Effects of diabetes self-management programs on time-to-hospitalization among patients with type 2 diabetes: A survival analysis model

Objective: This study compares time-to-hospitalization among patients that participated in a NIH funded study, employing a randomized controlled trial (RCT) of type 2 diabetes (T2DM) patients. We seek to determine whether diabetes delf-management interventions prolonged the time-to-hospitalization within any of the RCT groups, after controlling for relevant demographic and clinical variables.

Methods: Data were obtained from electronic medical records (EMR) and survey responses of 376 adults aged 18 or older who consented to participate in a randomized trial of T2DM self-management programs in Central Texas. All study participants had HbA1c &ge7.5% and were randomized into one of four study arms: personal digital assistant diabetes software (PDA), Chronic Disease Self-Management Program(CDSMP), combined PDA and CDSMP (COM), and controls. Consistent with survival analysis techniques, we measure time-to-hospitalization (survival time) as the interval between study enrollment and the occurrence of a hospitalization event relating to diabetes. For the purposes of our analyses, we define hospitalization as any acute hospital event relating to diabetes. If a subject did not experience any diabetes-related hospitalization, the subject was considered censored at the end of the 2-year study. We plot Kaplan-Meier survival curves stratified by the RCT study arms, gender, race, and identified comorbidities. Multivariate analyses employing a Cox proportional hazards model are used to model the data while controlling for baseline independent variables.

Results: Compared to subjects in the control arm, subjects in the CDSMP only arm had significant longer time-to-hospitalization(Hazard ratio:0.10; p=0.002). Subjects in the PDA and CDSMP combined arm showed no improvements in comparison to the control arm. Increasing age and higher HbA1c values were significantly associated with shorter time-to-hospitalization at the 0.05 significance level (hazard ratios: 1.03 and 1.21 respectively; p-values: 0.01 and 0.006 respectively). Greater educational attainment was associated with longer times to hospitalization (Hazard ratio: 0.6; p= 0.008). Compared to non-Hispanic whites, Hispanics were associated with significantly longer time-to-hospitalization (Hazard ratio: 0.5; p=0.02) while non-Hispanic blacks were associated with shorter time-to- hospitalization, though insignificant (Hazard ratio: 1.1; p=0.64). As the number of comorbidities increased, the time to hospitalization reduced significantly (hazard ratio: 1.8; p=0.003). There were no statistically significant differences in outcomes by gender.

Conclusions: The CDSMP diabetes self-management program was effective in prolonging time-to-hospitalization among patients with T2DM. Persons who are young, have lower HbA1c values, have more than a high school education and have fewer comorbidities, are more likely to experience longer time-to-hospitalization following enrolment in diabetes self-management programs.