The prevalence of diabetes in Germany is increasing, with costs exceeding 37 billion euros annually. Patients with prediabetes have a high risk of developing diabetes and its complications. This study aims to assess the disparity between G-DRG reimbursement, actual hospital costs, and clinical outcomes in diabetic, prediabetic, and non-diabetic inpatients at a Level 1 trauma center. Between October and November 2021, patients hospitalized for primary and revision surgery were consecutively included in the study. In addition to the patients’ medical history reports, HbA1c levels were determined to classify the patients as prediabetic, diabetic, or non-diabetic. Clinical and economic data were extracted from the hospital information system. This study examined data from 1186 patients. 36.2% (n = 429) were prediabetic, and 17.3% (n = 205) were diabetic, but about half of those with diabetes were not documented in the hospital system. Complication rates were higher in prediabetic (26.8%) and diabetic (42%) patients compared to non-diabetic patients (24.6%). The median length of hospital stays was longer for prediabetic (+1.0 days) and diabetic patients (+3.0 days). G-DRG reimbursement was inadequate, leading to deficits of € 42389 for prediabetic and € 224174 for diabetic inpatients. Consistent HbA1c screening identified that approximately half of the inpatients at the Level 1 trauma center had prediabetes or diabetes, which highlights the need for routine HbA1c screening. Furthermore, our data show that the treatment costs of prediabetes and diabetes are not adequately covered in the updated G-DRG system. As the number of inpatients with prediabetes and diabetes continues to rise, inadequate reimbursement can further contribute to the financial burden of acute care hospitals.

Undiagnosed diabetes and prediabetes are common among trauma patients and increase perioperative complication rates, yet routine glycemic screening at hospital admission is rarely performed. We compared 10 supervised machine learning (ML) algorithms for early detection of impaired glucose metabolism in 1789 trauma surgery patients (1095 normal; 694 prediabetes/diabetes) using questionnaire data and preoperative laboratory values within a nested cross-validation framework with hyperparameter optimization. On the combined feature set, XGBoost achieved the highest area under the receiver operating characteristic curve (ROC-AUC; 0.815), followed by multilayer perceptron (0.814) and Gradient Boosting (0.813); adding blood parameters significantly improved performance for nine of ten algorithms (DeLong’s test, p < 0.05). SHapley Additive exPlanations (SHAP) and XGBoost feature gains identified six overlapping predictors, yielding a simplified model with comparable mean discrimination. Inter-model differences were small, and a fully interpretable logistic regression model remained within one percentage point of XGBoost. The best-performing model demonstrated strong calibration and a higher net benefit than either universal testing or no testing across the evaluated threshold probabilities. In routine trauma practice, such a model could identify patients who would benefit from confirmatory glycated hemoglobin (HbA1c) testing, enabling earlier detection of dysglycemia, risk-stratified glucose management, and timely referral for lifestyle intervention in patients with prediabetes.