**3. HITs in glycemic control among patients with T2D**

A growing research attention has been given to evaluate HITs' impact on diabetes management, including the primary management goal, glycemic status, and major complications such as cardiovascular conditions. Previous reviews on this subject suggested that HITs have the potential to improve these disease outcomes [18–23]. However, effect size is specific to the main outcome; glycated hemoglobin (HbA1c) varied between studies with reported mean difference ranging from −0.20 to −0.57% [19–23]. **Table 1** presented the synthesized findings from the latest systematic reviews. Heitkemper et al. searched randomized control trials (RCTs) that studied the effect of HITs on HbA1c among medically underserved patients [21]. In this meta-analysis of 10 eligible trials, HITs were associated with significant HbA1c reduction at 6 months (pooled standardized difference in mean: −0.36, 95% CI −0.53, −0.19) with diminishing but still significant effect at 12 months (pooled standardized difference in mean: −0.27, 95% CI −0.49, −0.04). The authors also performed analyses by HIT type including computer software without Internet

**137**

**Author, year** Yoshida et al., 2018 [33]

Evaluating of effect of HITs on T2D glycemic control in general T2D patients, including mobile phone-based HITs, Web-based HITs, short message/

RCTs conducted from 1946 to December 2017

34 RCTs (40 estimation points); 3983 participants with T2D

text, and other HITs

Heitkemper

Evaluating of

RCTs

10 RCTs; 3257

Not reported

Standard mean

No

Up to

Internet-based HITs (standard

External validity

issue (only

focused on a

specific patient

group); mixed

participants

with type 1 and

2 diabetes

mean difference = −0.50, 95%

CI −0.69, −0.32 at 6 months and

−0.87, 95% CI −1.58, −0.21 at

12 months); cellular/automated

telephone HITs (standard mean

difference = −0.26, 95% CI

−0.49, −0.03 at 6 months and

not significant at 12 months);

telehealth (standard mean

difference = −0.37, 95% CI −0.68,

−0.06 at 6 months and not

significant at 12 months)

12 months

difference: −0.36,

95% CI −0.53,

−0.19 at 6 months

and −0.27, 95% CI

−0.49, −0.04 at

12 months

conducted

medically

from 2000

underserved

adults with

diabetes

effect of HIT

self-management

interventions on

to 2015

glycemic control

in medically

underserved adults

with diabetes,

including computer

software without

Internet, cellular/

automated

telephone, Internetbased HITs, and

telemedicine/

telehealth

et al., 2017 [21]

**Objective and intervention(s) under review**

**Inclusion criteria**

**Sample**

**HbA1c reduction (absolute difference in means)**

**HbA1c reduction (standardized difference in means and Hedges' g)**

**CVD risk factor assessment**

**Intervention period**

**HIT subgroup analysis**

**Major limitations**

−0.65% (95% CI

Standard mean

A separate

2–12 months

Mobile phone-based approaches

Did not provide

analysis at

different time

points

[Hedges' g = −0.66 (95% CI

−0.88, −0.45)]; SMS/text [Hedges'

g = −0.63 (95% CI −1.07, −0.19)];

Web-based [Hedges' g = −0.48

(95% CI −0.65, −0.30)]

analysis

focusing on

CVD risk

factors is

upcoming

difference: −0.57

(95% CI −0.71,

−0.43);

Hedges' g: −0.56

(95% CI −0.70,

−0.43)

−0.99, −0.64%)

*Health Information Technologies in Diabetes Management*

*DOI: http://dx.doi.org/10.5772/intechopen.83693*


*Health Information Technologies in Diabetes Management DOI: http://dx.doi.org/10.5772/intechopen.83693*

*Type 2 Diabetes - From Pathophysiology to Modern Management*

**2. The potential of HITs in chronic disease management**

**3. HITs in glycemic control among patients with T2D**

A growing research attention has been given to evaluate HITs' impact on diabetes management, including the primary management goal, glycemic status, and major complications such as cardiovascular conditions. Previous reviews on this subject suggested that HITs have the potential to improve these disease outcomes [18–23]. However, effect size is specific to the main outcome; glycated hemoglobin (HbA1c) varied between studies with reported mean difference ranging from −0.20 to −0.57% [19–23]. **Table 1** presented the synthesized findings from the latest systematic reviews. Heitkemper et al. searched randomized control trials (RCTs) that studied the effect of HITs on HbA1c among medically underserved patients [21]. In this meta-analysis of 10 eligible trials, HITs were associated with significant HbA1c reduction at 6 months (pooled standardized difference in mean: −0.36, 95% CI −0.53, −0.19) with diminishing but still significant effect at 12 months (pooled standardized difference in mean: −0.27, 95% CI −0.49, −0.04). The authors also performed analyses by HIT type including computer software without Internet

suggested steps to move forward.

facilitating patient self-management (through risk communication, Web portals, telemedicine, e-mailing, and secure messaging) [6–8]. In this chapter, we summarized the current findings on HITs in managing T2D, especially on glycemic control and CVD risks management. In addition, we discussed limitations in the current research in this area and implications for future research. Further, we presented challenges of applying HITs in T2D management in the real-world context and

HITs include a broad range of technologies, electronic tools, applications, or systems that provide patient care, information, recommendations, or services for promotion of health and health care [9]. The advantages of using HITs in health care have been well documented [10–13]. They have the potential to empower patients and support a transition from a role in which the patient is the passive recipient of care services to an active role in which the patient is informed, has choices, and is involved in the decision-making process [10]. They are also designed to promote communication and relationships between clinicians and patients and overcome geographical barriers and logistical inconvenience when seeking health-care services [11]. In the realm of chronic disease management, a variety of technologies have shown their positive effects. For examples, electronic health record system provides reminders at the point of care for providers to identify high-priority clinical areas for patients with complex chronic illness [14]; telemonitoring system provides asthma patients with continuous individualized help in the daily routine of asthma self-care [12]; Web-based applications increase knowledge, problemsolving skills, and social support via an interactive system for patients with cancers [13]; mobile technology devices such as personal digital assistants (PDAs) and cellular phones enable additional resources to care and change the location of care; and mobile phone short message service (SMS) were able to remind patients of scheduled visits, deliver test results, and monitor side effects of treatment [15–17]. The HIT-enabled self-care keeps evolving and attempts to address more challenging health-care issues, such as diabetes management where patients need comprehensive information and ongoing guidance as they work to develop a diverse knowledge

**136**

and skills.


**139**

**Author, year** Pal et al., 2014 [24]

Evaluating computer-based interventions in self-management in T2D patients.

RCTs conducted up until November 2011

16 RCTs; 3578 participants with T2D

Intervention delivered via clinics, the Internet, and mobile phone

Marcolino

Evaluating of effect

RCTs

13 RCTs; 4207

−0.44% (95% CI

Not reported

Yes. Only

6–18 months

Not reported

Mixed

participants

with type 1 and

2 diabetes; did

not provide

analysis at

different time

points

found

telemedicine

was associated

with reduction

in LDL

(−6.6 mg/dL,

95% CI −8.3,

−4.9 mg/dL)

Liang et al.,

Evaluating of effect

RCTs

22 trials

−0.5% (95% CI

Not reported

No

3–12 months

Not reported

Lumped

nonrandomized

and randomized

trials together

into evaluation

−0.3, −0.7%)

conducted

including 11

RCTs and 11

non-RCTs; 1657

of mobile phone

intervention for

from

diabetes on glycemic

January

2010 to

participants

with diabetes

February

2010

*Synthesized findings of effect of HITs on HbA1c and cardiovascular risk factors among diabetes patients.*

**Table 1.**

control

2010 [23]

−0.61, −0.26%)

conducted

participants

with diabetes

up until

April 2012

of telemedicine on

diabetes care

et al., 2013 [22]

**Objective and intervention(s) under review**

**Inclusion criteria**

**Sample**

**HbA1c reduction (absolute difference in means)**

**HbA1c reduction (standardized difference in means and Hedges' g)**

**CVD risk factor assessment**

**Intervention period**

**HIT subgroup analysis**

**Major limitations**

−0.2% (95% CI

Not reported

Yes. Did

8 weeks to

Mobile phone intervention (mean

Did not provide

analysis at

different time

points

difference: −0.5%, 95% CI −0.3,

12 months

not find

improvement

−0.7)

of blood

pressure,

lipids, or

weight due to

interventions

−0.4, −0.1%)

*Health Information Technologies in Diabetes Management*

*DOI: http://dx.doi.org/10.5772/intechopen.83693*


*Health Information Technologies in Diabetes Management DOI: http://dx.doi.org/10.5772/intechopen.83693*

*Type 2 Diabetes - From Pathophysiology to Modern Management*

**138**

**Author, year**

**Objective and** 

**Inclusion** 

**Sample**

**HbA1c** 

**HbA1c reduction** 

**CVD risk** 

**Intervention** 

**HIT subgroup analysis**

**Major** 

**limitations**

**(standardized** 

**factor** 

**period**

**assessment**

**difference in** 

**means and** 

**Hedges' g)**

**reduction** 

**(absolute** 

**difference in** 

**means)**

**criteria**

**intervention(s)** 

**under review**

Tao et al., 2017

Evacuating

RCTs

18 RCTs;

Not reported

Standard mean

No

Up to

Not reported

Lumped

all types of

HITs into

analysis; mixed

participants

with type 1 and

2 diabetes

60 months

difference:

−0.31, 95% CI

−0.38, −0.23;

glycemic control

was significant

at intervention

duration of 3, 6, 8,

9, 12, 15, 30, and

60 months

Faruque et al.,

Evaluating of effect

RCTs

111 RCTs;

−0.57% (95% CI

Not reported

No

3–68 months

The effect was the greatest in trials

Mixed

participants

with type 1 and

2 diabetes

where providers used Web portals

or text messaging to communicate

with patients [mean difference:

−0.35% (95% −0.56, −0.14) and

−0.28% (95% CI −0.52, −0.14)] at

4–12 months

−0.74, −0.40%)

at ≥3 months;

−0.28% (95%

−0.37, −0.20%)

at 4–12 months;

−0.26% (95%

−0.46, −0.06%)

at >12 months

conducted

23,648

from

participants

with diabetes

1946 to

November

of telemedicine

on glycemic

control, including

broad forms of

electronic forms

2015

communication.

Alharbi et al.,

Evaluating of effect

RCTs

32 RCTs; 40,454

−0.33%, (95% CI

Not reported

No

3–36 months

Electronic self-management

Did not provide

analysis at

different time

points

systems [mean difference: −0.50%

(95% CI −0.67, −0.43%)]; EHR

[mean difference: −0.33% (95%

CI −0.40, −0.26%)]; electronic

decision support system [mean

difference: −0.15% (95% CI −0.34,

−0.16%)]; diabetes registry [mean

difference: −0.05% (95% CI −0.15,

−0.19%)]

−0.40, −0.26)

conducted

participants

up until July

with T2D

2016

of HITs in glycemic

control in T2D

patients.

HITs included Webbased approaches,

telephone-based

system, mobile

phone-based system,

and telemedicine

2016 [19]

2017 [20]

conducted

participants in

up until July

trials ranged

from 14 to 1382

of effect of

consumer-oriented

HITs in diabetes

2016

management

[18]

**Table 1.**

*Synthesized findings of effect of HITs on HbA1c and cardiovascular risk factors among diabetes patients.*

(n = 2), cellular/automated telephone (n = 4), Internet-based (n = 4), and telemedicine/telehealth (n = 3). The Internet-based interventions demonstrated the greatest reduction in HbA1c at both 6 months (pooled standardized difference in mean: −0.50, 95% CI −0.69, −0.32) and 12 months (pooled standardized difference in mean: −0.87, 95% CI −1.58, −0.21). Cellular and automated telephone interventions showed the smallest reduction. In Tao and colleagues' systematic review on consumer-centered HITs, they identified a significant pooled reduction of −0.31 (95% CI −0.38, −0.23) in HbA1c from 18 RCTs [18]. Similarly, Alharbi et al. also found HITs were associated with a statistically significant reduction in HbA1c levels (mean difference: −0.33%, 95% CI −0.40, −0.26%) [19]. In addition, Alharbi and colleagues found studies focusing on electronic self-management systems demonstrated the greatest reduction in HbA1c (−0.50%), followed by those with electronic medical records (−0.17%), an electronic decision support system (−0.15%), and a diabetes registry (−0.05%) [19]. Faruque et al. identified 11 RCTs with specific focus on effect of telemedicine [20]. Telemedicine refers to the use of telecommunications to deliver health services, expertise, and information on glycemic control [20]. In this study, the authors demonstrated a significant reductions in HbA1c all three follow-up periods (mean difference at ≤3 months: −0.57%, 95% CI −0.74, −0.40%, at 4–12 months: −0.28%, 95% CI −0.37, −0.20%, and at >12 months: −0.26%, 95% CI −0.46, −0.06%). In another meta-analysis that specially focused on telemedicine, Marcolino and colleagues found telemedicine was associated with a statistically significant and clinically relevant decline in HbA1c level compared to control (mean difference = −0.44%, 95% CI −0.61, −0.26%) [22]. Pal et al. examined the effect of computer-based intervention in selfmanagement in adults with T2D. The authors found modest effect associated with the interventions (mean difference: −0.2%, 95% CI −0.4, −0.1%) [24]. Liang et al. assessed the effect of mobile phone intervention on glycemic control in diabetes self-management and found a significant common reduction of HbA1c (mean difference: −0.5%, 95% CI −0.3, −0.7%) among 22 trials over a median follow-up of 6 months [23].

Many of review studies including those mentioned above have shed light on the effect of HITs in glycemic control. However, these studies often included limited number of trials [21], lack of adherence to standard quantitative methods [25], inadequate attention to heterogeneity across studies [26], lumped nonrandomized and randomized trials together into evaluation [19, 23, 25, 27–29], mixed participants with type 1 or type 2 diabetes into analysis [18, 22, 25, 27–29], or restricted searching criteria to a particular patient population or a specific type of HIT [27, 30–32]. To address these limitations and to verify if and how much HITs impact glycemic control, Yoshida and colleagues recently conducted a meta-analysis to examine the most current state of evidence from RCTs concerning the effect of HITs on HbA1c reduction among patients with T2D [33]. From an analysis of 34 eligible studies (40 estimates) identified from multiple databases from January 1946 to December 2017, the study reported that introduction of HITs to standard diabetes treatment resulted in a statistically reduced HbA1c. The absolute mean difference in HbA1c pre- and postintervention between intervention and control group was −0.65% (95% CI −0.99, −0.64%). The pooled reduction (standardized difference in means) of HbA1c was −0.57 (95% CI −0.71, −0.43) (**Figure 1**). In addition, Yoshida et al. also found the reduction was significant across each of the four types of HIT interventions (i.e., mobile phone-based, Web-based technologies, SMS/text, or others) under review, with mobile phone-based approaches generating the largest effects [pooled reduction was −0.67 (95% CI −0.90, −0.45)] followed by SMS/text [−0.64 (95% CI −1.09, −0.19)], and Web-based [−0.48 (95% CI −0.65, −0.30)] [33].

**141**

**Figure 1.**

*Health Information Technologies in Diabetes Management*

HITs also have significant clinical impact in reducing HbA1c among patients with T2D. It is reported that every 1% decrease in HbA1c over a 10-year period is associated with a risk reduction of 21% for diabetes-related death and 37% of microvascular complications [34]. This reduction results from HIT interventions may be bigger than effects of many targeted pharmacological therapies. Oral antidiabetic agents reduced HbA1c levels of 0.5–1.25%, with thiazolidinedione and sulfonylureas showing the best reduction (1–1.25%) [35]. Biguanide reduced HbA1c by 1.0–2.0%; dipeptidyl peptidase 4 (DPP-IV) inhibitor, 0.5–0.8%; GLP-1 agonists, 0.5–1.5%; and TZD, 0.5–1.4% [36]. It is questionable that the effects on HbA1c yielded from the HIT trials were a mixed product of both HITs and standard diabetes care including medication adherence and lifestyle modifications. This concern was addressed in the systematic review of Yoshida et al. [33]. The authors conducted a subset analysis of 18 studies that exclusively compared the outcome between a combined HITs and standard care intervention group vs. standard care control group. The effect size estimated from this analysis was −0.63 (Hedges' g: −0.63 95% CI −0.84, −0.42), which is attributable to HIT tools in addition to the

*Pooled reduction of HbA1c due to HITs. Adopted from the study of Yoshida et al [33].*

*DOI: http://dx.doi.org/10.5772/intechopen.83693*


#### **Figure 1.**

*Type 2 Diabetes - From Pathophysiology to Modern Management*

(n = 2), cellular/automated telephone (n = 4), Internet-based (n = 4), and telemedicine/telehealth (n = 3). The Internet-based interventions demonstrated the greatest reduction in HbA1c at both 6 months (pooled standardized difference in mean: −0.50, 95% CI −0.69, −0.32) and 12 months (pooled standardized difference in mean: −0.87, 95% CI −1.58, −0.21). Cellular and automated telephone interventions showed the smallest reduction. In Tao and colleagues' systematic review on consumer-centered HITs, they identified a significant pooled reduction of −0.31 (95% CI −0.38, −0.23) in HbA1c from 18 RCTs [18]. Similarly, Alharbi et al. also found HITs were associated with a statistically significant reduction in HbA1c levels (mean difference: −0.33%, 95% CI −0.40, −0.26%) [19]. In addition, Alharbi and colleagues found studies focusing on electronic self-management systems demonstrated the greatest reduction in HbA1c (−0.50%), followed by those with electronic medical records (−0.17%), an electronic decision support system (−0.15%), and a diabetes registry (−0.05%) [19]. Faruque et al. identified 11 RCTs with specific focus on effect of telemedicine [20]. Telemedicine refers to the use of telecommunications to deliver health services, expertise, and information on glycemic control [20]. In this study, the authors demonstrated a significant reductions in HbA1c all three follow-up periods (mean difference at ≤3 months: −0.57%, 95% CI −0.74, −0.40%, at 4–12 months: −0.28%, 95% CI −0.37, −0.20%, and at >12 months: −0.26%, 95% CI −0.46, −0.06%). In another meta-analysis that specially focused on telemedicine, Marcolino and colleagues found telemedicine was associated with a statistically significant and clinically relevant decline in HbA1c level compared to control (mean difference = −0.44%, 95% CI −0.61, −0.26%) [22]. Pal et al. examined the effect of computer-based intervention in selfmanagement in adults with T2D. The authors found modest effect associated with the interventions (mean difference: −0.2%, 95% CI −0.4, −0.1%) [24]. Liang et al. assessed the effect of mobile phone intervention on glycemic control in diabetes self-management and found a significant common reduction of HbA1c (mean difference: −0.5%, 95% CI −0.3, −0.7%) among 22 trials over a median follow-up

Many of review studies including those mentioned above have shed light on the effect of HITs in glycemic control. However, these studies often included limited number of trials [21], lack of adherence to standard quantitative methods [25], inadequate attention to heterogeneity across studies [26], lumped nonrandomized and randomized trials together into evaluation [19, 23, 25, 27–29], mixed participants with type 1 or type 2 diabetes into analysis [18, 22, 25, 27–29], or restricted searching criteria to a particular patient population or a specific type of HIT

[27, 30–32]. To address these limitations and to verify if and how much HITs impact glycemic control, Yoshida and colleagues recently conducted a meta-analysis to examine the most current state of evidence from RCTs concerning the effect of HITs on HbA1c reduction among patients with T2D [33]. From an analysis of 34 eligible studies (40 estimates) identified from multiple databases from January 1946 to December 2017, the study reported that introduction of HITs to standard diabetes treatment resulted in a statistically reduced HbA1c. The absolute mean difference in HbA1c pre- and postintervention between intervention and control group was −0.65% (95% CI −0.99, −0.64%). The pooled reduction (standardized difference in means) of HbA1c was −0.57 (95% CI −0.71, −0.43) (**Figure 1**). In addition, Yoshida et al. also found the reduction was significant across each of the four types of HIT interventions (i.e., mobile phone-based, Web-based technologies, SMS/text, or others) under review, with mobile phone-based approaches generating the largest effects [pooled reduction was −0.67 (95% CI −0.90, −0.45)] followed by SMS/text [−0.64 (95% CI −1.09, −0.19)], and Web-based [−0.48 (95% CI −0.65,

**140**

−0.30)] [33].

of 6 months [23].

*Pooled reduction of HbA1c due to HITs. Adopted from the study of Yoshida et al [33].*

HITs also have significant clinical impact in reducing HbA1c among patients with T2D. It is reported that every 1% decrease in HbA1c over a 10-year period is associated with a risk reduction of 21% for diabetes-related death and 37% of microvascular complications [34]. This reduction results from HIT interventions may be bigger than effects of many targeted pharmacological therapies. Oral antidiabetic agents reduced HbA1c levels of 0.5–1.25%, with thiazolidinedione and sulfonylureas showing the best reduction (1–1.25%) [35]. Biguanide reduced HbA1c by 1.0–2.0%; dipeptidyl peptidase 4 (DPP-IV) inhibitor, 0.5–0.8%; GLP-1 agonists, 0.5–1.5%; and TZD, 0.5–1.4% [36]. It is questionable that the effects on HbA1c yielded from the HIT trials were a mixed product of both HITs and standard diabetes care including medication adherence and lifestyle modifications. This concern was addressed in the systematic review of Yoshida et al. [33]. The authors conducted a subset analysis of 18 studies that exclusively compared the outcome between a combined HITs and standard care intervention group vs. standard care control group. The effect size estimated from this analysis was −0.63 (Hedges' g: −0.63 95% CI −0.84, −0.42), which is attributable to HIT tools in addition to the

usual care [33]. This result suggests that HITs are the key to the effectiveness rather than tools or components of these trials. Additionally, pharmacotherapies often use motivated patients' sample and they cannot generate their full effects without patients' adherence to treatment and persistence in usage [33]. In this sense, HITs may add additional value in the effectiveness by addressing challenges in adherence of a pharmacological therapy or of behavioral interventions.
