Description
Study design
We will use a pragmatic hybrid effectiveness-implementation type III21 design to test our implementation strategy while observing and gathering information on the QUALI-DEC intervention’s impact on relevant outcomes. Using a quasi-experimental design (interruptedtimeseries and before-after study), we will assess effectiveness and safety outcomes 22,23. A process evaluation will be carried out using mixed qualitative and quantitative approaches 24. The evaluation will account for the policy context, remuneration and organization of care, hospital characteristics, readiness to change and acceptability among healthcare providers and organizations. The evaluation will also consider the societal context, women’s preferences for mode of birth, acceptability, knowledge improvement, birth experience and satisfaction among women, and out-of-pocket expenditures linked to maternity care. We used the Standards for Reporting Implementation Studies (STaRI) checklist to report our research protocol 25.
Context
The multifaceted intervention will be implemented in facilities in Argentina, Burkina Faso, Thailand and Vietnam. These four countries illustrate various degrees of rates in LMICs (Table 3) and present specific challenges for QUALI-DEC implementation. Within these four countries, Argentina has the highest level of CS at national level and, more generally, of the biomedicalization of childbirth. Thailand has very low fertility, which may add pressure in favor of CS. A favorable socioeconomic context may also facilitate the preference for CS. Vietnam is interesting for its demographic impact (size of the population) and its performance in health indicators given its level of national income. However, the national CS rate has been continuously increasing over the past few decades, exceeding any reasonable level for medical needs and large inequalities in the use of CS. Burkina Faso has a low CS rate at national level that may hide inequalities26 and that suggests a greatpotential for further increase and consequently represents an opportunity to prevent thephenomenon before it agravates. There is leverage power given the consequences of CS for nulliparous women in a context where fertility is so high. Risk of impoverishing expenditure for surgical care is high in Burkina Faso and Vietnam where financial protection is still developing. Evidence from our previous experience and the existing literature has shown that the skills and beliefs of healthcare providers, access to information and emotional support of low-risk women, leadership and commitment to use evidence-based guidelines at the organizational and system levels are important drivers of the CS rate in these four countries 9.
Targeted sites and participants
The study will be conducted from January 2020 to December 2024 in 32 healthcare facilities (8 per country) with high CS rates. Facilities were selected purposely with country investigators to reflect the range of contexts, such as secondary and tertiary levels of care, public and private hospitals, andteaching and nonacademic facilities (Table 4). There is a wide variation in the number of deliveries among the different institutions. In Vietnam, the number of annual births in some hospitals exceeds 20,000 births per year, while in other countries, the number ranges between 1200 and 7500 births per year. Facility-based CS rates range from 21% (district hospital in Burkina Faso) to 62% (tertiary hospital inArgentina) across countries.
The intervention directly targets healthcare providers involved in obstetric care and all women who give birth in the participating hospitals during the study period. We have defined providers as obstetricians and nurses/midwives working inthe maternity ward in the study facilities. Women will be eligible if they give birth to a newborn (birthweight ≥ 500 g in Argentina and Vietnam or ≥ 1000 g in Burkina Faso and Thailand), alive or dead, and with or without malformations. The intervention does not target patients admitted for abortion or miscarriage or those who delivered at home or in another facility that is not a participating hospital.
Intervention
A multifaceted intervention was developed based on existing evidence (Table 1) and WHO recommendations on nonclinical interventions to reduce unnecessary CS20. Baseline formative research27 informed by the ecological framework9 will be conducted to improve our understanding of the different levels of factors affecting CS rates and to adapt the multifaceted intervention to each country. The four components of QUALI-DEC will be implemented simultaneously in each participating hospital during the 2-year implementation period (Figure 1):
Component 1 – Opinion leader (OL) – One OL in each facility will be identified by peers and local authorities. OLswill be gynecologists-obstetricians with proven communication skills and a reputable influence on their colleagues. OLs will lead the process of selection of evidence-based clinical protocols for CS decision-making and train on-site collaborators charged with implementing audit and feedback, decisionaid and companionship during labor.
Component 2 – Audit and feedback (A&F) – A&F involves the analysis of medical practices by the practitioners themselves (physicians, nurses/midwives) according to the evidence-based clinical protocols selected by the providers in each facility and subsequent advice as to whether the decision for CS was the most appropriate. The audit concerns several medical records, and as an output, it provides a conclusive analysis that will be presented to the rest of the medical staff (feedback). Audit cycles will be implemented monthly by the local committees following the different steps presented in Figure 2.
Component 3 – Decision analysis tool (DAT) – The DAT is adapted to each country and developed to be used during antenatal care (after 28 weeks of gestation) by women with a singleton pregnancy, without a previous CS and eligible for a trial of labor. It includes two sections: (i) an information section, providing a description and an explanation of the risks and benefits of each mode of birth (planned vaginal birth vs. planned CS); and (ii) an exercise section, allowing women to clarify and summarize their values and preferences with their clinician and indicate what aspects of the mode of birth are important to them. The DAT will be available as a paper booklet and an interactive web/smartphone application.
Component 4 – Companionship during labor – This component will be implemented using a tailored labor companionship model that will include information on (i) eligibility criteria for women and companions, (ii) identification of health care providers who will invite the chosen and eligible labor companion from the waiting area into the labor room, (iii) identification of healthcare providers who will deliver the messages to the laboring woman and her companion, (iv) how many people are allowed and when they are allowed to act as companions, (v) how physical space of the labor ward may need design modifications to accommodate a companion, and (vi) educational tools for companions on how to support women during labor and birth.
Pregnant women may benefit from the DAT, which can help them become more informed and active in their care and may reduce the likelihood of healthcare workers' preferences dictating their care pathway28. Women in labor may benefit from the support of a companion who can enhance their feelings of control and competence, which might reduce their reliance on medical interventions, such as pain relief 11. Healthcare providers who receive feedback from the audit on CS may be able to change their practice and clinical management in response to the feedback, with the support of local OLs 14. Providers may also benefit from the labor companion intervention, as the labor companion will help them better support women11. They may also benefit from the DAT, as it could help them discuss and clarify women’s perspectives and preferences on the mode of birth28.
Implementation strategy
The implementation strategy is aligned within the usual model of care in participating healthcare facilities. The main implementers are the local OLs and healthcare providers who are involved in the program and are supported by the country-level study coordinator.Formative research inthe baseline period will assess the main drivers and barriers, and a meeting will be held among all stakeholders to discuss implementation issues. Parliamentarians and representatives of women’s associations will be involved in this meeting to consider women’s views. Then, the intervention will be introduced in each country with a 5-day training workshop addressed to OLs to mobilize their awareness of the power/interaction model of interpersonal influence and to help them learn and practice intervention tools and techniques. The 5-day training session represents the beginning of the implementation phase of the QUALI-DEC project. A re-fresher 3-day training session will take place after one year to update OLson the process of the intervention, discuss their roles, share their experiences and confirm their capacity to provide leadership in their clinical settings. OLs will receive financial incentives during the intervention period to compensate for the loss of revenues related to the decrease in their clinical activities.TrainedOLs will create and train the local QUALI-DEC committee in each hospital charged to implement the intervention.The composition of this committee will be defined during the 5-day training workshop. It will include the OL, the data collector for the Robson classification (Ten Group Classification System)29and the audit of CS indication, representatives of physicians, nurses/midwives and representatives of the hospital’s administration.
The QUALI-DEC committee will launch the audit cycles, DAT and companionship in their hospital. They will encourage antenatal care providers to deliver the DAT booklet to eligible pregnant women. This will require a series of on-site meetings in all relevant facilities to inform and motivate providers and to obtain their formal commitment. In addition, a DAT application will be developed for smartphones and made available in thesettings in which it is considered culturally appropriate and most acceptable and convenient for women. Posters will be displayed on the wall of the waiting room of antenatal care centers with the QR code to access the web/smartphone application. Other information, educational and communication (IEC) materials, such as flipcharts or posters,will be developed to facilitate the briefing of healthcare providers, companions and laboring women. These IEC materials will include reminders about the importance of labor companionship, the role of companions and the regulations of the labor wards. The country-level coordinator will conduct quarterly visits toeach participating hospital during the 2-year implementation period to identify further barriers for the implementation process and possible strategies to overcome those barriers, verify data quality and document and report on the study’s progress.
Methods: Evaluation
Outcomes
The primary endpoint measure is the monthly CS rate in participating hospitals among women with a singleton pregnancy, with a fetus in cephalic presentation and at least 37 weeks of gestation, and with no previous CS (Groups 1-4 of the Robson classification). We will use the Ten Groups Classification System (TGCS) (also known as the Robson classification) to monitor CS rates at the hospital level.The TGCS classify women into prospective mutually exclusive and totally inclusivegroups of women based on a few obstetric variables which are easily obtained and most women themselves would know. Trained data collectors will gather information abouteach eligible woman using existing routine health information systems (paper-based or electronic records).The information includes five main variables to classify each woman into the TGCS: parity, fetal presentation, previous CS, number of fetuses and gestational age. Once each woman is classified, the data collector will enter themonthly number of vaginal and cesarean deliveries across the ten groups into a secure web application for building and managing online surveys and databases (REDCap). The system will automatically produce monthly hospital summary statistics according to the TGCS classification: group size, CS rate by group and the absolute contribution of each group to the overall CS rate. We will consider the monthly rate of CS before the onset of intervention (12-month period), during the implementation phase (24-month period), and after the implementation phase (24-month period) to assess the effects of the intervention.
As secondary endpoints, the following outcome measures will be assessed: assisted vaginal delivery; time of CS (before or during labor); third- or fourth-degree perineal laceration; antibiotics and uterotonics use; transfusion; admission of the mother or the newborn to intensive care unit; uterine rupture, hysterectomy, maternal or neonatal death; time of breastfeedinginitiation; woman's satisfaction with care and her birth experience; payment for medical care; indirect costs of care for childbirth (e.g., cost of transportation to hospital); and loss of earnings. A cross-sectional survey among a representative sample of postpartum women will be established at two time points: at baseline and at the end of the intervention period. All births occurring in the participating hospitals during two weeks in Argentina and Burkina Faso and one week in Thailand and Vietnamwill be covered in each survey. The data collection includes a face-to-face interview with women after childbirth and before they leave the maternity ward (facility-exit interview) and thecollection of information from the women’smedical records, including socioeconomic characteristics of the mother, reproductive history, antenatal and intrapartum care, time and indication of CS, if any, satisfaction with birth experience, breastfeeding practices, out-of-pocket costs, maternal and neonatal outcomes.
Process evaluation
We will use the UK Medical Research Council (MRC) process evaluation framework24 to describe how the intervention works (or does not work) along the pathway of implementation, including the internal dynamics of the four components of the QUALI-DEC strategy. The process evaluation also explores the roles, perceptions and coping strategies of actors, adaptation ofthe interventions based on the local context, and any unintended effects, with a view to understand the mediating effect of the context30. Figure 3presents the key functions of the QUALI-DEC process evaluation and the relations among them, while Figure 4 shows the data collection and analysis methods.
To align the intervention and the implementation strategy to the local context, we conduct qualitative research, document review at the country level and a readiness assessment ofeach participating facility. The qualitative research will include semistructured interviews with women, potential companions and healthcare providers to obtain a comprehensive understanding of the health system and societal context in each country27. Additional interviews with policy-makers and representatives of women’s and professional associations (gynecologists-obstetricians and nurses/midwives) will allow us to complete the stakeholder mapping and analysis. Following the context assessment and inconsultation with QUALI-DEC developers and implementers, we will define assumptions on what may need to happen (mechanisms of change), and we will hypothesize about how change will happen at the individual level (healthcare providers, women and companions), at the organizational level (healthcare facility) and through the interaction of participants. Discussions and meetings will be held by videoconference with country researchers and during face-to-face meetings with high-level stakeholders and local OLs in each country. To further assess the potential scalability of the intervention within the countries, we will include a participative scalability assessment in thein-country meetings to summarize early opportunities and challenges for scale up31.
We will construct a theory of change to guide the process evaluation. We will define indicators of fidelity (whether the intervention was delivered as intended), dose (the quantity of intervention implemented) and reach of intervention (whether women and providers came into contact with each relevant component of QUALI-DEC)24. These indicators will be measured at the individual and organizational levels in all facilities. During the quarterly monitoring visits at each hospital, the coordinator will use a checklist to rate the level of implementation for each component of QUALI-DEC and identify barriers/facilitators and adaptation strategies to overcome these barriers. Using data from the postpartum cross-sectional survey, we will be able to assess the proportion of women who used the DAT during antenatal care or chose a companion during labor. Interviews with providers and women will provide more detailed information on the perceptions and views of both stakeholders.
We will conduct in-depth case studies in a subset of four hospitals per country to investigate the details of what worked, why and why not. Study sites will be purposively selected based on indicators of fidelity, dose and reach to reflect the diversity of implementation across participating hospitals. Structured observations of each site and on-site meetings will be held with the members of the local committee, maternity ward staff and facility administrator. We will conduct in-depth individual interviews (IDIs) with key stakeholders at the end of the intervention period (months 24-30). The study instrument for IDIs with healthcare providers will be a semistructured interview guide covering the following topics: communication; interprofessional interaction; acceptability of the CS audit and feedback, DAT and labor companionship; and decision-making, including aspects of position/seniority, gender, weighing of alternatives and their implications, and information-sharing. The study instrument for IDIs with women and their companions will be a semistructured interview guide covering the following topics: process of and factors affecting the decision-making to use DAT; process of and factors affecting decision-making to use labor companionship; perceptions of the DAT and labor companionship related to knowledge, experiences, and support in choosing the mode of childbirth; perceptions and experiences of the relationship between themselves and providers; perception and experiences of how use of the DAT and/or labor companionship influenced trust, self-esteem, empowerment, and the relationship with providers. All interviews will take place in a private setting and will be audio recorded.
Economic evaluation
The cost-effectiveness of the QUALI-DEC intervention and the financial risk protection provided are important factors for decision-makers considering implementing new strategies to reduce unnecessary CS. Therefore, an economic analysis of the QUALI-DEC strategy will be conducted to inform decision-makers about the implementation of the QUALI-DEC intervention in relevant healthcare facilities.The impact of QUALI-DEC will be estimated in three domains across women in distinct wealth strata: (i) health gains (e.g., reduced CS rates), (ii) women’s out-of-pocket (OOP) expenditures averted by reducing unnecessary CS, and (iii) total net cost of the intervention to the implementer.
We will use an extended cost-effectiveness analysis (ECEA)approach to evaluate the cost-effectiveness of the QUALI-DEC intervention32. The ECEA follows the same principles as the traditional cost-effectiveness analysis; however, ECEA allows policy-makers to account forboth health and nonhealth outcomes when making decisions and thus to more effectively directscarce healthcare resources towards specific policy objectives. Furthermore, ECEA assesses the health and financial consequences of policies, including financial risk protection and other disaggregated outcomes per a specific population stratum of interest.
Information on intervention costs will be collected at the national and facility levels. Information on individual costs will be collected from women in relation to the postpartum survey.Intervention costs will be assessed at two levels: (i) the national level (development of guidelines and training materials, quarterly visits, and coordination costs) and (ii) the healthcare facility level (time spent by health professionals and OLs on training activities, A&F, consumables, equipment, overheads, and capital costs) and collected during the intervention implementation. OOP costs will also be collected from women in relation to activities performed in evaluating effectiveness outcomes (postpartum survey) and will include the following: the formal and informal direct payment for medical care, direct nonmedical costs (i.e., transportation costs to seek care), and indirect costs (i.e., time and productivity losses, which can be translated into wages and foregoneincome). Costs for the eventual scale-up of the intervention will also be calculated. Health gains will be determined by evaluating changes in CS practices and perinatal outcomes and fed into the ECEA.
The ECEA will provide two main outputs: the incrementalcost-effectiveness ratio (ICER) – the net costs per reduction in CS among low-risk women – will be computed; and OOP averted – this output can be referred to as ‘expenditures crowded out’ in the sense that the implementation of the QUALI-DEC intervention may lead to the ‘crowding out’ of these individual OOP costs. The distributional consequences (i.e., outputs) will be analyzed across distinct strata of the populations (e.g., socioeconomic status and geographical setting).
Sample size
Limited guidance is available on sample size calculation for time series analyses, and many of the recommendations focus on the need for sufficient time points pre- and postintervention to precisely ascertain trends and levels 33,34. Our analysis on the primary outcome (monthly CS rates among low-risk women) will include 12 time points preintervention, 24 time points during the intervention phase, and 24 time points during the follow-up phase.
The approach for calculating the required number of participants for the postpartum cross-sectional survey is that used for a “before-after” noncontrolled study design. We estimated that a sample of 470 women at baseline and 470 women at the end of the intervention period will ensure 90% statistical power to detect an effect size of 0.3 standard deviations or greater in satisfaction scores with a two-sided 5% significance level 35. The calculation accounted for the clustered nature of the data by hospitals with a design effect of 2. Allowing for a 20% nonresponse rate, we aim to recruit 564 women in each phase. The proposed sample size (i.e., 564) can be achieved with two weeks of data collection in Argentina and Burkina Faso and with one week of data collection in Thailand and Vietnam.This sample size will be for each country and will allow to draw conclusions independently for each country and produce individual country interpretations. We estimated that 3980 births will occur during this data collection period in all participating hospitals of the four countries. This overall sample size for the cross-sectional survey will ensure accurate measurements of other secondary outcomes (maternal and perinatal morbidity, time of breastfeedinginitiation). Estimations of outcomes at each time point will fluctuate within a 95% confidence interval with the following bounds: 10% rate ± 1% (example: postpartum hemorrhage); 20% rate ± 1.3% (example: second degree perineal trauma); 40% rate ± 1.5%; (examples: overall CS rate, breastfeeding within one hour of birth).
Analysis
Quantitative analysis
We will use different methods to evaluate the effectiveness of the QUALI-DEC intervention22,23. For the primary outcome, interrupted time series analysis (ITSA) based on segmented regression will estimate the mean changes in the level (immediate change) and trend (sustained change) of monthly CS rates across all participating hospitals in relation to their baseline level and pre-existing trend. For secondary outcomes, we will use a before and after cross-sectional design that will include medical records and women’s interviews. We will compare the outcomes between the two periods of the cross-sectional survey, adjusting for hospital and woman characteristics, to evaluate changes in satisfaction with thebirth experience, breastfeeding and medical practices, and maternal/perinatal morbidity. For implementation outcomes, checklist ratings during monitoring visits will be used to compute average scores of the fidelity, dose and reach of the intervention. Parametric tests will be used to assess changes over time for each facility and differences between study sites.
Qualitative analysis
Qualitative data from observations and IDIs will be analyzed and interpreted using a thematic analysis approach36. Interview transcripts will be analyzed in the local language at the countrylevel. Final themes and key quotations will be translated into English for sharing with the researchers of the QUALI-DEC consortium. Framework analysis will be used to provide an in-depth understanding of acceptability37,38 by providers and women/companions and the empowerment39,40 of both stakeholders to act on CS decision-making. We will define acceptability as the perception ofproviders and women that the QUALI-DEC intervention is agreeable, entailsan acceptable burden, is ethical and economically feasible, and leads to positive outcomes. We aim to understand the extent to which each component of the intervention and its process are both socially and technically accepted in each context. Empowerment can be understood as a process but also as an outcome to assess whetherthe intervention has helped providers and women act on CS decision-making. The analysis will focus on the individual empowerment of women and providers. We anticipate that our intervention will enhance providers’ empowerment by presenting them with monthly statistics promoting reflexivity on their practices, and this increased awareness of clinical practices will, in turn, enhance obstetricians’ and midwives’ sense of agency and self-determination in deciding on interventions during labor and delivery. From the women’s point of view, the study will detail the self-empowerment and professional support provided to women to choose the mode of delivery that better suits their needs. We will analyze the effect of the intervention on women’s self-esteem, knowledge and sense of empowerment when deciding on the mode of delivery. In our analysis of empowerment,particular attention will be paid to the gender dimension, including the gender dynamics between different categories of provider and between providers and women and how gender norms shape values and decisions related to childbirth.
Subgroup analyses
The integration ofquantitative process measures into outcome datasets will contribute to understanding how implementation variability affects outcomes (on-treatment analyses) and to testing hypotheses arising from qualitative analyses. For example, time series models with multigroup comparisons will enable us to conduct formal statistical tests comparing the level and slope of the primary outcome between different categories of healthcare facilities reflecting different levels of implementation, thereby quantifying the variation of effect size between subgroups41 and revealingthe mechanisms of impact. Additionally, outcomes between women in different socioeconomic categories (in terms of education, place of residence, place of birth and wealth index) and between periods will be compared to assess the equity of the QUALI-DEC intervention.
Knowledge transfer
In consultation with key stakeholders in each participating country, we will develop an innovative evidence-based knowledge transfer strategy 42,43, adapted to each context. The key ingredients of this strategy will be training, implementation and evaluation of a knowledge broker in each country who will facilitate the adaptation, dissemination and exploitation of QUALI-DEC findings by key stakeholders (ref). As the implementation of knowledge brokering is very innovative, it will be the subject of an in-depth evaluation in order to generate knowledge about its processes and effectiveness.A specific research protocol for this part of the QUALI-DEC project will be published later.