Home telehealth is used to support the management of type 1 diabetes in a number of ways. Its most frequent embodiment is as a remote monitoring service. Patients upload their blood glucose values to a provider-accessible server and receive feedback that assists them in managing their condition. Target populations tend consist of patient with sub-optimal glycemic control, but the emphasis is generally on supporting self-management rather than on enabling remote management. The feedback mentioned above, for instance, may simply be a graphic display of blood glucose trends. The provider can access this display and provide advice, but does not respond to every data transmission with an explicit recommendation.
The typical telehealth intervention was designed to improve glycemic control through patient education and/or the transmission of blood glucose values from patient to provider. Transmission was generally accomplished through use of a mobile phone, a blood glucose monitor with wireless transmission, or a combination. Data were then stored on a secure server and made accessible to both provider and patient on a password-protected website. Interfaces often displayed short- and/or long-term blood glucose trends in graph or chart form. Provider feedback to patients was delivered by through standard telephone calls, online messages, or mobile phone text messaging. Of 12 interventions, 10 fit this basic model (Albisser et al., 2007; Benhamou et al., 2007; 2010; Everett & Kerr, 2010; Farmer et al., 2005; Gomez et al., 2008; Jansa et al., 2006;Rigla et al., 2008;Rossi et al., 2009, 2010).
Patient time commitment was reported in only 2 studies. Patients in Albisser et al. (2007) needed an estimated <10 minutes per day for data entry. In Cox et al. (2008), a series of online educational units took an average of 26.4+/-16.3 minutes each. At the beginning of the study, time commitment per patient was estimated at 1-2 hours per week.
Human Resource Requirements
Descriptions of the types of providers involved in home-based telehealth for type 1 diabetes were scarce. When given, they tended to be cursory. Findings from other chronic diseases would suggest that nurses are usually responsible for patient contact and case management, but the evidence does not allow us to state this with certainty. Physicians appeared to be the central providers in an unusually high number of cases (Gomez et al., 2008; Rigla et al., 2008; Rossi et al., 2009, 2010).
Information on time requirements was reported in 5 studies (Benhamou et al., 2007, 2010; Everett & Kerr, 2010; Farmer et al., 2005; Jansa et al., 2006). Providers in Benhamou et al. (2007) were asked to use text messaging to acknowledge receipt of patient data transmissions and provide management recommendations at weekly intervals. Providers, who were allocated 15 patients each, required an average of 4.5 minutes per week per patient. Providers in Everett and Kerr (2010) report an average of <10 minutes per day to respond to patient messages. In Farmer et al. (2005), the study nurse contacted patients once every 2 weeks on average for the duration of the intervention. Contact time averaged <10 min. In Jansa et al. (2006), providers carried out appointments through teleconsultation. The time needed was heavily dependent on whether technical problems occurred; with no technical problems (70% of connections), appointments took an average of 20-30 minutes. The remaining 30% averaged 1 hour. When this 30% was excluded, the diabetes team as a whole spent 3 hours total in following up with the intervention group, vs. 6 hours with the control group.
These results, while certainly of some value, highlight the current inadequacy of research into the nature and extent of provider tasks associated with home telehealth interventions. Time commitment clearly varies drastically with the design of the intervention. Telehealth costs cannot be accurately calculated without taking this into account. Providers’ perceptions of this commitment – whether it is viewed as manageable or burdensome – are rarely addressed. Comparisons with the demands of providing usual care are also lacking. Given widespread limitations on health human resources and frequency of provider ‘burnout’, more research into provider experiences with home telehealth is recommended.
Interventions using home-based telehealth for type 1 diabetes made unusually high use of mobile phones, typically loaded with customized diabetes software (Benhamou et al., 2007, 2010; Farmer et al., 2005; Rossi et al., 2009, 2010; Vespasiani et al., 2009). Several possible explanations for the frequent use of mobile devices come to mind. Patient populations in studies of home-based telehealth for type 1 diabetes tend to be younger adults, and may be perceived as being more receptive to the technology. Most carry glucometers, and are therefore accustomed to having a device on them at all times. Finally, these patients tend to have few functional limitations and might find a schedule of home-based transmissions more difficult to adhere to than those who are home- or bed-bound.
As a general rule, this software’s primary function was to transmit blood glucose measurements to a provider-accessible server. This was accomplished with various levels of technical sophistication; some generated graphic displays of blood sugar trends, while other systems did no more than transmit numbers. Interestingly, we found only 1 instance of a system that made automatic dosage recommendations (Benhamou et al., 2010). An examination of home-based telehealth systems for type 2 diabetes (see Section C.8.2.3) shows very different trends.
As a general rule, mobile phones were used at the patient end only. Providers viewed data on an internet-enabled computer. An exception was Benhamou et al. (2007), in which care providers replied to data uploads through text messaging.
Little specialized equipment was required in Albisser et al. (2007), Cox et al. (2008), Jansa et al. (2006), and Wangberg (2008). Cox (2008) and Wangberg (2008) were educational interventions in which an internet-enabled computer was the only technical requirement. The remaining 2 studies were remote monitoring interventions that allowed patients to transmit data using their home telephone lines.
Benhamou et al. (2007) was one of the few studies to report the means of patient-provider communication. While findings in other areas suggest that simple telephone contact is still the default mechanism, this was a reporting gap in the type 1 diabetes literature.
Departures from Basic Model
The majority of interventions fit the basic model, although there was some variation in the technologies used in delivery. Benhamou et al. (2010) was unique in its use of smartphones and dosage prediction software, while Everett and Kerr (2010) was the only intervention to use videoconferencing. Cox et al. (2008) and Wangberg (2008), however, were exceptions. These studies were unusual in their focus on education rather than remote monitoring. In one case, this was an online blood glucose awareness program that had historically been delivered through outpatient sessions (Cox et al., 2008). In the other, participants completed a series of internet modules on various areas of self-care (Wangberg, 2008).
A scan of material from 2011-2012, a time period not covered by our initial searches, found 1 additional study in which smartphones were used (Charpentier et al., 2011). See Charpentier et al. (2011) for more details.