Introduction

Diabetes is a widely prevalent long-term condition in which raised blood glucose levels over time contribute to complications that are costly in terms of healthcare use and impact adversely on quality of life. Blood glucose monitoring (BGM) has an important part to play in the management of blood glucose in diabetes and the reduction of risk of serious secondary complications [1]. The first BGM meter was introduced in 1970 [1], and available technology has evolved markedly since then. In today’s technology-rich environment, self-monitoring of blood glucose is an integral part of effective diabetes management [2], allowing patients to monitor their glycemic status and adjust their therapy accordingly.

Modern healthcare requires that long-term conditions are managed in a shared and collaborative way between the patient and all stakeholders involved in their care, rather than solely by the patient or healthcare professionals (HCPs). Despite clear demonstrations of the benefits of BGM meters that provide feedback to patients [2], and diabetes being almost universally managed in the UK by NHS professionals, there is an apparent lack of clarity within the NHS regarding emerging BGM technologies and their potential to positively impact service delivery and increase resource efficiency while supporting patient empowerment. This was demonstrated in a survey of over 1000 BGM users carried out by Diabetes UK [3], with the responses suggesting that NICE (National Institute for Health and Care Excellence) guidance is not always implemented as recommended for type 1 [4] and type 2 [5] diabetes.

This consensus steering group (all members are co-authors to this paper) has sought the real-life perspectives of NHS staff involved in diabetes care to gain an understanding of their attitudes to BGM technology and its value for people with diabetes who require intensive BG monitoring. This was felt necessary because the speed of introduction of technology was perceived to move more quickly even than centralized expert guidance. In order to fulfil the requirements for fully collaborative shared care, it is important that optimum technology is utilized and a ‘big data’ approach is adopted, in order to increase resource efficiency.

Methods

This article does not contain any studies with human participants or animals performed by any of the authors.

The group used the Delphi [6] methodology to validate the views of a large number of peers regarding BGM in the current UK National Health Service (NHS) setting, with the goal of better understanding their views regarding changing demographics and BGM technology. The recommendations derived from the responses seek to support HCPs in assessing new technologies available to them and provide decision-support for choice of technology. As such, these consensus views and recommendations which the group felt were justifiable from the responses are offered in order to provide a reference point for the use of BGM for both commissioners and providers. In addition, the consensus process aimed to identify regional variation in the responses to statements and is intended to inform the paradigm shift required to adopt new technologies for BGM.

A group of clinicians and payers involved in diabetes care met to review the current adoption of BGM systems in the UK and determine a framework for consensus. Seven key topics were initially identified, and consensus statements developed for each (Table 1). These statements were collated into a questionnaire, which was circulated widely over a 3-month period to clinicians involved in diabetes management by in-person cascade and by e-mail from group members and with the help of LifeScan, Inc. (London, UK) representatives. Responses were collated and analyzed in confidence by Triducive Ltd (St Albans, UK). Respondents were required to use a 4-point Likert scale to rate their agreement with each statement. Responses were: ‘strongly disagree’; ‘tend to disagree’; ‘tend to agree’; and ‘strongly agree.’ The questionnaire also asked respondents for their locality, their speciality, and the department in which they work. While personal details were not used for reporting results, clinical background and locality were used to assess potential differences in responses between professions and across the UK.

Table 1 Consensus topics and statements
Full size table

Completed questionnaires were then collated, and the individual scores for each statement analyzed in order to produce a full arithmetic agreement score for each statement. The responses were then sub-grouped by locality and specialty to identify variances in the respondent’s agreement scores by either geography or role.

The steering group had predefined the threshold of agreement for consensus at ≥ 66%. Consensus was defined as ‘high’ at ≥ 66% and ‘very high’ at ≥ 90%.

Due to the high levels of agreement with all but three of the statements, demonstrating very strong consensus views, the group elected to avoid a further round of questionnaires to generate a larger number of responses, and instead chose to work with the responses to the original statements.

Results

Completed questionnaires were returned by 222 respondents. The questionnaires were analyzed to score the total level of agreement with each of the 38 statements. Of the 38 statements, 35 (92%) achieved an agreement score exceeding the 66% threshold, and 26 of these (68%) exceeded 90% agreement (Table 2, Fig. 1).

Table 2 Consensus statements and agreement scores
Full size table
Fig. 1
figure 1

Plot of total consensus agreement scores

Full size image

Only three of the statements achieved an agreement score of < 66% or > 33% and thus were defined as not achieving consensus (Table 3).

Table 3 Statements with agreement scores of between 33 and 66%
Full size table

Respondents were analyzed by region. The number of respondents from each region is shown in Table 4. The sample was dominated by respondents from England (n = 136) while only four respondents were from Wales. Eighteen respondents did not indicate their region.

Table 4 Respondents by region
Full size table

While the majority of statements were scored similarly, irrespective of the respondent’s region (Fig. 2), nine statements showed larger regional variation in score (Table 5).

Fig. 2
figure 2

Consensus agreement scores by region

Full size image
Table 5 Statements with largest regional variation in agreement scores
Full size table

Responses were also analyzed according to the respondent’s role (Table 6).

Table 6 Respondents by role
Full size table

While the largest single respondent groups by role were diabetologists and diabetes specialist nurses (DSNs), 39 respondents did not share their role on the questionnaire and thus remain unknown and 19 had other roles. Comparison between the responses of diabetologists, DSNs and the group overall is shown in Fig. 3.

Fig. 3
figure 3

Consensus agreement scores by role. DSN Diabetes specialist nurse

Full size image

With the exception of statements 3, 4, 5, 16, 17, 27, 30, and 37, the views of DSNs were well aligned with those of the wider group. The variation in agreement scores for these statements according to role is shown in Table 7.

Table 7 Statements with largest variation in agreement scores by role
Full size table

Discussion

In this paper we describe the consensus views of a large group of clinicians working the UK NHS, a government-funded, cost-limited healthcare system that is substantially free at the point of use. As such, the recommendations and views presented here may not reflect the views of HCPs in other healthcare systems.

For patients with a need to optimize their glycemic control, many factors may influence their choice of BGM system. Understanding the value of a system to a patient may be a more important measure than simply looking at the acquisition cost of that device. In light of this, sharing the acquisition cost with the patient may be a valid option, as patients have demonstrated that when a device is recognized as offering important advantages, they may be willing to contribute to the cost [7].

Effective BGM is recognized by 99.1% of respondents as having the potential to reduce hypoglycemia and ketoacidosis in patients with type 1 diabetes. Effective BGM may be regarded as achieving tight glycemic control (according to the type of diabetes) in line with personalized goals.

Respondents strongly agree with the assertion that flash glucose monitoring has the potential to improve the use of NHS resources (88.9% agreement) and should be considered as an investment rather than a cost (96.3% agreement).

Influencing Guidance

Unsurprisingly, respondents strongly agree with statements 12 and 13 concerning the need for real-world evidence when making recommendations and the need for clarity in the evidence base from guidance used in decision-support. There is strong agreement (98.2%) endorsing the need for digital solutions, which may support the objective of increasing personal management of long-term conditions. In addition, 96.8% of respondents agree that patients should have access to real-time data enabling them to act.

In total, 98.2% of delegates support the view that guidance should support the value of digital solutions when it comes to formulary inclusion. However, BGM formulary decisions are still firmly focused on devices and the acquisition cost of the device test strips, despite the NHS’s intention to move to solutions that are orientated around digital, big data and evidence that demonstrates an impact on health outcomes.

Access to BGM Data

All statements relating to BGM data are very strongly supported by respondents. Having consistency and reproducibility of data is considered to be of paramount importance (99.1%), and supporting type 1 patient’s treatment decisions in real-time is strongly supported (98.2%).

The need for patients to have real-time reports of their personal BGM data is supported by 94.5% of respondents. It has been shown that automated pattern algorithms can assist in the avoidance of hypoglycemic episodes [8, 9]. it has been reported that 32.3% of measured hypoglycemic episodes are preceded by an observed pattern of low blood glucose [10]. Data should always support treatment decisions for people with type 1 diabetes or patients using insulin (96.8%).

Data offer patients the insight to manage their blood glucose with confidence. Respondents strongly agree that decisions made by patients should always be supported by data (98.2%). Access to data promotes engagement with trends and patterns within the data to inform the patient in a way that complements the role of the HCP.

Patient Education

There is evidence to show that good BGM results in better outcomes and that people should be supported and educated in the interpretation of data to achieve these better outcomes. In our survey 100% of respondents support the importance of a common understanding of how to interpret and act upon BGM data, trends, and patterns. Given the nature of diabetes, it is vital that carers are actively involved with patients, and access to real-time data is an important consideration. Diabetes impacts the whole household [11].

Among respondents, 99.5% agree that effective BGM is a necessary component of patient education; in addition, the importance of access to BGM data and of understanding trends and patterns is strongly supported (statements 23 and 24). Access to feedback on their progress helps patients to take action to improve their diabetes control or continue with previous positive actions [2]. There is strong agreement that virtual data and applications can support education at home. Remote support in the patient’s home and work environments is important so that patients are not constrained by the need to attend a clinic.

Statement 27 achieved consensus (66.8% agreement) despite the inference that resources could be saved by replacing face-to-face education. In reality, respondents may better support the assertion that the two should complement one another. This would offer a synergistic benefit. There is 97.2% agreement among respondents that BGM offers a means to offer positive reinforcement in support of patient decisions, reflecting recent published data [2, 12].

Data Integration

Respondents strongly support that the notion that data systems should be integrated and inter-operable across the NHS (97.2% agreement). Eventually, integration of data should involve a standardized approach in order to improve access and utility (statements 31 and 32). Data are of no use unless it can be effectively used and applied.

Members of the consensus group agreed that high-quality data are essential for constructive consultations; therefore, access to data will improve resource utilization and avoid wasted clinic sessions. There was 98.6% agreement among respondents that patients’ BGM data should be accessible by every HCP involved in their care. This result is supported in the literature, with published data showing improvement in glycemic control by patients using integrated BGM data [8, 12]. Cloud-held data may lead to better data-utility among patients and HCPs (98.1%) as long as it is easily accessible as per regulatory guidelines. It may also facilitate remote consultations and can alleviate the problem of forgotten logbooks or devices in face-to-face consultations.

Accessible cloud-held data can also be shared/socialized with caregivers (e.g., elderly patients) and loved ones (e.g., parents of children). Studies have demonstrated improvements in outcomes when data is used in this way [8].

Use of Resources

Respondents strongly agree that manufacturers should provide quality-adjusted life-year data to support their products (89.9%). There is a need for clear and consistent guidance regarding BGM, as evidenced by the lack of clarity from respondents regarding statement 17.

Almost all respondents (99.1%) agree that effective BGM will reduce the risks of hypoglycemia and ketoacidosis in patients with type 1 diabetes. New technology has the potential to improve the use of NHS resources and should be considered as an investment rather than a cost (96.3% agreement). New technologies that facilitate effective glycemic control are strongly supported by respondents (88.9% agreement to statement 18, which is specific to flash glucose monitoring and 96.3% agreement to statement 20). Solutions with automated pattern algorithm integration can improve HCP efficiency and accuracy [13].

Evidence and Outcomes

Respondents strongly indicate (92.7%) that evidence is important in improving patient care and that it should inform choices relating to BGM. Recent evidence [2] shows that switching patients to BGM systems featuring color range indicators was associated with improvements in glycated hemoglobin levels compared to subjects using a wide selection of marketed BGM systems without color support [14]. Despite this, respondents support the assertion (78.5%) that there is a lack of evidence that more expensive BGM systems improve outcomes, suggesting that respondents may not be fully aware of recent innovations and published data. Data are available providing direct comparison between solutions that help the patient and HCP with insights and interpretation of the number versus traditional BGM [14].

While the preference of the HCP is not regarded by the respondents as being more important than evidence (34.9%), patient preference is strongly supported (71.8%) as more being important than evidence alone in choosing the right BGM system for the patient. Respondents strongly agree that evidence should include patient-reported outcome measures (PROMS) (97.2%), such as real-world usability (93.7%).

Patient Choice

A majority of respondents concurred that formulary availability should not be driven by a single factor alone, with general agreement that neither patient choice (55.3%) nor acquisition cost (92.2%) is sufficient to effectively make formulary choices. Interestingly, respondents are accepting of patient co-payment (77%) as a means of accessing novel technology and supporting individualized patient choice. This is a paradigm shift and may reflect changing professional attitudes. However, many patients have traditionally self-funded their BGM regimen in the past, and the 2017 Diabetes UK Survey recently revealed that BGM is less available than patients’ requirements [3].

There are 98.2% of respondents who agree that the choice of BGM system should support the achievement of the patients’ own goals and 84.3% who agree that the willingness of the patient to engage with a particular BGM device should be the driving factor in choice of system. Recent data suggest that BGM supported by a web application and OneTouch Verio (OTV; LifeScan Inc., Milpitas, CA, USA) meter improved blood glucose control [12]. In addition, recent studies indicate that connected BGM solutions that provide access to cloud-held data allow patients to feel more secure in the knowledge that they can access their diabetes information at any time [15].

Overall, the strong support from a large respondent group underlines the need to ensure that patients have access to appropriate BGM systems in line with the needs and expectations of their health care professionals. New BGM technologies should be regarded as an investment rather than a cost, as long-term cost-effectiveness may be achieved due to a reduction in negative outcomes and admissions and in greater patient satisfaction. This opportunity reflects the wider shift towards value-based healthcare in the NHS.

Limitations of Study

The cascade circulation of the questionnaire from the group to their network of colleagues and the willingness of the 222 clinical professionals to respond spontaneously may have introduced bias; nevertheless, the statements produced a high level of clinical consensus in areas of uncertainty around BGM. The professionals who responded covered a wide range of disciplines and geographical spread within the UK, but despite this there was a great deal of agreement in their responses.

Conclusions

The conclusions drawn by the consensus group after consideration of the questionnaire responses are as follows:

  • Informed patient choice should be regarded as the prime factor in the effective use of a BGM system.

  • Patient co-payment should be supported where appropriate as this will allow more individualized patient care and empower self-management.

  • Data integration is urgently required for BGM system access.

  • Real-time BGM data should be made available to patients and provide positive reinforcement.

  • Guidance should be inclusive of evidence, patient preference, and outcomes data.

  • Wider factors, such as patient choice, should inform guidance rather than cost alone.

  • New technology for BGM should be regarded as an investment rather than a cost.

  • Structured education should be complemented by access to real-time feedback data.

  • The choice of BGM system should support the achievement of the patients’ own goals.

  • Shared ownership of the responsibility for self-care is critical.