Background: Age specific mortality rate (ASMR) of 16-60 years is very high in Indian tribal areas due to communicable and non-communicable diseases.

Objectives:

The primary objective is to reduce the ASMR in 16-60 years age group by at-least 30% from 16 tribal villages of Melghat. The secondary objectives are to reduce cause-specific mortality rates (CSMR) in 16-60 years age group because of diarrhoea, malaria, pneumonia, etc.

Design: Community-based, single-centered, parallel-arm, cluster randomised controlled trial.

Setting: 36 tribal villages/clusters (research phase: 2004-2015) and 44 new villages (replication phase: 2016-2022) were externally randomized by lottery method to intervention arm (IA) or control arm (CA).

Participants: All persons in age group of 16-60 years from inaccessible 80 tribal villages.

Interventions: Trained VHWs in IA, provided behaviour change communication, treatment and referral. Except principal investigator, other study staff and participants were double blinded.

Community-based Management of Adult Diseases for Reducing Mortality in Economically Productive age-group: Cluster Randomised Control Trial in India: 2004 -2022.

INTRODUCTION : The adult mortality rate, ASMR (16 to 60 years) was 140 globally (WHO), 179 in India during 2015-2020, increased to 184 globally, by 2023. Within India, ASMR (15-59 years) shows wide disparities: 320 (rural India), 230 (urban India), High adult mortality threatens families, communities, and nations, intensifying poverty. Despite being very important priority of global health, WHO, MDG and SDG-3, adult mortality was neglected by policy makers. 70% of adult deaths are home deaths in Indian tribals. The primary objective is to reduce the ASMR in 16-60 years age group by at-least 30% from 16 tribal villages of Melghat. The secondary objectives are to reduce cause-specific mortality rates (CSMR) in 16-60 years age group because of tuberculosis, cardio-vascular diseases (CVDs), diarrhoea, malaria, pneumonia and COPD.

METHODS Study design: Community based, cluster randomised controlled trial (CRCT) with parallel arm, open label, was conducted in 34 tribal villages/clusters of Melghat during research phase (2004-2015).

Melghat is difficult to access, impoverished, hilly, forest, tribal area in Maharashtra with grossly inadequate medical services. Due to low health seeking behaviour and poor socioeconomic conditions, Melghat has very high baseline ASMR (16-60 years) (>500 per 100,000 population). Hence there is need of community-based program through VHW to reduce ASMR. MAHAN's Independent Institutional Ethics Committee approved the study protocol in 2002-2003, (approval number: MAHAN 2/2002).

This study followed CONSORT guidelines for CRCT.

Patient Public Involvement (PPI):

Methods The study engaged >60% of village adults through community meetings to understand issues, co-create research aims and methodology, and secure written informed consent.

Door-to-door surveys and PPI identified accessible, free, culturally acceptable community-based adult care design, based on communities' demand and cultural practices, utilized trained sensitive tribal female VHWs, selected by the community, to provide treatment and BCC in the local dialect to reduce adult mortality.

Recruitment in the study Local tribal VHWs recruited study participants. VHWs and key community members were involved in reporting of the trial to participants and communities during quarterly gramsabhas.

Participants All individuals aged 16-60 years residing in the study villages who gave written informed consent/assent were included as participants. Out-migrants >6 months, were excluded from analysis. De-facto and de-jure methods were used during research phase and de-facto method for replication phase. Randomisation and Masking Dharni block with 160 villages, was divided into five zones. External person randomly selected eight clusters from each zone using the close envelope lottery method.

An external statistician and local person generated the sequence and randomly assigned clusters to the intervention arm (IA) and control arm (CA) of 18 villages each, by lottery method, respectively with allocation concealment in sealed envelopes. They were not involved in the trial. Co-investigators, data team, statisticians, outcome assessors and participants, were masked to group allocation and interventions, but due to visible nature of the intervention, PI was not masked. Co-investigators and statistician were blinded from all study procedures. VHWs enrolled the participants, assigned a unique identification number (UID) and provided the intervention, and statistician was given anonymised data with UID.

Procedures

Study phases:

Research phase: 2004 to 2015.

1. Baseline phase: January -December 2004

   Microplanning:

   Community participatory meetings were conducted at the onset of the study to introduce the project, identify health resources, secure written consent from villagers in both arms, and community engagement.

   The activities included: (a) village mapping, (b) census, (c) vital statistics data collection, including births, deaths, and (d) verbal autopsies to determine mortality causes and (e) recording anthropometric data throughout the research and service phases.

   Trainings: All VHWs underwent initial monthly residential training of two days per month on village mapping, census, data collection, verbal autopsy, and blood pressure measurement over two months. The trained supervisors and program managers provided classroom and on-site mentoring and quality control.

   Baseline data collection in IA and CA (January 2004-December 2004):

   VHWs conducted a village mapping, comprehensive door-to-door census, and baseline survey covering demographic data, births, deaths (ages 16-60 years), maternal health and adult healthcare practices.

   Participants of 16-60 years of age were included in the study using the complete enumeration method. VHWs collected real-time deaths data in the two arms by prospective household visits within 24 hours. Verbal autopsies for deaths (ages 16-60) were carried out by team, and then reviewed by two physicians following standardized methods. Home-based BP records were collected by complete enumeration method.

2. Intervention phase: January 2005 to December 2015. The VHWs enrolled the participants and provided behaviour change communication (BCC) and treatment in IA. Data were collected and participants in both arms received referral and usual care provided by government health services.

   Training: January 2005 to December 2015. Training topics were based on the results of the survey. Monthly residential trainings for IA-VHWs were arranged from January 2005 to December 2005.

   BCC & Referral Phase: January 2005 to December 2015. We started BCC about diseases, in phase wise manner. The VHWs advised appropriate referrals to the patients in both arms where they received standard hospital care.

   Treatment phase (January 2006 to December 2015):

   Once the VHWs received 75% or more marks in exams in training centre and field, they were certified by well experienced physician for treatment. VHWs treated the patients of specific disease independently under trained supervisors in villages, in sequential/phase-wise manner.

   VHWs treated HTN with tab. Amlodipine 5mg or tab. Enalapril 5mg OD. VHWs started treatment of diarrhoea, malaria, pneumonia, asthma and COPD by ORS, tablets Norfloxacin, Chloroquine, Azithromycin, and Salbutamol Rotacaps, tablet Derriphylline respectively .The VHW referred suspected TB patients to the hospital.

3. Replication phase: (September 2016 - August 2023):

Government and MAHAN replicated the study in new 20 villages, as IA and new 24 villages as CA,. We followed the same procedures during the replication phase.

All treated patients were followed up daily by VHWs till recovery, for adverse drug reactions.

There was no treatment related death.

Statistical analysis:

Sample size was based on significant reduction in ASMR. All randomized participants were included in analysis.

Simple linear regression of ASMR against year for each arm was performed for evaluation over time and comparison of the slope coefficients (trend in intervention vs control).

Relative risk ratio (RRR) were obtained as a ratio of RR of IA vs CA in each year. We report RRR trends with 95% CIs and p-values for each year's comparison.

Difference-in-differences (DID) of ASMR from baseline to endline for each arm and then the difference of these changes between two arms, was calculated.

Chi-square test was used to compare baseline vs endline CSMR within the IA for each cause. Relative risk ratios were used for comparison of combined CSMR of treated diseases with non-treated diseases .

Data safety and monitoring committee consists of government officers of tribal welfare and health departments.

Role of the funding source:

The study was funded by Stitching Geron, the Netherlands, Caring Friends, Mastek Foundation and Tribal Development Department, Government of India.

"The funders of the study had no role in study design, conduct, data collection, data analysis, data interpretation, writing of the report or decision to submit for publication."

Inclusion Criteria:

• All individuals aged 16-60 years residing in the study villages who gave written informed con-sent/assent were included as participants.
• De-facto (only in study village or hospital) and de-jure methods (outside study villages) were used during research phase.
• De-facto method for replication phase.

Exclusion Criteria:

• Out-migrants >6 months, were excluded from analysis.

The de-identified individual data that support the findings of this study (including mortality da-ta, verbal autopsy details, and intervention logs) will be made available from the corresponding author (drashish@mahantrust.org )upon reasonable request. The study protocol and statistical analysis plan are also available on request. Data will be shared beginning 6 months after publi-cation of this article to researchers whose proposed use of the data has been approved by an independent review committee, for the purpose of academic analysis. For this sharing, a signed data access agreement is needed.

Data will be shared beginning 6 months after publication of this article upto 2 years after publication

Data will be shared to researchers whose proposed use of the data has been approved by an independent review committee, for the pur-pose of academic analysis. For this sharing, a signed data access agreement is needed.

Co-investigators, biostatistician, data team,