A Better Way to Identify Type 1 Diabetes in Young People Using Medical Records

Type 1 and Type 2 diabetes are different diseases that require different management approaches. But in young people, distinguishing between them using electronic health records (EHR) alone—diagnosis codes and medication lists—can be surprisingly difficult.

Accurate classification is critical not only for individual patients getting the right care, but also for researchers trying to understand how common each type of diabetes is in different populations and how well treatments work.

A team developed and tested an algorithm using health records from over 71,000 adolescents and young adults (ages 10-30) in Northern California between 2007 and 2024.

The algorithm gathers information from multiple sources in the medical record: diagnosis codes, lab results, and medications. It then uses a step-by-step approach that gives special weight to pancreatic autoantibody results—a specific blood test that shows whether the immune system is attacking insulin-producing cells, a hallmark of Type 1 diabetes.

This combination of data sources is more powerful than relying on any single piece of information.

Researchers validated their algorithm by having it classify a random sample of 100 patients, then checking those classifications against medical records reviewed by trained clinicians who didn't know what the algorithm had predicted.

For Type 1 diabetes, the algorithm was 100% sensitive (caught every case) and 98% specific (correctly ruled it out when it wasn't present). For Type 2 diabetes, it was 98% sensitive and 95% specific.

Notably, including autoantibody data prevented the algorithm from mislabeling about 6% of patients—roughly 609 individuals in the full group—who might otherwise have been incorrectly classified as having Type 2 diabetes.

This algorithm offers a validated tool for health systems and researchers to more reliably identify and track Type 1 versus Type 2 diabetes in young people using existing electronic health records.

Better classification can lead to more accurate epidemiology—understanding who has what disease and why—and to research that more reliably reflects real outcomes. For patients, it supports the foundation of getting the right diagnosis and treatment path from the start.

Source: Diabetes, obesity, and cardiometabolic CARE. Evidence type: PubMed indexed literature. Type1Cure is an information and intelligence hub, not a medical advice service. This article summarizes published research and does not provide diagnosis, treatment, or personal medical guidance. Always talk to your own care team before changing anything about your Type 1 diabetes management.