A newly developed hypoglycemia risk stratification tool may assist clinicians in identifying patients’ 12-month risk for hypoglycemia-related emergency department (ED) visits or hospital utilization in type 2 diabetes (T2D), according to a study published by JAMA Internal Medicine.
A sample of 206,435 patients from Kaiser Permanente Northern California were used to develop and test the classification tree internally; external testing was conducted in more than 1.3 million patients from the Veterans Health Administration and 14,972 patients from Group Health Cooperative.
Participants were classified as either high, medium, or low risk based on 6 patient-specific variables: total number of prior episodes of hypoglycemia-related ED or hospital utilization, number of ED encounters for any reason in the past 12 months, insulin use, sulfonylurea use, presence of severe or end-stage kidney disease, and age younger than 77 years.
Patients identified as high risk based on the risk assessment tool were 5 times more likely to report a severe hypoglycemic episode vs patients categorized as low risk (49.7% vs 9.2%; P <.001). Patients in the internal validity groups categorized as high (2%), medium (10.7%) and low (87.3%) risk had observed 12-month utilization rates of 6.7%, 1.4%, and 0.2%, respectively. During internal validation testing at Kaiser Permanente, the investigators noted good discrimination (C statistic, 0.83); similarly, good discrimination was also identified in both external validation groups (Veterans Affairs C statistic, 0.81; Group Health Cooperative C statistic, 0.79)
The investigators concluded that this new hypoglycemia risk stratification tool may assist clinicians in the early identification and management of patients with T2D at high risk for either ED visits or hospitalizations caused by a severe hypoglycemia event. Early identification and intervention in patients with T2D may reduce severe hypoglycemia events and improve patient safety and overall quality of life.
Reference
Karter A, Warton EM, Lipska KJ, et al. Development and validation of a tool to identify patients with type 2 diabetes at high risk of hypoglycemia-related emergency department or hospital use [published online August 21, 2017]. JAMA Intern Med. doi:10.1001/jamainternmed.2017.3844