**11. Economic effectiveness**

It was establish that standard, technology unsupported, diabetes interventions are costeffective. Effective base therapy typically costs up to \$50,000 per each quality adjusted life year

gained [117]. Activities that focus on intensive lifestyle changes, universal opportunistic screening for undiagnosed type 2 diabetes, intensive glycemic control, annual screening for diabetic retinopathy were proven to cost ≤\$25,000 per life year gained or per quality-adjusted life year, what categorizes them as very cost-effective [118]. If there is to be expected that mHealth interventions will be introduced in everyday diabetes patients' care, they need to show themselves to be more cost-effective than standard treatment. In other words, they should cost less than an amount that we are already willing to pay for diabetes treatment. Current diabetes cost-effectiveness studies are sparse, but promising. Findings of one such study demonstrated annual cost decrease by using mHealth glucose meter with a support of disease management call center that outweighed higher program costs by several-folds. Implementation of technology supported care meant \$50 per patient per month higher expenses than standard care, however in a year's time it was possible to register \$3384 cost decrease compared with an increase of \$282 among those with previously established course of treatment [119]. Immediate cost reduction after implementation of telehealth is primarily due to the absence of transportation costs per patient visit to outpatient clinic and productivity savings, because of eliminated need for frequent work absences. More substantial medical savings can be seen with a long term use [120]. Furthermore, it is reasonable to predict lowering of medical cost with growing number of diabetes patient included in automated telephonelinked interventions. For illustration, delivery of mHealth solution to 10,000 patients instead of 1000 can reduce expenses from £444 per patient to £301 [121]. In other economic evaluations, new management methods were determine to be associated with higher cost per quality adjusted life year and not cost-effective addition to standard care. This economic model argues that even with 80% reduction in equipment cost and full utilization of the telehealth service the probability of cost-effectiveness would only reach 61% at the £30,000 threshold of willing‐ ness to pay [122]. Still, individual research results are too heterogeneous to enable extraction of significant meaning regarding a possible medical expenditure reduction with continuous use of mHealth solution [123].

## **12. Future trends**

Qualitative interviewing and exploring how young people with diabetes type 1 make use of technology in their lives and in relation to their condition and treatment, was made. On that basis, many suggestions to develop apps were found including issues such as alcohol and diabetes, hypoglycemia and diabetes, illness and diabetes, and Twitter use for diabetics. All listed suggestions were taken forward for prototyping, with alcohol and diabetes being developed as clinically approved app. There were also many other issues suggested, that were

In UK, a competition for teams including at least one young person with diabetes to develop an app, that might help this group of patients in preparation for their diabetes appointments, was conducted. After the development, other young people with diabetes were invited to choose, test and review new apps. The competition proved successful, showing the app

Insulin calculator apps for patients with diabetes were scrutinized, because self-medication errors are recognized source of avoidable harm. Users are at risk of both catastrophic overdose and more subtle—suboptimal glucose control. In a research, considering input, only 59% calculators included clinical disclaimer and only 30% documented the calculation formu‐ la. 91% lacked numeric input validation, problems were also with calculation with missing values, ambiguous terminology, even with numeric precision. Considering output, 67% of calculators carries a risk of inappropriate output dose recommendation that either violated basic clinical assumptions or did not match a stated formula or correctly update in response to changing user input. It is advised, that health care professionals should exercise substantial caution in recommending unregulated dose calculators to patients and take care for proper

Little attention has been paid to physicians' intentions to adopt mobile diabetes monitoring technology. Japan study showed that overall quality (system quality, information quality, and service quality) assessment does affect doctors' intention to use this technology, but only indirectly through perceived value. Net benefits (both ubiquitous control and health improve‐

Combined smartphone-based logging of different health parameters (e.g., blood sugar logging, insulin dose logging, bread unit logging, activity logging) can of course help doctor (diabetologist) in solving glycemic control problems. With these data, diabetologist can make

It was establish that standard, technology unsupported, diabetes interventions are costeffective. Effective base therapy typically costs up to \$50,000 per each quality adjusted life year

ment) seem to be also a strong driver in both a direct and indirect manner [115].

individualized recommendations for every patient [116].

designers and developers a need to develop a range of new functions [113].

not prototyped [111].

44 Mobile Health Technologies - Theories and Applications

education about possible threats [114].

**10. Doctors' involvement**

**11. Economic effectiveness**

Clinical decision support systems are active knowledge systems using two or more items of patient's data to generate case specific advice. It is in majority of cases standalone technology or integrated in provider's information system and is used by doctors or other medical staff [124]. Many mobile decision support software apps for smartphones are now available for diabetes and are intended to assist patients to make decisions for themselves in real time without having to contact their health care provider. For many minute-to-minute decisions, the questions are not sufficiently significant to warrant contacting a health care provider and there is insufficient time to wait for a reply. Mobile decision support apps can be helpful to assist patients to identify data patterns and make it easier for them to come to an immediate decision on their own [52]. With the advent of minimally invasive subcutaneous continuous glucose monitoring increasing academic and industrial effort has been focused on the devel‐ opment of SC-SC (subcutaneous-subcutaneous) artificial pancreas systems, using continuous glucose monitoring coupled with continuous subcutaneous insulin infusion. Next step is use of mobile system as user interface which is controlled by the patient. The interface is based on patient's commercial mobile phone [125].
