Influences on Technology Use and Efficacy in Type 1 Diabetes

Abstract

The majority of people with type 1 diabetes have suboptimal glycemic control, increasing their complication risk. Technology to support diabetes self-care has advanced significantly and includes self-monitoring of blood glucose (SMBG), insulin pump therapy (IPT), continuous glucose monitoring (CGM), and sensor-augmented pump therapy (SAPT), which are stepping stones toward the “artificial pancreas” using closed-loop technology. Use of these technologies improves clinical outcomes for patients with the appropriate skills and motivation. This review addresses the psychosocial factors that influence both technology provision and clinical outcome and also how technology impacts on psychological outcomes. Optimal use of the various diabetes self-management technologies is influenced by previous self-care behaviors, demographic and psychological factors. Provision of IPT is also influenced by the same factors. Despite technology increasing the complexity of treatment, the lack of evidence for adverse psychological outcomes is reassuring. Treatment satisfaction is high, and discontinuation rates are low. However, technology will widen the health inequality gap if its use is limited to motivated patients who demonstrate good self-care behaviors. Pivotal to the success of the various technologies is provision of appropriate education at initiation of the technology, regular ongoing contact for treatment adjustments and trouble-shooting device issues plus access to psychological support when required. Additional support strategies may be required to help patients struggling with their diabetes to benefit from the available technology, recognizing that they may have most to gain.

Since 1921 technology has existed to replace insulin and treat diabetes. Although a definitive link was established between HbA1c and complication risk over 20 years ago,¹ only a minority achieve HbA1c targets despite intensified insulin regimens and new technologies.²

This narrative review focuses on how psychosocial functioning affects uptake and use of technologies and the impact of technology on psychosocial outcomes in adults and children with type 1 diabetes and their caregivers. It examines evidence from large meta-analyses to small qualitative studies which provide detailed insight into technology use.

Self Monitoring of Blood Glucose

Frequent blood glucose testing should be a routine part of diabetes self-care^3,4 as it is integral to intensive insulin therapy¹ and improves glycemic control.⁵ However, studies of adult patients reveal low testing frequency,^6,7 which is influenced by psychosocial factors, with increased testing in patients of older age, female gender, longer diabetes duration, on basal bolus therapy compared with less intensive regimens, with decreased hypoglycemic awareness, higher “fear of hypoglycemia” and positive self-care personality traits⁶ and less testing in ethnic minorities, smokers, those with low education,⁸ low income and language barriers.⁷ In adolescents, testing is more frequent in younger age groups, higher social status, and higher self-efficacy and less frequent in those with lower self-esteem, high-stress life events, lower parental support,^9,10 from single parent, lower-income, and ethnic minorities, and when given self-care autonomy before reaching cognitive and social maturity.¹¹ In children, family functioning¹² and parental education¹³ predict adherence to diabetes self-management and glycemic control.

Qualitative studies in adults show that some perceive self-monitoring of blood glucose (SMBG) as a tool to achieve good control and a normal life, but others see SMBG as a burden, citing practical difficulties, testing pain, and frustration with high results.¹⁴ A 5-step bio-psychosocial model of SMBG facilitates understanding of testing barriers:¹⁵ Step 1 is the decision to test: Some patients having extreme coping styles, with “blunters” who skip testing to avoid getting poor results and “monitors,” who test excessively, driven by fear and anxiety. Step 2 is the actual test: Barriers include pain and needle phobia. Step 3 is the interpretation of the results: Negative feedback from high readings leads to anger, frustration, and hopelessness. Step 4 is to act on the readings. Step 5 involves appraisal of the experience: Negative self-appraisal can decrease testing.

Despite being integral to diabetes care, SMBG is frequently not performed as recommend and is influenced by psychosocial variables. Recognizing and addressing barriers to SMBG will help support patients to achieve optimal diabetes outcomes.

Blood Glucose Meters With Automated Bolus Calculators

Intensive insulin therapy with carbohydrate counting requires insulin dose calculations based on preprandial blood glucose testings (BGTs), carbohydrate counting, variable insulin to carbohydrate ratios, insulin sensitivity, target BG levels, and active insulin. Such calculations are time-consuming and complex, particularly for patients with poor numeracy,¹⁶ increasing the self-management burden. Automated bolus calculators (ABCs) programmable with patient specific parameters provide insulin dose advice and improve carbohydrate counting competency,¹⁷ testing frequency,¹⁸ and glycemic control^17,19 in children and adults.

ABCs improve lifestyle flexibility, reduce fear of hypoglycemia,^17,18 improve confidence in bolus calculations and treatment satisfaction in children and adults.^17,18,20 ABCs are valued by adults with poor numeracy and when concentration is affected by hypo/hyperglycemia.²¹ Adult patients initially report double-checking the ABC dose advice, but this wanes as trust in the device increases.²¹ However, reliance on ABCs can have negative consequences if patients become deskilled, forget their ratios, stop recording data, and fail to recognize when changes are required.²¹

Bolus calculators improve outcomes and reduce self-care demands. However, to ensure ongoing benefit, patients need support to regularly review their data and make appropriate adjustments.

Pump Therapy

Meta-analyses show that insulin pump therapy (IPT) in adults and children improves glycemic control, without increasing severe hypoglycemia.^22-25 IPT in children improves complex cognitive task performance,²⁶ behavior, and mood.^26,27

Early studies of IPT reported increased diabetic ketoacidosis:²⁴ As pumps infuse only rapid acting insulin, any supply interruption increases blood glucose and ketones. Over 95% of pump problems are thought to be due to patient error,²⁸ therefore robust structured education is required at pump initiation, to minimize problems with IPT.^28,29

IPT is complex, demanding, and involves being permanently connected to a medical device, so could affect psychosocial functioning.^24,30 Two meta-analyses of pump versus basal bolus therapy (BBT) established that few studies assess psychosocial functioning, and different assessment tools precluded comparison between studies.^24,25 However, in all age groups there is no evidence of psychological harm and some benefit to pump therapy: Adults have improved or similar levels of depression, anxiety, family functioning, self-efficacy, and quality of life (QOL),^24,25,31,32 adolescents have no change in QOL, depressive symptoms or locus of control²⁴ and parent carers have reduced stress and fear of hypoglycemia.³³

Many studies report high levels of satisfaction amongst pump users and their caregivers particularly for the flexibility it confers.^34-38 Pump patients also report a more collaborative relationship with their health care professional (HCP), which in combination with the pump facilitates better diabetes control.³⁴ However, pumps can create additional work and obligations for families because of the potential to do frequent correction doses and set temporary basal rates.³²

Patient self-care behaviors predict glycemic control on pumps: Higher SMBG and bolus frequency, bolus calculator and temporary basal rate use, and lower pre-pump HbA1c are associated with improved control.^39,40 The provision of pump therapy shows disparity, being strongly influenced by race, socioeconomic status, family functioning, and diabetes self-care behaviors.^41-43

However a meta-analysis identified that patients with the poorest control on MDI had the greatest reduction in HbA1c and therefore may have the most to gain from IPT.²³

Psychological variables also influence outcomes: Adult patients with a high internal locus of control feel responsible for the success of IPT, have greater treatment satisfaction and better HbA1c values than patients with a high external locus of control who ascribe outcomes to luck.⁴⁴ Qualitative research supports these findings: “Responders,” patients with good control report active involvement with their pump, realize it is simply a tool that requires regular attention and are cognizant of its limitations. Conversely patients with poor control tend to have more negative perceptions about pumps, take a passive approach, adhere poorly to pump self-management tasks, and have unrealistic expectations of IPT leading to discouragement and frustration.³⁸

Pumps offer flexibility and the potential of a more normal lifestyle, but are a constant reminder of diabetes and could affect body image. Body image concerns are more pronounced in women, who report self-consciousness from wearing their pump and the inevitable questions that arise.^35,38 Adolescent girls and women report practical challenges of wearing the pump unobtrusively,^35,38 which may explain higher rates of pump discontinuation amongst adolescent girls.⁴⁵ Social acceptance seems less of an issue for men, who report that widespread use of technological devices makes them more at ease with their pump.³⁸

Pump discontinuation is infrequent at about 4%,^25,45,46 probably reflecting high patient satisfaction and patient selection. Psychological reasons stated for stopping IPT include a greater sense of illness, worse mood, shame, and fear.⁴⁶

IPT is a useful tool in diabetes self-care and can improve QOL and glycemic control, but provision and outcomes are dependent on patients’ self-care behaviors and sociodemographic factors. Strategies are therefore needed to improve IPT access and outcomes for all.

Continuous Glucose Monitoring and Sensor-Augmented Pump Therapy

Continuous glucose monitoring (CGM) sensors measure interstitial glucose every few minutes, and have evolved from systems that give retrospective data when removed and downloaded, to sensor-augmented pump therapy (SAPT) with pumps that display real-time data and alarm with hypo- and hyperglycemia. Pumps are now available with low glucose and predictive suspend functions that temporarily stop insulin infusion when glucose levels drop, to reduce the frequency and duration of hypoglycemia.

SAPT is associated with improved HbA1c without increasing hypoglycemia in patients with T1D when used consistently (≥6 days per week).^47-52 However, young adults and children are less likely to use sensors consistently, so do not attain the same glycemic control benefits as adults.^49,52 CGM use is higher in patients who do frequent SMBG⁵³ and those motivated to achieve good control,⁵⁴ and lower in patients with a high baseline HbA1c.^55,56 CGM users report high satisfaction overall,^55,57-61 but satisfaction is higher in frequent users ( days per week) and lower in adolescents.⁵⁵

CGM users can be categorized into “responders” who achieved an HbA1c reduction of 0.5%/5 mmol and “nonresponders” who do not.⁶² Qualitative research identified that nonresponders tended to have an emotions-based coping style and get frustrated with CGM. Conversely responders reported a self-controlled coping style, could problem-solve issues, and review data to identify patterns.⁶² Responders were also more likely than nonresponders to report their partner was interested in their CGM.⁶²

Several studies reported no QOL differences with SAPT,^57,61,63 but others found that SAPT improved diabetes-specific HRQOL (Hypoglycaemia Fear Survey Worry and Behaviour subscales)^58,59 and health and social functioning.⁶⁰ This is reassuring as there have been concerns that CGM data overload could have negative effects on QOL.^63,64 However its impact varies with age group, with trait anxiety higher in adolescents and lower in adults using CGM, compared with SMBG alone,⁶⁵ suggesting that CGM data may feel overwhelming in adolescents but empowering in adults.

Fear of nocturnal hypoglycemia is common, leading to regular overnight testing.⁶⁶ Parents may chose SAPT in hope of reducing sleep disturbance, but sleep may actually worsen if they wake frequently to check displayed glucose readings to see if adjustments are required.⁶⁷ SAPs with a low glucose suspend (LGS) function reduce hypoglycemia, without increasing mean glucose levels or risk of hyperglycemia^31,68,69 or diabetic ketoacidosis.⁶⁹ LGS is associated with high levels of parental satisfaction and makes parents feel their child is safer overnight.⁶⁹

The main barriers limiting CGM use are intrusive and irrelevant alarms,^54,62 skin irritation and sensor adherence^62,70 and body image concerns by making patients feel self-conscious and “robotic.”^54,55,62

CGM improves glycemic control when used consistently, but this can be difficult to achieve. CGM use and outcome is predicted by age, self-care behaviors and personality traits. Strategies are therefore needed to help patients optimize CGM use.

Artificial Pancreas—Closed-Loop Glucose Control

An “artificial pancreas” or closed-loop system contains a computer algorithm that continuously adjusts insulin delivery based on CGM readings. Closed-loop technology is evolving, from nocturnal closed-loop, to hybrid systems requiring some patient intervention, until ultimately the system is fully automated.⁷¹ True closed-loop systems could be in clinical use in the next decade.⁷² Patients perceive that closed-loop has the potential to optimize glucose control, reduce SMBG and disease burden and increase QOL.⁷³

Fear of nocturnal hypoglycemia can be a barrier to good control. Many patients or carers do routine overnight testing leading to disturbed sleep, anxiety, exhaustion and poor cognitive functioning.⁶⁶ Overnight closed-loop use reduces nocturnal hypoglycemia^74,75 and increases time spent with BGTs in target.^74-77 Patients and parents in overnight closed-loop studies reported the technology provided reassurance, that they trusted the device and experienced improved control. Although technical problems occurred (calibration, alarms and discomfort), patients reported that the psychological and physical benefits outweighed the difficulties. Importantly, patients using the system reported feeling “normal” rather than “diabetic.”⁷⁸

Great advances have been made with closed-loop at home studies underway,^72,76 and results of safety, clinical and psychological outcomes for patients and their carers are eagerly anticipated.

E-health Technologies to Support Diabetes Self-Management

Diabetes self-management is complex, requiring life-long behavior modification to optimize glycemic control and reduce complications. Behavioral support interventions typically require significant patient commitment and HCP resources, thus rarely become incorporated into routine clinical practice, limiting their impact.⁷⁹ In comparison, E-health technologies have the potential to deliver low-cost, convenient, tailored support to empower diabetes self-management^80-82 using the internet, mobile phone, decision support tools or telemedicine. Systematic reviews of a range of e-health interventions,^83,84 telemedicine,⁸⁵ and mobile phones⁸⁶ report improvements in clinical outcome, self-care and satisfaction.

A meta-analysis of the internet to promote diabetes health behavior change showed effect on behavior was small overall, but interventions incorporating behavior change theory showed larger effect sizes.⁸⁷

End-user satisfaction is evaluated in <50% of studies.⁸⁴ Post study technology adoption is rarely reported (≈4% of studies) and in those that do, it is only about 60%.⁸⁴ The Unified Theory of Acceptance and Use of Technology (UTAUT) identifies key factors affecting uptake of technology: perceived ease of use; perceived usefulness and perceived expectations of others. Therefore if technical difficulties occur or patients already have good control, adoption will be low. Conversely, IT interventions that are personalized, are multifaceted, promote self-care, and are integrated into existing infrastructure have greater success.⁸⁴

Addressing Barriers to Technology Use and Strategies to Optimize Outcomes

Currently availably diabetes technologies are not a panacea, but simply tools to support diabetes self-management. To achieve optimal control patients must have motivation to perform high levels of self-care including frequent BGTs and boluses, interpret data, make timely appropriate adjustments to therapy and in IPT, utilize the advanced pump functions.

SMBG is the simplest of the diabetes self-management tools, widely available yet performed as recommended by a minority. Patients need to be empowered to perform SMBG and use their results to optimize control, therefore addressing barriers to self-testing is a priority.^8,15 Questionnaires that assess fear of testing can be used to identify patients who may benefit from psychological therapy.⁸⁸ Problem solving techniques¹⁵ and motivational interviewing are useful strategies to increase adherence with SMBG.⁸⁹ Minimizing pain with alternative lancing devices,⁹⁰ alternative site testing,^91,92 or flash glucose monitoring systems,⁹³ may encourage testing. Encouraging patients to view results as “high” rather than “bad” increases testing frequency and improves HbA1c in adults and young people.⁹⁴ Paper-based diaries are useful but require patient commitment, and data interpretation is time-consuming for the HCP.⁹⁵ Technology can provide solutions for this: A systematic review of use of mobile phones to transmit data and send patients automated text/graphical or HCP feedback showed improved glycemic control in most studies, even those without HCP feedback, suggesting that this technology could be used to support self-management.⁹⁶ Bolus calculators also increase patients’ ability to use their SMBG results by providing “actionable advice”⁹⁷ empowering self-management.⁹⁸

To make optimal use of the available technologies, HCPs must ensure patients have realistic expectations³⁴ and reduce frustrations by addressing technological limitations, such as the lag-time between blood and interstitial glucose levels in CGM and providing coping strategies for managing alarms.^62,70,99

It must be recognized that the available technologies increase patients’ need for HCP support and education: Success with ABCs relies on blood glucose pattern recognition and making appropriate adjustments to the calculator settings when required. Similarly patients using pumps and CGM need a structured approach to interpret large volumes of data and make appropriate adjustments^21,62,70 and ongoing support to trouble-shoot device issues and sustain optimal use over time.^70,99

The critical role of HCP support was evident at the end of the DCCT study, when glycemic control deteriorated in the intensive group, probably because of decreased HCP contact.¹⁰⁰ Likewise, success of IPT versus BBT in randomized studies may be due to increased quality and quantity of education received by the pump patients.¹⁰¹ Outcomes of the REPOSE study, which provided equal structured education to adult patients randomized to both the pump and MDI arms, will be published soon and will provide insight into the role of education.¹⁰²

As patients with the highest HbA1c may benefit most from pump therapy,²³ it is important they are not excluded for being unsuitable. Otherwise diabetes technologies could continue to widen the health inequality gap, as race, socioeconomic status, family functioning, and diabetes self-care behaviors predict provision of pump therapy.^41-43 The REPOSE study included a qualitative evaluation of HCPs’ perceptions of which patients benefit most from IPT.²¹ HCPs reported they had previously selected patients with the personal and psychological attributes they felt were suited to pump therapy, but results of the REPOSE study challenged this, as they recognized that patients previously thought unsuitable for pump therapy had good outcomes and for some was a “tipping point” for increased engagement.²¹ It may therefore be appropriate to broaden patient selection criteria for IPT, and develop evidence-based interventions to help patients struggling with their diabetes⁴³ or with literacy and numeracy issues,¹⁰³ to benefit from IPT. Technology combined with appropriate psychological support may be a powerful motivator of health behavior change.²³ However although psychological support is a recognized standard of diabetes care,¹⁰⁴ it is not available to all^105,106 so there is also an imperative for all diabetes HCPs to become behavioral experts.⁸

Greater parental involvement is associated with better glycemic control in children.^12,107,108 Encouraging parents to remain involved in SMBG and interpretation and reducing family conflict during adolescence may prevent deterioration of glycemic control common in this age group.^12,109 Diabetes technologies could inhibit parental involvement by exposing the “digital gap” between young people who have grown up in the digital generation and their parents who find technology less intuitive to use.¹¹⁰ Good parental knowledge of pump functions is related to better outcomes, whereas parents less skilled with pumps secondary to the “digital gap,” are more likely to transfer responsibility for diabetes self-management to their children, with worse glycemic outcomes.¹¹¹ HCPs should therefore encourage parents’ pump competency to ensure they can optimally support their children. Similarly, as partner understanding of CGM predicts better outcomes, HCPs should invite partners to educational programmes.⁶²

Ongoing research and development will help address barriers to device use. For example an integrated pump cannula with CGM may improve sensor use and glucose control.¹¹² Including usability and intuitiveness evaluations during device development is paramount, as unsurprisingly patients report preferences for pumps and sensors they find easier to use.^113,114 Remote monitoring systems for CGM and IPTs that wirelessly transmit data to a monitor in the parents’ room addresses parental fear of nocturnal hypoglycemia that is often a barrier to optimal overnight control.¹¹⁵ Closed-loop technology is close to a clinical reality and will reduce the human behavioral elements that are a barrier to optimal diabetes management,²⁸ and is eagerly anticipated for its potential to improve glycemic control and QOL whilst reducing disease burden.^73,78

IT interventions have the potential to deliver low-cost accessible self-management support, but to be successful they should be theoretically grounded and deliver a comprehensive intervention. Incorporating user-centered sociotechnical design⁸⁴ will help create applications that fit with patients needs.⁸⁵ Their effectiveness should be evaluated in terms of motivation, adherence, cost, adoption, satisfaction and clinical outcomes.⁸⁴ IT applications incorporating HCP feedback are an important motivation for self-care,⁸⁴ so research should focus on the impact of different feedback types (including automated electronic feedback) on self-care and outcomes.⁸⁵ Developing interventions for tablets and mobile phones may increase participant uptake and reduce attrition, particularly in adolescents⁸³ and low-income countries. Likewise use of gaming technology^116,117 and mobile phone apps provide formats that may increase patient engagement. However, a review of apps showed only a few improved glycemic outcomes compared with controls,¹¹⁸ and had been tested for safety and efficacy or obtained regulatory approval, which is concerning if they are used as decision support tools.¹¹⁹ Developing guidelines and legislation for E-health would protect HCPs and patients alike.⁸³

Conclusion

Closed-loop technology may overcome many of the barriers to optimal clinical outcomes. However, until it is routinely available there is a need for evidence based interventions and further studies which include representative patients with intervention and control groups that receive equal education and support, to address applicability to normal clinic populations. Studies of emerging technologies should also incorporate standardized psychosocial measures to evaluate QOL and treatment satisfaction in addition to glycemic control.

Our common goal must be to devise and deliver diabetes self-management support so that all patients have access to technology and can optimize its use to improve their glycemic control and QOL, without increasing burden of treatment.