<![CDATA[Diabetes mellitus (DM), a disease defined by consistently high blood sugar levels, is recognized as a significant global health burden and holds the seventh position among causes of death in Malaysia. This underscores the need for affordable risk stratification. Complete blood count (CBC), a simple and widely available test, provides useful biomarkers, yet its role among Malaysian diabetic patients in Sabah remains understudied. This research aimed to examine CBC biomarkers for mortality risk stratification among diabetic patients in Sabah. This retrospective cross-sectional study utilized data from 10,672 diabetic patients retrieved from medical records at Queen Elizabeth Hospital 1, Sabah. Logistic regression (LR) was applied to estimate the probability of mortality (Alive = 0, Deceased = 1) from demographic and CBC parameters. Model performance was evaluated using calibration and discrimination metrics. Significant mortality risk factors included advanced age, inpatient status, higher neutrophil counts, platelet-to-lymphocyte ratio (PLR), red cell distribution width (RDW), neutrophil-to-lymphocyte ratio (NLR), and monocyte-to-lymphocyte ratio (MLR), as well as lower lymphocyte levels and hemoglobin. The LR model showed excellent calibration and discrimination, with non-significant Spiegelhalter Z-tests, low Brier scores, and strong performance across balanced accuracy, Matthews correlation coefficient, and F1-score. These findings highlight that CBC biomarkers can be integrated into clinical models as a low-cost approach for early detection and management of diabetes risk in resource-limited settings.]]>
