Tayside Diabetes MCN Handbook
Screening for Diabetes

 

Population Risk Assessment and Screening - Risk Score Tool and Letter

 

High Risk Screening for Diabetes

 

Patients with Impaired Glucose Tolerance, Impaired Fasting Glycaemia, a past history of Gestational Diabetes or ‘stress’ hyperglycaemia are at particularly high risk of developing Type 2 Diabetes and should be recalled annually for a fasting venous (plasma) glucose measurement, with the results interpreted and acted on as per the Diagnostic Algorithm.

Other patient groups at high risk of diabetes should be screened opportunistically, preferably by fasting venous glucose measurement (see below). These include:

  • Subjects from non-Caucasian ethnic groups
  • Those with a family history of Type 2 Diabetes
  • Those who are obese, in particular with central adiposity (waist 31.5 inches or over for women and 37 inches or over for men)
  • Women with polycystic ovarian syndrome

 

Population Risk Assessment and Screening for Diabetes

 

  • Recent Scottish data has shown that 19.3% of patients have retinopathy at diagnosis (1.9% sight-threatening).
  • It is known that retinopathy only develops after at least 3 - 5 years of consistent hyperglycaemia in the diabetic range.
  • In 2013, The Scottish Public Health Observatory calculated that ~ 3,400 patients in Tayside had undiagnosed diabetes (0.8% of the Tayside population) - around 17% of people with diabetes are undiagnosed.
  • The NHS Scotland National Diabetes Improvement Plan 2015-17 has tasked Diabetes MCNs “to establish and implement approaches to support the prevention and early detection of Type 2 Diabetes”.
  • The National Institute for Health Research (NIHR) Health Technology Assessment Programme has concluded that systematic population-screening for diabetes is not cost-effective but that targeted screening of those at high risk is.
  • There is good evidence that healthcare professional support and ‘brief interventions’ to encourage exercise and dietary modification in patients at moderate and high risk of diabetes can be effective in producing weight loss and thus in preventing the development of diabetes.
  • There is good evidence that capillary ‘finger-prick’ glucose testing is NOT an appropriate screening tool for diabetes.
  • NICE public health guidance 38 - Preventing type 2 diabetes: risk identification and interventions for individuals at high risk (Issued: July 2012)  recommends that GPs and other health professionals and community practitioners in health and community venues, including pharmacists, voluntary organisations and employers should implement a two-stage strategy to identify people at high risk of type 2 diabetes and those with undiagnosed type 2 diabetes. First, a risk assessment should be offered using validated self-assessment or validated web-based tools. Second, where necessary, a blood test should be offered to confirm whether people have Type 2 diabetes or are at high risk (see NICE flow-chart and further guidance – below).
  • The version of this tool and the Patient Results Letter recommended for use in Tayside can be viewed and downloaded from here: (Validated Risk Score & Risk Score Letter)
 General Practice Teams in Tayside are asked to:
  • note that diabetes risk assessment, using validated questionnaires, is being carried out by a variety of organisations within our community, including Dundee Healthy Living Initiative and Diabetes UK Road-Shows.
  • note that patients calculated as being at moderate or high risk are being signposted to Practice Nurses for lifestyle advice and, where appropriate, diagnostic blood testing.
  • consider supporting patients in this process and so to support the prevention and early detection of diabetes (see below for guidance).
  • consider embedding in their IT System and using and a validated web-based diabetes risk-assessment tool [http://www.nhs.uk/Tools/Pages/Diabetes.aspx] with their patients.

 

 Second Stage Screening Process

Patients found to be at Moderate or High Risk on the risk-assessment tools should have further fasting blood glucose testing carried out, with the results interpreted and acted on as per the Diagnostic algorithm.

 

If screening has been carried out by taking a fasting venous (plasma) sample for laboratory analysis, perhaps as part of a Keep Well or Cardiovascular health check-up, and this is normal i.e. < 6.0 mmol/L, then there is no need to repeat for 5 years unless other indications develop. If this > 6.1 mmol/L, but < 11.1 mmol/L, then two further plasma samples (or one if patient is symptomatic) should be taken in line with the diagnostic algorithm for diabetes to confirm or refute a diagnosis.

 

In patients where the first random screening sample is >11.1 mmol/L, only one further sample (preferably fasting) is required.

 

The use and interpretation of random blood glucose measurements for ‘screening’

Urinalysis or capillary blood glucose meter/ test strip analyses are not considered to be accurate screening tools. Abnormal findings using these methods should always be followed up by repeating with a venous (plasma) screening sample.

 

While convenient, the use and interpretation of random venous blood glucose measurement for screening is problematic:

  • A result below 6.1 mmol/l gives a very high probability that the patient does not have diabetes.
  • A result above 11.0 mmol/l gives a very high probability that the patient does have diabetes, though in an asymptomatic patient a further fasting or random blood glucose is required to confirm the diagnosis.
  • The range between 6 and 11 is particularly difficult to interpret:    Studies have shown that (i) a result of 6.9 mmol/l is 93% specific (i.e. 7% false positives) and 41% sensitive (i.e. 59% false negatives) for a diagnosis of diabetes & (ii) a result of 7.2 mmol/l is 87% specific (i.e. 13% false positives) and 63% sensitive (i.e. 37% false negatives) for a diagnosis of diabetes.

 

The 2005 JBS 2: Joint British Societies’ guidelines on prevention of cardiovascular disease in clinical practice (Heart Volume 91 Supplement V December 2005) suggested that a pragmatic strategy for assessing glycaemia in clinical practice is to measure random glucose as part of an initial cardiovascular risk assessment. If random glucose is normal (6.0 mmol/l or less) there is no need to repeat this measurement for five years. If the random glucose is potentially abnormal (6.1 mmol/l or higher) but not indicative of diabetes (≥ 11.1 mmol/l) then this should be repeated fasting. If this fasting glucose is ≥ 6.1 mmol/l but < 7.0 mmol/l, the fasting glucose measurement should be repeated.

The more recent NICE public health guidance 38 - Preventing type 2 diabetes: risk identification and interventions for individuals at high risk (issued July 2012) has moved away from recommending this approach, to the initial use of validated risk assessment questionnaires and tools, followed if necessary by a fasting venous blood glucose measurement. (see detailed Guidance above).