Ia: Baseline cross-sectional survey

A cross-sectional survey was initially conducted to quantify the burden and awareness of hypertension and examine how the barriers to diagnosis and management differ between settings. Recruitment for this study started in January 2014 and was completed in December 2015. Approximately 14 500 adults living in the study areas have been selected to provide data for the baseline survey. The role of gender, socioeconomic deprivation and education on the diagnosis and management of hypertension in each of the three rural areas is being explored. The data have been used to design the intervention component of the study.

Population censuses (specifically completed for this study or existing polling booth registers) at each site were used to randomly select potential participants. In Trivandrum and West Godavari, sampling was stratified by age and sex using this approach. However, due to structural factors related to distance between hamlets and demography, this was not feasible in the Rishi Valley region. Residents of the PSU aged at least 18 years were eligible for recruitment to the cross-sectional survey.

Trivandrum: Among the 14 districts in Kerala, Trivandrum district was selected based on its proximity to the collaborating institute (Sree Chitra Tirunal Institute for Medical Sciences and Technology). Chirayinkizhu taluk was selected randomly from the four taluks in Trivandrum district. Of the 22 Panchayats (local administrative body in rural areas) within Chirayinkizhu taluk, 10 were randomly selected. From each of the selected Panchayats, one ward (the smallest geographic unit of a Panchayat) was randomly selected. In each ward, using the polling booth list, the total number of individuals was divided into 12 age and sex groups (18–24, 25–34, 35–44, 45–54, 55–64 and 65+). From each group, 30 individuals were randomly selected to get a total sample of 360 in each ward. Thus, 3600 participants were selected in the Trivandrum district to participate in the cross-sectional survey. Additional sampling was conducted (10 from each age and sex band) to replace those participants who had migrated, died or refused to participate.

West Godavari: A sampling frame was developed by mapping all 17 Primary Health Centres (PHCs) within a 50 km radius around the town of Bhimavaram. We excluded those PHCs participating in other studies conducted by the George Institute and randomly selected 10 of the remaining PHCs. All the villages serviced by each PHC were included in the list. Villages with fewer than 3000 residents were excluded. One village was then randomly selected from each PHC, resulting in inclusion of 10 villages. Mapping all the selected villages enabled generation of an age and sex population list with house addresses. Population lists from the mapping were used to randomly select 4500 individuals over the age of 18 years from each participating village. Additional sampling was conducted to replace those participants who had migrated, died or refused to participate. Sampling was stratified by 12 groups defined by age (18–24, 25–34, 35–44, 45–54, 55–64, 65+) and sex with the goal of including comparable numbers of individuals from each group.

Rishi Valley region: The study population comprised the six villages of the Kurabalakota Mandal in the Chittoor District, Southern AP. These villages comprised 221 hamlets or small habitations.

Hamlets were stratified by population size (small, medium and large) and then 139 were randomly selected in accordance with the sampling strategy using computer-generated random numbers (generated at Monash University). This was to ensure sampling of approximately equal numbers of hamlets from each size stratification. Six hamlets were excluded due to migration of population. We also excluded the hamlet in which the Rishi Valley Rural Education Centre was located because the population was largely transient, comprising teachers and students who reside in the hamlet only during school time. A study centre was set up in a communal area of the hamlet convenient for all the residents. All residents aged at least 18 years were invited to participate in the cross-sectional survey. Research officers ensured that all residents were informed of the presence of the study team in the habitation by house-to-house notification and encouragement to attend.

The instruments and measurements chosen for this project are based on recommendations from the WHO STEP-wise approach to disease surveillance (WHO STEPS)²⁶ and other validated tools for quality of life, SEP and barriers to changes in alcohol and tobacco behaviour as listed in online supplementary table S1. The list of measures comprise (1) basic demographic information, including age, income, gender, marital status, religion, number of children and type of work undertaken; (2) lifestyle-related factors such as physical activity, tobacco use and alcohol consumption, dietary factors, including cooking practices and use of salt, stress and overcrowding; (3) knowledge about hypertension and its risk factors, awareness of hypertensive status and reports of the timing and outcome of prior BP measurements; and (4) further details about the use of medications (allopathic and AYUSH or other traditional therapies), barriers to treatment, including access, cost, adoption of lifestyle factors and compliance with medication use (see online supplementary table S1 for full list of variables). Questionnaires were developed in English and then translated into the site-specific language (Telugu (AP), Malayalam (Kerala)) and back translated to detect and correct errors.

Standardised clinical measurements are collected as follows; arterial BP and heart rate are measured after the participant has sat quietly for at least 15 min. BP is measured at least three times at 3 min intervals using the appropriate cuff size and a Digital Automatic Blood Pressure Monitor (OMRON HEM-907, OMRON Healthcare Company, Kyoto, Japan) according to the WHO STEPS protocol, modified only by using the right arm for all measurements.²⁶ Measurement continues until two consecutive readings differ by <10 mm Hg systolic and <6 mm Hg diastolic, with a maximum of five measurements. The mean of the last two consecutive measurements are used to define hypertensive status. Height is measured to the nearest 0.1 cm using a portable stadiometer (213, Seca, Hamburg, Germany). Weight is measured to the nearest 0.1 kg using a portable digital weighing scale (9000SV3R, Salter, Kent, UK). Waist and hip circumference is measured using a spring-loaded tension tape (Gulick M-22C, Patterson Medical, Illinois, USA) in a private setting. In accordance with the WHO STEPS protocol,²⁶ waist circumference is measured at the midpoint between the lowest rib and upper point of the iliac crest and at the end of normal expiration and hip circumference is measured at the maximum protrusion of the buttocks.

To ensure standardisation, data collectors are trained in collection of anthropometric and BP measurements in accordance with the WHO STEPS protocol.²⁶ This training, conducted by the project manager for at least 5 days, is to ensure consistency of data collection between sites. Training is provided in a similar manner to ensure that questionnaire administration is also consistent across all sites. A study-specific training manual containing step-by-step procedures for all data collection (anthropometric and survey administration) is provided to each data collector. Data collection at each site is further monitored by site supervisors. Follow-up training by the project manager and/or site supervisor is also undertaken at each site ∼1 month after initiation of data collection to ensure that data collection methods are implemented according to the protocol.

Definitions: Participants whose measured mean SBP is ≥140 mm Hg and/or mean DBP is ≥90 mm Hg or who are taking medication for lowering BP are defined as hypertensive.⁷ Waist circumference is deemed high when >80 cm in women and >90 cm in men. A body mass index (BMI) ≥25 to 29.99 kg/m² is defined as overweight and BMI ≥30 kg/m² as obese according to the revised BMI classification.²⁷

Sociodemographic and economic characteristics of each population are determined, including gender, SEP, education, income and expenditure. Information on usage of healthcare, physical activity, tobacco use and consumption of alcohol are also collected. Primary outcomes for Phase Ia of the study are prevalence, awareness, treatment, control, knowledge of hypertension and associated risk factors. We compare knowledge of hypertension, previous measurement of BP and barriers to treatment and management (such as cost, education and SEP) of hypertension across study sites. Good (BP <140/90 mm Hg) or poor (SBP ≥140 mm Hg or DBP ≥90 mm Hg) control of hypertension is assessed in relation to lifestyle factors (physical activity, use of tobacco and consumption of alcohol), dietary factors (including salt intake) and healthcare usage. Attitudes to healthy behaviour change are also assessed.