Evaluation of the Ambuja Cement Foundation Noncommunicable Disease Program in Bathinda District of Punjab, India

Overview

This is an academic evaluation of a program being implemented by the Ambuja Cement Foundation (ACF). The aim of the evaluation is to determine the effect of the ACF noncommunicable disease (NCD) program on (1) fasting blood glucose (a biomarker of diabetes), (2) systolic blood pressure, (3) diastolic blood pressure, and (4) body weight. ACF will be implementing a program on NCD management. Due to resource constraints, the program will be implemented stepwise (e.g. four villages in the first 6 months then an additional four villages added every 6 months over the next 12 months). In order to improve the interpretation of the evaluation data, the team will randomize the order in which the program is implemented in the 12 villages. The design is therefore a stepped wedge randomized cluster trial. This design is ideal for (1) minimizing the practical, logistical, and financial constraints associated with large-scale project implementation, (2) control for the effect of time, and (3) ensure that all villages in the project are eventually offered the intervention.

Evaluation visits will be conducted at the participant's home. At baseline, the ACF field team will first administer a screening questionnaire to all adults in the household to confirm eligibility. For those adults who meet all eligibility criteria, trained enumerators will (1) complete two brief surveys, (2) measure their waist circumference and body weight, and (3) measure their blood pressure using an automatic digital blood pressure machine. The surveys will cover demographics, family history of disease, tobacco and alcohol use, physical activity, and standard clinical symptoms. The field team will then schedule a follow-up visit within the next week to measure fasting blood glucose using a point-of-care device. For those adults who meet criteria for being high-risk for diabetes and hypertension (criteria defined further in the next section), trained enumerators will also administer an additional survey on this second day that covers dietary intake and clinical history including medication use. For high-risk adults, after the baseline visit, three additional follow-up visits will be completed every 6 months for a total duration of 18 months. Each follow-up visit will include: two brief surveys covering the same content summarized above for baseline, as well as measurements of waist circumference, body weight, blood pressure, and fasting blood glucose. The visits will be scheduled such that participants are fasting on the day of the visit, thus allowing all visit activities to be completed in 1 day. All surveys will be completed via interviews by trained field enumerators who will enter participant responses into password-protected tablets (Qualtrics offline mobile app). The total population of these 12 villages is 60,701 according to the 2011 Census. All adults ≥30 years old meeting eligibility criteria will be enrolled, which is estimated to be \~40% of the total population or \~24,280 people. Based on ACF's current programming in 2 other villages in this district, it is expected that 40% of those screened (Group 1 participants) will be high-risk (Group 2 participants) or \~9,720 people (\~810 per village). For fasting blood glucose, a sample of 12 clusters in a complete stepped-wedge cluster-randomized design with 4 time periods (including the baseline), 3 steps, 4 clusters switching from control to treatment at each step, and an average of 500 participants per cluster with an average of 125 participants per cluster per time period (for a total sample size of 6000 subjects) achieves 90% power to detect a difference between means of 0.32 mmol/L (4% of anticipated mean baseline level). The test statistic used is the two-sided Wald Z-Test. The intra-class correlation coefficient (ICC) is 0.010, and the significance level of the test is 0.050. Given that near to 9000 participants are expected to be enrolled, the team is exceptionally well powered to detect significant effects in this evaluation even if there is loss to follow up.