One in four children and young adults aren’t diagnosed with T1D until they have potentially fatal diabetic ketoacidosis – when the body doesn’t have enough insulin to allow blood sugar into cells for use as energy.
Thanks to funding from Diabetes UK, a team led by Dr Julia Townson of Cardiff University has developed a predictive tool using artificial intelligence that looked through health data from more than a million children to identify patterns that could help identify T1D much earlier. Patterns such as urinary infections or bedwetting, family history of type 1 diabetes and being prescribed certain antibiotics, were found to be associated with the condition.
On average, the new tool would have allowed children to be diagnosed – and started on life-saving insulin therapy – 11 days earlier which could be the difference between life and death, according to results from studies of the tool presented at Diabetes UK’s 2022 Professional Conference.
Beth Baldwin from Cardiff whose 13 year-old son Peter died from undiagnosed diabetes in 2015, who has campaigned tirelessly for more innovation and T1D awareness, has welcomed the project. ‘Life without Peter is incredibly hard, but in our work stops even one family going through the experience of losing a child to Type 1 diabetes, it will have been worth it,’ she said.
Currently, the best way to spot early Type 1 diabetes is to be aware of its four most common symptoms: frequent weeing or bedwetting in a previously dry child, excessive thirst, or fatigue and unexplained weight loss.
