence of hyperglycaemia with a serum blood glucose of 11 to 33.2 mmol/L (approximately 200 to 599 mg/dL), acidosis with a venous pH of less than 7.3 or a bicarbonate of < 15 mmol/L. 3,4 As serum and urine ketones may not be obtained at presentation, ketone measurement was not an inclusion criterion.
Subjects with DKA + SHG met criteria for DKA, but also had a blood glucose at presentation that was > 33.3 mmol/L (600 mg/dL).
Subjects with DKA + HH were those with DKA complicated by both severe hyperglycaemia and hyperosmolality with a serum osmolality of > 320 mmol/kg at presentation. When serum osmolality was not measured, it was calculated based upon the calculation serum osmolality (mmol/kg) = 12 : Data collected included patient demographics, personal and family medical history, hospital course and laboratory and clinical assessment at presentation. Corrected sodium for hyperglycaemia was calculated based upon a 1.6 mmol/L increase in sodium for every 5.55 mmol/L (100 mg/dL) serum glucose was above 5.55 mmol/L (100 mg/dL). 2 Presence or absence of altered mental status (AMS) was determined by manual chart review of documented physician assessment and neurologic exam.. A positive psychiatric history was defined as any documented psychiatric disorder except for attention deficit hyperactivity disorder. AKI was defined the 2012 clinical practice guideline for AKI as a creatinine 1.5x above baseline. 13 For patients without a known baseline creatinine, the discharge creatinine was used as a proxy. Other medical complications were determined by manual chart review.

| Ethics
This study was exempted by the University of Alabama at Birmingham Internal Review Board.

| Patient characteristics
There were 1027 patients identified after initial query of the medical record, with 48 admissions in subjects age between 9 and 18 meeting inclusion criteria. Figure 1. On average, patients were non-Hispanic, African-American, male, governmentally insured, adolescents with obesity. Developmental delay was uncommon, but a positive
No linear association was found between serum osmolality and pH, bicarbonate, haemoglobin A1c or urine ketones. Figure 2.

| Concurrent infection
Three patients had soft-tissue infections at presentation. Patients were not on insulin pumps, and infections were not associated with insulin injection sites. Two patients with isolated DKA had soft-tissue infections. One had an abscess was on the buttock, the other had a labial pustule; both were complicated by surrounding cellulitis. They were treated surgical drainage and systemic antibiotics. One patient with DKA + SHG had cellulitis of the mons pubis that required oral antibiotics. The frequency of infection was 67% in isolated DKA, 33% in DKA + SHG and 0% in DKA + HH (P = .22).

| Other medical complications
Three patients had medical complications other than soft-tissue infections, AKI or AMS. One patient with DKA + SHG required intubation due to severity of AMS and suffered a bladder haematoma.
Another patient with DKA + SHG had pancreatitis with triglycerides of 1014 mg/dL. One patient with DKA + HH had rhabdomyolysis, parotiditis, a deep vein thrombosis and remained on renal replacement with dialysis at discharge. Rate of complications was 0% in isolated DKA was 0%, 67% in DKA + SHG, and in 33% in DKA + HH (P = .013).

ACK N OWLED G EM ENTS
We would like to thank the patients and their families. We would like to thank the ICU staff for their care of these patients.

CO N FLI C T O F I NTE R E S T
Authors have no conflicts of interest to declare.

AUTH O R CO NTR I B UTI O N
JS conceptualized and designed the study, acquired the data, assisted with data analysis, drafted the initial manuscript and approved the final manuscript. AKMFR performed data analysis and reviewed the final manuscript. AA assisted with study design and critically revised the manuscript. All authors read and approved the final manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data supporting the findings of this study are not publicly available due to ethical considerations and inclusion of protected health information.