Promoting Respectful Maternity Care for Adolescents In Ghana: a Quasi-Experimental Research Study Protocol

Intra-partum mistreatment by healthcare providers remains a global public health and human rights challenge. Adolescents, who are typically younger, poorer and less educated have been found to be disproportionately exposed to intra-partum mistreatment. In Ghana, maternal mortality remains a leading cause of death among adolescent females, despite increasing patronage of skilled birth attendance in health facilities. In response to the the World Health Organisation Human Reproduction Programme (WHO-HRP) recommendations to address mistreatment with Respectful Maternity Care (RMC), this study aims to generate evidence on promoting respectful treatment of adolescents using an intervention that trains health providers on the concept of mistreatment, their professional roles in RMC and the rights of adolescents to RMC. Methods This study will employ a pre-test post-test quasi-experimental design. At pre-test and post-test, quantitative surveys will be conducted among adolescents who deliver at health facilities about their labour experience with mistreatment and RMC. A total target of 392 participants will be recruited across intervention and control facilities. Qualitative interviews will also be conducted with selected adolescents and health professionals for an in-depth understanding of the phenomenon. Following the pre-test, a facility-based training module will be implemented at intervention facilities for the facility midwives. The modules will be co-facilitated by the principal investigator and key resource persons from the district health directorate Quality of Care teams. Training will cover the rights of adolescents to quality healthcare, classications of mistreatment, RMC as a concept and the role of professionals in providing RMC. No intervention will occur in the control facilities. Descriptive statistics, logistic regressions and difference in differences analyses will be computed. Qualitative data will be transcribed and thematically analysed. mistreatment during facility-based births. The study seeks to test the outcome of the proposed intervention in promoting RMC during facility-based deliveries for parturient adolescents. This study also presents an opportunity in addressing some existing gaps in knowledge and information for policy and programming for adolescent maternal health and SRHR.

Mistreatment during childbirth remains a well-documented worldwide challenge, especially among poorer, younger and less educated women. In Ghana, the most vulnerable groups include adolescents, who also happen to bear a disproportionately high risk of maternal mortality. Despite the increasing rates of facilitybased births which are attended to by trained obstetric professionals, maternal mortality persists as a public health challenge in Ghana and sub-Saharan Africa in general. Following landscape studies which suggest that mistreatment during delivery may be a strong predictor for poor birth outcomes, the WHO has recommended that this issue be addressed with respectful maternity care. This study, therefore, aims to determine the effectiveness of a training intervention for health providers targeted at reducing mistreatment of adolescents during childbirth. This will contribute to knowledge and inform programming and policy on improving maternal healthcare and reducing maternal mortality and morbidity, especially among adolescents.

Background
Mistreatment suffered by women who deliver within health facilities is an extensively documented phenomenon [1][2][3] . The most reported forms of facility-based mistreatment which often occur in a continuum are categorized as physical abuse, non-consented care, non-con dential care, non-digni ed care, discrimination based on speci c patient attributes, abandonment of care, and detention in facilities 4 . Whilst intrapartum mistreatment traverses geographical boundaries and socio-economic levels, 1,2,5-7 certain factions of women appear to bear a markedly high burden of the phenomenon [8][9][10] . As a group of interest, adolescent women appear disproportionately predisposed to disrespectful treatment attributable to their young age, lower educational attainment, lower socio-economic status, lack of autonomy, as well as provider and social prejudices about their engagement in pre-marital and early sex [11][12][13][14][15] . Mistreatment in service provision for pregnant adolescents is problematic as it compounds the already existing burden of social and health risks 11, [16][17][18] . Currently, maternal mortality remains one of the leading global causes of death among adolescents, especially in sub-Saharan Africa and is considered a pressing public health challenge 19 . Approximately 11% of births worldwide occur among adolescent girls aged between 15 and 19 years old, majority of which occur in low to middle income countries 20 . In Ghana, this proportion of 14% is approximately a third higher than the global average 21,22 . Despite the absence of nationally representative data on what proportion of maternal deaths are adolescents, research suggests adolescents may constitute a larger proportion of maternal mortality and morbidity gures since they bear a higher risk of pregnancy-related complications than their older counterparts 23 . A WHO Multi-country Survey on Maternal and Newborn Health revealed that adolescents 19 years and below were twice as likely to develop eclampsia than women over 20 years of age 24 . The same study reported a 50% higher probability of systemic infections among adolescents and a 13% higher likelihood of severe maternal outcomes as compared to women over the age of 20 years. Studies investigating e cacious strategies which may successfully alleviate this burden have identi ed appropriately-timed intrapartum interventions to be the most bene cial means of reducing maternal mortality and morbidity compared to antenatal and postpartum interventions 25-27 . These include promoting facilitybased births and as a corollary, increasing skilled birth attendance to enable early identi cation and management of complications 27,28 . Accordingly, interventions in Ghana have targeted overcoming barriers to facility access especially geographic and nancial factors. Financial access has been improved by waiving antenatal care and delivery fees whilst the establishment of Community-Based Health Planning Service Compounds (CHPS-Compound) has improved provision of maternal care to the last mile within communities 29-31 . Nonetheless, the proportion of facility-based births in Ghana at 78% is currently less than the global average of 81% 22 . Additionally, the national maternal mortality rate of 308 per 1,000 births is still markedly higher than the global average of 211 per 1000 births and far short of the SDG target of 70 per 1000 births 21 . Within the ve years preceding the 2017 Ghana Maternal and Health Survey, only 71% of pregnant adolescents utilized facility-based deliveries 22 . This has been linked to the mistreatment parturients face during facilitybased births 32, 33 . This data suggests that attempts to improve maternal mortality and morbidity may be hindered by the mistreatment of adolescents during childbirth.
Consequently, the Human Reproduction Programme of the World Health Organization (WHO-HRP), has prescribed recommendations on addressing mistreatment of women during childbirth with respectful maternity care 34 . Respectful Maternity Care (RMC) refers to the organization and management of health systems in a manner that ensures respect for women's sexual and reproductive health and human rights 34 . Indeed, the absence of RMC has been identi ed as a key disincentive to accessing facility-based births; even in contexts where nancial and physical access are not necessarily problematic 14,32,35 . Studies have also reported adverse clinical outcomes like prolonged labour, unnecessary pain, haemorrhage, mental distress and in extreme cases, death [36][37][38][39] during facility based births where RMC is not practiced. This underscores the need for RMC not only to promote uptake of facility-based deliveries, but also to improve birth outcomes and reduce complications.
Considering that the sexual reproductive health and rights (SRHR) of Ghanaian adolescents are constitutionally entrenched 40 , backed by research-based policy 41 and heavily invested in by the government of Ghana and other development partners, improving the quality of maternity care for adolescents should be a national priority. Moreover, whilst studies have strongly linked mistreatment to poor outcomes, there is relatively little insight on speci c interventions that successfully promote RMC for adolescents. Against this background of limited information and knowledge about adolescents' experiences and outcomes with mistreatment during childbirth, this study seeks to characterize the experiences and outcomes of adolescents who face mistreatment during facility-based births. The study seeks to test the outcome of the proposed intervention in promoting RMC during facility-based deliveries for parturient adolescents. This study also presents an opportunity in addressing some existing gaps in knowledge and information for policy and programming for adolescent maternal health and SRHR.

Study Aim
The overall aim of the study is to determine the effectiveness of a training intervention in promoting respectful maternity care for adolescents during facility-based delivery.

Speci c Objectives
The speci c objectives of this study are to 1. Synthesise the available evidence relating to the mistreatment and the provision of respectful maternity care for adolescents during facility-based childbirth using a systematic review approach 2. Estimate the prevalence of mistreatment during facility-based childbirth among adolescents 3. Identify the risk factors and outcomes of intra-partum mistreatment during facility-based childbirth among adolescents 4. Document and characterize the experiences of adolescents who suffer mistreatment during facility-based childbirth 5. Determine the most and least common forms of Respectful Maternity Care offered to adolescent parturients 6. Determine the effectiveness of an intervention in promoting respectful maternity care and increasing satisfaction with facility-based births and skilled birth attendance for parturient adolescents

Conceptual Framework
The proposed study is diagrammatically represented in Figure 1. The conceptual framework is theoretically based and developed from the review of existent literature on intrapartum mistreatment and RMC.

Study Settings
The study intervention will be sited in three regions in Ghana: Eastern, Central and Greater-Accra regions. The intervention sites are public health facilities in the Lower Manya Krobo, Upper Manya Krobo district, and the Awutu-Senya East Municipality. The control sites will be in the Ada East, Shai-Osudoku and Effutu Districts.
These districts have an average of 11% of their population being adolescent females 42 and reported adolescent pregnancy rates higher than the national average of 14% 22 . The intervention and control sites are selected health facilities which are matched on the type of facilities being district health centres and district hospitals. For each of the intervention sites, a senior member of the district quality of care team will be included as a resource person in the RMC trainings.

Study design
The study is a non-equivalence pre-test post-test quasi-experimental study. Both qualitative and quantitative methods will be used in data collection.

Quantitative study population
Study participants in the quantitative phase will be adolescent females between the ages of 10 and 19 years old who deliver at the participating facilities.

Qualitative study population
The rst group of participants in the qualitative phase of this study are adolescent females between the ages of 10 and 19 years old who deliver at the participating study sites. The second group of participants are health providers in full time employment at the study sites. Adolescents and health professionals alike will be invited to participate in the study based on the study inclusion criteria and will be enrolled based on their consent to participate. Adolescents will be invited to participate in In-Depth Interviews (IDIs) and the professionals will participate in Key Informant Interviews (KIIs).

Inclusion criteria
Adolescents will be deemed eligible for participation if they satisfy the following criteria The minimum sample size for adolescents is obtained observing the assumptions that the estimated proportion of women who encounter mistreatment during childbirth is 18% 43 and the proposed intervention may reduce this by 10 percentage points. Additionally, a 5% margin of error, a 95% signi cance level and 80% power are estimated.
N= estimated sample size for one arm Za=Critical value at con dence interval of 95% (1.96) Zb=Critical value at power of 80% (0.842) d= Actual difference in proportion between control and intervention groups (0.1)

P= Average of difference in proportion of both groups
An intra-cluster design effect of 1.11 and a non-response adjustment of 10% is factored to obtain a minimum sample size of 392.
A minimum of 10 healthcare participants and 10 adolescents will be included in qualitative Key Informant Interviews It is anticipated that the number of interviews will allow a sample size large enough to allow data saturation to understand the phenomenon of mistreatment from both provider and client perspectives.
Sampling procedure Multistage, strati ed and systematic sampling will be used to recruit study participants from the facilities where they attend. Using data available from the 2010 Population and Housing Census, three out of the formerly ten regions in Ghana with adolescent pregnancy rates higher than the national average (14%) were randomly selected. Subsequently, six districts were purposefully selected from these regions considering appropriateness for the intervention, available resources, and geographical non-proximity between intervention and control facilities. Facilities in the intervention and control arms of the study were matched based on the number of adolescent pregnancies delivered annually using the Ghana Health Service District Health Information Management System (GHS-DHIMS) data from 2018 as a guide.
All eligible adolescent participants who attend the participating intervention and control facilities for antenatal care in their third trimester and delivery will be invited to enrol in the study by a trained research assistant. The nature of the study and all ethical concerns will be explained to the candidate prior to their enrolment. Enrolment will be run on a cumulative bases until the minimum sample size is satis ed. Some adolescents will be purposively selected to participate in the adolescent IDIs based on their characteristics which is aimed at achieving maximum variability of participants. Each district will have a health professional who works in administrative capacity and in a caregiving capacity invited to participate in the Key Informant Interviews.

Measurement tools
Quantitative data will be collected with the aid of an electronically-based structured questionnaire adapted from Bohren and colleagues' study 44 which has been modi ed to suit the local context. The tool is designed to collect data on the socio-demographic characteristics of adolescents and their knowledge on health and sexual reproductive rights. The tool will additionally collect data on the labour experience including the provider-patient encounter and relationships, encounters of mistreatment and RMC, and health facility factors.
Qualitative data collection will be done with the aid of interview guides adapted from the studies by Bohren et al 44 and Afulani and colleagues 45 . The tool is speci cally developed to collect data on the varying forms of mistreatment within the local context. Sections of the guide probe for an in-depth discussion on the themes of knowledge on health and sexual reproductive rights, provider-patient interactions, encounters of mistreatment and RMC and health facility factors. For professionals, the interview will probe for the role of healthcare providers in the labour experience especially in the administration of RMC.
Data collection procedure The quantitative interviews will be administered by trained research assistants. Interviews will be administered both at baseline and endline (from day 1 following the intervention). They are estimated to last between 30 and 45 minutes per participant and will be held at the home of the participant or a convenient private venue outside the health facility within two weeks after delivery. This venue arrangement is considered important to create an enabling environment which allows free and candid dialogue with the participant which may not be possible at the health facility due to fear of retaliation from providers. This time interval between delivery and the interview is also considered adequate for the parturient to recuperate without introducing signi cant recall bias. Interviews with adolescent participants will be in the most commonly spoken languages in the district which are English, Twi, Dangme or Krobo, based on the preference of the participant. This will allow the participants to express themselves freely in the language they are most comfortable with.
Qualitative IDIs will be conducted with selected adolescent participants and KIIs with health professionals. Interviews will be audio-taped, with prior permission from the participant. Additionally, notes will be taken to capture non-verbal nuances and expressions which may aid in better analysis and interpretation of the audio information. The number of interviews will continue until a saturation of information is reached. Each interview is estimated to last between 45 and 60 minutes. For adolescents, IDIs will take place at a location convenient to the participant outside the facility within two weeks after delivery. Similar to the quantitative interviews, the qualitative interviews will be conducted within a post-recovery timeframe which mitigates the probability of recall bias. Based on the preference of the participant, IDIs will be conducted in English, Twi, Ga-Dangme or Krobo as done in the quantitative interviews. The KIIs will be conducted within the facilities were the staff work and will be conducted in English.

Data quality assurance activities
The quality of the study and data collected will be assured via the following means 1. Intervention-Control Allocation: The intervention and control facilities are selected based on the similarity of their characteristics. This includes the type of facility, proportion of adolescent deliveries per year and the socio-demographic characteristics of the community. The geographic non-proximity of the facilities was also highly considered to avoid contamination of the control communities. The age group and sex of the assistants will help establish rapport between the assistants and participants. This will also help minimize response desirability bias on the side of participants. The educational and technical criterion are necessary to ensure that the assistants understand and effectively carry out their roles in the study to collect quality data. The assistants will undergo a two-day training on the theoretical background of the study, enrolment and obtaining consent from participants, administration of questionnaires and facilitation of interviews, as well as ethical issues. The principal investigator will always closely supervise the research assistants to ensure their observance of study protocols and ethical codes 3. Data Collection Instruments: The data collection instruments are adapted from a previous multi-country study (which included Ghana) which included parturient women of all ages 46 instruments will be adapted for use with adolescents. The tools will be pretested on 5% of the sample size in a pilot facility in the Kpone-Katamanso district and modi ed accordingly based on ndings from the pilot. The pilot district was selected based on its similarity to study sites in having an adolescent birth rate of greater than 11%.

Research Assistants
4. Data Collection Timing and Location: Data collection at baseline will be over a 3-month period. Endline data collection will start from day 1 after the intervention until three-months post-intervention. The venue and time interval between delivery and the adolescent surveys will allow participants express themselves freely without fear of victimization by health providers. This will also allow the participants enough recovery time from childbirth with minimal recall bias.

Data
Monitoring: Data collection will be conducted in real time electronically. The data will be monitored daily to allow early detection of discrepancies and resultant remedial action. Transcripts from audio interviews will be reviewed by an independent listener who is uent in the language and did not conduct or transcribe the interview to audit for accuracy and completeness.

Data processing and statistical analysis
Both baseline and endline data will be reviewed and edited for completeness and consistency daily. The primary outcome variable of interest is de ned as disrespectful intrapartum care which is measured as a response of "Yes" to any of the nine mistreatment questions. The explanatory variables of interest include socio-economic status, educational level, marital status and knowledge of health rights. Data will be collected on the KoBoCollect platform, synced onto a web server and downloaded in Ms-Excel format to the principal investigator's computer. The data will be exported in STATA version 16 for cleaning and analyses. Descriptive statistics as well as binary and multiple logistic regressions will be computed. Additionally, Difference in Differences (DID) will be calculated to determine the net RMC training intervention effect on mistreatment and RMC. Finally, statistical signi cance of the RMC training intervention effect will be evaluated using adjusted prevalence ratio, adjusted odds/risk ratio at 95% con dence interval (CI) and P-Value of less than 0.05.
Audiotapes from qualitative interviews will be translated and transcribed verbatim into Ms-Word format in English. Themes will be identi ed and coded in Ms-Excel and exported into NVivo for analyses. Quotes will be narrated and cited in participant codes.
Data analysis for qualitative and quantitative data will be conducted separately. Subsequently, inferences from both qualitative and quantitative results will be jointly used to draw conclusions.

Intervention
The proposed intervention aims to educate and inform midwives about the concepts of mistreatment and RMC. The training additionally seeks to help participants explore and shift their current values and beliefs about mistreatment of adolescent parturients towards more positive behaviours. It also aims to encourage the practice of RMC by promoting adherence to positive professional codes of conduct, ethics, roles, values and attitudes. The proposed intervention is a training module facilitated with the aid of a resource package adapted from the White Ribbon Alliance/ USAID "Respectful Maternity Care for Healthcare Workers: Tackling Disrespect & Abuse During Facility-Based Childbirth" project 47 . Additionally, the structure and content of the module is in uenced by evidence from previous studies and systematic reviews on mistreatment and RMC [48][49][50][51][52][53][54] . This course will be presented in the format of presentations, case studies, role plays, discussions and evaluations. The content will cover the adolescent and their reproductive health rights, the Ghana Health Service (GHS) code of professional ethics, Nurses and Midwives Code of Conduct, de nitions and typologies of mistreatment and RMC as well as their outcomes to the woman, neonate, professional and health system. The content will also cover the professionals' role in the delivery of RMC. The training workshops will be implemented by the principal investigator supported by a facilitator who is member of the District Quality of Care team. Each workshop will be a two-day event. The rst day will cover a theoretical introduction of the concepts of mistreatment and RMC and the role of the obstetric professional. The second day will be a series of case studies and role plays to apply the theoretical basis from day 1. The facilities will additionally be provided with visible posters and materials as a constant reminder of the contents of the training programme.
There will be no intervention conducted in the control facilities.

Expected Outcomes
It is anticipated that at the end of this proposed study, the following outcomes will be achieved per each objective as summarised in Table 1 Insert Table 1 here

Discussion
This study will rstly establish the prevalence of mistreatment among parturient adolescents. Several studies worldwide have reported the prevalence of mistreatment among women in general 3,55−58 . This proposed study, however, is among the rst of its type to focus exclusively on the mistreatment of adolescents. Furthermore, the study will determine the effectiveness of an intervention in promoting RMC in Ghana and reducing mistreatment. Previous studies have employed interventions that target facility users such as holding maternity open birth days and community workshops to improve public awareness about mistreatment 6,43 . Mediation and resolution of past incidences of mistreatment whilst preventing future events has also been managed by educating communities on their SRHR whilst providing a means of legal redress 59 . Whilst these interventions may be helpful for older women, they may not be ideal for younger adolescents as they may not possess the autonomy required to properly exercise these facilities. Some interventions have also targeted providers with evidence-based information by providing access to the World Health Organization Reproductive Health Library (RHL) 60 and promoting advocacy for birth companions 61 . A study by Abuya and colleagues in Kenya recorded a 7% reduction in mistreatment following the implementation of a multi-component intervention at the policy, facility and community levels 43 . However, there was no special focus in these studies on adolescent sub-groups and their peculiar experiences or needs. This study therefore seeks to provide insight on interventions that may better address the peculiar maternal care needs of adolescent parturients.

Limitations
There are a few limitations in this study which measures have been taken to address. Firstly, it was not possible to randomly assign the study sites due to geographical and health system constraints. It is therefore possible that there will be observed and unobserved differences in the baseline measures of the intervention and comparison groups that may account for differences in outcomes. However, the study sites were selected based on their similarities in characteristics such as type of facility and socio-geographic factors. Furthermore, the use of difference-in-differences analytic methods in the analyses of data will allow the account of both observed and unobserved confounders. Another limitation that may be faced is the employment of new staff into intervention facilities after the intervention has been conducted. To mitigate this limitation, we shall take note of new staff and encourage the training module to be shared during their staff orientation.

Dissemination
Dissemination strategies have been drawn for results emanating from this study. Oral and poster presentations of the results will be made to the participating facilities as well as district and regional directorates of the Ghana Health Service. Furthermore, the results of the speci c objectives in this proposal will be published in peer review journals. The data management procedures, study materials including questionnaires and implementation frameworks of the study will be made upon reasonable request to the corresponding author. be obtained from all participants after explaining to them the direct risks and bene ts of their participation as well as their freedom to withdraw their participation at any time in the study.

Consent for publication N/A
Availability of data and materials The data management procedures, study materials including questionnaires and implementation frameworks of the study will be made upon reasonable request to the corresponding author.

Competing interests
The Conduct a baseline assessment of mistreatment across adolescent participants in intervention and control facilities. Speci c Objective 3: Identify the risk factors and outcomes of intra-partum mistreatment among a cross-section of adolescents Interview health professionals to explore their perceptions on the associated risk factors of mistreatment and its outcomes on adolescents, health professionals and the national health system. Speci c Objective 4: Document and characterize the experiences of a group of adolescents who suffer mistreatment during childbirth Interview adolescents on their detailed perceptions and personal experiences with intrapartum mistreatment Speci c Objective 5: Determine the most and least common forms of Respectful Maternity Care offered to adolescent parturients Conduct an evaluation of the most and least commonly occurring RMC practices offered in the selected intervention and control facilities Speci c Objective 6: Determine the effectiveness of an intervention in promoting respectful maternity care and increasing satisfaction with facility-based births and skilled birth attendance for parturient adolescents Implement an RMC training intervention across the selected intervention communities Conduct an endline assessment across intervention and control facilities to determine the prevalence of mistreatment Compute the differences in mistreatment and RMC between intervention and control facilities at baseline and endline. Figure 1 Conceptual Framework