An extensive ICU database analysis has found that elevated Stress Hyperglycemia Ratio (SHR) is strongly associated with higher mortality in patients hospitalized with hyperlipidemia. SHR is a measure of stress-related blood sugar elevation. 
The findings suggest SHR could be a valuable tool for improving risk prediction in this population. The analysis was published in Diabetes Research and Clinical Practice.
The study analyzed data from 4,883 first-admission hyperlipidemia patients in the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. The mean age of the participants was 70 years. Patients were divided into quartiles based on SHR levels, and mortality outcomes were assessed for ICU stay, hospital stay, and 28-day follow-up.
Higher SHR was independently linked to worse outcomes. The ICU mortality had an HR of 1.29, the in-hospital mortality had an HR of 1.24, and the 28-day mortality had an HR of 1.32. The study found a linear relationship between SHR and mortality risk. 
There was a significant improvement in the accuracy when SHR was added to standard predictive models. LightGBM indicated SHR as a key predictor.