INA–Follow up
Researchers in the United States have developed a new artificial intelligence-based tool capable of predicting hypoglycemia (low blood sugar) in hospitalized patients up to 24 hours in advance, giving doctors sufficient time to intervene and prevent complications.
The tool analyzes patient data, including laboratory test results, medications, dietary information, and electronic health records, to identify patients most at risk of developing hypoglycemia the following day.
Hypoglycemia is a serious condition that can lead to complications such as loss of consciousness, seizures, or cardiac arrhythmias, particularly among patients with diabetes, those in intensive care, or individuals preparing for surgery.
The researchers tested the model using data from more than 143,000 hospital admissions across three US hospitals between 2014 and 2025.
The results demonstrated the tool’s ability to accurately predict cases of hypoglycemia, enabling medical teams to take preventive measures before the condition develops.
The researchers expect that the use of this technology will help reduce the number of daily hypoglycemia cases in hospitals and improve patient safety. They also emphasized that the tool is designed to operate using routinely collected hospital data, making its future implementation easier.
