These sections briefly explain why the committee made the 2021 and 2022 recommendations and how they might affect practice.
The committee wanted to highlight the importance of distinguishing between type 1 diabetes and other diabetes types because these conditions are treated differently (particularly in terms of insulin use).
It is less harmful to be diagnosed with type 1 diabetes when the person actually has type 2 diabetes. However, there are still harms, including:
the long-term effects and costs of unnecessary insulin therapy
the missed opportunity for oral diabetes therapies
the psychological impact of misdiagnosis.
There was no new definitive evidence on clinical features for identifying diabetes type, so recommendation 1.1.1 remains unchanged from the 2015 guideline. Because of this lack of new evidence, the committee made a recommendation for research on identifying diabetes type.
The evidence showed that no single clinical feature had a sufficient predictive value to make a diagnosis by itself. The committee were particularly concerned that age and body mass index (BMI) might be used in isolation. They noted that the average BMI in people with type 1 diabetes is increasing, and the age at which people are diagnosed with type 2 diabetes is decreasing. This means these clinical features are becoming less useful on their own to differentiate between the subtypes. Despite this, the committee agreed that these characteristics are still useful for making an initial working diagnosis of diabetes subtype in many people. However, further testing is increasingly needed, as previously 'atypical' features of type 1 and 'uncertain' classifications become more common.
In a change from the 2015 guideline, the committee agreed it was important to encourage the use of diabetes-specific autoantibody testing at diagnosis, to avoid misclassifying diabetes subtype. They also clarified that autoantibody testing is appropriate for people with suspected type 1 diabetes.
There was no high-quality evidence on tests to distinguish type 1 from type 2 or other types of diabetes, so the committee based the recommendations on the timings when the tests might be most useful (autoantibody testing is best used at the time of presentation, and C‑peptide is best used with increasing time from initial presentation) rather than which tests were most accurate.
Based on clinical experience, the committee were confident that measuring autoantibodies in people with suspected type 1 diabetes would be cost effective. This is because:
autoantibody tests are cheap when compared with the much higher costs associated with inaccurate diagnosis
misclassification using clinical criteria alone results in additional costs, both from the use of ineffective treatments and from clinical harm.
Further, because autoantibody tests are more accurate when done at the time of presentation rather than later at a clinical review, they would also be more cost effective at that time rather than later, since the cost will be the same but more useful information will be obtained from the test at the time of diagnosis.
The committee noted that using autoantibody testing also means that healthcare professionals do not have to rely on characteristics alone when people first present. This can help avoid assumptions about links between ethnicity and diabetes type (for example, assuming that people in Black, Asian and other minority ethnic groups are more likely to have type 2 diabetes).
The committee could not recommend routine non-fasting serum C‑peptide testing because of a lack of high-quality and clinical evidence, and this would be a significant and costly change in clinical practice. However, they thought it would be an appropriate test if clinical presentation and autoantibody testing did not provide a clear classification of diabetes (for example, if clinical features were consistent with type 1 diabetes but autoantibody results were negative).
The committee noted that serum C‑peptide is more appropriate in individual clinical diagnosis settings as it can be paired with blood glucose, while urine C‑peptide is mainly used in epidemiological studies.
Because of the lack of high-quality evidence, the committee made a recommendation for research to examine the effectiveness of C-peptide at correcting misclassification of diabetes at initial diagnosis and the optimal timing for this test in distinguishing between subtypes of diabetes.
It is likely that these recommendations will lead to increased autoantibody testing in people presenting with suspected type 1 diabetes. This will increase testing costs, but this increase is not expected to be substantial, because of the low costs of the tests themselves. An increased use of serum C‑peptide testing in the classification of diabetes is less likely, because the guideline only recommends this be done if there is still diagnostic uncertainty after the use of autoantibody testing.
Although there is a cost associated with some of the new recommendations and concerns about the availability of the tests in all settings, the committee felt this would be balanced out by benefits from reducing the misclassification of diabetes at presentation and ensuring early appropriate treatment of type 1 (and in some cases type 2) diabetes. This will avoid unnecessary tests, treatment and side effects from prolonged insulin use.
The committee agreed that there was enough evidence in key outcomes, such as HbA1c, time in range and severe or nocturnal hypoglycaemia, to demonstrate that both real-time continuous glucose monitoring (rtCGM) and intermittently scanned CGM (isCGM) provide clinical benefits over standard self-monitoring of blood glucose. However, they considered that the evidence for rtCGM compared with isCGM was not good enough in terms of quality or sample size to clearly show clinical benefits of 1 technology over the other.
The committee also acknowledged that CGM technologies were changing very quickly, with increasing overlap between rtCGM and isCGM as features such as predictive alerts are added to newer isCGM devices.
The health economic modelling found that, when the benefit of reduced fear of hypoglycaemia with CGM was included, both technologies were cost effective for the full population of adults with type 1 diabetes compared with standard self-monitoring of blood glucose.
Based on the above factors and the evidence, the committee agreed that there was no advantage to recommending 1 specific device over another compared with standard self-monitoring of blood glucose. They concluded that the specific functionality of isCGM versus rtCGM devices should be discussed between the person and their healthcare professional. In particular, finding the right device for each person is likely to improve adherence, which means the device will provide more benefits and so will be more cost effective.
The committee did not make a recommendation on using specific devices because CGM technologies are changing very quickly, and this recommendation would soon be out of date. Local healthcare services are better placed to assess which devices are evidence-based and suitable for use at any given time.
There are benefits of providing a choice of different CGM devices because the most suitable device would vary for each person. The committee produced a list of factors to consider when choosing a CGM device with people. The committee agreed that this freedom of choice is beneficial, particularly because adherence to the technology is likely to be higher if the device is matched to the person's needs and preferences.
The committee retained the 2015 recommendation on providing people with support from a healthcare specialist team with expertise in diabetes and the use of CGM. Community-based specialist teams are now available and are no longer always based in secondary care, so 'centre' was changed to 'team' to make this clearer.
Despite the positive recommendations on CGM, the committee were concerned that existing health inequalities may still lead to lower uptake of CGM in some groups of people. To address this, the committee made a recommendation outlining actions for commissioners, providers and healthcare professionals.
Given the rapid advances in technology, the committee made a recommendation for research on using routinely collected real-world data to examine the effectiveness and cost effectiveness of CGM. If routine healthcare data is collected, it can show the direct effect of implemented technology on the population, rather than it being interpreted through the results of clinical trials. Increased monitoring of routine healthcare data including registries and audits would ensure that the findings from a broader population is captured.
These recommendations are likely to result in broader access to isCGM and rtCGM devices, as opposed to a binary decision on access based on stringent criteria. This should reduce inequalities and enable more people to use CGM. Currently, people with more time and knowledge to self-advocate are often more likely to gain access to these devices.
Some people have insulin insufficiency because of other conditions. The committee noted that these people would get the same care as people with type 1 diabetes, so they should have access to CGM in the same way.
There is likely to be an increase in costs from more use of CGM devices. A number of different devices are available, so if more than 1 device would be appropriate for a person and would meet their needs and preferences, using the lowest cost device among those options would help to reduce the cost impact.
Evidence showed that there were fewer severe and nocturnal hypoglycaemic events with insulin detemir twice daily compared with neutral protamine Hagedorn (NPH). Insulin detemir twice daily was also found to be the most cost-effective treatment strategy in the economic analysis. Based on this evidence and their clinical experience, the committee recommended twice-daily insulin detemir as a basal insulin therapy for adults with type 1 diabetes.
The committee agreed there were situations in which an insulin other than twice-daily insulin detemir might be preferred, and set out specific clinical scenarios where alternative long-acting insulins could be used. This includes if the person is already using an insulin regimen that is working well for them and helping them meet their treatment goals.
Some people may not be able to tolerate insulin detemir, or for some, a once-daily regimen may be necessary (either because the person has a strong preference for once-daily injections or there are circumstances that make twice daily not practical). Glargine (100 units/ml) was found to be the most cost-effective once-daily insulin (particularly when the costs of glargine biosimilars were considered) so it was recommended as the appropriate alternative in these cases.
People on insulin therapies can still have hypoglycaemic episodes. This can be a cause of concern, particularly if they have nocturnal hypoglycaemic events. Evidence showed a lower proportion of nocturnal hypoglycaemic events with degludec (100 units/ml), when compared with other once-daily insulins. Degludec (100 units/ml) is an ultra-long-acting insulin, which means it has a longer duration of action compared with long-acting insulins. The committee agreed that once-daily degludec could therefore be considered as an alternative basal insulin therapy if there is a particular concern about nocturnal hypoglycaemia.
The committee agreed that once-daily ultra-long-acting insulin regimens, such as insulin degludec (100 units/ml), may also be needed by people who need support from a carer or healthcare professional to administer injections, for example, because they are frail or have a physical or mental health condition, or learning disability. Insulins that offer flexibility in dosing time, such as insulin degludec (100 units/ml), have a long duration of action and may be particularly useful because they give more flexibility in when the dose can be given. Insulin glargine (300 units/ml) is another example of an ultra-long-acting insulin. Healthcare professionals should also refer to NHS England's patient safety alert on the risk of severe harm and death from inappropriately withdrawing insulin from pen devices.
Biosimilars have the potential to offer the NHS considerable cost savings. To gain approval for use, biosimilar medicines have to be shown to be safe and as effective as the original reference medicine, and have the same quality. Based on this understanding, the committee noted it was appropriate when starting a new prescription of an insulin where a biosimilar is available, to use the one with the lowest cost.
Additionally, people may be using an insulin for which a lower cost biosimilar is available. In such cases, the committee recommended discussing with people the possibility of switching to the biosimilar. This could happen at the person's routine review. They also agreed that switching to the biosimilar should be carefully planned, taking into consideration the dose-switching protocols, monitoring and the person's concerns about switching from their existing regimen, and a shared decision reached. Healthcare professionals should also refer to the summary of product characteristics for further information when considering switching to biosimilars.
The committee retained the recommendation from the 2015 version of the guideline on considering the use of other basal insulin regimens not covered by other recommendations. Based on their clinical understanding, they added comorbidities (such as renal function), risks of hypoglycaemia and diabetic ketoacidosis and concerns about adherence to the factors to take into account when considering alternative regimens. To support pharmacovigilance and patient safety, the committee also recommended that insulins should be prescribed by brand name.
Use of long-acting insulins varies across the country, with some centres offering twice-daily insulin detemir to people who are newly diagnosed, whereas other centres start with once-daily regimens. A major shift in practice is unlikely but the recommendations do set out scenarios where other insulins such as ultra-long-acting insulins and biosimilars may be useful and cost effective.
The evidence showed that people with diabetes are at increased risk of periodontitis, and that non-surgical periodontal treatment can improve diabetic control. Although most of the research was focused on type 2 diabetes, the committee thought that the evidence on the bidirectional link between increased HbA1c and periodontitis was also applicable to people with type 1 diabetes.
In the committee's experience, people with diabetes are often unaware of the link between diabetes and periodontitis and may not be having regular oral health reviews. To address this, the committee recommended routinely discussing the risk of periodontitis at annual reviews, alongside eye disease and foot problems.
The evidence also showed that periodontal treatment is cost effective for people with type 1 diabetes, assuming improvements in HbA1c are maintained. This was tested with health economic modelling in a range of different scenarios. The only situation in which treatment would not be cost effective is if the analysis only considered up to the first 10 years of periodontal treatment. However, the committee did not think this was realistic, as this excludes the benefits from reducing diabetic complications, which often happen later in life.
For oral healthcare professionals, the long-term impact of the recommendations is uncertain. The recommendations specify that people should follow existing NICE guidelines on oral health. However, the recommendations may also increase awareness of periodontitis, leading to a possible short-term increase in the number of oral health reviews. Any increase in the number of oral health reviews will potentially impact on services, as NHS dental services already have capacity issues.
A short-term increase in the number of oral health reviews will also lead to a short-term increase in costs. However, there is likely to be a larger long-term reduction in costs from the improvement to oral health and diabetes control.
Oral healthcare and dental teams will need clear advice on what they need to do for people with diabetes. They will need clear care pathways to improve quality of care and service delivery, in line with the NHS England commissioning standard on dental care for people with diabetes.
mothers with babies under 1 year old
on low income benefits, or under 20 and dependent on someone who is receiving low income benefits
having treatment in an NHS hospital by the hospital dentist.
The recommendations may encourage more people with diabetes to have regular oral health reviews. Combined with proactive engagement and enhanced support for people with diabetes, this may broaden access to dental and oral healthcare and help to reduce oral health inequalities.