Mistreatment, discrimination, and poor psycho-social support during childbirth at health facilities are common in lower- and middle-income countries. Despite a policy directive from the World Health Organisation (WHO), no operational model exists that effectively demonstrates incorporation of these guidelines in routine facility-based maternity services. This early-phase implementation research aims to develop, implement, and test the feasibility of a service-delivery strategy to promote the culture of supportive and dignified maternity care (SDMC) at public health facilities.

Guided by human-centred design approach, the implementation of this study will be divided into two phases: development of intervention, and implementing and testing feasibility. The service-delivery intervention will be co-created along with relevant stakeholders and informed by contextual evidence that is generated through formative research. It will include capacity-building of maternity teams, and the improvement of governance and accountability mechanisms within public health facilities. The technical content will be primarily based on WHO's intrapartum care guidelines and mental health Gap Action Programme (mhGAP) materials.

A mixed-method, pre-post design will be used for feasibility assessment. The intervention will be implemented at six secondary-level healthcare facilities in two districts of southern Sindh, Pakistan. Data from multiple sources will be collected before, during and after the implementation of the intervention. We will assess the coverage of the intervention, understanding and attitude of maternity team, and implementation challenges faced. Additionally, we will also gather women's maternity experiences and psycho-social well-being that will also inform the success of the intervention.

Evidence from this implementation research will enhance understanding of health systems challenges and opportunities around SDMC. A key output from this research will be the SDMC service-delivery package, comprising a comprehensive training package (on inclusive, supportive and dignified maternity care) and a field-tested strategy to ensure implementation of recommended practices in routine, facility-based maternity care. Adaptation, Implementation and evaluation of SDMC package in diverse setting will be way forward.

Objectives:

Design a contextually-informed (or evidence-based) service-delivery model through a participatory, consensus-driven process for inclusive, supportive and dignified maternity care; and Implement and test the feasibility of the intervention model for its content and operationalisation in public health facilities

Study phases:

Implementation phases:

The implementation of this study will be divided into two phases: (1) Development of intervention: grounded in formative research around supportive and dignified care, participatory approach will be used to co-create service-delivery intervention package; and (2) Implementing and testing feasibility: a two-level feasibility assessment will be carried out - health systems and staff assessment, and women's experiences around SDMC.

We will use the framework of human-centred design (HCD) to guide implementation of this study (see figure 3). Human-centred design is an emerging concept and being used increasingly in global health . The approach is classified in three phases: inspiration, ideation, and implementation and testing.

Objective 1 - Design a contextually informed (or evidence-based) service-delivery model through a participatory, consensus-driven process for inclusive, supportive and dignified maternity care:

Phase 1: Inspiration:

The first step is 'empathise' - whereby, through a formative research we will understand prevalent behaviours, challenges and opportunities around SDMC. It will be supplemented by review of national and international guidelines, and best practices around the topic in focus. We will begin with the participatory approach to engage with key stakeholders, including practitioners, programme managers, policy makers, and community representatives to ensure contribution and ownership of the project.

Formative research:

Focus Group Discussions (FGDs):

A total of five to eight FGDs will be conducted with postpartum women who have delivered in public health facility during the last three months. These women will be identified from a list frame to be obtained from the selected health facility. The FGDs will broadly explore: (a) attitudes towards facility-based births; (b) motivations to give birth at the facility; (c) expectations regarding maternity care; and (d) perceived quality of care with according to the Hulton framework for quality of care in maternity services . The framework focuses on both 'provision of care' and 'experience of care' (human and physical resources, cognitive support, respect and dignity, equity, emotional support). A trained sociologist/anthropologist will conduct FGDs who will be accompanied by note-taker. Informed consent will be taken from participants of FGDs and all discussions will be audio-recorded. The data generated from FGD will inform development of SDMC intervention package and contextualised structured tools to assess experiences of SDMC at healthcare facilities.

Health facility assessment:

The assessment of health facility will comprise of: a) interviews with health system staff (clinical and non-clinical service providers and administrators), b) review of management information system, and c) observation of client-provider interactions.

• Indepth interviews (IDIs) with staff: IDIs will be conducted with clinical and non-clinical maternity staff and administrators to broadly understand challenges and opportunities around SDMC. The COM-B behavioural framework will guide the the development of interview guide and data synthesis to understand individual and systemic factors around motivation, capabilities and opportunities for supportive and dignified maternity care.

Observations:

• Review of management information system (MIS): We will review the kind of information documented on inpatient records and assess mechanisms to use this data to make informed decisions. Understanding MIS will enable the investigators to create or strenthen mechanism for record keeping on incidents of any mistreatment as a measure of accountability.

Review of relevant national and international material:

We will conduct a thorough review of existing national and international material that is pertinent to the supportive and dignified maternity care. This would broadly include: existing national service delivery guidelines for maternity care during labour and childbirth; WHO recommendations for intrapartum care; mhGAP material for neurological disorders and substance abuse; existing training material on respectful maternity care.

Phase 2: Ideation:

In second phase, data synthesis will generate numerous issues and their possible solutions around supportive and dignified maternity care. These solutions/ideas will then be discussed with stakeholders to select the ones that are most relevant and applicable in the given context, around which the prototype will be formally developed. The prototype will undergo provisional testing through a collective review with stakeholders - who will reflect on its relevance and operational feasibility in their routine practice.

Objective 2: Implement and test the feasibility of the intervention model for its content and operationalisation in public health facilities

Phase 3: Implementation and testing:

The third phase of HCD fall within the scope of second study objective. In phase three, the finalisation of prototype will be followed by pilot testing in real-life settings over a period of three months. It is important to note that this implementation is will be considered pilot as the field learnings will be fed back into design of intervention package.

Evaluation design:

A mixed-method, pre-post design will be used to test feasibility of the intervention prototype.

Study settings:

The study will be implemented at six public health facilities - with three facilities in each Thatta and Sujawal districts of southern Sindh province, Pakistan. These contagious districts are located approximately 100 kilometres away from megacity, Karachi.

Study population:

The primary target-audiences will be women in childbirth and service providers (clinical and non-clinical staff) of the labour room / maternity wards.

Health system and staff assessments:

Quantitative methods:

Baseline assessment will follow the capacity-building of clinical and support staff of the health facility in training modules. The maternity team will receive comprehensive 3-day training on SDMC that is tailored to their contextual settings. Senior staff members will become master trainers, responsible for providing supportive supervision to ensure adherence to set standards. Feedback from staff will be sought during the training about the relevance of content and its implementation feasibility.

The knowledge and attitude of training participants (clinical and non-clinical staff and administrators) regarding SDMC will be assessed through pre-and post-training tests. The tools will be developed in accordance with the intervention package, after it is finalised.

Qualitative methods:

During the intervention period quality audits (or supportive supervision visits) will be routinely conducted on all selected health facilities to assess adoption of, and adherence to, quality standards through observation of women-provider interactions, staff coordination, and facility assessments. The maternity staff will be provided with customised feedback based on the observation findings.

Post-intervention, in-depth interviews will be conducted with practioners and programme managers to explore intervention fidelity, dose and coverage, penetration of SDMC practices across cadre, and individual, interpersonal, and system-level facilitators and barriers that were encountered during delivery of the intervention in routine practice. We will also gather their suggestions for improvement in the SDMC intervention package.

Women's experiences:

Quantitative methods:

Pre-post intervention assessments will gather information on women's experiences of SDMC at health facilities, and suffering of postpartum depression within 42 days of childbirth. Unlike health system assessment, two independent group of women will be enrolled in baseline and endline. Recently published structured questionnaire on mistreatment during childbirth will be contextually adapted and used for these pre-post assessments . Postpartum depression and anxiety will be assessed using Patient Health Questionnaire - 9 and Generalised Anxiety Disorder - 7, respectively.

Sample size calculation:

A total of 308 postpartum women will be required to detect a statistical difference of 2.5-units decrease (or 10% reduction) in the mean score of mistreatment (from 25 to 22.5) with standard deviation of 6 , 90% power, 5% level of significance, 2 design effect, and 10% lost-to-follow-up (LTFU) and/or non-response. We decided to proceed with a sample of 308 as it gives sufficient precision to perform range of additional analysis such as estimation of prevalence of different types of mistreatment and the effect of intervention on each of these types of mistreatment, while will not negatively affecting study timelines.

Sampling strategy:

A convenience sampling technique will be used for the selection of study participants. All women giving births at the selected hospitals during the data collection period will be invited to participate in the survey, until the desired sample is achieved. During the consent process, participants will be informed and asked whether they would be interested in participating in a follow-up interview in their homes at six weeks of postpartum. Women who showed interest in home-based interview and gave consent will be recruited in the survey. Participants who lived outside of the study district or even in hard-to-reach rural areas of the district will be excluded from this sample due to logistical constraints.

Data collection and management:

Standardised and contextually-adapted tools will be used in this study. All data collection tools will be translated in the relevant local language and pre-tested in similar settings prior to being used. Data collectors with prior experience will be hired and trained for baseline and endline assessments, and routine audits. Indepth interviews (IDIs) and focus group discussions (FGDs) will be conducted by trained sociologist and/or anthropologist. All data collectors will receive a comprehensive and standardised training on both theoretical and practical aspect of study instruments.

Qualitative interviews will be audio-recorded for later transcription and translation. All quantitative data will be collected electronically on tablets. The software application, developed in Epicollect5, will have built-in validation checks to minimize data collection and entry errors. All software applications will be tested prior to their use in the study. Unique IDs will be given to each study participant which will be consistent (where applicable) in across assessments for synchronisation.

Data analysis:

Descriptive statistics will be used to describe the characteristics of study participants such as age, level of education, socio-economic status etc. In order observed change in women's experiences of SDMC, advanced statistical techniques will be used to estimate changes in outcome indicators (SDMC experience) by means of multivariable linear (continuous) and logistic (binary) regression models, accounting for intra-class correlation within health facilities. We will create a composite score for women's experiences of SDMC which will treated as outcome for pre-post comparison using linear regression. Depending on distribution of the variable, we will consider dichotomising it to see pre-post intervention changes in prevalence of SDMC experiences using logistic regression. Paired t-test (for continuous) and McNemars test (for binary) will be used to observe change in knowledge and attitudes of health facility staff before and after training. Stata version 16.1 (StataCorp LP, Texas, United States) will be used for all analyses, and p-values of 0.05 will be considered statistically significant.

For the qualitative enquires of formative research, we will use bottle-neck analysis approach to identify prevalent issues that impedes provision of SDMC in health facilities, and also determine the opportunties that can be leverage to promote SDMC. These findings will inform the development of intervention package. The qualitative data to be collected post-intervention, will be analysed to understand feasibility and acceptability of SDMC intervention, and related barriers and facilitators of implementation. We will also analyse suggestions of study participants to make improvements in the intervention package.

Postnatal women

Inclusion Criteria:

• Gave birth in study site (health facility) during data collection period
• Gave informed consent

Exclusion Criteria:

• Reside outside of the demarcated catchment area

Health facility staff

Inclusion Criteria:

• Clinical and non-clinical staff working in maternity section of selected health facility
• Gave informed consent

Exclusion Criteria:

• Recently appointed at the health facility (< than 6 months)

Study datasets for the Supportive and Dignified Maternity Care will be made available on LSHTM Data Compass, a research data repository operated by the London School of Hygiene and Tropical Medicine.

Two years after the completion of study.

To protect participant confidentiality, the de-identified datasets are made available through a controlled access approach. Researchers wishing to access datasets will be asked to apply for access via the repository's data request form, providing information on the variables they wish to access and details of their analysis plan. The request will be sent to the study team (that performed the research and have greatest understanding of the data) and the LSHTM Research Data Manager (who acts as an independent advisor). If the data analysis can be performed in compliance with the study's ethical and legal requirements, the study team will produce a derived dataset that contains the requested variables and work with the applicant to help them to understand the data. If there remains a recognisable risk that the derived dataset contains potentially identifiable information, the applicant will be asked to sign a Data Sharing Agreement before being provided with the dataset.