Objective: We examined the 30-day hospital readmission rates and their association with the admission diagnosis and the length of stay (LOS) in patients with diabetes versus those without diabetes mellitus (DM) in an urban teaching hospital.

Methods: In this retrospective study, we compared the 30-day readmission rates in patients with DM (n=16,266) versus those without DM (n=86,428) at an urban teaching hospital between January 1, 2013, and September 30, 2015. In individuals with a secondary diagnosis of DM, we analyzed the relationship between readmission rates and the ten most common Medicare Severity Diagnosis Related Groups (MS-DRGs). Additionally, we examined the relationship between the LOS and readmission rates in patients with diabetes and those without DM.

Results: The 30-day readmission rates adjusted for age and gender were higher in patients with DM compared to those without DM (15.3% vs. 8.4%, respectively, <0.001). The increased risk of readmissions was present both in patients with a primary or a secondary diagnosis of DM. For the secondary diagnosis of DM, statistically significant difference was present for two out of the ten most common DRGs (DRG # 313 [chest pain], and # 392 [esophagitis, gastroenteritis, and miscellaneous digestive disorders], p=0.045 and 0.009, respectively). There was a direct correlation between LOS and readmission rates in both patients with diabetes and those without DM (p<0.001 for both).

Conclusions: The 30-day readmission rates are higher in patients with DM compared to patients without DM. DM is an independent risk factor for hospital readmissions. The readmission rates correlate directly with LOS in both patients with diabetes and those without DM.