Prevalence and outcomes of patients with diabetes mellitus in Haematology and Oncology Units: 18 months of the Austin Health Diabetes Discovery Initiative     — ASN Events

Prevalence and outcomes of patients with diabetes mellitus in Haematology and Oncology Units: 18 months of the Austin Health Diabetes Discovery Initiative     (#350)

Wei-Ling Chiu 1 , Peter Lee 2 , Alvin Kong 3 , Peter Davey 4 , Raymond Robbins 4 , Harvey Sutcliffe 5 , Que Lam 6 , Natalie Nanayakkara 1 , Leonid Churilov 7 , Andrew Lee 8 , Adrian Lio 8 , Jeffrey David Zajac 1 9 , Elif Ilhan Ekinci 1 9
  1. Department of Endocrinology, Austin Health, Heidelberg, Victoria
  2. Austin Health, Heidelberg, Victoria
  3. University of Melbourne, Parkville, Victoria
  4. Clinical Information Analysis & Reporting, Austin Health, Heidelberg, Victoria
  5. Pathology IT Service, Austin Pathology, Heidelberg, Victoria
  6. Department of Pathology, Austin Health, Heidelberg, Victoria
  7. Statistics and Decision Analysis Academic Platform, The Florey Institute of Neuroscience & Mental Health, Heidelberg, Victoria
  8. Clinical Informatics Unit, Austin Health, Heidelberg, Victoria
  9. University of Melbourne (Austin Campus), Parkville, Victoria

Background:Use of HbA1c measurement to diagnose diabetes in hospital has potential benefits, as there is no need for fasting and HbA1c is less likely to be affected by stress hyperglycemia. Glucocorticoids are frequently used in the management of haematological and oncological conditions but can exacerbate hyperglycemia in patients with preexisting diabetes and cause new-onset diabetes.

Aim:To determine the prevalence of diabetes mellitus and to assess hospital outcomes of patients with diabetes in Haematology and Oncology Units.

Methods:Using the Cerner Millennium® Health IT Platform, routine HbA1c testing was performed in all patients aged ≥54 years admitted to Austin Health as part of the Diabetes Discovery Initiative between June 2013 and January 2015. Diabetes was diagnosed in patients with an HbA1c≥6.5%. Clinical and demographic characteristics as well as glucocorticoid use were recorded.

Results:Over the 18-month study period, 1076 inpatients above aged 54 years were admitted to Haematology (n=298) and Oncology (n=778) units. 29% (315/1076) of these patients had diabetes and of these 315 patients, 73% (n=231) had previously known diabetes while 27% (n=84) had unrecognised diabetes. Patients with diabetes (known and unrecognised) were significantly older, with lower estimated-glomerular-filtration-rate (eGFR) and lower haemoglobin level (Hb), compared to those with no diabetes (Table 1). After adjusting for specialty unit and comorbidity score, patients with known diabetes had significantly longer length of stay (LOS), compared to those with no diabetes (IRR=1.18; 95%CI: 1.01-1.36; p=0.032), and no statistically significant difference was observed between those with unrecognised diabetes and with no diabetes. After adjusting for comorbidity score, age, Hb and eGFR, patients with known diabetes had significantly higher readmission rates compared to those with no diabetes (167/231 (72%) vs 500/761 (66%); IRR=1.4; 95%CI: 1.0-1.9; p=0.047).  No significant differences in mortality rates during the study period between patients with or without diabetes were observed.

Conclusions:Approximately a third of Haematology and Oncology inpatients had diabetes, and about a third of these patients had unrecognised diabetes. Patients with known diabetes had longer length of stay and higher readmission rates compared to those with no diabetes. Further work is currently underway to assess the rates of glucocorticoid exposure in this cohort. 

2206-Table%201.png

  1. d'Emden MC, Shaw JE, Colman PG, et al. The role of HbA1c in the diagnosis of diabetes mellitus in Australia. Med J Aust. 2012;197(4):220-1.
  2. Clore JN, Thurby-Hay L. Glucocorticoid-induced hyperglycemia. Endocr Pract. 2009;15(5):469-74.