Our findings suggest that telemedicine is a safe method of delivering type 1 diabetes care to rural patients. Telehealth patients in our study experienced improvements overall in diabetes outcomes, although our findings were not significant. Patients also had an increased mean frequency of hypoglycemia. Our observation of increased hypoglycemic episodes is consistent with literature that suggests improved glycemic control, indicated by lower hemoglobin A1c levels, is correlated with an increased frequency of hypoglycemia.
Our findings are in line with those of other studies that suggest that diabetes care via telemedicine is comparable to in-person diabetes care. For example, in a recent randomized controlled trial of 282 diabetes patients, those who received telemedicine consultation had a −1.01% decrease in hemoglobin A1c compared with a −0.68% decrease in hemoglobin A1c in those receiving in-person consultation, although the change was nonsignificant. Our findings, which demonstrated a 0.6% decrease in hemoglobin A1c at 12 months of telemedicine follow-up consultation, complement this study's findings and growing evidence that suggests that telemedicine is a viable alternative for in-person care.
Previous studies also demonstrated telemedicine's effectiveness in delivering diabetes care to rural patients. Wood et al described telemedicine's use in pediatric type 1 diabetes care for patients in rural Wyoming, demonstrated equivalency between telemedicine and in-person visits, and found that patients received more follow-up visits after telemedicine's implementation. Similarly, Wagnild et al described the use of telecommunications for diabetes patients in Montana and found that patients showed improvements in hemoglobin A1c levels, blood pressure, and diabetes knowledge. Our findings are consistent with literature that suggests that telemedicine may effectively deliver diabetes care to rural patients.
Our study has limitations. First, the referring diabetes specialty provider at CAVHCS also independently manages the diabetes treatment of many of the patients enrolled in the telehealth clinic, in some cases just before referral to the telehealth clinic but mostly with select patients between telehealth visits as needed. Thus, telehealth patients' glycemic control before baseline visits and afterward may have been better than that of patients who receive care only from primary care providers. However, use of midlevel providers such as pharmacists and nurses is common across the VA health system, is an integral part of the VA-established Patient Aligned Care Team model, and may represent the patient-centered care model in use.
Another limitation was significant loss of follow-up. Many patients had follow-up visits that did not meet our study criteria of 6- and 12-month follow-up points. This apparent loss of follow-up may have been because the Atlanta VA Telehealth Endocrinology Clinic is available only once per week. As more patients enrolled in the clinic over time, the intervals between follow-up appointments necessarily increased. Therefore, some patients did not have an appointment scheduled at the 6-month point (5–7 months after baseline) or the 12-month point (11–13 months after baseline). Thus, if a patient had an appointment before 11 months or over 13 months after their initial appointment, they would not have been included for the 12-month follow-up analysis. Our follow-up data may have been further confounded by the possibility that patients with worse glycemic control needed more frequent follow-up and thus were more likely to have 12-month follow-up data.
Additionally, our study used convenience sampling of patients enrolled in the Atlanta VAMC Endocrinology Telehealth Clinic. Our findings may not accurately represent patients with type 1 diabetes in the general population because all our patients were veterans seen at the VA and most had insulin pumps, which are associated with better glycemic control compared with insulin injections. Furthermore, our evaluation of aspirin use may have been limited by inconsistent documentation of its use, because many patients purchase it over-the-counter at local drug stores, leading to an underestimation of its use.
Lastly, our limitations include self-selection bias and small sample size. Self-selection bias may have affected our satisfaction survey results because patients who prefer telemedicine may be more likely to enroll in telehealth clinics, whereas patients who prefer in-person care may be more likely travel to VA medical centers to receive treatment. Furthermore, our small sample size limited our statistical power and generalizability. However, these limitations were inherent in our study design, because we conducted a retrospective review of only patients enrolled in our telehealth clinic.
One of telemedicine's most important benefits is its ability to increase access to health care. Distance is a significant factor for many veterans living in remote and rural areas seeking health care, because travel distance is negatively correlated with use of outpatient services. The VA has mitigated this issue by providing travel reimbursement and bus services for patients, but telemedicine further promotes health care accessibility for rural patients. Another important aspect of telemedicine is its acceptance by patients and providers. Our study demonstrates that most patients are satisfied with telemedicine care, believe that telemedicine appointments are convenient, and would recommend telemedicine to other veterans. Our findings are consistent with those of studies that report that both patients and providers are highly satisfied with telemedicine.[21–24]
Lastly, our findings suggest that telemedicine leads to substantial cost savings and complement findings from studies that demonstrate telemedicine's cost-saving capacity in larger health care systems. For example, the use of telemedicine in 7 rural hospital emergency departments in Mississippi decreased the hospitals' expenditures from $7.6 million to $1.1 million during a 5-year period with no apparent effect on clinical outcomes. If the VHA implements telemedicine on a broader scale, veterans could receive more accessible patient-centered care, and the VHA could benefit from significant cost savings.
Our findings suggest that telemedicine delivers safe diabetes care to rural veterans and supports growing evidence that suggests that telemedicine is an effective alternative method of health care delivery. Additionally, telemedicine is associated with cost savings for the VHA, time savings for patients, high appointment adherence, and high patient satisfaction. Future studies with larger, more representative samples of patients with type 1 diabetes are needed to elucidate telemedicine's effectiveness in providing health care to broader patient populations.
This research was supported by the National Center for Advancing Translational Sciences of the National Institutes of Health under award no. UL1TR000454.
Prev Chronic Dis. 2018;15(1):e13 © 2018 Centers for Disease Control and Prevention (CDC)