Search

We evaluated the effectiveness of an intervention for patient-family dyads aimed at improving diabetes self-management and outcomes for adults with T2DM (AWD) at risk for complications. AWD were randomized with a family supporter to an intervention vs. usual care for 12 months. The intervention provided dyads: one health coaching session, biweekly automated phone calls to prompt action plans on new diabetes health concerns, coaching calls to prepare for medical visits, and medical visit summaries. 239 AWD-supporter dyads were enrolled at two Veterans Health Administration sites. AWD eligibility criteria included baseline HbA1c>8% or systolic blood pressure (SBP)>150mmHg. Intention-to-treat analyses of changes in outcomes were adjusted for baseline outcome levels, patient-supporter cohabitation, and patients’ baseline insulin use. 96% of dyads had complete 12-month outcome data. Enrolled patients were on average 60±9 years old and had HbA1c levels of 8.5±1.6% and SBP 140mmHg (IQR 128,150). 70% lived with their supporter. Over 12 months, Patient Activation Measure scores (pre-specified primary outcome) increased significantly more in the intervention group (intervention effect +2.6 points (95%CI 0.005, 5.149). There were no statistically significant differences by arm in diabetes-specific cardiac risk score, HbA1c, or SBP change. Patient self-reported adherence to healthy eating increased significantly more in the intervention group (intervention effect +0.7 days/week (95% CI 0.2, 1.2) but not for other self-management behaviors. 97% of patients and 94% of supporters in the intervention group agreed they would recommend the program to others. This family supporter-patient dyadic intervention successfully increased patient activation and one health behavior but not physiologic outcomes over a 12 month period. Improving sustained engagement of family supporters in care of diabetes for adults is feasible and could serve as an additional source of support for diabetes care. Disclosure A. Rosland: None. J. Piette: None. R.B. Trivedi: None. A.A. Lee: None. S.C. Stoll: None. M.K. Mor: None. D. Obrosky: None. E.A. Kerr: Advisory Panel; Self; Bind Insurance. M. Heisler: None. Funding Health Services Research and Development (IIR 14–074-1); Michigan Center for Diabetes Translational Research (5P60-DK09292); Michigan Claude D. Pepper Older Americans Independence Center (AG-024824); VA Ann Arbor Center for Clinical Management Research

Many adults with DM have a family supporter (FS) involved in their care. Informal support has been linked with better DM management and outcomes. We evaluated the effect of a patient and FS engagement intervention on FS roles in DM care, and the impact of those roles on HbA1c. Two-hundred and thirty-nine adult Veterans with type 2 DM and a FS were randomized to intervention vs. usual care for 12 months. Intervention dyads received one coaching session, biweekly automated calls to prompt action planning on new DM concerns, coaching calls to prepare for clinic visits, and summaries of primary care visits and DM complication risk. All models were controlled for baseline HbA1c, age, sex, insulin use, and FS cohabitation with patient. Over 12 months, participants in the intervention group were more likely to have increased involvement in several FS roles, including: deciding when to contact healthcare providers about concerns (AOR 2.080, 95% CI 1.137,3.804), remembering to go to appointments (AOR 2.852, 95% CI 1.503, 5.412), take (1.853, 95% CI 1.001, 3.428) and refill (AOR 2.193, 1.196, 4.019) medications, and perform (AOR 2.615, 95% CI 1.405, 4.867) and review (AOR 2.822, 95% CI 1.506, 5.290) home testing. Changes in help navigating the healthcare system, and encouraging healthy eating and exercise were not different by intervention arm. Help remembering to take medications at baseline was associated with 0.457% increase in HbA1c over 12 months (95% CI 0.107, 0.807, p=0.011), and no support roles were significantly associated with HbA1c decrease. In conclusion, this intervention increased FS involvement in diabetes-specific self-management, with less impact on general lifestyle support. Patients with support for taking medication prior to the intervention had significant subsequent increases in HbA1c. This may indicate family are recognizing which patients have challenges with medication taking, but need more training to effectively help improve subsequent patient outcomes. Disclosure M. Zupa: None. M. Heisler: None. J. Piette: None. R.B. Trivedi: None. S.C. Stoll: None. D. Obrosky: None. E.Z. Lovelace: None. M.K. Mor: None. A. Rosland: None. Funding VA Health Services Research and Development (IIR 14–074-1); Michigan Center for Diabetes Translational Research (5P60-DK09292); Michigan Claude D. Pepper Older Americans Independence Center (AG-024824); VA Ann Arbor Center for Clinical Management Research; University of Pittsburgh (5T32DK007052-45)

Many adults with diabetes have family or friend ‘supporters’ involved in their care. Adult-supporter dyads were recruited to a trial comparing usual care to a 12-month program to engage and train supporters. Potentially eligible patients with T2DM and either HbA1c>8% or high systolic BP were identified via medical records. Patients randomized to the program received biweekly automated calls prompting reports on diabetes status, with tailored feedback to patients and supporters. Before primary care appointments, health coaches conducted a ‘visit prep call’ to improve dyad engagement in the visit. We examined patient characteristics predicting enrollment, and patient, supporter, and call characteristics predicting visit prep and automated call completion via separate multi-level regression models. Of 1119 contacted, 239 dyads (21%) enrolled. Patient predictors of enrollment included younger age (M=60y enrolled vs. 62 not, p In conclusion, high BP predicted enrollment and engagement in a dyadic-based diabetes program more than high HbA1c. Programs emphasizing the importance of BP in preventing diabetes complications may increase participant engagement. Disclosure S.C. Stoll: None. J. Piette: None. R.B. Trivedi: None. A.A. Lee: None. D. Obrosky: None. M.K. Mor: Stock/Shareholder; Spouse/Partner; AbbVie Inc. M. Heisler: None. A. Rosland: None. Funding U.S. Department of Veterans Affairs (IIR 14–074-1); Michigan Center for Diabetes Translational Research (NIDDK5P60-DK09292); Michigan Claude D. Pepper Older Americans Independence Center (NIA AG-024824)

Background: It is unknown whether family members or friends (‘supporters’) tailor their assistance to the unique needs of patients with diabetes and low HL. Methods: Surveys were conducted among 239 VA healthcare system patients with type 2 diabetes who were 30-70 years old, had A1c>8% or high blood pressure, did not have dementia or need help with basic activities of daily living, and had an unpaid supporter who was involved in their care. Patients reported the frequency of supporter involvement with diabetes management roles (e.g., help filling medications). HL was assessed using two items from the well-validated Chew scale: 1) How often do you have problems learning about medical conditions because of difficulty understanding written information (0=none of the time, 4=all of the time), and 2) How confident are you filling out medical forms by yourself (0=extremely, 4=not at all). Those responding ‘3’ (somewhat) or greater were designated as low health literacy. Logistic regression models were used to identify the independent association between patient HL and types of supporter assistance. Findings: Patients with low HL were significantly more likely than patients with adequate HL to have a supporter with low HL (32% vs. 18% of supporters, p=0.02). Low patient HL was significantly associated with higher odds of supporters helping with: deciding when to call doctor (AOR=2.21, 95% CI: 1.10-4.42); remembering medical appointments (AOR=2.48, 95% CI: 1.24-4.98); navigating the health system (AOR=2.67, 95% CI: 1.34-5.32); and attending medical appointments with patient (AOR=2.11, 95% CI: 1.11-3.96). Patient HL was not associated with whether supporters helped with filling or taking medications, or managing diet and exercise. Conclusion: Supporters of patients with low HL were more likely to have low HL themselves, and to help patients with care coordination and equally likely to help with carrying out diabetes medication and lifestyle behaviors compared to patients with adequate HL. Disclosure B. Fields: None. A.A. Lee: None. J. Piette: None. R.B. Trivedi: None. D. Obrosky: None. M.K. Mor: Stock/Shareholder; Spouse/Partner; AbbVie Inc. M. Heisler: None. A. Rosland: None. Funding U.S. Department of Veterans Affairs (IIR 14–074-1); Michigan Center for Diabetes Translational Research (NIDDK5P60-DK09292); Michigan Claude D. Pepper Older Americans Independence Center (NIA AG-024824)