A systematic review of population health interventions and Scheduled Tribes in India

Population health interventions include policies or programs - either within or outside of the health sector - that address the underlying causes of ill health embedded in behaviours, contexts and systems of social stratification that increase risk and vulnerability. Population health intervention research encompasses "attempts to capture the value and differential effect of these interventions, the processes by which they bring about change and the contexts within which they work best."[[5], p.I-8] We systematically searched and assessed peer-reviewed literature of population health interventions research involving STs in order to inform policy and to identify important research gaps. The aim of this review was to identify the public health interventions or components of these interventions that are effective in improving population health outcomes (reducing morbidity or mortality rates and reducing risks of ill health) among ST populations in India.

Scheduled Tribes across India

According to India's most recent census in 2001, there are 84.3 million STs, which is 8.2% of the population [6]. The distributions of ST populations varies widely across India's states and territories. (Figure 1) In Mizoram and Lakshadweep, STs represent close to 95% of the population whereas in Kerala and Tamil Nadu STs represent only 1% of the population. Among the total ST population in India, the highest proportions are found in Madhya Pradesh (14.5%), Maharashtra (10.2%), and Orissa (9.7%). There are around 700 different tribes living across India, predominantly in remote areas: forests, hills, and rough terrain in plateau areas [3]. There is great heterogeneity across different tribal groups, including a sub-category of particularly vulnerable STs known as "primitive" tribes (due to the derogatory nature of this term, these groups are being renamed by the Government of India as 'particularly vulnerable tribal groups') [3].

The search was restricted to peer-reviewed articles published in English and French languages during the period of 1990 to 2009. The search was undertaken by the first author during the month of October 2009. Since the health of STs are both a public health issue and a development issue, we searched databases in both the health and social science: PubMed, EconLit, and Social Sciences Index and Abstract. The grey literature was not included for several reasons. First, given India's size and diversity (including languages), it would be extremely difficult to systematically identify and review the literature in a timely manner. Second, as the focus of the review is on intervention research, which requires rigorous methods, the quality of much of the available grey literature would likely be inadequate. Third, in our experience, local nongovernmental organizations have among the greatest expertise in working with ST populations, but they tend not to systematically document their knowledge. Tapping into this evidence base will take innovative approaches (e.g. collaboration of NGOs, ST communities, and academics in writing projects, which was done as part of Paniya Sadas, a forum held on marginalised STs in South India) that were beyond the scope of this review.

Four inclusion criteria were to be met for articles to be included in our review. These were: (a) evaluations or intervention studies of a population health intervention, including surveillance systems, health promotion or prevention programs, and treatment programs, (b) undertaken with an ST population or in a tribal area (where the majority of the population have ST affiliations) in India, (c) with a population health outcome(s), which included morbidity or mortality, and risks of ill health and (d) involving primary data collection. Studies that addressed genetic disorders or evaluated a drug or plant for therapeutic properties were excluded. The review of articles proceeded in four steps. First, we conducted our searches using the following abstract key words: (a) Trib*, (b), India (c) health, HIV, tuberculosis, malaria, nutrition, alcohol, tobacco, smok*. Since population health interventions can be defined using a variety of terms, we did not include a term for intervention. We searched the literature for articles that included Trib* (population) and India (setting) and one health outcome, leaving the search for intervention studies in the following step. Second, titles and abstract were screened to meet inclusion criteria, including whether the study assessed a public health intervention. Third, guided by Critical Appraisal Skills Programme (CASP), potential articles were further scrutinised for quality. Any study with a major flaw in design or methodology (the intervention failed to meet either the CASP screening questions or 2 or more of the detailed questions) was excluded. Fourth, the references of the selected articles were reviewed to identify other articles not found in our search. The first author reviewed and prepared critical appraisal summaries for the papers, then, the second author reviewed these summaries, followed by a discussion and synthesis of the findings by both authors.

Information was collected on the description of the study, including the intervention, objectives of the study, region of study, study design, methods and study population (including the specific tribal groups). We further abstracted information on ethical considerations of the study and the nature of ST participation in the research process. Finally, information on the main findings and conclusions were retained.

Despite the large disparities in health between ST and non ST populations in India, we identified only a small number of articles that examined interventions to improve the health of STs. Furthermore, the studies in this review included only disease-specific interventions; we did not identify any article that assessed a comprehensive health intervention for STs. Finally, assessing public health interventions requires rigorous methods and the majority of studies were either descriptive or had a weak study design, even after rejecting those that were more seriously limited. Implementing stronger quasi-experimental designs that address major biases, such as selection bias, would have provided stronger evidence [42]. However, experimental and quasi-experimental designs may not be feasible or appropriate to assess some interventions, notably complex interventions with multiple interacting components [43]. Researchers need to consider various 'trade-offs' that may arise in the pursuit of the best choice of design, given the type of intervention, the populations, the context, and the availability of resources [43, 44]. But regardless of study design, there are steps researchers can take to ensure their approach is rigorous and the study is of good quality [45]. In our review, we found the evidence collected to be limited in its quantity, scope, and methodological rigour, and thus should be interpreted with caution. Reviews of the effectiveness of public health interventions in other contexts with indigenous populations have similarly found an 'underdeveloped' evidence base in terms of both the number of available studies and the rigour of the studies [46–48].

The evidence compiled in this review revealed three issues that promote effective public health interventions with STs. First, there is a need to develop and implement interventions that are low-cost, give rapid results and can be easily administered. This addresses the challenge of delivering effective interventions to ST communities who predominantly live in remote areas, where there is limited access to health care, few laboratory facilities, inadequate surveillance, poor vital registrations, and few skilled workers to implement highly technical tasks. Innovative technologies, such as rapid diagnosis tests that do not require laboratories and GIS, offer new opportunities. But by moving one step further and connecting innovative technologies with ST culture and harnessing STs' capacity to use these technologies, could increase opportunities for STs to develop new skills and increase their control over interventions.

Second, even though the study included interventions that addressed a specific disease or a single health need, a multi-pronged approach was advanced by several authors based largely on limitations to the intervention design identified by their studies. ST populations face multiple risks of ill health and are exposed to a number of diseases. Combining two or more activities in a single intervention will likely improve health outcomes. This has been demonstrated among other indigenous groups and vulnerable populations in low-income countries [49]. Given the high levels of health need there is a need for a public health (as opposed to a medical or disease approach) to improve the health of STs and reduce the disparities in health that exist between STs and other social groups. For policy-makers considering new intervention designs, the challenge is to retain the 'quick-wins' of low-cost and easy administration, but always with an eye to adding more elements to a program to make it more comprehensive and capable of responding to health determinants residing in the contexts of STs' cultural, geographic and economic environments.

Third, the involvement of ST populations in the intervention was advocated to help address the isolation of many ST communities, the cultural specificities of STs, the limited resources available and to promote community control. Involving ST populations may improve current programs that have demonstrated poor performance. For example, an increase in the prevalence of TB between 1986 and 2002 was observed in Car Nicobar (the administrative headquarters for the Nicobar district in the Andaman and Nicobar Islands), despite the implementation of a national TB program [38]. The authors argue that this increase may be attributed to the lack of a district level TB control program which contributed to insufficient TB control in the area. A district level program as part of the national one could allow for ST participation in its design, improving its community acceptance by, and cultural appropriateness for, the local population. The participation of indigenous populations in public health interventions has been found to be an effective strategy in other contexts [47].

The findings of this review identified three effective strategies for improving health outcomes among ST populations in India: low-cost, rapid results, and easily administered programs, multi-pronged approaches, and including ST participation in the intervention. The evidence base, however, is insufficient. There is a need for a better understanding of how to improve their health by pursuing public health intervention research appropriate for ST populations.