MyDiabetesMyWay

Abstract

MyDiabetesMyWay (MDMW) is an award-wining national electronic personal health record and self-management platform for diabetes patients in Scotland. This platform links multiple national institutional and patient-recorded data sources to provide a unique resource for patient care and self-management. This review considers the current evidence for online interventions in diabetes and discusses these in the context of current and ongoing developments for MDMW. Evaluation of MDMW through patient reported outcomes demonstrates a positive impact on self-management. User feedback has highlighted barriers to uptake and has guided platform evolution from an education resource website to an electronic personal health record now encompassing remote monitoring, communication tools and personalized education links. Challenges in delivering digital interventions for long-term conditions include integration of data between institutional and personal recorded sources to perform big data analytics and facilitating technology use in those with disabilities, low digital literacy, low socioeconomic status and in minority groups. The potential for technology supported health improvement is great, but awareness and adoption by health workers and patients remains a significant barrier.

The use of online interventions for diabetes care is increasing, alongside a political drive toward e-health initiatives,¹ despite limited robust evidence relating to clinical outcomes. E-health is a term coined in the late 1990s to encompass convergence of the Internet and health care. The definition is constantly changing but generally covers user interaction with health data through online systems, cross transfer of data (often between institutions or systems) and user to user communication. Traditional users of e-health are health institution staff, but a new wave of e-health initiatives directed at patients brings exciting possibilities.²

More cost-effective diabetes management is needed, and technology could be key to care delivery and diabetes prevention. Around 8-9% of the global population has diabetes.³ Health expenditure for diabetes is increasing dramatically. The total health care costs of a person with diabetes in the United States are 2 to 3 times those for people without the condition.⁴ In the United Kingdom, annual spending on diabetes is expected to reach 17% of the entire NHS budget over the next 20 years.⁵

Around 40% of the world population is now Internet-connected⁶ rising to around 84% in higher economic countries, like the United Kingdom (2015).⁷ Mobile phone and tablet use in the United Kingdom has increased from 30% in 2013 to 41% in 2014⁷ associated with increased use of apps, remote monitors, fitness devices, and environmental sensors.^8,9 Health care system data quantity is also increasing exponentially, although it often remains in silos and its true potential thus never realized. Good data linkage and timely data availability to frontline health staff have the potential to transform care delivery.^10-12 This vision is highlighted in key political reports^13,14 making data and technology use in diabetes care a key priority. As informatics systems mature in their ability to connect and transform data, their outputs can become increasingly personalized, driving individualized support, advice and real-time decision making.

This will become more apparent with the advent of the “Internet of Things”¹⁵ opening up further linkages to community devices containing sensors and electronics. This forthcoming tsunami of health related data will move into the realm of genuine “big data,” requiring “data science” approaches to extract meaning and useful knowledge.¹⁶ Artificial intelligence techniques will increasingly be used to predict future events from data patterns, allowing early intervention. This level of informatics maturity could have particular benefits for self-management and complications prevention in long-term conditions.^17,18

Aims

In this article, we discuss

1. the evidence for e-interventions in the management of diabetes

2. the development and evaluation of MyDiabetesMyWay (MDMW), Scotland’s national electronic personal health record (ePHR) and self-management platform for people with diabetes

Background Literature: Diabetes Web-Based Interventions

Web based interventions for people with diabetes are further categorized here, however in reality many interventions are complex utilizing multiple technology approaches:

• m-health (abbreviation for mobile health) has increased the reach of e-health and describes use of mobile and wireless technologies to support the achievement of health objectives.¹⁹ M-health uptake is largely dictated by mobile phone and wireless signal availability. In diabetes care, interventions range from simple SMS texting to more sophisticated app based products.

• Social media (computer-mediated tools that facilitate the sharing of information) have transformed the way people live, work, shop, and socialize and represent an emerging aspect of e-health.²⁰ There is a very active online diabetes social media community utilizing global and local networks.²¹

• Telemedicine utilizes remote video consultation/monitoring increasingly provided through software based tools

• Remote glucose monitoring utilizes data sharing from Internet-linked blood glucose meters with or without external feedback.

• ePHRs are data repositories used to securely store and manage personal medical information for tracking and monitoring purposes. ePHRs come in 3 main categories:

  • Stand-alone: unconnected to any external system, containing solely patient-recorded health information
  • Tethered: a view into an electronic medical record or a subset of its data, in some cases containing functionality to request medication and appointments and to communicate with physicians²²
  • Interconnected/integrated: collecting data from multiple data sources while incorporating bi-directional communications and functionality (such as MyDiabetesMyWay).²³

Evaluation of Internet Tools

Systematic reviews and meta-analyses of e-interventions in the management of diabetes have demonstrated acceptability and improvement in glycemic control of participants, although the effects are often modest and variable across studies.^24-27 These reviews contain heterogeneous interventions, which are often outdated by the time of publication, but suggest that goal setting, personalized coaching, interactive feedback, mobile use, and online peer support may predict success when applied to management of type 2 diabetes mellitus. Evaluation of different categories of e-interventions is discussed below.

m-health

Self-management interventions delivered through mobile technology tend to be more successful than browser based systems.^24,26 Systematic reviews suggest m-health interventions can improve HbA1c,²⁸ although impact is variable.²⁹ A systematic review (2014) of available diabetes applications, of which 656 were reviewed in full, showed apps to be feasible with moderate to good usability in patients over age 50.³⁰ Apps can support self-management of patients with diabetes mellitus type 1 or 2^31,32 and aid glycemic improvements.³³

Social Media

Gilbert et al found that online communities are a valued resource by type 1 diabetes patients for peer support.³⁴ Story telling, sharing of experiences and tips for self-management are common themes. While health care staff are often concerned about sharing of personal data online, patients are generally happy to share with peers for mutual benefit.³² Little research has been conducted evaluating impact on glycemic outcomes.^21,31,35

Telemedicine/Virtual Clinics

Freeman et al³⁶ found low-cost online video services such as Skype, to be feasible for providing online virtual clinics,³⁷ allowing more flexible communication, while maintaining the patient-provider relationship, a strong predictor of medical adherence.³⁶ Since January 2011, clinics in Newham (England, UK) have been trialing Skype consultations.³⁸ Early indications reported reduced HbA1c levels, uptake across a spectrum of ages and ethnic groups, fewer accident and emergency attendances and modest efficiency savings. Those in employment or education or with carer responsibilities reporting greatest benefits.³⁸ Robust quantitative clinical outcome evidence is, to date, lacking.

Remote Glucose Monitoring

Improved clinical outcomes (including reductions in HbA1c, LDL cholesterol, and triglyceride levels) have been demonstrated in studies delivering remote health care recommendations based on daily or weekly³¹ or biweekly³⁹ online review of uploaded blood glucose readings.²³ Furthermore, there is evidence that such interventions may be feasible even among low-income diabetic patients with a high risk for poor diabetes outcomes.³⁹ A meta-analysis revealed that using self-monitoring blood glucose data, stored in and/or shared through a personal data assistant, logbook, the Internet and other technologies, worked well alongside consistent feedback from a health care professional. This resulted in enhanced glycemic improvements and reduced hospitalizations.⁴⁰

Electronic Personal Health Record

A Nuffield Trust report⁴¹ summarized a debate from key stakeholders involved in implementation of early ePHR systems, stating “ePHRs have the potential to improve communication between providers and patients by sharing information, to enhance the quality of records by highlighting inaccuracies, and to reduce the burden of care by engaging patients in managing their own health and illness.” A recent systematic review of electronic patient record access reported improved patient satisfaction, improved self-care and better communication and engagement with clinicians but did not evaluate clinical outcomes. “ePHR services positively impacted on patient safety, although there were variations in record access and use by specific ethnic and socioeconomic groups. Professional concerns about privacy were unrealized and those about workload were only partly so.”⁴²

MyDiabetesMyWay: A National Online Self-Management Platform

Against this background of emerging online interventions, MyDiabetesMyWay (MDMW) was developed as a national self-management portal for diabetes patients in Scotland. It a now encompasses (1) an education resource website, (2) an ePHR for diabetes patients, (3) patient decision support (with data-driven information tailoring), (4) goal-setting functions, (5) social media including user peer-peer discussion groups, and (6) a linked remote glucose monitoring system, and it can be accessed on mobile devices.

MDMW was launched in October 2008 as an open access web site with patient educational materials, which were, and continue to be, curated, digitized, and updated by the MDMW team. These include traditional information leaflets, interactive educational tools, videos, animations. Some of these resources originated from the Scottish diabetes community (eg, Scottish Diabetes Group/ NHS trusts across Scotland), while others have been commissioned.

Since December 2010, an ePHR has been available, allowing patients access to their own clinical information through a secure login (via the Scottish government “my account” citizen portal).⁴³ Data for MDMW is extracted from SCI-Diabetes, a national disease registry for Scotland, functioning as a comprehensive national diabetes electronic health record. It collates information from multiple sources including laboratories, primary care records, pharmacy/prescribing sources, screening services, and secondary care information systems. Data derived from SCI-Diabetes are used for regular regional and national audit and epidemiological research.⁴⁴

Patients use MDMW to access information about their condition including weight, blood pressure, blood results (HbA1c, creatinine, eGFR, cholesterol), eye and foot screening results, medication, clinical letters, and appointment records. Information is displayed with patient-friendly data visualizations and explanations (see Figures 1-4), for example:

Figure 1.

A pop-up screen grab from the MDMW platform offering an explanation for measurement terms, in this case body mass index (BMI).

Figure 2.

A screen shot from the MDMW platform (fictitious patient) mapping an individuals care parameters to a national standard (the Diabetes UK 15 care measures). This can be printed as a PDF for discussion in a clinic consultation.

Figure 3.

A screen shot from the MDMW platform (fictitious patient) demonstrating HbA1c displayed in graphical format showing change over time and includes target ranges.

Figure 4.

A screen shot from the MDMW platform (fictitious patient) demonstrating data (HbA1C, cholesterol, blood pressure and BMI) displayed in a target chart format.

• Data item definitions and explanation of normal ranges

• Target charts and graphs for HbA1c, cholesterol, blood pressure (BP), body mass index (BMI)

• Care standard visualizations to map personal care against national targets, such as Diabetes UK 15 Health Care Essentials⁴⁵

Patients can contribute to their care record through the addition of data such as weight, smoking status, blood pressure and upload blood glucose readings (either manually or through integration with a third-party product DIASEND).⁴⁶

Personalized Tailored Education

Individual patient data drive tailored links to specific educational resources to aid personal self-management. Examples include foot care information tailored to a person’s foot risk, drug administration guidance dependent on personal prescribing records, and links to specific sections of a retinopathy educational video based on individual retinal screening results. In addition patients can view tailored lifestyle advice dependent on dynamic changes in anthropometric and biochemical readings.

MyDiabetesMyWay: Use, Effectiveness, and Future Developments

Usage

From January to September 2015, the main MDMW website has averaged 60 792 page accesses per month, with a peak of 69 841 in July 2015. Up to the end of September 2015, 15 877 people have registered with MDMW electronic record access system and 6482 people with diabetes have accessed their records (Figure 5). This represents a 108% increase since September 2014 (n = 3119) and a 296% increase since September 2013 (n = 1636).

Figure 5.

Ongoing increase in active users of MDMW over time, which still represent only a small percentage of the overall Scottish diabetes population of 276 430 people.²

Usage analysis (data from all MDMW registrants at the end of 2014) demonstrates that those registering to access MDMW are generally younger and more recently diagnosed. 35.7% have type 1 diabetes, compared to the NHS Scotland prevalence of 10.9%. A high proportion of registrants have a reported ethnicity of white and its subcategories. A significant proportion of registrants come from least deprived areas, whereas most of the background population come from the most deprived areas. The “test results” screen was the most popular summary section of the website. The most utilized longitudinal history graph was, unsurprisingly, HbA1c.⁴⁷

MDMW Patient Reported Outcomes and Feedback

Regular feedback is received from users through targeted qualitative and quantitative evaluation questionnaires. These cover patient experiences, usability issues, reported benefits and concerns, and invite suggestions for future developments. Anecdotal feedback received through the website and social media has demonstrated high levels of satisfaction. Examples include:

• “I wish there were systems like this for other conditions as it would save me having to see doctors regularly.”

• “I can see when my diabetes was poor and when it improved, a constant battle.”

• “Immensely satisfied with the system. Really amazing to see my results online. Real motivator.”

• “I don’t have to make appointments to see GP or nurse which is good as I work and don’t like taking time off work.”

Key findings from a recent evaluation survey of 1098 MDMW users who had logged in at least once prior to January 2015 (27.5% response rate) include that 89.6% felt it helped them make better use of their consultation, 88.2% felt it helped them manage their diabetes better, and 90.3% and 89.3% respectively felt it improved their knowledge and motivation.

Clinical Improvements and Cost Effectiveness

Unpublished observational data (from users up to end September 2015 with complete data) suggest modest improvements in most clinical parameters including HbA1C (n = 2611), weight (n = 2297), diastolic BP (n = 2552), and cholesterol (n = 2026) (all P < .01) 1 year after first login for the MDMW platform. However confounders, including lifestyle and metabolic changes around time of diagnosis, which may coincide with platform registration, need to be taken into account. More detailed analysis is planned.

Costs to run the MDMW platform are very low. Costs for 2015 are around £5.40 per registrant per year, and could drop under 85p per person per year if 50% of the Scottish diabetes population enrolled. Full cost effectiveness analysis has not been carried out, but considering the low running costs is expected to be positive.

Funding and Recognition

The funding for this project is predominantly from the Scottish Government and the platform supports key national strategies improving information availability, supporting health and wellbeing and the needs of people with long term conditions.⁴⁸ MDMW has received many quality and innovation awards including the 2015 Diabetes UK Annual Self Care Award Winner, and UK Quality in Care Diabetes 2013 award for best initiative supporting patient care

Barriers to Uptake

Despite positive feedback, uptake and usage remain very low, with only 5.7% of patients with diabetes in Scotland having registered (end Sept 2015). A survey of 175 diabetes patients across primary and secondary care (September 2015) (106 paper surveys [48% response] completed in clinic and 69 online surveys [25% response] completed by a cohort registered with the Scottish Diabetes Research Network [SDRN] registry [email invitation]) was performed. Barriers to using MDMW included lack of awareness (48.9%), difficulties using computers (28.0%), not owning a computer (24.7%), remembering passwords (16.6%), concerns about privacy (14.3%), no interest (13%), and having other priorities (11%). Difficulties with the registration process was also raised as a concern. Work by Ronda and colleagues reported similar difficulties with recruiting and retaining patients onto a diabetes patient web portal in the Netherlands.^49-51

User Guidance and Future Developments

In August 2015, preferences for features in a future mobile application was assessed through user questionnaires⁵² and readiness for further online communication tools was assessed in the MDMW patient user group (n = 206), and SDRN registry (n = 73), using online (MDMW) and paper-based (SDRN) questionnaires. In the latter group, which is more representative of the wider diabetes community, 50% would like to access their health records online, 47% would be willing to engage in online forums, and only 37% are willing to consider online consultations despite 74% having Internet access.⁵³ While there is support for digital tools among some diabetes patients, this is not uniform across the diabetes community.

Discussion

There is increasing evidence that digital interventions may have the potential to improve diabetes care. This evidence base, combined with user feedback has driven changes in the MDMW platform from a web based information website to an ePHR now encompassing a broader diabetes self-management platform with communication tools, remote monitoring, peer support and tailored self-management advice. While the platform is unique in providing a national data driven resource, there remains significant challenges and barriers. MDMW platform usage is low, with disproportionately lower use in older populations, lower socioeconomic groups, and minority ethnic groups, and there remains a risk that digital innovations may contribute to health inequalities. Despite a national advertising campaign and good national clinical networks, awareness of MDMW remains suboptimal. More proactive promotion by health care teams particularly in primary care may reap results. One GP was able to recruit 23.4% of their caseload within 2 weeks of writing to them individually.⁵⁴ Further developments such as increasing use of remote communication tools was welcomed by some but not all patients,⁵³ suggesting that traditional models of care are still likely to be required for the foreseeable future.

Proving clinical benefit and cost effectiveness for any digital intervention is difficult, but increasingly necessary for funding justification and wider adoption. While MDMW is a low-cost intervention and has extensive evidence of perceived benefit from user feedback, it has not yet demonstrated definitive clinical outcome improvements or cost-effectiveness.

MDMW Current and Future Developments

The MDMW platform evolution has been driven by patient preference and the wider published evidence base. Further planned functionality includes

• an associated m-health app

• enhanced automatic data collection and integration from environmental and activity sensors

• algorithm development and machine learning to enable timely tailored “push” decision support to mobile devices and timely warnings of acute deterioration to patients and clinicians

These latter developments would align with a high level of health informatics maturity with potential for personalized, real-time medical support systems. There is huge potential to deliver online self-management to empower and increase knowledge and self-management capabilities of patients.

The marketplace is flooded with stand-alone tools to support diabetes management ranging from mobile apps to remote monitoring and information prescription systems. Most do not integrate with institutional records, and work is required to break down barriers in support of big-data analytics. Computers are capable of delivering sophisticated data-driven advice; however, for many people with long-term conditions, health management competes with other priorities.

Widespread, low-cost web access is lessening the economic, geographic, and demographic barriers for use. Technology approaches however will work only if patients are willing, motivated, and able to use them. Technology must be intuitive, and include automaticity where possible, and must ease rather than add to daily care activities. Further understanding of user requirements and motivation is required. Unless technology is accessible across the range of socioeconomic, ethnic, disability, and age spectrums it will continue to be the pursuit of limited numbers of motivated patients, and will fail to greatly impact on population health outcomes.