12397 Predictors Of Hospital Readmission Within Three Months Among Patients Admitted For Diabetic Ketoacidosis (dka)

Abstract

Abstract Disclosure: O. Onwudiwe: None. L. Yerashevich: None. C. Onwudiwe: None. B. Baral: None. A. Acharya: None. H. Nguyen: None. Purpose: Diabetic ketoacidosis (DKA) represents a critical, life-threatening complication of diabetes, with higher rates of re-admission noted among individuals from lower socioeconomic backgrounds, those lacking insurance coverage, and those with suboptimal follow-up care. This study seeks to identify predictors of hospital readmission within three months among patients admitted for DKA in a community hospital setting. Methods: Patient data were extracted from electronic medical records from 2019 to 2021 for individuals admitted due to DKA using ICD codes. DKA diagnosis was based on the presence of hyperglycemia, high anion gap metabolic acidosis, and ketonemia. Demographic information, results of routine biochemical tests, insurance status, follow-up with primary care providers (PCP) or endocrinologists, and clinical progression were retrospectively collected from patients' records. Multivariable logistic regression analyses were conducted to identify predictors of readmission. Results: A total of 222 unique DKA admission visits were analyzed. The mean age (standard deviation) of the patients was 48 (16) years, and 58% of them were male. Among these visits, 39% had type 1 diabetes, 43% had type 2 diabetes, and 18% were newly diagnosed diabetes cases. The majority of patients were overweight, with a mean body mass index (BMI) of 27, and the mean glycated hemoglobin level was 10.8. Common triggers for DKA included insulin non-adherence (52%), infection (37.8%), and alcohol use (8.6%). Logistic regression analysis revealed that lack of follow-up significantly increased the odds of readmission within three months (B = 2.025, p = .003, Exp(B) = 7.576, 95% CI [1.952, 29.401]), as did having type 1 diabetes (B = 1.473, p = .019, Exp(B) = 4.364, 95% CI [1.272, 14.978]). However, the association between insulin adherence and readmission was not statistically significant at the conventional alpha level of 0.05 (B = 1.329, p = .112, Exp(B) = 3.777, 95% CI [0.733, 19.458]). Other demographic (age, gender, employment status, insurance status, medication compliance,) and clinical [insulin adherence, presence of infection, blood glucose level at diagnosis of DKA, and level of glycated hemoglobin (Hba1c)] variables did not significantly predict readmission within 3 months. Conclusion: DKA, being preventable with adequate insulin adherence, is also influenced by factors such as infection and lifestyle choices. The findings highlight the critical role of post-discharge follow-up care in reducing readmission rates among DKA patients. Healthcare providers should prioritize ensuring timely and comprehensive follow-up, particularly for patients with type 1 diabetes, to optimize post-hospitalization management and prevent avoidable readmissions. Presentation: 6/3/2024