Risk factors for hospitalization in youth with type 1 diabetes: Development and validation of a multivariable...

Area of Science:

• Pediatrics
• Endocrinology
• Diabetes Research

Background:

• Type 1 diabetes (T1D) management requires proactive identification of patients at risk for severe complications.
• Diabetic ketoacidosis (DKA) and hyperglycemia with ketosis are significant causes of hospitalization in pediatric T1D.
• Predictive models can aid in resource allocation and preventative care strategies.

Purpose of the Study:

• To develop and validate a multivariable prediction model for identifying pediatric patients with T1D at high risk of hospitalization due to DKA or hyperglycemia with ketosis within 12 months.
• To assess the model's ability to discriminate between high-risk and low-risk patient cohorts.

Main Methods:

• Retrospective analysis of clinical data from pediatric T1D patients (<17 years) at a major children's hospital.
• Generalized estimating equations were used to predict 12-month hospitalization risk based on prior 12-month data.
• Data from 2014-2016 served as the training set, with 2017-2019 data used for validation.

Main Results:

• Significant predictors for hospitalization included prior year admissions, elevated hemoglobin A1c (HbA1c), and non-commercial insurance.
• A multivariable model identified approximately 8% of patients at a 5-fold increased risk of hospitalization (42% collective risk) compared to the remaining 93%.
• The model demonstrated consistent predictive performance in the validation dataset.

Conclusions:

• A validated multivariable prediction model effectively identifies pediatric T1D patients at increased risk for DKA or hyperglycemia with ketosis-related hospitalizations.
• This model can facilitate targeted interventions and personalized management plans for high-risk individuals.
• The findings underscore the importance of historical admission data and glycemic control (HbA1c) in risk stratification for T1D complications.