Quantitative component.

Married women aged 18–49 years residing in the selected study block will be selected for quantitative household survey. In addition, all public health facilities within the study block (like: Community Health Centres (CHCs), Primary Health Centres (PHCs), Rural Hospitals, and Sub-District Hospitals (SDHs)) will be included for facility assessments.

Qualitative component.

For the qualitative data collection, married women aged 18–49 years who have experienced an abortion within the preceding two years (identified from the household survey) will be purposively selected to explore their experiences in availing abortion services. Additionally, key stakeholders across the health system will be included to understand facilitators and barriers to abortion service delivery. These will include healthcare professionals such as Community Health Officers (CHOs), Auxiliary Nurse Midwives (ANMs), Gynaecologists, Medical Officers, Staff Nurses, District Health Authorities, Facility Managers, and community health workers including Accredited Social Health Activists (ASHAs) and Anganwadi Workers (AWWs).

Quantitative sample size.

Six MRHRUs were purposively selected to represent diverse geographic regions of the country. For the community-based survey, the sample size was calculated assuming a 50% prevalence of awareness of safe abortion services due to the absence of relevant Indian literature, with a 95% confidence level and 5% absolute precision. This yielded a base sample size of approximately 384 married women. After accounting for a conservatively assumed design effect of 2, specified a priory for sample size planning, and a 10% non-response rate, the final target sample size was estimated at 854 married women aged 18–49 years per site. As abortion is a sensitive subject, We are conducting a household based survey, some eligible participants may decline participation or may not feel comfortable responding during household-based interviews. To address this possibility and ensure adequate sample size, a conservative non-response rate of 10% was considered. A multistage cluster sampling design will be used. The sampling frame will be prepared using Accredited Social Health Activist (ASHA) lists of married women aged 18–49 years, which will serve as the measure of size (MOS) for each village. Based on the calculated total sample size and the decision to conduct the study across 20 clusters, it was determined that 42 interviews would need to be conducted per cluster to achieve the required overall sample size.

In the first stage: 20 villages (clusters) will be selected using probability proportional to size (PPS) sampling, employing a random start and fixed sampling interval. For, MRHRU-Khumulwng (Tripura), hamlets will be considered as cluster, as the hamlets serves as independent settlement units. In the second stage: a complete household listing will be undertaken in each village to identify household with at least one eligible women. A systematic sampling technique will be used to select 43 eligible women per village using a random start and fixed interval. Households will be approached using the right-hand rule, and when more than one eligible woman present, one will be selected using the Kish grid or next birthday method [18]. In villages with fewer than required number of sample, all will be included and the remaining sample will be drawn from the nearest village/hamlet or an adjacent village within the same Gram Panchayat.

For the facility assessment survey, a total enumeration sampling strategy will be applied, whereby all public health facilities located within the study blocks will be included.

Qualitative sampling.

Participants for the qualitative data collection will be selected using purposive sampling with a maximum variation approach to ensure representation across facility types, provider cadres, and geographic contexts. The sample will include healthcare providers and women aged 18 –49 years who have either accessed abortion services from the public health facilities or availed abortion care elsewhere within the past two years. At each MRHRU site, approximately 25 healthcare providers and 15 care recipients or seekers will be sampled. Sampling will be guided by the principle of thematic saturation [18,19] within individual sites and across six MRHRU sites representing different regions. Overall, approximately 232–240 qualitative interviews are proposed, with final sample size guided by thematic saturation and contextual requirements across sites. In addition, 18 focus-group-discussion (three per site) comprising Accredited Social Health Activists (ASHA) and Anganwadi Workers (AWW) will be conducted to capture frontline health worker’s perspectives. This recruitment method will facilitate varied samples and allow for adaptability. The detailed distribution of the qualitative sample is provided in Table 2.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Sample size for qualitative data collection across six MRHRU Sites.

https://doi.org/10.1371/journal.pone.0358274.t002

Quantitative data.

Before collecting data, all the study participants will be informed about the purpose of the study and written informed consent will be obtained. The research instruments will be administered by trained field staff across all sites. A structured household survey questionnaire will be used to collect information on socio-demographic characteristics (such as age, residence, education, occupation, socio-economic status), reproductive history (i.e., gravida, parity, number of living child and, age of living child), pregnancy outcomes (such as: spontaneous abortion, induced abortion, stillbirth), knowledge and awareness of abortion services and options, utilization of abortion services, knowledge about danger signs, post abortion care, contraceptive use, and barriers faced in accessing care.

The questionnaire underwent a multi-step development and validation process, including internal review by the Principal Investigators of each study site, external expert review by national-level public health and qualitative research specialists, and validation through a one-day in-person workshop. Following these reviews, the tools were pilot-tested in the field to assess clarity, comprehensibility, and contextual relevance. Feedback from piloting was incorporated, and final refinements were made to ensure cultural and linguistic appropriateness across all study sites. The finalized tools were then translated into local languages using forward and back-translation to maintain conceptual equivalence and consistency. All quantitative data will be captured electronically using KoboCollect and stored on a secure, password-protected cloud server with restricted access, in accordance with data protection and confidentiality protocols.

Qualitative data.

The qualitative data will be collected through Key informant interviews (KIIs), in-depth interviews (IDIs), and focus group discussions (FGDs). Along with this, observation method will be used to document the roll out of the Comprehensive Abortion Care (CAC) guideline at facility level. Separate semi-structured interview guides for each category of service providers have been developed using a staged process informed by the Comprehensive Abortion Care (CAC) Guidelines (2023). The guides covered components such as provider roles, service availability, referral mechanisms, post-abortion care, community responses, barriers to access, and service improvement. Along with KIIs, a separate Focus Group Discussions (FGDs) guide (comprising knowledge and awareness about abortion, prevalent abortion care practices and the social norms and stigma in the community) has been prepared to collect data from the Accredited Social Health Activists (ASHAs) and Anganwadi workers who act as the community health worker and are responsible for providing counselling and referral to health care facilities. The study tools were developed using a systematic and rigorous approach informed by an extensive literature review, consultations with national-level experts in the subject area, and qualitative research. Key literature findings that informed guide development included decision-making processes, information pathways, care-seeking trajectories, quality of care, stigma, and other contextual barriers and facilitators influencing access and experiences of care. The CFIR framework also informed the development of the guides to ensure that relevant implementation domains and contextual factors were adequately captured. All tools underwent internal review, external expert review, validation through an in-person workshop, piloting, and contextual refinement. Guides were translated into local languages using forward and back-translation. Pretesting of the tool was done in a separate block.

Trained researchers (preferably female) will collect qualitative data with minimum graduate and post graduate degree in social science, or public health or related discipline, and demonstrated experience in qualitative research and sensitive interviewing. Prior experience in community based data collection will be considered during recruitment. A multi-level capacity-building approach will be adopted for the selection and deployment of qualitative data collection teams across study sites. In the first phase, investigators and study coordinators from all participating sites will undergo orientation and capacity building workshop. Subsequently, the trained personnel will recruit the research team at their respective sites.

Following recruitment, a comprehensive two-day in-person training workshop will be conducted for the research teams from all sites at a central location. The teams will be trained in qualitative research methods, interviewing techniques, ethical conduct, informed consent, confidentiality, non-judgmental interviewing techniques, probing techniques, and management of sensitive conversations and participant distress. Training will be delivered by a multidisciplinary team with significant expertise in sensitive research areas, including reproductive health and maternal mental health, with backgrounds in community medicine, social anthropology, and ethnographic research. Following initial training, the research team will continue to receive regular orientation, mentoring, and support through reflective debriefing sessions conducted both online and onsite throughout the data collection period to strengthen consistency, address field challenges, and enhance reflexive practice.

The interviewer will be familiar with the local language and socio-cultural context but will have no prior relationship with study participants before or during the data collection process, thereby minimizing potential bias and facilitating independent and ethically appropriate engagement.

To ensure participants’ full participation, comfort, privacy and confidentiality, only the study team and respective participants will be present at the time of data collection. Presence of non-participants will be avoided during the interview unless it seems essential due to the safety of the researcher or the researched. Interviews and FGDs will be conducted in private settings and at the place of participant’s choice to minimize any reputational risk. In cases where women may not be able to complete the interview due to privacy or any other reasons and wishes to continue later, repeat interviews will be carried out as per their convenience. Written or verbal informed consent will be obtained. Discussions will be audio-recorded with permission; otherwise, detailed field notes will be maintained. Reflexive field notes will be documented throughout data collection.

All interviews will range between 45–60 minutes on average depending on willingness and availability of the participants and will be audio recorded after obtaining either a written or verbal consent, depending on the comfort and willingness of the participant. Participants across study groups will be provided a participant information sheet in both English and the regional language. Throughout the data collection and analysis period, the research team will be engaged in bracketing exercise to reflect on and acknowledge their own assumptions, prior knowledge, values, and positionalities on issues such as abortion and abortion care. Qualitative data collection and analysis will occur concurrently, allowing iterative reflection on emerging findings and enabling refinement of data collection processes and discussion guides where necessary throughout the study period. An audit trail will be maintained throughout the research process to systematically document the entire research process (including the decisions related to research design, tool preparation, sampling, data collection, coding and interpretation) to enhance the credibility and transparency of the findings. Similarly, regular debriefing and reflective discussions will be conducted throughout data collection to refine discussion guides where necessary. All modifications in the study tool will also be documented in the audit trail.

Quantitative analysis.

Quantitative data from facility assessments and household surveys will be retrieved from the data management system and analysed using statistical software like STATA (or an equivalent statistical software). Initially, data cleaning and validation will be done using built-in system checks and manual verification. Data will be analysed at both an overall and site-specific level to observe the variation within and across study sites in knowledge, awareness, experiences, and attitudes related to CAC services. In addition to identifying the common barriers and facilitators, the analysis will also explore how these factors vary across study sites and influence service delivery and utilization within different local contexts. Descriptive statistics will be used to summarize sociodemographic characteristics, reproductive history, and service-related variables. Continuous variables will be reported as means with standard deviations (SD) or medians with interquartile ranges (IQR), as appropriate. Categorical variables will be reported as frequencies and percentages. The study will use survey design commands and appropriate sample weight to account for village-level clustering. Prevalence estimates will be reported with 95% confidence intervals (CIs).

A composite facility readiness index will be constructed a priori using predefined indicators derived from the Comprehensive Abortion Care (CAC) Guidelines (2023) and the Comprehensive Abortion Care: A Health Facility Assessment Tool (2023) by WHO. Each indicator will be assessed through direct observation or facility records and coded as binary (available and functional on the day of assessment = 1; not available or non-functional = 0), and aggregated to generate both domain-specific scores and an overall readiness classification. Facilities meeting all minimum readiness criteria will be classified as “ready”, if they demonstrate availability and functionality on the day of assessment for the following minimum criteria: 1. Minimum one CAC trained provider; 2. Availability of the essential drugs and consumables; 3. Functioning equipment for the surgical method of abortion; 4. Availability of proper method of abortion in the facility; 5. A functional referral system. A facility lacking any of these essential components will be considered as “non-ready”. The proportion of ready facilities will be estimated with 95% CIs and compared across facility types using appropriate statistical tests.

Provider and women’s knowledge scores will be computed as summative indices, with each correctly answered item assigned a score of 1 and incorrect or “don’t know” responses assigned a score of 0. Item scores will be summed to generate total knowledge scores. Facility-based abortion utilization and post-abortion contraceptive uptake will be analyzed as binary outcomes.

Missing data will be assessed for each variable. For the primary outcome, complete case analysis will be used if missingness is below 5%; otherwise, multiple imputation by chained equations (MICE) will be performed under the missing at random assumption.

Quantitative findings will inform interpretation and integration with qualitative results during mixed-methods analysis.

Qualitative analysis.

The methodological orientation of the study will include a combined analysis using phenomenology and thematic analysis which allows for a systematic identification and interpretation of patterns across the study sites and stakeholder groups (providers, administrators, community actors, and women). All the interviews will be first transcribed verbatim by the same interviewer in the language in which the interview was conducted following the process of de-identification (by removing name, address and any other information revealing the identity). In the second stage, the transcripts from all regional languages will be professionally translated to English. In the third stage, translated interviews will be coded and analysed using qualitative data management software (NVivo or equivalent softwares). A combination of deductive and inductive thematic analysis will be employed to generate comprehensive insights. An initial broad coding framework will be developed to guide analysis across sites and participant groups and will be iteratively refined throughout the analytic process to incorporate emerging themes and contextual variations. The preparation of the deductive codebook will be guided by the Consolidated Framework for Implementation Research (CFIR), which offers a structured lens to explore multilevel determinants of implementation. Specifically, codes will be drawn from CFIR domains including Intervention (CAC interventions) characteristics, inner setting (intervention deliverers), outer setting (in which inner setting is lying), characteristics of Individuals (recipient and deliverers of the interventions), and process (implementation of intervention), capturing aspects such as perceived complexity of CAC service delivery, organizational culture, external policy influences, provider knowledge and attitudes, and the role of training and engagement. This approach will ensure systematic examination of implementation barriers and facilitators. In parallel, inductive codes will be developed by two lead researchers who would first read and reread transcripts to gain familiarity with the data following which they will code two transcripts from each site and prepare the inductive codebook after discussing any discrepancies, collapse codes and adding any new codes based on emergent themes. The entire research team involved in analysis will be oriented about the definition of each code in the revised codebook for consistent application across all the transcripts. In the last stage, all the emerged themes will be analysed to identify similar or different patterns across the sites. Following this, thematic analysis will be conducted to interpret meanings and relationships relevant to all four research objectives. Throughout this process, the coding and analytic decisions will be checked for consistency, and representative quotations will be identified to support key themes in the findings for a transparent and rigorous analysis of the qualitative data. To address site specific issues, the analysis will examine how common barriers and facilitators operate across different contexts, with particular attention to site-specific variations that may influence implementation processes and outcomes. Similarly, variations across participant groups and sites will be examined using three data sources, including data from different participant groups, contextual information, and field observations where necessary.

Integration and model development.

Findings from the quantitative and qualitative components will be triangulated to generate a comprehensive understanding of the current status of CAC services and their determinants. Converging evidence will be used to inform the development of a preliminary implementation framework (“Model 0 + ”), rooted in national CAC guidelines but refined based on empirical insights. The Model “0+” will map identified barriers (e.g., supply gaps, provider training, and stigma) and facilitators (e.g., community engagement, referral efficiency) to potential strategies. This model will be shared with stakeholders in a consultative workshop, allowing for validation and refinement based on field realities and policy relevance. The finalized model will be co-designed involving the research team and government stakeholders.