Childhood stunting affects about 148·1 million children under 5 years of age worldwide, hindering their overall development⁽¹⁾. India accounts for approximately one-third of the world’s total population of stunted children^{(Reference Menon, Headey and Avula2)}. The predominant challenges faced to reduce stunting in children are poor maternal health, inappropriate infant and young child feeding practices^{(Reference Vir and Suri3,Reference Adhikari, Acharya and Upadhya4)} , infections^{(Reference Brown, Cairncross and Ensink5)}, sanitation issues^{(Reference Brown, Cairncross and Ensink5,Reference Mbuya and Humphrey6)} , poor coverage, delivery, access and utilisation of maternal, neonatal and child health and nutrition services provided by the government. Existing government services struggle to address these challenges^{(Reference Modi, Gopalan and Shah7,8)} . After the initiation of the National Rural Health Mission in 2005⁽⁸⁾, an evaluation of the community health workers, that is, Accredited Social Health Activists (ASHA), was conducted in 2011, and many shortcomings in training and monitoring were identified⁽⁹⁾. Therefore, to achieve the desired health impact in the community, there is need for targeted training and effective supervision of ASHA to ensure delivery of community health services and implement interventions.

mHealth interventions that utilise mobile phone technology hold the promise of improving healthcare delivery. Healthcare service coverage can be augmented by reducing outreach time^{(Reference Hurt, Walker and Campbell10–13)}. A systematic review has shown widespread acceptance of mobile technology in developing countries^{(Reference Hurt, Walker and Campbell10)}. In India’s healthcare system, separate components of mobile phone technology like apps to aid health workers, text or voice messaging, call centres, etc., have been used. However, it is rare to find an integrated system that incorporates all these separate components in one intervention package^{(12–Reference Murthy, Chandrasekharan and Prakash15)}.

Systematic assessments of gaps in the public health systems and knowledge of the cultural and social practices that may affect acceptance of the intervention by the community and the provider are needed to develop an effective intervention. We therefore conducted a series of pilot studies to understand the extent of mobile phone use in the community and women’s acceptance for mHealth intervention during antenatal and postnatal period in Maharashtra. Our first pilot was conducted in 2010 in four urban maternity hospitals, which tested mobile phone-based lactation counselling and health messaging via short text messages (SMS) to promote appropriate infant and young child feeding practices. The results demonstrated that in the mobile-based intervention group, breast-feeding initiation within an hour of birth was significantly higher (36·9 % v 23·6 % P < 0·001). The rates of exclusive breast-feeding were consistently above 95 % at all follow-up visits. Appropriate complementary feeding was observed in 99·6 % of the intervention group v 73·1 % in the control. Infants in the intervention had higher weight than infants in the control group at all follow-up visits. Additionally, 92·3 % of intervention participants expressed satisfaction with the mobile-based breast-feeding counselling^{(Reference Patel, Kuhite and Puranik16)}. Subsequently in a rural community-based qualitative research, we found that women and families were willing to use mobile phones and be contacted for counselling^{(Reference Khan, Patel and Puranik17)}. We then assessed the feasibility of using a CommCare app on JAVA phones by ASHA and its implementation challenges⁽¹⁸⁾. The pilot studies helped to design the comprehensive Behaviour Change Communication (BCC) intervention ‘Mobile Solutions Aiding Knowledge for Health Improvement’ or ‘M-SAKHI’ to impact health behaviours of pregnant women and mothers of infants in the selected rural communities to reduce infant stunting and to improve their development.

The description and methodology of evaluation of the effectiveness of M-SAKHI intervention in 2501 pregnant women using cluster-randomised controlled trial to reduce infant stunting was recently published^{(Reference Patel, Kuhite and Alam19)}. The women were enrolled before 20 weeks of pregnancy and followed up through delivery till their infants were 18 months old. The five components of M-SAKHI intervention were (i) an ‘ASHA’ app for real-time data collection and face-to-face counselling of participants during ASHA’s monthly home visits, (ii) text messages (thrice a week) and voice messages (once a week) sent to the participants by a server, (iii) automated delivery of alert text messages to participants, ASHA and study auxiliary nurse midwife (trained as counsellor), (iv) mobile phone to mobile phone direct counselling of participants every fortnight by study auxiliary nurse midwife counsellor and (v) a field supervision app to monitor ASHA field activities. The outcome indicators, data collection methods and sources are elaborated in Table 1.

Table 1 M-SAKHI outcome indicators

M-SAKHI, Mobile Solutions Aiding Knowledge for Health Improvement; FRO, field research officers; AFRO, field research officers specially trained to collect anthropometry, child development and dietary data; FS, field supervisor; ASHA, Accredited Social Health Activists.

* All data collection will be collected for both intervention and control arms, except the knowledge and performance of ASHA which will be conducted only in the intervention arms.

SOURCE: Patel AB, Kuhite PN, Alam A, et al. M-SAKHI – mobile health solutions to help community providers promote maternal and infant nutrition and health using a community-based cluster randomised controlled trial in rural India: A study protocol. Matern Child Nutr. 2019; e12850. https://doi.org/10.1111/mcn.12850.

In this manuscript, we explain the Program Impact Pathway (PIP) for the development of M-SAKHI using Theory of Change (ToC) approach and the methods for Process Evaluation. The ToC provides a structured method for mapping cause-and-effect relationships in social interventions, illuminating how specific activities drive desired outcomes^{(Reference Rogers20,21)} . The ToC approach identifies pivotal triggers for behavioural change and enables meticulous planning and evaluation of interventions. The PIP frameworks play a vital role in providing a structured framework for planning, monitoring and evaluating the implementation of complex interventions. The PIP ensures that activities are carried out as intended and helps to track the progress of the intervention. It helps to provide informed decisions for improvements in future implementations^{(Reference Le Port, Zongrone and Savy22)}. The Process Evaluation forms an integral part of any PIP analysis. It helps to identify which components of the intervention are likely to work, what will not work and which areas need to be strengthened^{(23,Reference Wynn, Dutta and Nelson24)} . BCC interventions involving mHealth are often resource-intensive and need continuous monitoring and evaluation to be successful. Publications describing the PIP for complex BCC interventions are scarce^{(Reference Briaux, Martin-Prevel and Carles25)}. The description of how the impact pathways are developed may contribute towards designing impact assessments of similar interventions.

Methods

The detailed description of how we developed the PIP using the ToC approach for M-SAKHI are provided below.

Step 1: identifying gaps in the health systems and strategies

We conducted a systematic review of literature and searched electronic bibliographic databases like PubMed, Embase, Web of Science and Scopus using key words like nutrition/ health/ mHealth intervention /rural India /ASHAs /ASHA evaluation /ASHA workload/gaps in service delivery, etc., to retrieve publications^{(Reference Modi, Gopalan and Shah7,9,Reference Guha, Raut and Maliye26–Reference Dongre, Deshmukh and Garg31)} . Additionally, the results from the series of pilot studies helped assess the cultural and social practices that may affect acceptance of the intervention by the community and the provider^{(Reference Patel, Kuhite and Puranik16–18)}. Table 2 enlists the major gaps in health system service delivery and factors responsible for infant stunting and how we planned to address these gaps through the M-SAKHI intervention.

Table 2 Gaps in existing health system and factors responsible for stunting (identified through literature review) and strategies to address these using M-SAKHI intervention

M-SAKHI, Mobile Solutions Aiding Knowledge for Health Improvement; ASHA, Accredited Social Health Activists; IFA, iron and folic acid.

Step 2: developing the Theory of Change model for the M-SAKHI intervention

The ToC model or the pathway to change was developed by ‘backward mapping’ from the desired goal of achieving reduction in rates of infant stunting^(32,33) . The pathway includes the outcomes needed to achieve the goal (of reducing stunting), the change in behaviour (the outputs) that enables the desired outcomes, the contextual factors at individual, societal and system level that need to be considered for attaining these outcomes, and the processes that need to be adopted to obtain the identified solutions (Table 3). The M-SAKHI outcomes were specific, reliable, measurable and time-bound indicators that are described in Table 1. The contextual factors at the individual and societal level were the extent of health awareness of the participant and cultural acceptance of the intervention by the family and community. At the health system level, barriers and facilitators for implementation were identified. It was assumed that if these contextual factors are addressed, then there will be an acceptable level of participation of the individual, the family and the ASHA^{(Reference Anderson34–Reference Taplin, Clark and Collins36)}. The processes and inputs were the five components of the intervention and the resources required for developing and implementing the intervention respectively.

Table 3 The Theory of Change hypothesising behaviour change in communities through M-SAKHI intervention

M-SAKHI, Mobile Solutions Aiding Knowledge for Health Improvement; ASHA, Accredited Social Health Activists; IFA, iron and folic acid.

Step 3: consultative process with the stakeholders for designing the Program Impact Pathway

The preliminary framework for the PIP was conceptualised before the start of programme by the study team at Lata Medical Research Foundation and University of Sydney. The framework was developed by using lessons from the previous pilot studies^{(Reference Patel, Kuhite and Puranik16–18)}, consultations with international experts in PIP, dialogue with key stakeholders, implementers and the recipients of the programme. The stakeholders were private mobile phone companies, the technology providers, public health systems, public health academicians and experts and the donor organisations. This draft was then refined with the help of programme implementation experts at the Nutrition Embedding Evaluation Program – of PATH (global health organisation), global meeting held in Barcelona, Spain^{(Reference Borja37)}. Components of the programme contributing towards the goal were enumerated, and their pathways to achieve the impact were identified. The intervention components were further refined contextually in the consultative process, based on community-based qualitative research. A Process Evaluation plan aligning with the PIP was developed to evaluate the impact of the M-SAKHI intervention.

Results

The M-SAKHI Program Impact Pathway framework

This framework (Fig. 1) was developed using the steps mentioned above. The ‘Inputs’ included resources needed to set up the systems for the intervention. The ‘Processes’ were the activities that were enabled by the intervention such as ASHA app for face-to-face counselling, server delivered text messages, voice messages, automated delivery of alert text messages and mobile phone counselling by study auxiliary nurse midwife counsellor. These processes are expected to achieve the ‘Outputs’, that is, improved awareness of mothers regarding nutrition, water, hygiene and sanitation, improved awareness regarding her health, and improved maternal and infant immunisations. The outputs for the community health workers included improved awareness and motivation of ASHA to help mothers adopt healthy behaviour and facilitate timely referrals to healthcare centres when needed. The improved awareness regarding health behaviour and nutrition is expected to result in ‘Outcomes’, that is, improvement in practices pertaining to maternal nutrition and maternal weight; improved infant and young child feeding practices and thus improved infant weight; improved hygiene and sanitation practices; early recognition of illnesses/danger signs; improved health-seeking behaviour and utilisation of health services (antenatal and postnatal care); decreased maternal and infant morbidities; and decreased preterm births and low birth weight babies. For the community health workers, the outcomes were improved knowledge and skills and improved implementation and service delivery by ASHA.

Fig. 1 M-SAKHI Program Impact Pathway framework. M-SAKHI, Mobile Solutions Aiding Knowledge for Health Improvement

As shown in Fig. 1, the ‘Inputs, Processes, Output and Outcomes’ are expected to result in the final ‘Impact’ of reducing stunting and improving development of infants at the age of 18 months through five pathways (represented by dark blue coloured boxes in Fig.1). The pathways are (1) improving maternal and infant nutrition, (2) early recognition of maternal and infant danger signs, (3) improving access and utilisation to healthcare services, (4) improving hygiene, sanitation and immunisation practices and (5) improving implementation and service delivery of community health workers through their training, monitoring and supervision in real time.

The Process Evaluation plan

This plan is to evaluate the following seven attributes of the M-SAKHI intervention: (i) fidelity – this is the extent to which the intervention will be implemented as planned. (ii) Dose delivered – this will be evaluated by assessing the number of activities/deliverables. (iii) Dose received (exposure) – this will be evaluated by assessing the extent to which participants actively engage with, interact with and/or use M-SAKHI programme components and activities. (iv) Dose received (satisfaction) – this will be evaluated by assessing the participants’ satisfaction with the programme and the interaction with those who deliver it. (v) Reach – this will be assessed by proportion of target group that participated in the programme. (vi) Recruitment – this will be assessed using methods of communicating and recruiting participants, and maintaining their participation. (vii) Contextual factors will consider the elements of the physical, social, cultural and political environment that influence implementation or outcomes. All seven attributes will be assessed for each of the implementation channels of the intervention, namely – use of ASHA app for face-to-face counselling, phone counselling by study auxiliary nurse midwife counsellor, text, alert messages and voice messages and the field supervisory app. The data sources to be used will be qualitative based on interviews or quantitative based on data regarding field implementation activities, that is, available on the server.