Targeted screening for prediabetes and undiagnosed diabetes in a community setting in India

Abstract

Background and objectives

Data to support the use of risk scores in screening programs to detect people with prediabetes and undiagnosed diabetes in low- and middle-income countries are limited. We evaluated a targeted screening program involving a diabetes risk score in a community setting in India in terms of its uptake, yield, and costs.

Methods

In the Kerala Diabetes Prevention Program, 2586 individuals (age 30–60 years) without known diabetes were screened using a two-step procedure. Step 1: screening with the Indian Diabetes Risk Score at participants’ homes by trained non-medical staff. Step 2: oral glucose tolerance test (OGTT) among those with IDRS score ≥60 (“screen-positive”) at community-based clinics. Screening costs were expressed in 2013 US dollars.

Results

96.3% of those invited for the IDRS screening consented and 79.1% of screen-positives attended clinics for an OGTT. Older age and male gender were associated with higher IDRS uptake. Female gender, higher monthly household expenditure, and higher IDRS score were associated with higher OGTT uptake. The number needed to screen (yield) to detect one person with prediabetes and undiagnosed diabetes was two and six, respectively. The average screening cost of identifying one person with prediabetes and undiagnosed diabetes was $33.8 and $116.5, respectively.

Conclusion

This targeted screening program had a high uptake and high yield for prediabetes and undiagnosed diabetes in a community setting in India. Alternative strategies are likely required to enhance the uptake of screening in certain groups.

Introduction

Type 2 diabetes has emerged as a major public health problem worldwide [1]. According to the International Diabetes Federation, in 2017, there were 425 million people with diabetes, and the global healthcare expenditure on diabetes was estimated at 727 billion USD [1]. People with type 2 diabetes may remain undiagnosed for many years, and undiagnosed diabetes is associated with micro- and macro-vascular complications [2]. Early diagnosis and treatment of type 2 diabetes are likely to reduce cardiovascular morbidity and mortality [3]. The natural history of type 2 diabetes is well understood [4], and effective treatment is available for those with established diabetes [5]. There is compelling evidence to show that type 2 diabetes can be prevented with lifestyle interventions in people with prediabetes who are at high risk of developing type 2 diabetes [6]. Concurrent screening for type 2 diabetes and prediabetes, with appropriate intervention for people with prediabetes, is the most cost-effective strategy [7]. These have provided a strong rationale to screen for prediabetes and undiagnosed diabetes.

Mass screening for prediabetes and type 2 diabetes with an oral glucose tolerance test (OGTT) is generally not recommended because such a procedure is invasive, expensive, and inconvenient to individuals and healthcare providers [8]. Several international organizations and expert groups advise that screening with an OGTT should be targeted to high-risk individuals identified using a diabetes risk score [5,9,10]. Studies have shown that screening a population with a diabetes risk score followed by an OGTT would cost less [11], and have higher uptake and yield compared to mass screening with an OGTT [12]. Furthermore, modeling studies suggest that such a targeted screening program is the most cost-effective way of identifying people with prediabetes and type 2 diabetes [13].

Much of the evidence on the uptake, yield, and costs of screening programs involving diabetes risk scores have come from studies conducted in clinical settings of high-income countries [12,[14], [15], [16], [17], [18], [19], [20]]. These may not be necessarily extrapolated to low- and middle-income countries (LMICs), where the burden of prediabetes and diabetes is substantial, and people have limited access to healthcare facilities [21]. The Kerala Diabetes Prevention Program (K-DPP) was a cluster-randomized controlled trial of a peer-support lifestyle intervention for the prevention of type 2 diabetes in India [[22], [23]]. K-DPP involved two phases: a screening program to identify high-risk individuals (diabetes risk score followed by an OGTT) and an intervention program for the identified high-risk individuals. In this paper, we aimed to examine the screening program in terms of the following: (1) uptake and factors influencing the uptake of each step of the screening program; (2) yield of the screening program, measured as the number needed to screen (NNS) to detect one person with prediabetes and undiagnosed diabetes; and (3) average cost to identify one person with prediabetes and undiagnosed diabetes.