Comparison between Mass Community Screening and Targeted Health Screening in detection of Type 2 Diabetes Mellitus (T2DM) in Singapore
Ho, Emily TL1,2; Chew WJ1 ( 1. Singhealth Service Pte Ltd, 2. Department of Endocrinology SGH)

Background: Type 2 Diabetes Mellitus (T2DM) is a growing problem in Singapore. In 2010 11.3% of residents aged 18 to 69 years had diabetes and of these, 51.4% were newly diagnosed. Early screening and prevention efforts are therefore important in tackling the diabetes epidemic. Effective models of screening are needed.

Aims: To describe and compare the outreach and detection of pre-diabetes and T2DM in residents who participated in two different models of community based chronic disease screening.

Methods:
The two different screening models are mass community screening and targeted screening at general practitioner clinics. Mass community screenings are held at local residential and community sites on weekends recruiting 80-100 people per event. Residents age 40 years old and above without preexisting chronic diseases are invited through local publicity by the community centers. In the targeted health screening model, the aim was to screen the higher risk individuals with a family history of DM. Through wider media publicity, younger residents aged 30 years and above, with a first degree relative with T2DM are eligible and encouraged to go for screening at participating General Practitioner (GP) clinics at their convenience. All screenings are heavily sponsored and include a lifestyle questionnaire, measurements for body mass index (BMI), blood pressure, fasting plasma glucose and a lipid profile. Individuals with abnormal fasting glucose above 6.0mmol/L are offered a confirmatory test through a repeated fasting glucose or an oral glucose tolerance test. Individuals with abnormal results are counselled, offered lifestyle intervention programs and clinical follow up at the GPs.

Results:
In mass community screening, from July 2015 to January 2017, 897 participants (93.8% Chinese, 3.7% Indian, 1.8% Malay and 0.8% others) were screened across 11 community sites. Characteristics of participants were 58.9% females, 41.1% male; mean age 59.89. For the targeted screening cohort 550 participants were screened by GPs over a 14 month period (77.6% Chinese, 11.8% Malay, 7.8% Indian and 2.7% others). Similarly most participants were female (65.3% versus 34.7% males) but targeted screening attracted younger participants (mean age 48.13). Comparing the mean age of the two programmes, targeted screening attracted a significantly younger screening population (t = 19.395, p < 0.05).

Mass screenings attracted a lower socio-economic status and less educated cohort compared to targeted screening. 39% of mass screening participants stayed in 1-3 rooms public housing flats compared to 14% of the targeted screening cohort; 26% of participants in mass screening reported an average income of below S$2,000 compared to 13% in targeted screening; 15% of mass screening participants are college educated compared to 33% in targeted screening. Both screening programmes attracted predominantly Chinese ethnic residents, consistent with the ethnic distribution of the residents. A larger proportion of Malays were screened from the targeted screening cohort, 12% compared to just 2% for the mass screenings. These mostly came with relatives indicating that banding families with history of diabetes in the Malay community is an effective way of promoting screening.

Despite initial postulation of a much higher detection rate of DM and pre-DM in the targeted group (positive family history), percentages of participants screened with abnormal fasting glucose (above 6mmol/L) are not significantly different between the targeted and mass screening programmes (t = 1.460, p > 0.05). A positive family history of DM (first degree relative) did not significantly increase the risk of a positive screen for DM (t = 0.519, p > 0.05). By
comparing detection rates between fasting glucose test against the abnormal screen rates of each of the three other tests (BMI, BP and Lipids), preliminary evidence suggests that BMI (r = 0.153, p < 0.05) and BP (r = 0.103, p < 0.05) results are positively correlated with that of fasting glucose, suggesting that participants with high blood glucose will generally have high BMI and hypertension.

Discussions: Mass community screenings attracted the older (> 60 years old) residents from a lower socio-economic group. Targeted screening which required participants to present themselves to a GP clinic at their own convenience drew the younger educated participants of a higher income bracket. A more targeted approach, where participants are risk stratified using a calculated localized risk scoring incorporating variables such as BMI may increase screening yield for DM in the general population.

Comparison between mass screening and targeted health screening in detection of type 2 diabetes mellitus in Singapore