Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Finding type 2 diabetics in primary care


A concern about the obesity epidemic is the increased numbers of people with adult-onset diabetes. Chance finding of frank diabetes or pre-diabetic hyperglycaemia is often a major trigger for lifestyle changes of less but better food, more exercise, and lost weight. Early detection and better control could ameliorate problems associated with diabetes.

This smacks of screening. Screening is a word fraught with danger, because in any set of circumstances there are three camps: the small numbers of enthusiasts who are either for it or against it, and the great mass of normal professionals whose main reaction is profound cynicism about another target. A study that shows that real-world targeted screening can work and might make sense [1] is a welcome relief.


The study was conducted in 16 practices in Somerset and Devon, randomly selected from 42 volunteer practices. They had to have over 3,500 patients and have good (>60%) recording of BMI. Each practice was asked to sample 100 patients, 25 from each of four groups with different entry criteria (Table 1) relating to age and BMI. Selection of patients within the practices was done randomly. Patients could be selected for more than one group. Those with previously diagnosed diabetes were excluded, and only Caucasians were screened.

Trained practice nurses ran the screening clinics. Patients were sent a provisional clinic appointment, followed up by telephone reminder. Weight, height and age were recorded, and a fasting venous blood sample taken for plasma glucose measurement. Those with fasting plasma glucose over 6 mmol/L were invited for repeat testing. Diabetes was defined as plasma glucose of 7 mmol/L or more on both occasions. Impaired fasting glycaemia was defined as levels of 6.1-6.9 mmol/L on both tests.


The response rate to invitation to attend the screening clinic was 61%. That meant 1,287 people attended, and, as some were in more than one group, there were 1,644 data points for analysis. BMI information was available for 77% of the over-50 population, and 20% of these were out of date or inaccurate compared with clinic measured BMI. Self-reported age differed from practice computer in 27/1,287 cases by more than one year. Of the 1,287 who attended for screening:

The numbers of patients needed to be screened to detect one case of type 2 diabetes or impaired fasting glycaemia was low (7-13, Table 1), and reasonably flat across the groups.

Table 1: Screening group characteristics and results of screening

Characteristics of groups
Age (years)
over 70
over 65
over 60
over 50
BMI (kg/sq m)
over 32
over 30
over 28
over 26
Percentage of population in each group
Number of people tested
Percent with established type 2 diabetes in each group
Results of screening
Number (%) with new type 2 diabetes
14 (4.7)
28 (5.7)
18 (3.8)
10 (2.6)
Number (%) with new impaired fasting glycaemia
25 (8.3)
41 (8.4)
39 (8.3)
20 (5.2)
Number needed to test for type 2 diabetes
Number needed to test for type 2 diabetes or impaired fasting glycaemia


These screening strategies discovered substantial numbers of people with previously undiagnosed type 2 diabetes. Undiagnosed diabetes rates were about 20% of those already diagnosed. For those with impaired fasting glycaemia, a glucose tolerance test might have been appropriate. Better recording of BMI and an expert computer system (they do exist!) could identify people at risk relatively simply. Practices could choose what criteria they might wish to adopt based on perceived workload, and on resources available. Lower age and BMI criteria should identify people early enough for lifestyle changes to be effective, especially in those with impaired glycaemia.


  1. CJ Greaves et al. A simple pragmatic system for detecting new cases of type 2 diabetes and impaired fasting glycaemia in primary care. Family Practice 2004 21: 57-62.

previous or next story