Effects of Telemedicine on the Management of Diabetes

Abstract

Objective: To evaluate the effects of various telemedicine interventions on managing diabetes.

Methods: Four databases were searched using keywords telemedicine, management, chronic illness, adults, and diabetes to generate literature on using telemedicine in managing diabetes.  Six studies were included in this paper.  Each study offered a unique intervention in promoting better self-management and education in an effort to better manage a patient’s diabetes.  After providing the intervention, researchers assessed patients’ hemoglobin A1c levels and additional secondary parameters to evaluate the effectiveness of the intervention.

Findings: Each study found that participants in the intervention groups had lower hemoglobin A1c levels after receiving the intervention than patients in the control groups.  In addition to hemoglobin A1c, many studies found that the intervention groups also saw decreases in blood pressure, low density lipoprotein cholesterol, and body mass index.

Limitations:  Studies had small and very specific sample sizes making it difficult to generalize findings to larger populations.  One study found high attrition rates due to difficulty using the technology, which may be a potential barrier for many patients when using telemedicine.  Some of the technology involved is extremely in depth and expensive and may not be feasible to apply to a larger population due to cost.

Conclusion: Many studies have shown that a variety of telemedicine interventions can be an effective means to help control diabetes.  Because telemedicine is a fairly new concept, more research is necessary to evaluate the cost-benefit of telemedicine and its effect on overall morbidity and mortality.

Themes

  • The prevalence of diabetes in the United States is increasing and telemedicine has emerged as a potential mechanism in which healthcare providers can help better promote patient self-management and education regarding diabetes.
  • Telemedicine can include interventions as simple as automated phone calls reminding patients about appointments, medications, or lab draws, or can be quite extensive in installing devices in patients’ homes that allow patients to check their blood sugar, blood pressure, and weight regularly, all of which can be transmitted to providers to help adjust treatment plans accordingly.
  • Telemedicine has shown to be effective in managing diabetes through lowering patients’ hemoglobin A1c levels.
  • In addition to lowering hemoglobin A1c levels, many studies have shown that telemedicine can also be effective in lowering patients’ blood pressure, low-density lipoprotein cholesterol, and body mass index, all of which may contribute to diabetic complications if left unmanaged.
  • Telemedicine is a new concept that has not been extensively studied.   No long-term studies have evaluated the effectiveness of telemedicine, so no data has been generated regarding its overall effect on morbidity and mortality or whether or not it is cost-effective in the long-term management of diabetes.

Quotes

  • "As of 2012, 29.1 million Americans were diabetic, with 8.1 million of those being undiagnosed (Centers for Disease Control and Prevention [CDC], 2014)."
  • "In 2012, the estimated total cost of care for diabetics in the United States was $245 billion with diabetics having medical expenses that were 2.3 times more than those of an individual without DM (CDC, 2014)."
  • "Each study had a unique intervention with a wide range of services from automated phone calls to laptops and technological devices that were installed in patients’ homes."
  • "Despite the wide range of interventions provided throughout these six studies, each study found a decrease in HbA1c levels in the intervention groups when compared to control groups"
  • "Carter et al. (2011), who as previously mentioned had a very in depth intervention with African American diabetics living in Washington D.C., found a significant decrease in HbA1c in the intervention group from a baseline of 9.0 percent to 6.82 percent ."
  • "Shea et al. (2009) also evaluated BP over the five year period and found that the intervention group had a significant decrease in mean SBP of 4.51 mm Hg and mean DBP of 4.22 mm Hg, compared to the control group that had a decrease in mean SBP of 1.70 mm Hg and mean DBP of 2.06 mm Hg." 
  • "Thus, it can be questioned if the only users of a voluntary telemedicine intervention for which logging into an internet portal for educational purposes to improve self-management would be those who are more motivated about managing their DM." 
  • "Participants reported frustration and difficulty using the telemedicine device as causes of dropping out of the study."
  • "Finally, one limitation of telemedicine that none of the studies addressed is the legal aspect regarding situations in which a patient is in one state and a provider, whose license is potentially state regulated, is in another state." 
  • "With much of the technology involved in telemedicine being expensive, it would be beneficial to have information regarding the cost effectiveness of such in depth interventions and whether or not they are more cost effective than regular care and feasible for the long-term management of DM." 
  • "One area for future research would be to implement a long-term study and evaluate if better management of DM through the use of telemedicine decreased the incidence of diabetic complications and comorbidities and overall mortality." 

Effects of Telemedicine on the Management of Diabetes

Diabetes mellitus (DM) is an epidemic across the United States.  As of 2012, 29.1 million Americans were diabetic, with 8.1 million of those being undiagnosed (Centers for Disease Control and Prevention [CDC], 2014).  As the prevalence of DM continues to increase, so does the financial burden.  In 2012, the estimated total cost of care for diabetics in the United States was $245 billion with diabetics having medical expenses that were 2.3 times more than those of an individual without DM (CDC, 2014).  In addition to healthcare costs, diabetics are also at increased risk for numerous complications of DM and additional comorbidities, such as diabetic retinopathy and blindness, neuropathy often leading to lower extremity amputations, diabetic nephropathy leading to kidney disease, and cardiovascular events such as heart attacks and strokes . 

In 2010, DM was the seventh leading cause of death in the United States (CDC, 2014).  This epidemic will likely worsen if healthcare providers do n ot find additional ways to effectively manage DM and promote better patient self-management.  With the rapid expansion of technology , telemedicine is one area that has emerged as a potential mechanism to help improve the overall management of DM.  This paper evaluates previous studies that have implemented telemedicine interventions to assess their effect on managing DM.

Review of Literature - Methodology

To find existing literature regarding telemedicine and its role in managing DM, the databases PubMed, PsycINFO, Scopus, and CINAHL were searched using the keywords telemedicine, management, chronic illness, adults, and diabetes.  These search engines were used so that peer reviewed literature from scientific journals would be found.  As detailed in Figure 1, the initial search yielded 81 articles.  After eliminating duplicates and evaluating titles and abstracts of each article, additional cuts were made to irrelevant sources so that 26 articles remained.  Articles were excluded for the following reasons:

  • a lack of specificity to DM;
  • no measurable outcome or specific goal for DM was mentioned;
  • they were not written in English;
  • the study involved only one patient;
  • the focus was on electronic medical records (EMR) rather than telemedicine;
  • they lacked any component of telemedicine, or systematic reviews; or,
  • they included studies in which the people supporting the telemedicine could not intervene medically or provide any educational benefit to the patient because they had no access to a medical record, no communication with a primary care provider (PCP), or no ability to treat the patients themselves.

A n additional article was found in PubMed that was not found in the initial search.  Three of the articles were excluded as they discussed the same study.  And the two oldest articles were excluded , thus six articles were included in this paper.

Review of Literature - Results

As shown in Table 1, of the six remaining studies, four were randomized controlled trials and two were retrospective studies.  Each of the studies were similar in that they evaluated some aspect of telemedicine and its effect on managing DM.  To measure the effectiveness of telemedicine, each study measured a hemoglobin A1c (HbA1c) at baseline and at different intervals thereafter and evaluated the trend in the HbA1c.  Where studies varied was whether they evaluated additional secondary parameters that can contribute to an increased risk of comorbidities and complications of DM, such as blood pressure (BP), low-density lipoprotein cholesterol (LDL), and body mass index (BMI).  Shea et al. (2009) evaluated HbA1c, BP, and LDL and set specific goals for HbA1c for patients based on their life expectancy or risk of being unaware if severely hypoglycemic and set goals for LDL based on risk for cardiovascular events.  In addition, Carter, Nunlee-Bland, and Callender (2011) also evaluated HbA1c and BP with specific goals, and included patients’ BMI with a goal BMI between 18.5 and 24.9.  Holbrook et al. (2009), further evaluated patients’ feet for ulcers or neuropathy and microalbuminuria.    

While each study was somewhat unique in the parameters they measured to evaluate the effectiveness of telemedicine on DM, they also differed in population, sample size, length of study, and how telemedicine was actually provided.  One study evaluated medically underserved older diabetics in rural and urban areas of New York who were Medicare beneficiaries (Shea et al., 2009).  Carter et al. (2011) only included African-American patients with type 2 DM who were 18 years or older, living in Washington D.C., and who had at least an eighth grade reading level.  Wakefield et al. (2011) studied patients with PCPs at a VA Medical Center in Iowa City who had a landline telephone and had both hypertension and type 2 DM.  Finally, Holbrook et al. (2009) included patients in Ontario, Canada who were 18 years or older with type 2 DM who were fluent in English.   

In addition to different patient demographics, the sample size and lengths of the studies varied.  One study, which lasted nine months, was extremely small and included only 47 participants with 26 in the intervention group and 21 in the control group (Carter et al., 2011).  The largest study evaluated 1,665 patients in total, with 844 in the intervention group and 821 in the control group.  This study was also the longest in duration, with the intervention lasting five years (Shea et al., 2009). 

Finally, there was no single intervention that seemed to define telemedicine amongst the six studies.  Each study had a unique intervention with a wide range of services from automated phone calls to laptops and technological devices that were installed in patients’ homes.  In one study, researchers provided each patient with a laptop and different attachments such as a scale, a BP cuff, and a glucometer so that patients could weigh themselves and monitor their BP and blood glucose (BG) levels.  Information was uploaded onto a server where a nurse could review and send the data to the patient’s PCP for possible updated treatment plans.  In addition, the laptop enabled the patient to video conference with the nurse to help develop an action plan based on the PCP’s treatment plan.  The patient could also ask questions and get real- time verbal feedback.  Also, the patient had access to a portal which provided a self-management module where the patient could view their health record and action plan, an educational module which included videos, health education websites, and information on diet, exercise, weight loss, and stress, and finally, a social networking module so patients could exchange information and strategies with others in the intervention group (Carter et al., 2011).

In another study, patients were assigned a case manager who called the patient regularly to develop an individually tailored care plan regarding their DM, to provide counseling on medication adherence and their individual treatment goals, and to support lifestyle changes (Jordan, Lancashire, & Adab, 2011).

Holbrook et al. (2009) provided an I nternet-based tracker that connected to the provider’s EMR so that the patient would receive automated phone calls with reminders about medications, doctor’s appointments, and lab draws. 

Lau, Campbell, Tang, Thompson, & Elliott (2014) provided patients with an email address to voluntarily access an I nternet portal which provided medical educational material, personal lab information, and messaging systems to providers.

Finally, Wakefield et al. (2011) provided a more in- depth intervention where a device was provided to each patient that transmitted BG and BP readings to a nurse.  The device asked the patients automated questions and provided reinforcement and additional education.  Each day, a nurse reviewed the data and messages from the patient, allowing the nurse to closely monitor the patient and report any abnormal BG parameters to the patient’s provider so that a change in treatment could occur prior to an appointment.

Findings

Despite the wide range of interventions provided throughout these six studies, each study found a decrease in HbA1c levels in the intervention groups when compared to control groups.  Carter et al. (2011), who as previously mentioned had a very in- depth intervention with African American diabetics living in Washington D.C., found a significant decrease in HbA1c in the intervention group from a baseline of 9.0% to 6.82%.  The control group in this study still saw a decrease in HbA1c, but not nearly to the same extent, decreasing from 8.8% to 7.9%.  In addition, a larger decrease in BMI was seen in the intervention group , from 35.4 to 23.8, achieving a BMI below their goal.  The control group’s BMI still decreased from 36.1 to 26.5, but did not get below their target.

Jordan et al. (2011), whose intervention provided phone calls for support and counseling, also found larger decreases in HbA1c, BP, and BMI when compared to their control group.  They did not find as large of an effect of the intervention on LDL.  In addition, participants were divided into subgroups depending on their baseline HbA1c and found that the higher the baseline HbA1c (greater than or equal to 8.0%), the greater the reduction they saw over the course of the study.

Shea et al. (2009) looked at the effects of an in- depth intervention over five years using a home telemedicine device that connected to a patient’s phone line and offered video conferencing between a nurse and patient and home glucometers and BP cuffs so BG and BP measurements could be uploaded to a database for providers.  After following-up with patients each year, they found that although the mean HbA1c was comparable between the intervention group and the control group at the onset of the study, the group receiving telemedicine support had a larger overall decrease in HbA1c of 0.34 percent between the baseline measurements and those taken at the fifth year follow-up compared to the control group that had a decrease of 0.07 percent . 

Shea et al. (2009) also evaluated BP over the five year period and found that the intervention group had a significant decrease in mean systolic blood pressure (SBP) of 4.51 mm Hg and mean diastolic blood pressure (DBP) of 4.22 mm Hg, compared to the control group that had a decrease in mean SBP of 1.70 mm Hg and mean DBP of 2.06 mm Hg.  Finally, LDL was evaluated in this study and was initially found to have a larger decrease in the intervention group between years one through four, however, this decrease evened out and at year five; the decreases in LDL were comparable between the intervention group and the control group with mean LDL decreasing by 15.51 mg/dL in the intervention group compared to 13.14 mg/dL in the control group .

Finally, Wakefield et al. (2011) did something slightly different in that they evaluated patients at three points: at baseline, after receiving the intervention for six months, and then again after six more months without the intervention.  At baseline, the control and intervention groups had comparable HbA1c scores.  At six months, the intervention group saw a significant decrease in HbA1c while the control group did not see much of a change.  However, after the intervention was removed, the intervention group’s HbA1c actually increased and while still lower than baseline, the level of change was not nearly as great as it was while receiving the intervention.  

Limitations

Although these studies consistently found lower HbA1c measurements in patients enrolled in telemedicine intervention groups, many limitations were evident amongst the studies that may affect the widespread use of telemedicine in managing DM long-term.  For example, one limitation in the study by Lau et al. (2013), was that the study provided an intervention in which patients voluntarily signed up for a service.  Of the 1,957 patients who were given a username and password to the online portal, only 411 patients actually logged in at least one time, providing only a 21 percent rate of participation .  While only 157 patients of the 1,957 met criteria to be included in the study, only 50 of the 157 participants were considered actual users of the online portal.  Thus, it can be questioned if motivation is a key factor when users of a voluntary telemedicine intervention log into a I nternet portal for educational purposes to improve self-management .  Did this study find that the intervention group saw a larger decrease in HbA1c because the users of the portal were more motivated to manage their DM than those in the control group? 

In addition, many of these studies included extremely specific populations in their sample.  Shea et al. (2009) evaluated an older population of Medicare beneficiaries residing in rural and urban medically underserved areas, many of whom stated they were not proficient in using a computer.  Also, Carter et al. (2011) evaluated only African Americans living in Washington D.C.  Having such specific samples within studies makes it extremely difficult to generalize the results to larger populations. 

An additional limitation was that in the larger study conducted by Shea et al. (2009), high attrition rates were reported, especially in the intervention group.  Participants reported frustration and difficulty using the telemedicine device as major reasons for dropping out of the study.  Even patients who remained in the study reported needing a fair amount of training and support in using the telemedicine device.  Thus, when implementing telemedicine in older, less educated, and less computer literate populations, these factors need to be taken into consideration when deciding on the type of telemedicine intervention to employ, as some of the more in -d epth computer-based interventions may not be appropriate or effective for all populations. 

Another limitation is that many of these interventions are not feasible in a larger setting or population due to cost.  It would be extremely expensive to provide telemedicine devices equipped with glucometers and BP cuffs for every diabetic in need of better management across the country.  None of these studies addressed any sort of cost analysis of these interventions making the generalizability of some of the more in- depth interventions to larger populations somewhat limited. 

Finally, one limitation of telemedicine that none of the studies addressed is the legal aspect regarding situations in which a patient is in one state and a provider, whose license is potentially state regulated, is in another state.  The aspect of where care is technically being provided will need to be addressed as telemedicine continues to expand. 

Recommendations for Clinical Practice

After looking at these six studies, it is evident that telemedicine can have an effect on the management of DM in lowering HbA1c levels in addition to improving other secondary parameters related to diabetic complications.  With the aging of the baby boomer generation, PCPs are in increasingly high demand.  While telemedicine is still new, technology continues to expand rapidly, and telemedicine will follow that trend.  By embracing telemedicine, clinicians can offer a more collaborative treatment plan between patient and provider that encourages better self-management and education about DM.  Thus, as clinicians, it is important to be open- minded about patients enrolling in telemedicine programs.  By encouraging patient participation and learning more about getting involved as providers, clinicians can begin to help tackle the growing epidemic of DM in the United States.

Recommendations for Research

Telemedicine is an extremely new concept in medicine with very few research studies done to date.  Future research on this topic is critical to further evaluate its effectiveness in managing DM, especially for long-term use.  One recommendation for future studies would be to include larger sample sizes.  As mentioned previously, many of the studies evaluated in this paper were extremely small, making them hard to generalize to a larger population.  In addition to larger sample sizes, studies need to be longer in time duration.  The longest study included in this paper was five years.  Long-term benefits or drawbacks cannot be evaluated when long-term data is not available.

In addition, research would benefit from studies on cost analysis.  None of the six studies evaluated cost benefits regarding telemedicine interventions.  Since much of the technology involved in telemedicine is expensive, it would be beneficial to have information regarding the cost- effectiveness of such in- depth interventions and to compare whether or not they are more- cost effective than regular care, thus feasible for the long-term management of DM. 

Also, as mentioned previously, Jordan et al. (2011) found that patients with a higher baseline HbA1c saw bigger improvements throughout the study.  Shea et al. (2009) also noted larger decreases in HbA1c in patients with a higher baseline value.  However, they also cited that they did not subdivide participants due to baseline HbA1c values, so inferences about interventions on certain subpopulations could not be made.  One area for future research would be to categorize patients into subgroups based on HbA1c at the onset and assess if the need and effectiveness for telemedicine varies by group.  This may be an important finding in implementing telemedicine in the management of DM in the future, especially if certain subgroups may benefit more than others.

Finally, few studies have addressed the long-term effects of telemedicine on morbidity and mortality.  Shea et al. (2009) saw a comparable number of patients die between the intervention group and control group over the five years, however, they did not find that all-cause mortality changed.  Also, none of the studies addressed whether better control of DM directly led to fewer diabetic complications.  One area for future research would be to implement a long-term study and evaluate if better management of DM through the use of telemedicine decreased the incidence of diabetic complications, co-morbidities, and overall mortality. 

Conclusion

Diabetes Mellitus is extremely prevalent across the United States and is continuing to increase.  With the aging baby boomer population and increasing demand on PCPs, clinicians across the country must look to additional resources and methods to provide increased care to combat this epidemic.  Telemedicine is an area of health technology that is beginning to be used to help provide greater support to patients to promote self-management of DM.  Four randomized controlled trials and two observational studies evaluated the effectiveness of telemedicine in managing DM. While the type of telemedicine that was employed differed among each study, ranging from automated phone calls to more complex home devices, each study saw a decrease in patients’ HbA1c, no matter the size of the intervention group or the demographics.  While these studies had limitations, more research needs to be done to evaluate the long-term use of telemedicine in managing DM.  In the meantime, healthcare providers should embrace the use of such technologies as potential methods to help better manage DM with their patients .

References

Carter, E. L., Nunlee-Bland, G., & Callender, C. (2011). A patient-centric, provider-assisted diabetes telehealth self-management intervention for urban minorities. Perspectives in Health Information Management, 8(1b). Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3035826/

Centers for Disease Control and Prevention (CDC). (2014). National diabetes statistics report, 2014: Estimates of diabetes and its burden in the United States. Retrieved from http://www.cdc. gov/diabetes/pubs/statsreport14/national-diabetes-report-web.pdf

Holbrook, A., Thabane, L., Keshavjee, K., Dolovich, L., Bernstein, B., Chan, D., Gerstein, H. (2009). Individualized electronic decision support and reminders to improve diabetes care in the community: COMPETE II randomized trial. Canadian Medical Association Journal, 181(1-2), 37-44. doi:10.1503/cmaj.081272

Jordan, R. E., Lancashire, R. J., & Adab, P. (2011). An evaluation of Birmingham Own Health telephone care management service among patients with poorly controlled diabetes: A retrospective comparison with the General Practice Research Database. BMC Public Health, 11(707). doi:10.1186/1471-2458-11-707

Lau, M., Campbell, H., Tang, T., Thompson, D. J., & Elliott, T. (2014). Impact of patient use of an online patient portal on diabetes outcomes. Canadian Journal of Diabetes, 38(1), 17-21. doi:10.1016/j.jcjd.2013.10.005

Shea, S., Weinstock, R. S., Teresi, J. A., Palmas, W., Starren, J., Cimino, J. J., . . . Eimicke, J. P. (2009). A randomized trial comparing telemedicine case management with usual care in older, ethnically diverse, medically underserved patients with diabetes mellitus: 5 year results of the IDEATel study. Journal of the American Medical Informatics Association, 16(4), 446-456. doi:10.1197/jamia.M3157

Wakefield, B. J., Holman, J. E., Ray, A., Scherubel, M., Adams, M. R., Hillis, S. L., & Rosenthal, G. E. (2011). Effectiveness of home telehealth in comorbid diabetes and hypertension: A randomized, controlled trial. Telemedicine and e-Health, 17(4), 254-261. doi:10.1089/tmj.2010.0176

Keywords: 
telemedicine, diabetes, hemoglobin A1c, self-management, Technology