Screening: the test

The UK National Screening Committee use the following criteria to assess the test to be used within a screening programme: (1) There should be a simple, safe, precise and validated screening test; (2) The distribution of test values in the target population should be known and a suitable cut-off level defined and agreed; (3) The test should be acceptable to the population; and (4) There should be an agreed policy on the further diagnostic investigation of individuals with a positive test result and on the choices available to those individuals [1]. In terms of type 2 diabetes there are a number of tests which can be employed within a screening programme, some require blood tests, other utilise non-invasive risk scores. Below we discuss both invasive and non-invasive screening tests, followed by a recommended approach to screening for type 2 diabetes which involves a combination of the two.

Invasive tests

Blood tests which have been shown to have reasonable performance when evaluated against recommended diagnostic criteria include random glucose, fasting glucose, the glucose challenge test, and 2-h post-challenge glucose. The recommended diagnostic criterion for type 2 diabetes incorporates two tests, the oral glucose tolerance test (OGTT) or HbA1c. The OGTT involves an overnight fast, a fasting glucose measure, followed by a glucose load, a two hour wait and a repeat assessment of the blood glucose level (2-h post-challenge glucose). Although this is the recommended screening test for diabetes it is inconvenient, expensive and impractical when large numbers of people are being screened. The OGTT has also been reported as a barrier to uptake of screening [2]. In 2011, the WHO recommended that HbA1c can be used as a diagnostic test for type 2 diabetes, with a cut-off value of ≥6.5% [3]. Before this recommendation, HbA1c had been used in many large scale screening studies as the screening test, followed by an OGTT.

HbA1c is a good indicator of chronic hyperglycaemia and long term complications and is less affected by concurrent physical and emotional stress levels than plasma glucose levels. Measurement of HbA1c is standardised and has low inter-test variability. The HbA1c test can be completed in a non-fasted state and is therefore convenient [4]. There are disadvantages associated with the use of HbA1c which include misleading results in those with hemoglobinopathies, iron deficiency, hemolytic anaemias, and severe hepatic and renal disease which makes HbA1c unsuitable for screening in these groups [5]. There is also data to suggest that HbA1c is systematically higher in particular ethnic groups and that it can increase with age, which may also affect its interpretation [6].

Random glucose testing usually involves a capillary measure and is attractive as it can be carried out opportunistically and does not require an overnight fast, which may increase uptake. However it is not widely used as it has high variability and poor sensitivity, particularly in low risk groups [7]. Very high results are a good indicator of IFG/IGT, but lower ranges of 6-10 mmol/l may need to be re-screened using a fasting test. Fasting glucose requires an overnight fast so is not as appealing or practical as the random glucose but these disadvantages are outweighed by its stability and sensitivity. However this test may miss people who are carbohydrate intolerant, and whose hyperglycaemia is only manifest after a carbohydrate load.

The 50g glucose challenge test has been used mostly in screening for gestational diabetes and there is little data regarding its use as a screening test for type 2 diabetes. It can be used without the need for fasting, which makes it more convenient than tests which require this. While the testlooks promising, more research is required before it can be routinely recommended [8]. Postprandial urine testing has low sensitivity for detecting type 2 diabetes, although it may have a place in low resource settings where no other procedure is possible [9].

Non invasive

Non-invasive risk scores can be applied to individuals as a self-report questionnaire (so called self-assessment) or to a population by using routinely available data. These scores increasingly being used to stratify populations before inviting those at high risk to attend for blood glucose testing. Risk scores in the form of a questionnaire can be used for opportunistic screening and maybe a means to engage with groups who would not routinely attend a screening appointment or engage with the health service. These scores only include risk factors which would be known to the individual and not risk factors which require clinical assessment. For example they may ask if the individual has ever been diagnosed with hypertension, but not the actual value of their systolic blood pressure.

The FINDRISC score is an example of a self-assessment questionnaire [10]. It was developed for use in Finland to detect individuals at risk of developing type 2 diabetes in the future. It includes eight questions on age, BMI, waist circumference, blood pressure, history of high blood glucose, family history of diabetes, physical activity and consumption of vegetables, fruits or berries. Each response is allocated a number of points and the total score reflects the individual’s diabetes risk and they are signposted appropriately depending on their score. Other similar scores detect this risk of current undiagnosed diabetes rather than the risk of developing the disease in the future. One example is the UK Leicester Self-Assessment score [11], which detects undiagnosed type 2 diabetes and impaired glucose regulation. This score was developed using data from a cross sectional screening study where all participants received an OGTT. The score contains similar variables to the FINDRISC but with the addition of ethnicity which reflects the multi-ethnic population within the UK.

Risk scores can also be applied to a population rather than an individual. Here risk scores are used to interrogate routinely stored data, such as primary care electronic medical records, to find those at highest risk for invitation to a screening appointment. These scores are usually applied using a piece of software and do not rely on an individual calculating the total risk score. Therefore a more sophisticated scoring algorithm can be used which improves the ability of the score to discriminate between individuals with and without diabetes. The UK Leicester Practice Risk Score was developed for this purpose [12]. It is similar to the self-assessment score described earlier but does not contain waist circumference as this is not well recorded in primary care in the UK. The score does not categorise continuous risk factors, but rather multiples the level of the risk factor by an inflation factor. In an external validation study comparing Leicester self-assessment and practice risk scores in a young south Asian cohort, both scores performed well but which a slight increase in discrimination and calibration for the practice risk score [13].

To date many risk scores have been developed within the field of type 2 diabetes, with one review from 2011 reporting 43 scores for type 2 diabetes [14]. One reason for this is that scores which have been developed for a particular population tend to have low sensitivity and specificity when used in other populations. Although many scores exist, relatively few are actually used in clinical practice [15][16][17]. There are many possible reasons for this, including health care professional attitudes toward risk scores, impracticality of using the risk scores, and lack of reimbursement and regulatory support [18]. One review reported that many scores are not developed with an intended use in mind [19].

The PREDICT-2 group have summarised all of the currently available risk scores world-wide. This information is hosted on the International Diabetes Federation website (http://www.idf.org/risk-prediction-tools-predict-2) [20]. A number of risk scores have been developed for use in European countries including the Canary Isles, Denmark, Finland, France, Germany, The Netherlands, Portugal, Spain and the UK. Additionally the FINDRISC has been validated for use in Greece, Bulgaria, Italy, Spain and Sweden. The PREDICT-2 project is also collecting data from completed screening studies from around the world to develop a globally applicable risk score for use in countries which do not have the infrastructure available to develop a country specific score [20]. Risk scores incorporating invasive measures also exist, i.e. those including biomarkers or genetic factors. These do not generally out perform their non-invasive counterparts and are not therefore routinely used [21][22].

Two stage approach

Figure 1 (Click to enlarge)
Figure 1 (Click to enlarge)
Given the growing number of people at risk of type 2 diabetes, many international bodies, including NICE and the Europe-wide IMAGE project [23][24], now recommend a two stage screening programme. All individuals have their risk of diabetes calculated using a non-invasive risk score, which can be used to stratify the population in terms and those at highest risk are invited for an invasive blood test (see Figure 1). This approach offers many benefits over using a blood testing in the entire population. Firstly a two stage approach has been shown to reduce the cost of screening. A recent study estimated that a one-step screening strategy where everyone undergoes an HbA1c test costs around €1084 per each case of diabetes detected [25]. This is reduced to around €658 per case if a two stage strategy employing a non-invasive risk score is used. Secondly a two-step approach increases the number of people with the disease identified. A two stage approach was used in two diabetes prevention studies to identify cases with impaired glucose regulation (IGR). Of those who attended screening 25.7% were found to have IGR, with 4.2% having undiagnosed type 2 diabetes [26]. These rates were significantly higher than when a one-stage screening approach was taken in the same population [27]. Thirdly, using a risk score which incorporates questions on risk factors engages people with their own personal level of risk, in a way that an abnormal blood test result maynot.

References

  1. ^ National Screening Committee, Evaluation of Type 2 Diabetes Mellitus Screening against the NSC Handbook Criteria. 2009.

  2. ^ Eborall, H., et al., Influences on the uptake of diabetes screening: a qualitative study in primary care. Br J Gen Pract. , 2012. 62(596): p. e204-11.

  3. ^ World Health Organisation, Use of glycated haemoglobin (HbA1c) in the diagnosis of diabetes mellitus, World Health Organisation, Editor 2011: Geneva.

  4. ^ Gholap, N.N., et al., Diagnosing type 2 diabetes and identifying high-risk individuals using the new glycated haemoglobin (HbA1c) criteria. British Journal of General Practice, 2013. 63(607): p. e165-e167.

  5. ^ Gallagher, E.J., Z.T. Bloomgarden, and D. Roith, Review of hemoglobin A1c in the management of diabetes. J Diabetes 2009. 9(17).

  6. ^ Herman, W.H., Y. Ma, and G. Uwaifo, Differences in A1C by race and ethnicity among patients with impaired glucose tolerance in the Diabetes Prevention Program. Diabetes Care, 2007. 2453(2457).

  7. ^ Qiao, Q., et al., Random capillary whole blood glucose test as a screening test for diabetes mellitus in a middle-aged population. Scand J Clin Lab Invest, 1995. 55(1): p. 3-8.

  8. ^ Waugh, N., et al., Screening for Type 2 Diabetes: literature review and economic modelling. Health Technology Assessment, 2007. 11(17): p. 144.

  9. ^ World Health Organization, Screening for Type 2 Diabetes. Report of a World Health Organization and International Diabetes Federation meeting, 2003.

  10. ^ Lindström, J. and J. Tuomilehto, The Diabetes Risk Score. Diabetes care, 2003. 26(3): p. 725-731.

  11. ^ Gray, L.J., et al., The Leicester Risk Assessment score for detecting undiagnosed Type 2 diabetes and impaired glucose regulation for use in a multiethnic UK setting. Diabetic Medicine, 2010. 27(8): p. 887-895.

  12. ^ Gray LJ, et al., Detection of impaired glucose regulation and/or type 2 diabetes mellitus, using primary care electronic data, in a multiethnic UK community setting. Diabetologia, 2012. 55(4): p. 959-66.

  13. ^ Gray, L.J., et al., External validation of two diabetes risk scores in a young UK South Asian population. Diabetes research and clinical practice, 2014. 104(3): p. 451-458.

  14. ^ Collins, G.S., et al., Developing risk prediction models for type 2 diabetes: a systematic review of methodology and reporting. BMC Medicine, 2011. 9: p. 103.

  15. ^ Glümer, C., et al., Risk Scores for Type 2 Diabetes Can Be Applied in Some Populations but Not All. Diabetes care, 2006. 29(2): p. 410-414.

  16. ^ Chien, K., et al., A prediction model for type 2 diabetes risk among Chinese people. Diabetologia, 2009. 52(3): p. 443-450.

  17. ^ Al-Lawati, J.A. and J. Tuomilehto, Diabetes risk score in Oman: A tool to identify prevalent type 2 diabetes among Arabs of the Middle East. Diabetes research and clinical practice, 2007. 77(3): p. 438-444.

  18. ^ Dhippayom, T., N. Chaiyakunapruk, and I. Krass, How diabetes risk assessment tools are implemented in practice: A systematic review. Diabetes research and clinical practice, 2014. 104(3): p. 329-342.

  19. ^ Noble, D., et al., Risk models and scores for type 2 diabetes: systematic review. BMJ, 2011. 343(d7163).

  20. ^ Lee, C.M. and S. Colagiuri, Risk scores for diabetes prediction: the International Diabetes Federation PREDICT-2 project. Diabetes Res Clin Pract. , 2013. 100(2): p. 285-6.

  21. ^ Gray, L.J. and K. Khunti, Type 2 diabetes risk prediction—Do biomarkers increase detection? Diabetes research and clinical practice, 2013. 101(3): p. 245-247.

  22. ^ Echouffo-Tcheugui, J.B., S.D. Dieffenbach, and A.P. Kengne, Added value of novel circulating and genetic biomarkers in type 2 diabetes prediction: a systematic review. Diabetes Res Clin Pract., 2013. 101(3): p. 255-69.

  23. ^ Chatterton, H., et al., Risk identification and interventions to prevent type 2 diabetes in adults at high risk: summary of NICE guidance. BMJ, 2012. 12(345): p. e4624.

  24. ^ Paulweber, B., et al., A European Evidence-Based Guideline for the Prevention of Type 2 Diabetes. Horm Metab Res 2010. 42(Suppl. 1): p. 42.

  25. ^ Khunti, K., et al., A comparison of screening strategies for impaired glucose tolerance and type 2 diabetes mellitus in a UK community setting: A cost per case analysis. Diabetes Research and Clinical Practice, 2012. 97(3): p. 505-13.

  26. ^ Gray, L.J., et al., Implementation of the automated Leicester Practice Risk Score in two diabetes prevention trials provides a high yield of people with abnormal glucose tolerance. Diabetologia, 2012. 55(12): p. 3238-44.

  27. ^ Webb, D.R., et al., Screening for diabetes using an oral glucose tolerance test within a western multi-ethnic population identifies modifiable cardiovascular risk: the ADDITION-Leicester study. Diabetologia 2011. 54(9): p. 2237-46.

Comments

Nobody has commented on this article

Commenting is only available for registered Diapedia users. Please log in or register first.