Objective:To investigate the incidence and risk factors of coma in patients with hypoglycemia (≤3.9 mmol/L).Methods:A retrospective study was conducted. Patients aged 20 years and older with blood glucose levels ≤3.9 mmol/L, and measured by emergency physicians from January 2020 to December 2022 were collected. Baseline patient data, clinical values collected on-site, and treatment outcomes were analyzed. The Glasgow Coma Scale (GCS) was used to determine if patients were comatose, with GCS ≤8 classified as the coma group and GCS >8 as the non-coma group. Further analysis was conducted on the resuscitated coma group to identify factors affecting patient recovery. Patients were divided into eight age groups, seven time periods within 24 h, and six blood glucose level groups to calculate the incidence of coma. A multivariate logistic regression model was constructed to analyze independent risk factors for coma in hypoglycemic patients.Results:A total of 754 patients with blood glucose levels ≤3.9 mmol/L were collected, with 425 cases of coma and 329 non-coma cases, resulting in a coma probability of 56.37% (95% CI: 52.82%-59.91%). Patients in the coma group were older ( P<0.001) and had a higher prevalence of diabetes compared to the non-coma group (82.12% vs. 67.78%, P<0.001). The age of all patients was (73.05±15.20) years, with the 61-90 years age groups being the most prone to hypoglycemia and coma. In terms of time distribution, the high-incidence periods for hypoglycemia and coma were 0-6 o’clock, 6-9 o’clock, and 14-18 o’clock. The primary causes of hypoglycemia included reduced energy intake after insulin injection (12.07%), improper use of insulin (6.37%), and reduced energy intake (6.23%), with 71.09% of cases having unknown causes. Additionally, 18.44% of patients used insulin before the onset of hypoglycemia, with a higher proportion in the coma group compared to the non-coma group (22.12% vs. 13.68%, P=0.003). The initial blood glucose level of all patients was (2.13±0.85) mmol/L, with lower levels observed in the coma group compared to the non-coma group ( P<0.001). The probabilities of coma occurrence corresponding to blood glucose levels were: 1.1-1.5 mmol/L (72.97%), 1.6-2.0 mmol/L (68.90%), 2.1-2.5 mmol/L (54.10%), 2.6-3.0 mmol/L (38.20%), 3.1-3.5 mmol/L (37.50%), and 3.6-3.9 mmol/L (19.40%). Multivariate logistic regression analysis indicated that age ( OR=1.021, 95% CI: 1.010-1.033, P<0.001), insulin use before onset ( OR=1.948, 95% CI: 1.142-3.323, P=0.014), and blood glucose concentration ( OR=0.426, 95% CI: 0.347-0.522, P<0.001) were independent predictors of coma in hypoglycemic patients. The investigation revealed that after intravenous injection of 50% glucose solution, 215 of 425 coma patients regained consciousness (50.58%), and the recovery time was (18.43±9.09) min. Patients in the recovery group were younger and had lower initial blood glucose levels compared to the non-recovery group (both P<0.05), while recovery group re-measured blood glucose levels were higher than those in the non-recovery group ( P=0.002). Conclusions:The probability of coma in hypoglycemic patients was high, with insulin use being a common trigger. Proper use of insulin is essential to prevent hypoglycemia and coma.