A pilot study to evaluate home-based screening for the common non-communicable diseases by a dedicated cadre of community health workers in a rural setting in India

Noncommunicable diseases (NCDs) are the leading causes of mortality across the globe, accounting for 39.5 million deaths in 2015 [1]. Government of India, as part of its commitment to reduce premature mortality from NCDs, reorganized the NCD control program in 2016 aligned to the WHO Global NCD Action Plan 2013-2020 and introduced screening for the common NCDs [2]. The program guidelines stipulate screening of males and females (age 30 years and above) for hypertension, diabetes and oral cancer at the nearest community clinics (health sub-centers) and screening of females for breast and cervical cancer at the next levels of health service delivery [primary health centers (PHC) and district hospitals]. A major challenge to implement the program is that the sub-centers are under-staffed, poorly equipped and are rarely visited by the residents [3]. Long distance, inadequate transport facilities, high travel cost and loss of wages are major barriers for the rural population to seek preventive health services at the PHCs or higher centers [4]. The problem of ensuring equitable access to preventive health care in the rural and the remote areas is not unique to India but to all the low and middle income countries (LMICs) implementing NCD screening programs [5]. The community health workers (CHWs) are well-recognized to deliver different preventive health care services at the doorsteps of the beneficiaries, especially in the LMICs. The female CHW (known as ‘Accredited Social Health Activist or ASHA’) in India works as an interface between the community and the primary health care. Each ASHA is selected from the village she caters to and usually has a target population of 1,000 to cover. She receives a performance-based incentive for promoting universal immunization, referral and escort services for reproductive and child Health and other healthcare programs.

Our pilot study aimed to evaluate a novel service delivery approach to provide the early detection facilities for the common NCDs (hypertension, diabetes, breast, cervical and oral cancers) at home by trained CHWs in a rural setting. The feasibility and efficacy of the CHW-driven service delivery model were assessed by estimating the coverage of the targeted households, the compliance of screen-positives to further evaluation, the detection rates of the targeted diseases and the positive predictive values of the tests. The study provided an opportunity to study the common risk factors for diabetes and hypertension in the rural community.

The study, conducted between January 2017 and December 2017, was approved by the ethics committees of International Agency for Research on Cancer (IARC), France and GBH American Hospital, India.

OVE was abnormal in 8.0% men and 0.4% women. HPV test was positive among 8.6% women. None of the women participating in the study refused or failed to collect the vaginal samples. Only 32 samples were found to be inadequate due to leakage of the contents during transport; 30 of them could be re-collected from the women. Only 28 (0.6%) women complained of breast lump or nipple discharge while they were visited by the CHWs.

VIA was performed on 76.5% (328/429) HPV-positive women. Only 46 women (14.0%) were VIA-positive and were biopsied. Histopathology reports were normal in 26; CIN 1 in 9; CIN 2 in 6; CIN 3 in 4; and invasive squamous cell cancer in 1. Thermal-ablation was performed on all 26 women eligible for ablative treatment. Women with CIN2+ diagnosis on histopathology and not treated by thermal-ablation were treated at GBHMCH.

A single case of fibroadenoma was detected among the 28 women with breast symptoms. Rest did not have any significant abnormalities on clinical examination and no further investigations were advised. None of the 181 men or women examined for positive OVE had any growth or ulcer in the oral cavity. They were counseled to quit tobacco and alcohol.

For the diabetes endpoint, the risk significantly increased with increasing age (ORs ranging from 1.6 to 4.6 compared to those aged 30-34 years) with a significant dose-response relationship (p-value for trend <0.001). The risk was significantly higher in the overweight participants (OR=2.1; 95% CI=1.4-3.1) and the obese (OR=4.2; 95% CI=2.7-6.3) compared to those having normal weight, again with a significant dose response relationship (p-value for trend=<0.001).

Detection of a large number of previously unknown hypertensives and diabetics, successful screening and treatment of women for cervical neoplasias, low refusals to screening and high compliance of the screen-positives to further assessment validate our model of providing NCD screening services at home by the CHWs. However, a major challenge to scale up such an approach is that the CHWs in the existent health services are engaged in providing various other services, mostly related to reproductive and child health. While the CHWs in our study could dedicate full time to the screening activities, the CHWs in the routine health services already over-burdened with many health programs will have burn out effects. It is important for the health services to make some vertical investments and build a dedicated work force providing NCD services. Tamilnadu is the only state in India that has been able to successfully implement mass-scale NCD screening, supported initially by the World Bank. The program invested heavily in engaging and training dedicated nurses, multi-purpose health workers and auxiliary nurse midwives, not only to provide NCD services but also for community mobilization, record maintenance and data management [7].

The access to the primary care in the rural areas in the LMICs is limited by a combination of factors such as poor physical infrastructure, lack of trained providers, inadequate financing and scarcity of transport. The facilities designated to provide the basic care at the community level are often under-staffed and under-equipped and have low uptake of the services due to sub-optimal quality. The Family Health Survey (2015-16) in Rajasthan, India (where our study was implemented) reported that only 3% of the rural households visited the sub-centres during any illness [8]. While people may be ready to accept the extra cost and hardships and loss of wages to reach the higher centres situated far away during sickness, the apparently healthy individuals targeted for screening are unlikely to bear the cost and the hardships for the preventive services [9].

The problem of accessing health care is even more complex for the women. Women in rural India like in many other LMICs are the victims of health inequity as a consequence of discriminatory beliefs and practices, lack of awareness (knowledge of women, their families and the health care providers about the existence of a health problem) and acknowledgement (recognition that something should and can be done about the health problem) [10]. The NCD screening program to succeed in LMICs need to deliver the services at the doorsteps of the beneficiaries. To deliver NCD prevention and early detection services at the doorsteps, the CHWs are the most key members of the healthcare team.

CHWs have been successfully utilized in many countries to facilitate access to care, increase quality of service and improve healthcare outcomes [11]. A systematic review demonstrated that the task-shifting from physicians to health workers, with appropriate health system re-structuring is viable as well as cost-effective in improving access to healthcare for NCDs in the LMICs [12]. While in other studies, the utilization of CHWs has been limited to screening for hypertension and diabetes, we have demonstrated how the cancer early detection services can also be delivered by them at the doorsteps. The recently launched WHO guideline on health policy and system support to optimize CHW programs highlights the evolving role of CHWs over time with changes in the epidemiological profile of the diseases and the requirements of the health system [13]. The services of the CHWs need to be aligned to the growing global burden of the NCDs and it may be worthwhile to consider building a separate cadre of ‘NCD workers’ to improve the quality and reach of the NCD control services in the community.

Our approach to screening for oral, breast and cervical cancers was different from the recommendations of the Indian guidelines. The Indian guidelines recommend oral screening for all, breast cancer screening by clinical breast examination (CBE) and cervical cancer screening by VIA. We limited OVE to the ever-users of tobacco/alcohol, since oral cancer screening is ineffective in the non-users [14]. We made the women aware of early symptoms of breast cancer rather than screening them with CBE. CBE screening has demonstrated significant downstaging of breast cancer but there is inadequate evidence for its impact on mortality reduction [15, 16]. We followed the early diagnosis approach (breast awareness combined with ready access to diagnosis and treatment) that has been recommended by World Health Organization as simple, effective and feasible in all LMICs [17]. We used HPV detection test that has proved to be significantly more efficacious than VIA and allows screening to be performed at home [18, 19].

Ours was the first Indian study to perform HPV testing of self-sampled vaginal specimens collected at home setting. In the earlier studies the women provided the samples in hospital-based or community-based clinics [20, 21]. The HPV positivity (8.6%) was comparable to the test positivity (7.7%) reported in the other Indian study that used CareHPV^TM to test self-collected samples. The compliance of the women was high, there were very few operational problems and the technology transfer was easy.

The comparability of our screening test results with other Indian studies and the high positive predictive values (PPV) of further assessment to detect hypertension or diabetes (42.3% for high BP and 35.0% for high sugar) reinforce the validity of our approach. The Integrated Disease Surveillance Project (2007-2008) by the Government of India reported high BP ranging from 18% to 26% among rural population (age 15-64 years) across different provinces [22]. Another nationally representative, population-based study screened a total of 1.3 million Indian adults (>18 years) for hypertension and diabetes [23]. The reported prevalence of hypertension (25.3%) and diabetes (7.5%) were similar to the values we observed.

If high referral rate for high BP (32.6%) is a concern, the rate could be drastically reduced by modifying the referral threshold to 160/90 mm, which also resulted in significant improvement in the PPV. In our study only 4.0% (277/6,985) participants had BP >160/90 mm at screening, and the PPV of high BP at screening at this threshold was 75.4% for the final diagnosis of hypertension. Benefit of revising the referral threshold has been observed by earlier studies and may be considered for population screening [24].

There are several unanswered questions in our study, especially related to the integration of the CHW-driven strategy in the larger program, e.g. improving compliance to further assessment, monitoring and quality assurance mechanism and cost-effectiveness in a scaled-up program. The missing follow-up data could potentially alter the test characteristics. We did not assess the impact of awareness or study the adherence to the recommended treatment, which are important factors for success of such a program. We have conducted a separate qualitative study to identify the structural barriers to implement CHW-driven NCD control in the study setting and also the knowledge, attitude and perceptions of the men and the women that could influence uptake of services. The results will be published separately. The generalisability of the study findings in India may be questioned given the heterogeneity in population characteristics and the health system organization. Considering the facts that 75% of the Indian populations still live in villages with problems in accessing health care, the CHWs are integral part of the health work force and have high acceptance in the community, the population has limited awareness of preventive health care - the CHW-driven approach has high relevance across the country, especially in the context of the revamped national NCD control program. Further evaluation studies are needed with an appropriate ‘standard of care control’ to quantify the benefits and harms of our approach, with an emphasis on patient-relevant long-term outcomes such as morbidity and mortality reduction and also cost-effectiveness.

To conclude, our pilot study demonstrated the feasibility of delivering NCD prevention and screening services at home by CHWs and the high acceptance of the services at least among our study population. The effectiveness of such an approach and the implications to the health services need to be carefully evaluated further.