LY Lange, Center for Health Care Evaluation, VA Palo Alto Health Care System; JD Piette, Center for Health Care Evaluation, VA Palo Alto Health Care System; MN Lee, Center for Health Care Evaluation, VA Palo Alto Health Care System; SR Tobin, Center for Health Care Evaluation, VA Palo Alto Health Care System; LJ Trotter, Center for Health Care Evaluation, VA Palo Alto Health Care System

Objectives: In a prior study, we compared VA and non-VA diabetes patients' perceived access to care. Although VA patients reported fewer financial and non-financial barriers than county-clinic patients, a large number of patients in both groups still reported access problems (23% vs. 37%, respectively). In this study, we investigated the potential consequences of these perceived access problems. Specifically, we: 1) determined whether a link exists between perceived access barriers and health outcomes, 2) identified specific access barriers that are especially predictive of poor outcomes, and 3) determined whether access-health relationships observed among VA patients were unique to VA or consistent with relationships observed among county-clinic patients.

Methods: Data were from two controlled trials of bi-weekly automated telephone assessments with telephone nurse follow-up. A total of 570 diabetes patients were enrolled from general medicine clinics and randomized, with 520 patients (272 VA; 248 county) patients providing 12-month follow-up data. As part of their 12-month follow-up, patients were asked to report whether they experienced each of six financial and non-financial access barriers during the prior six months. In each sample, the relationships between access problems and 12-month outcomes (e.g., symptoms, satisfaction, HbA1c) were examined controlling for patients' intervention group and baseline values for the outcome measure.

Results: VA patients who reported one or more access problems reported experiencing on average more symptoms at follow-up (5.0 vs. 3.8, p<.004), even when controlling for baseline symptoms and intervention group. Those with access problems also reported less satisfaction with multiple dimensions of care. Non-financial barriers had a greater impact on VA patients' health outcomes than financial barriers and were related to greater follow-up symptom reports, lower perceived general health, greater diabetes health worries, and less satisfaction with their choice of doctors and access to care. Non-financial barriers to urgent care and telephone advice were especially predictive of more symptoms and lower perceived health (ps<.05). Not knowing how to get medication was associated with lower perceived general health and reduced overall satisfaction with care (p<.03 and p<.05, respectively). Neither financial nor non-financial access problems were associated with higher endpoint HbA1c values among VA patients. In general, all of these relationships were corroborated in analyses of county clinic patients. However, county patients with one or more non-financial access problems also had worse follow-up HbA1c levels (8.6% vs. 8.1%, p=.03).

Conclusions: Access problems among VA patients lead to significant health problems and worse satisfaction with care. Non-financial barriers to urgent care, medicine obtainment, and medical advice over the telephone are especially predictive of poor outcomes.

Impact: Improving access to health care has the potential to significantly improve health among diabetes patients treated in VA as well as county clinics. Although VA patients report experiencing few financial barriers to care, the presence of non-financial barriers appear to inhibit their seeking medical care and advice, especially for urgent health problems. Attention should be given to reducing these barriers in the VA through greater access to information and medical advice over the telephone.