Healthcare providers perspectives on labor monitoring in Nigeria and Uganda: A qualitative study on challenges and opportunities

First published online: 7 December 2017 DOI: 10.1002/ijgo.12379 QUALITATIVE STUDY Healthcare providers perspectives on labor monitoring in Nigeria and Uganda: A qualitative study on challenges and opportunities Fan Yang 1 Meghan A. Bohren 2, * David Kyaddondo 3 Musibau A. Titiloye 4 Akinpelu O. Olutayo 5 Olufemi T. Oladapo 2 João Paulo Souza 2 A. Metin Gülmezoglu 2 Kidza Mugerwa 6 Bukola Fawole 7 1 Department of International Health, Social and Behavioral Interventions Program, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA 2 UNDP/UNFPA/UNICEF/WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproduction, Department of Reproductive Health and Research, WHO, Geneva, Switzerland 3 Child Health and Development Centre, Makerere University, Kampala, Uganda 4 Department of Health Promotion and Education, College of Medicine, University of Ibadan, Ibadan, Nigeria 5 Department of Sociology, Faculty of Social Sciences, University of Ibadan, Ibadan, Nigeria 6 Department of Obstetrics and Gynecology, Makerere University, Kampala, Uganda 7 Department of Obstetrics and Gynecology, College of Medicine, University of Ibadan, Ibadan, Nigeria *Correspondence Meghan A. Bohren, UNDP/UNFPA/UNICEF/ WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproduction, Department of Reproductive Health and Research, WHO, Geneva, Switzerland. Email: bohrenm@who.int Funding Information This paper was developed as part of the World Health Organization (WHO) BOLD ( Better Outcomes in Labour Difficulty ) project, which was funded by the Bill and Melinda Gates Foundation, the United States Agency for International Development (USAID), and UNDP- UNFPA-UNICEF-WHO-World Bank Special Programme of Research, Development and Research Training in Human Reproduction (HRP), a cosponsored program executed by the WHO. Abstract Objective: To explore current practices, challenges, and opportunities in relation to monitoring labor progression, from the perspectives of healthcare professionals in low- resource settings. Methods: Thematic analysis of qualitative data (in- depth interviews [IDIs] and focus group discussions [FDGs]) obtained from a purposive sample of healthcare providers and managers in selected health facilities in Nigeria and Uganda. Results: A total of 70 IDIs and 16 FGDs with doctors, midwives, and administrators are included in this analysis. Labor monitoring encompasses a broad scope of care jointly provided by doctors and midwives. A range of contextual limitations was identified as barriers to monitoring labor progression, including staff shortages, lack of team cooperation, delays in responding to abnormal labor observations, suboptimal provider patient dynamics, and limitations in partograph use. Perceived opportunities to improve current practices included streamlining clinical team cooperation, facilitating provider client communication, encouraging women s uptake of offered care, bridging the gaps in the continuum of monitoring tasks between cadres, and improving skills in assessment of labor progress, and accuracy in its documentation. Conclusion: Healthcare providers face many challenges to effective monitoring of labor progress in low- resource settings. This analysis presents potential opportunities to improve labor monitoring practices and tools in these contexts. KEYWORDS Clinical decision making; Facility-based delivery; Healthcare providers perspective; Labor monitoring; Nigeria; Quality of care; Uganda 2017 International Federation of Gynecology and Obstetrics The World Health Organization retains copyright and all other rights in the manuscript of this article as submitted for publication. Int J Gynecol Obstet 2017; 139 (Suppl. 1): 17 26 wileyonlinelibrary.com/journal/ijgo 17

18 Yang ET AL. 1 INTRODUCTION Sub- Saharan African countries share an enormous burden of maternal deaths. As part of the Sustainable Development Goals (SDGs), world leaders adopted the target of reducing global maternal mortality to less than 70 per 100 000 live births by 2030. 1 To achieve this global target, countries with baseline maternal mortality ratios (MMRs) of less than 420 should reduce their MMR by at least two- thirds by 2030, while all other countries should target an MMR of less than 140 deaths per 100 000 live births. In 2015, the national MMRs in Uganda and Nigeria remained high at 343 and 814 per 100 000 live births, respectively. 2 For Uganda and Nigeria to achieve the target of lowering MMR and to ultimately end preventable maternal deaths, highquality intrapartum care must be ensured for all women. Labor monitoring is essential to reduce maternal, fetal, and neonatal mortality and morbidity through early detection and management of complications. Ideally, as a woman progresses through labor, maternal and fetal well- being (including fetal heart rate, maternal blood pressure, temperature, uterine contractions, urine output, and cervical dilation) should be closely monitored and appraised against clinical standards by skilled birth attendants. While not all intrapartum complications can be identified through labor monitoring, assessing maternal and fetal well- being and labor progress offers the health provider the best chance for early identification of physiological and labor abnormalities associated with complications and poor outcomes. Regular assessment of the laboring woman can also identify infection, sepsis, hypertensive disorders, and other complications. In the present study, the operational definition of labor monitoring covers the continuous monitoring of women and their unborn babies throughout labor, as well as health professionals actions in response to women s emerging needs. The monitoring package includes the partograph, which is widely recommended as a tool for graphically assessing labor progression, so that any deviations from normal can be quickly detected and responded to. The modified WHO partograph begins with the onset of the active phase of labor 3 6 and continues through to childbirth, recording information about cervical dilation, number and duration of uterine contractions, fetal heart rate, maternal vital signs, and status of amniotic membranes. Health professionals such as midwives, obstetricians, medical officers, and interns are deemed as the frontline health workers (FLHWs) responsible for labor monitoring in health facilities, including use of the partograph. Consequently, the optimum development of innovative tools to improve the quality of intrapartum monitoring depends on understanding both FLHWs current practices, and their views on the challenges and opportunities in this area. 7,8 Previous qualitative studies in this area conducted among health professionals have generally focused on specific aspects of intrapartum assessments, such as the monitoring of fetal heart rate, 9 12 use of the partograph, 13 and labor monitoring provided to a particular subgroup of women, or by a certain cadre of health workers. 7,8,14 In particular, use of the partograph for labor monitoring is problematic: studies conducted in Nigeria concluded that the lack of knowledge about and training in the use of this tool limited the correct use and often resulted in nonuse among health workers. 15,16 Similarly, a study conducted in Uganda found that in two large referral health facilities only 23.9% and 18.3% of births were monitored using a partograph. 17 Apart from partograph use, much of the evidence on intrapartum monitoring is based on high- income countries. Knowledge in this area would be strengthened by a comprehensive analysis of all aspects of labor monitoring as an integral part of high- quality intrapartum care in low- income settings, where most of the intrapartum maternal and neonatal morbidity and mortality events occur. The Better Outcomes in Labor Difficulty (BOLD) project was launched by the WHO with the goal of reducing intrapartum- related maternal, fetal, and newborn mortality and morbidity. 18 Specifically, the BOLD project focused on the development of: (1) a Simplified, Effective, Labor Monitoring- to- Action tool (SELMA) to assist healthcare providers with labor monitoring and decision- making processes; and (2) tools to generate demand and promote women s access to respectful, quality care during childbirth ( Passport to Safer Birth ). 18 SELMA is envisioned as an electronic tool designed to enhance human expertise and optimize the outcomes of labor and childbirth. 19 SELMA aims to assist FLHWs with recording and interpreting complex information about labor and taking appropriate clinical actions, thus enhance the efficiency of labor monitoring and improve the capacity of health workers. 19 Formative research was conducted as part of the BOLD project to develop an intervention strategy to overcome barriers to adequate labor monitoring and improve the overall quality of intrapartum care. The findings reported in the present paper are part of the BOLD project formative research. The aim was to explore healthcare providers perspectives on current labor monitoring practices, and their perceptions of local challenges to, and opportunities for, the provision of high- quality monitoring and timely, safe, and effective interventions during labor and childbirth. 20 2 MATERIALS AND METHODS 2.1 Study design and setting This analysis is nested within the BOLD formative research project with a specific focus on health professionals perspectives. 18 20 In- depth interviews (IDIs) and focus group discussions (FGDs) were conducted with healthcare providers and health facility administrators to explore labor monitoring practices, challenges, and opportunities from both clinical and managerial perspectives. 20 2.1.1 Study setting The study took place in the Federal Capital Territory (Abuja) and Ondo State, Nigeria, and Kampala, Uganda, from December 2014 to April 2015. Low density of health workers at less than 0.5 per 1000 population poses challenges to providing high- quality maternal and obstetric care to women in both countries. 21 23 In Nigeria, the three- level health service delivery system consists of primary, secondary, and tertiary levels of health care. Usually, specialist obstetric and gynecological care is provided only in the secondary and tertiary hospitals, while primary care is delivered at community- based facilities staffed by midwives and

Yang ET AL. 19 community health extension workers (CHEW). In Uganda, health services are delivered in a pyramid of seven levels, from Health Center I to Health Center IV, district, regional to national levels. Only Health Center III and above levels have the capacity to provide maternity services. More details of the study settings can be found in the study protocol 20 and country- specific analyses in this Supplement. 24,25 Facilities were chosen based on the following criteria: (1) having at least 1000 births per year; (2) being a major healthcare facility in the region; (3) not being a primary health unit; (4) having relatively stable access to skilled birth attendants, and to the provision of cesarean delivery, augmentation of labor, assisted vaginal birth, and good intrapartum care practices. 20 2.1.2 Domains of interest Guides for semistructured IDIs and FGDs were developed based on the aims of the project, and tested prior to the data collection. The discussion guides included the following domains (full discussion guides are available as Supporting Information S1): (1) exploring the meaning of quality childbirth care; (2) expectations and needs to provide quality childbirth care; (3) barriers and facilitators to the provision of quality childbirth care, focusing on labor monitoring; (4) potential changes to enhance the provision of quality childbirth care, across women, community, provider, facility, and health system levels; and (5) perceived expectations and needs of women seeking facility- based intrapartum care. This analysis focuses on the data relating to labor monitoring, as this was the aspect that was of specific importance to the conceptualization of SELMA. 18 20 2.1.3 Sampling and data collection Eight health facilities were selected as sites in Nigeria and Uganda. 26 Within each selected facility, participants were stratified by cadre: (1) doctors (including obstetrician- gynecologists); (2) midwives; and (3) health facility administrators. We formed the discussion groups within each cadre of providers, such that the participants from the same cadre were similar in terms of their clinical or managerial roles. For IDIs, participants within each cadre were further selected based on their age and work experience to maximize the variation and to increase the representativeness of the sample. 18 20 Face- to- face IDIs and FGDs were conducted in English at the health facilities where the privacy of the discussions was assured, after obtaining written consent from the participants. Each IDI or FGD lasted approximately 40 90 minutes, and each was audio- recorded. Data were collected by the local teams, consisting of public health professionals with maternal health and qualitative research experience, and with postgraduate credentials. Prior to data collection, the teams received two days of training on the study protocol, interview guides, and research ethics. Data were transcribed on an on- going basis and non- English language words were translated to English verbatim during that process. Data collection was completed when saturation was reached (i.e. where no additional new information emerged). Field notes were taken simultaneously with data collection to ensure the credibility and trustworthiness of the data. 20 2.2 Data analysis We conducted thematic analysis using a combined inductive and deductive approach to developing a codebook. This analysis focuses on four key themes and topical codes related to intrapartum care that emerged from the discussion guides: (1) description of labor monitoring; (2) decision- making in response to observations of labor progress; (3) challenges to monitoring labor and deciding to intervene; and (4) opportunities to improve labor monitoring and decision- making. We then completed an iterative appraisal of the data on those themes and identified key findings. When subthemes emerged, we integrated them into the existing code structure as response codes. Finally, we synthesized the key findings and selected quotations based on their representativeness and relevance to the aim of this analysis. 2.3 Ethics approval Scientific and technical approval was obtained from the UNDP/ UNFPA/UNICEF/WHO/World Bank Special Program of Research, Development and Research Training in Human Reproduction (HRP) Review Panel on Research Projects (RP2). Ethical approval was obtained from the WHO Ethical Review Committee (protocol A65878), the Makerere University School of Health Sciences Research and Ethics Committee, Uganda (protocol #SHSREC REF 2014-057), the Federal Capital Territory Health Research Ethics Committee, Nigeria (protocol FHREC/2014/01/43/27-08- 14), and Ondo State Government Ministry of Health Research Ethics Review Committee, Nigeria (AD 4693/168). 3 RESULTS 3.1 Overview A total of 70 IDIs were conducted in both countries: 23 with doctors, 35 with midwives, and 12 with health facility administrators. In addition, 16 FGDs were conducted, seven with doctors and nine with midwives. The distribution of group discussions and interviews according to cadre and country of participants is summarized in Table 1. Our analyses yielded several salient subthemes within the data. First, in describing labor monitoring practices, emergent subthemes included a woman s path through labor monitoring and use of the TABLE 1 Distribution of study participants. Interviews (n=70) Focus groups (n=16) Country Uganda 33 8 Nigeria 37 8 Cadre Midwives 35 9 Doctors 23 7 Administrators 12 0 Total 70 16

20 Yang ET AL. partograph. Second, when complications arose and decisions were needed, midwives and doctors clinical responsibilities were frequently discussed side- by- side. Third, on challenges during labor monitoring, subthemes included staff shortages, delays in response, suboptimal clinical team cooperation and provider client communication, and limitations of labor monitoring tools. Fourth, the perceived opportunities to improve labor monitoring practices included subthemes on human resource optimization, women s preparedness enhancement, and individualized intrapartum care and tool improvement. These themes and subthemes are summarized in Table 2 and are presented in detail in the following sections. 3.2 Descriptions of labor monitoring practices Participants reported that labor monitoring starts from the moment the woman is admitted in labor to the time that her third stage of labor is complete. They reported that midwives and doctors shared labor monitoring tasks, and that they complement each other s roles. Specifically, they noted that midwives are stationed in labor and antenatal wards, and that they take the responsibilities for continuous monitoring by bedside, while providing routine labor care to the women. The specific monitoring tasks mentioned by midwives included checking vital signs, fetal heart rate, blood pressure, and ruling out the possibility of fetal distress. In contrast, participants reported that most doctors were not stationed within the labor wards, but that they could be reached in the casualty ward or theatre when needed. Participants reported that the medical staff carried out intermittent checks on women, and undertook interventions in the event of complications. TABLE 2 Overview of key themes. Themes 1. Descriptions of labor monitoring practice 2. Decision- making related to labor progress 3. Perceived challenges during labor monitoring 4. Perceived opportunities to improve labor monitoring Subthemes 1.1. A woman s path through labor 1.2. The use of labor monitoring tools 2.1. Midwives clinical responsibilities 2.2. Doctors clinical responsibilities 3.1. Staff shortages 3.2. Delays in responding to abnormal labor observations 3.3. Suboptimal clinical team cooperation 3.4. Ineffective provider client communication 3.5. Limitations of the labor monitoring tools 4.1. Optimizing human resources and clinical roles 4.2. Increasing women s preparedness before labor 4.3. Individualizing care in labor monitoring and actions 4.4. Developing/improving labor monitoring tool 3.2.1 A woman s path through labor Participants were asked to describe a woman s path throughout labor in their setting. They reported that pregnant women were admitted to the maternity units after an initial assessment by midwives and doctors (if admitted during the day), or by midwives only (if admitted at night). From there, the paths diverged into high- risk and low- risk cases depending on whether certain risk factors were present. High- risk cases were those with prior or recent obstetric problems that might require doctors immediate advice and engagement on proper interventions, while low- risk cases (that constituted the majority) were expected to progress through labor normally (without intervention), only requiring routine check- ups by a midwife. According to participants, assessment at labor admission in both Nigeria and Uganda usually included a physical examination, assessment of any abnormality (such as anemia and edema), and evaluation of the fetal heart rate. According to the cervical dilation status, women in the latent phase of labor (often based on cervical dilation <4 cm) were not admitted to the labor ward. In contrast, those deemed to be in the active phase of labor (based on cervical dilation 4 cm) were granted admission: When a mother comes to the labor ward, we receive her; we assess her from there depending on how far she is. If she is still in latent labor I will continue doing the assessment and at least do vaginal examinations every four hours or when she complains of increased contractions, then maybe I will break my four- hourly vaginal examinations and check to see how far she has gone as soon as she gets into active labor we move her to the labor ward and start plotting the partograph (Doctor, IDI, Uganda). 3.2.2 The use of labor monitoring tools The partograph was frequently mentioned. It appeared to be considered as the most important tool for labor monitoring by midwives, doctors, and administrators. Its original intention was to distinguish the abnormal from the normal in labor progression, and to enable decision- making around transfer from remote areas to facilities if labor progress is stalled. In practice, the data suggest that it also served to pass health information among midwives when handing over shifts, and between midwives and doctors when midwives report to or consult doctors for clinical advice. Again my partograph will guide me When I monitor this mother and she crosses the alert line to action line that means action has to be taken because she is not progressing and I have to call the doctor to re- examine and then opt for plan B. (Midwife, IDI, Uganda). One doctor commented that the key strength of the partograph is its objectivity, as opposed to the subjective nature of decision- making by clinicians.

Yang ET AL. 21 3.3 Decision- making related to labor progress In both Nigeria and Uganda, participants consistently reported that doctors were responsible for making major medical decisions for the management of labor and birth, based on critical information passed by the midwives to the doctors, either verbally, or with the aid of a plotted partograph depicting the woman s labor progression. However, if actions were needed while waiting for medical input, midwives argued that they had the agency to decide what to do, and to act accordingly, to prevent poor birth outcomes: Usually, if we get a case that we cannot manage, as we wait for the doctor, we can still do the needful We take the observations, we put up the IV line like normal saline, and we get blood for grouping and matching So, we just don t sit (saying) that we cannot do anything for the patient because the doctor is not there. (Midwife, FGD, Uganda). Some midwives expressed that they felt empowered to make clinical decisions on medical interventions that they believed to be within the scope of their skillset and knowledge, such as labor augmentation, and recommending and administering commonly used essential medications. 3.3.1 Midwives clinical responsibilities Midwives responsibilities were reported to be a continuous process, encompassing four dimensions: (1) routine assessment of uterine contractions, fetal heart rate, cervical dilation, descent of the fetus, pulse, maternal blood pressure, etc.; (2) documenting the progress of labor; (3) consulting with doctors about abnormalities or concerns; and (4) actively communicating with women and their family members and providing support. Communicating with women and their families involved constant exchange of information with the woman, explanation of the next planned activities, enquiries about the women s emotional state, and providing reassurance that the woman was in safe hands : We monitor her using a partograph and we start a partograph when a mother is 4 centimeters dilated. Every 4 hours we have to review the fetal part and maternal part. The fetal heart of the baby, we do it every after 30 minutes, plus the contractions of the mother Then also you have to take their vital observations: the blood pressure is taken 2 hourly, temperature is taken 4 hourly, then pulse which is taken every 30 minutes. (Midwife, FGD, Uganda). 3.3.2 Doctors clinical responsibilities Doctors reported working at the health facilities during standard working hours (from 8:00 am to 4:00 pm) and staying on call duty for emergencies occurring after hours. During normal hospital hours, doctors noted that they discussed cases with midwives, conducted regular rounds of routine checks in the labor ward, provided guidance and supervision to midwives labor monitoring work, and made decisions about when to intervene. Respondent 3: On the partograph, aside from poor cervical dilation, there are some other indices that can make the doctor intervene in labor that will help you to make a good judgment on the type of intervention you need. Respondent 1: The doctor is responsible for every action in management of labor he takes responsibility whenever this is a complication. Respondent 6: Decision is taken by the leader of the team, who is the doctor. (Doctors, FGD, Nigeria). 3.4 Perceived challenges during labor monitoring Participants expressed concerns about the difficulties in monitoring labor. The commonly reported challenges included staff shortages, delays in responding to abnormal labor observations, suboptimal clinical team cooperation, insufficient provider client communication, and limitations of labor monitoring tools. 3.4.1 Staff shortages Staff shortages were viewed as a major barrier to effectively monitoring labor in both countries. Low provider- to- client ratios diminished midwives ability to attend to labor, especially during night shifts when few doctors were present and multiple needs for care arose simultaneously. Doctors from both countries felt strongly about the challenges brought by understaffing, which they considered accountable for incomplete recording of labor progress. Sometimes I feel that before she finished filling all those things (in a partograph) it would be time for the next monitoring and then we don t have one patient to one nurse. (Doctor, IDI, Nigeria). Staff shortages not only affected observing and recording, but also affected the readiness of tackling complications during labor. When multiple women with urgent medical needs arrived at an under- staffed facility demanding immediate care, the situation became dire. Staff shortages were not limited to obstetricians and midwives but also those from other specialties such as anesthesiologists, neonatologists, and nurses. You know, sometimes a patient may come, she has obstructed labor, and you may find you have three people with obstructed labor and you have one anesthetist, you see? Meaning you have to work on this one as another one has to wait. And that can prolong the time from decision making to intervention, you see? (Doctor, IDI, Uganda).

22 Yang ET AL. Administrators confirmed that the shortage of human resources hindered the effective functioning of labor monitoring and led to incorrect completion of the partograph: [when] you are like two midwives on duty at night, there could be six women [in labor]. Monitoring becomes difficult, so sometimes they lie and just plot. (Administrator, IDI, Uganda). 3.4.2 Delays in responding to abnormal labor observations Another challenge reported by the participants was the delay between the onset of danger signs and the provision of clinical intervention, which was felt to put the health of the woman and/or the baby at risk. Some delays were due to the separation of monitoring and decision- making processes embedded in the health facility structure, and others were attributed to other nonstructural reasons on the clients side. To me, sometimes it s not taking prompt action. Maybe during the course of your monitoring, you may see a woman who needs intervention; sometimes you call for the doctor. The decision may be taken but before you get to the theater, time would have been wasted before the woman is finally wheeled to the theater for necessary action, so there was delay. (Midwife, FGD, Nigeria). Midwives and administrators both noted that it was extremely difficult for the health workers to implement a clinical decision and deliver the care in a timely manner when women or their family members hesitate to consent to clinical decisions, especially when the hesitation was compounded by the preference for nonbiomedical remedies, religious reasons, or economic hardship: Women feel that if you are operated [on], you are not a complete woman. So, for her to make that decision, she has to consent. So, you can take a lot of time having a discussion and then they say that let us wait for the husband and then the husband says that let us wait for her mother. So, making serious significant decisions [is] commonly delay[ed] when you are giving care. (Administrator, IDI, Uganda). 3.4.3 Suboptimal clinical team cooperation Team cooperation among midwives and doctors was reported to shape the quality and the outcomes of labor monitoring. Seamless cooperation between doctors and midwives facilitated the process of handing- over cases and exchanging clinical information. One doctor stressed the importance of such cooperation and emphasized that labor monitoring is not a one- man show. Another doctor mentioned: And I feel that the junior doctor did not realize the need to involve the other people By the time we reached the theatre, the patient had lost so much [blood] and she did not survive. (Doctor, IDI, Uganda). 3.4.4 Ineffective provider client communication A word commonly used by FLHWs to describe challenges in interacting with women in labor was uncooperative. Participants concluded that it was essential for providers to take time to communicate with women and their families in a way that they understood, and to build collaborative relationships with women and their families to effectively provide intrapartum care. Otherwise, tension may occur: You know somebody is yelling and you do not know what is happening inside with the baby and then the patient may refuse to be examined and says I have had enough of this, no more vaginal examinations. They can walk away, start shouting, abusing you yet you think you are doing your best. (Administrator, IDI, Uganda). Inadequate provider client communication may lead to misunderstanding, which further fosters tension. A doctor described an instance where a recommendation for a cesarean delivery was understood as a forceful demand from the client s perspective: So I explained to this lady, I told her you can deliver normally but there are risks and I told her sometimes a cesarean may be needed she told me, my sister is there, so I called her and I also explained to her the same thing. The sister told me, let her deliver normally. I said Okay. We monitored her, the mother delivered, fortunately everything was okay but now according to the sister, what she told the other attendants is that we were forcing them to operate the woman. That is why I was saying that communication at times is perceived differently, because you explain to maybe the patient or attendant but she will understand different things. (Doctor, IDI, Uganda). 3.4.5 Limitations of labor monitoring tools Participants expressed their concerns with use of the partograph. Sometimes charts were completed retrospectively after childbirth to fulfil hospital protocol requirements. Doctors tended to believe that the nonuse or incorrect use of the partograph frequently occurred among midwives because they lacked knowledge of how to complete it, or that they failed to appreciate its usefulness and value in monitoring labor. In their opinion, even if midwives were trained to use the partograph, they might not truly understand the implication of using it. In comparison to midwives, doctors generally saw more benefits of using the partograph, partly because they interacted less with women in labor, and therefore they relied on the collection of clinical information, such as that recorded on correctly completed partographs.

Yang ET AL. 23 In contrast to the reports of some medical staff that midwives did not understand or complete the partograph effectively, many midwives argued that they used them consistently during labor monitoring, and that they recognized their value. Other midwives held their overwhelming workload accountable for partograph nonuse or misuse. They argued that suboptimal use could be related to the staff shortages, and to the heavy workload in labor monitoring, rather than owing to lack of knowledge: These days, the partograph is well filled. But understaffing may make that partograph not to be filled well because the mothers are many. However much you try, you will find that you fill only three partographs for those mothers. But they know how to do things but sometimes it s understaffing that makes documentation and filling not done. But we have the knowledge. (Midwife, IDI, Uganda). One administrator combined both aspects (lack of knowledge and insufficient time) in explaining why the partograph was not used optimally: First of all, people assume that people have knowledge about partograph use and yet it is not there. People themselves, the health service providers the knowledge about a partograph is not there, they lack knowledge. And the other thing, sometimes the environment we are working in, you deliver like 50 to 100 [babies], then you are two or three midwives, do you expect quality partograph use? (Administrator, IDI, Uganda). 3.5 Perceived opportunities to improve labor monitoring Four subthemes emerged from the data surrounding potential opportunities to improve labor monitoring: optimizing human resources and clinical roles, increasing women s labor preparedness, individualizing care in labor monitoring and actions, and developing or improving a labor monitoring tool. 3.5.1 Optimizing human resources and clinical roles Many participants recommended optimizing human resources to streamline the workflow and to improve the practices of labor monitoring. Specifically, they suggested the need to strengthen the overall teamwork, foster better communication within the team, reduce the number of clients per provider, and enhance the clinical capacity of midwives to improve the quality of care. If we all see ourselves as a team, not as individuals, then the work will be much better. [We should not] see it as: the doctors are supposed to do this, it is the nurses that are supposed to do that. We should work together. (Midwife, FGD, Nigeria). Several midwives suggested increasing autonomy for midwives in clinical decision- making, and in their capacity to intervene during labor. They argued that decentralizing certain intervention decisions from doctors to midwives could lead to more efficient decision- making and shorter time- to- action in addressing emerging clinical needs: I would like the midwives to be the ones to take in the interventions to decide because the doctor has not been there but this midwife has been with the mother and knows what exactly was happening there That is what I would change at the provider level. (Midwife, IDI, Uganda). 3.5.2 Increasing women s preparedness before labor Some doctors suggested that efforts should be made prior to labor to mentally and physically prepare women for childbirth. For example, if chronic maternal conditions such as diabetes and obesity could be detected and managed early on through preconception counselling, the burden of labor monitoring would be lessened. Some participants adopted a continuity of care perspective, and emphasized the role of antenatal care as essential early preparatory steps to achieve a positive labor experience for women. Participants also noted that by enhancing community awareness and educating women and their family about childbirth, issues surrounding provider client communication could be addressed beforehand. This could alleviate health professionals workload when monitoring labor, and also improve women s satisfaction throughout childbirth as a result of prior knowledge of, and preparedness for, physically intrusive assessments such as vaginal examination. 3.5.3 Individualizing care in labor monitoring and actions Respondents from all the professional groups that took part in the study also discussed the concept of a normal uncomplicated labor. Some doctors questioned standardized assumptions about the limits of normal labor, noting that a labor monitoring tool should allow certain deviation from the standardized alert line on a partograph to acknowledge individual variations. For example, one Ugandan doctor advocated for more individualized intrapartum care, where normal is defined not by comparison with other women or a fixed standard, but by a woman s natural labor progression: Those ones are supposed to be individualized depending on the stage of a woman s labor. For example, if somebody is just 4 cm [cervical dilation] and they don t have any complications, they don t have to be repeatedly checking the centimeters in the next four hours. (Doctor, FGD, Uganda).

24 Yang ET AL. In a demonstration that this belief was widespread, it was echoed by a midwife from Nigeria: depending on the type of patient you have, you care for the patient with respect to her case my care to my patient is individualized depending on what you know [about] the condition of the patient. (Midwife, IDI, Nigeria). 3.5.4 Developing/improving labor monitoring tools A few participants voiced a strong desire for a more effective tool than the current partograph. Apart from maintaining key functionalities of a partograph as discussed above, an improved labor monitoring tool should be user- friendly, feasible to deploy in the local context, reliable as a tool for detecting danger signs in a timely manner, and, ideally more automated/dynamic (e.g. by indicating possible courses of action relating to the needs of the individual woman/fetus as labor progresses) and less complex for FLHWs to use (see Box 1). 4 DISCUSSION 4.1 Findings The present study found that in the included sites, as in most facilitybased maternity care around the world, labor progress monitoring was, in general, divided into routine bedside care by midwives and as- needed examination and interventions by medical staff; however, there were exceptions to this pattern of working. A range of contextual limitations stood out as barriers to monitoring maternal and fetal well- being during labor, including staff shortages, lack of team cooperation, delays in responding to abnormal labor observations, suboptimal provider service user dynamics, and limitations in knowledge about and effective use of the partograph. Findings related to staff BOX 1 Goals that an improved labor monitoring tool should achieve according to healthcare providers suggestions. Foster better communication within a team. Optimize provider patient ratios by increasing efficiency and decision-making ability of individual FLHWs. Enhance the clinical and decision-making capacity of midwives. Individualized expectations of labor progression. User-friendly. Feasible to deploy in the local context. Effective and reliable in detecting danger signs in a timely manner. More automated/dynamic than the current partograph (for example, by indicating possible courses of action as labor progresses). Less complex and labor-intensive for the provider to use. shortages echoed those found in relation to low physician to patient ratios in both countries in previously published reports. 23 Some of the contextual barriers such as lack of respect and teamwork between obstetricians and midwives, heavy workloads, and challenges in maintaining continuity of labor monitoring and care by midwives have been reported by health professionals in both high- and low/middleresource settings. 27 30 Although some of the health systems challenges may not be amenable in the short term, these findings provide a clear understanding of the potential obstacles that any labor monitoring tool must overcome to function in a real- life setting. One challenge that stood out was the delay in the continuum of labor progress assessment, and the diagnosis of and action to address stalled progress. This directly impacted on labor outcomes at critical moments. Taking into account the findings presented in this paper, along with those presented in other papers in this Supplement, these delays could be chronologically categorized into the following stages: (1) delays in recognizing danger signs by midwives; (2) delays in reporting the case to the doctor; (3) delays in making a clinical decision in response to observed danger signs; (4) delays in initiating intervention owing to prolonged negotiation with the woman and her family members, or difficulty in obtaining consent; and (5) delays in intervening owing to lack of necessary equipment or timely referral (see Fig. 1). The classic three delays framework in maternal- care seeking as coined by Thaddeus and Maine 31 involves accessing facility care from the community. The third delay in this model is in receiving adequate care in the facility. The five delays noted in this study unpack the detail of this third delay. The accounts of respondents in the present study suggest that these internal delays may be partly due to separation of the clinical decision- making roles (usually undertaken by doctors) from continuous monitoring roles (usually undertaken by midwives), and that this might be further compounded by the perceptions, expectations, and consequent behaviors of childbearing women. Midwives in this study commonly emphasized the importance of the partograph in assessing labor progression and insisted that they used it consistently in documenting dynamic labor events such as cervical dilation and fetal heart rate. Conversely, doctors tended to report that midwives frequently misused the partograph, or didn t use it at all. The disparity in the perspectives of these two cadres on partograph use might be due to social desirability bias (a type of response bias where respondents answer questions in a manner that is favorable to others, such as over- reporting good behaviors or under- reporting bad behaviors). The concept of individualized care as a basis for assessing labor progress in the context of a particular woman, and the definitions of normal versus abnormal labor were discussed by some of the study participants. They suggested that the standardized criteria for assessing cervical dilation and judging normal versus abnormal labor progress in reference to a partograph alert line may neglect the variations among women with different characteristics and different labors. Such typology medicalizes slowly, yet normally progressing labor and may trigger unnecessary interventions, including labor augmentation and cesarean delivery. In settings with limited capacity to provide close intermittent fetal monitoring, unnecessary labor augmentation

Yang ET AL. 25 FIGURE 1 Delays in the process of labor monitoring. [Colour figure can be viewed at wileyonlinelibrary.com] may expose babies to an increased risk of fetal distress and contribute to poor maternal and newborn outcomes. Unnecessary labor augmentation may also lead to an increase in cesarean deliveries performed in settings where access to safe surgery is challenging. Safely avoiding unnecessary interventions also has the potential to reduce the burden on midwives and doctors as it reduces the additional workload associated with the interventions. An improved labor monitoring- to- action tool could usefully integrate a means of bridging the gaps between the activities, values, and attitudes of doctors, midwives, and women during labor. This could minimize delays in response to slow labor progress during facilitybased intrapartum care while maximizing individualized labor care, with the ultimate aim of averting preventable morbidity and mortality. Such a tool would need to pay attention to these interpersonal elements, as well as to the resource constraints that are endemic in low- income settings. 4.2 Strengths and limitations The study included findings from respondents from a range of professional groups and from more than one setting. While the data are situated in the local health systems, the contextual enablers and barriers are common to many low- resource maternity care facilities. The richness of data from multiple perspectives provided a comprehensive view on labor monitoring from the key stakeholders, and supplements the findings from other papers in this Supplement. 24,25,32 This study may have been limited by the self- reported practices from the health professionals, which can lead to social desirability bias, resulting in a narrative of ideal situation rather than the actual situation of labor monitoring. However, this is minimized by the triangulation of information sources in this study. This adds to the credibility of the qualitative data and the trustworthiness of the conclusions drawn. 4.3 Conclusion Labor monitoring encompasses a broad scope of care. In this study, assessment of labor progress was jointly provided by doctors and midwives, but they reported having different spheres of practice in this area, which did not always intersect smoothly. Health professionals faced several challenges in effective monitoring of labor progress and there were opportunities for improvement in the practices and the tools used in labor monitoring. The findings from this study suggest that, from the health professionals perspective, the ideal tool for labor monitoring needs to improve accurate, individualized labor progress assessment and its documentation; to be actionoriented in terms of indicating different possible actions relating to the individual woman/baby as the labor progresses; to accommodate contextual factors such as staff shortages in the local health system; to facilitate positive and effective communication between midwives, obstetricians, and laboring women; to ensure the optimal care pathway is followed for each individual woman and baby in collaboration with the woman and provider; and to bridge the gap in the continuum of monitoring and intervention tasks between healthcare cadres. AUTHOR CONTRIBUTIONS MAB, JPS, OTO, and AMG designed and planned the study with input from BF and KM. MAB, BF, KM, MAT, DK, and AOO led data collection and management. FY and MAB led the analysis with input from all authors. FY wrote the first draft of the manuscript with input from MAB, OTO, JPS, and AMG. All authors read, commented on, and approved the final manuscript. ACKNOWLEDGMENTS We would like to express our sincere gratitude to the providers who participated in this study and the data collectors for their hard work and dedication. Thank you also to Özge Tunçalp, Joshua Vogel, Rajiv Bahl, and the WHO BOLD Research Group for their invaluable feedback throughout the project. Fan Yang would like to personally thank the China Scholarship Council (funding No. 201407990005) for partially supporting her ongoing doctoral education. The authors alone are responsible for the views expressed in this publication and they do not necessarily represent the views, decisions, or policies of the institutions with which they are affiliated. CONFLICTS OF INTEREST The authors have no conflicts of interest to declare.

26 Yang ET AL. REFERENCES 1. World Health Organization. Targets and Strategies for Ending Preventable Maternal Mortality: Consensus Statement. Geneva: WHO; 2014. 2. World Health Organization. Trends in Maternal Mortality: 1990 to 2015: Estimates by WHO, UNICEF, UNFPA, World Bank Group and the United Nations Population Division. Geneva: WHO; 2015. 3. Levin K, Kabagema JD. Use of the Partograph: Effectiveness, Training, Modifications, and Barriers A Literature Review. New York: EngenderHealth/Fistula Care; 2011. 4. Mathews J, Rajaratnam A, George A, Mathai M. Comparison of two World Health Organization partographs. Int J Gynecol Obstet. 2007;96:147 150. 5. Philpott R. Graphic records in labour. BMJ. 1972;4:163. 6. Philpott R, Castle W. Cervicographs in the management of labour in primigravidae. BJOG. 1972;79:592 598. 7. Widmark C, Leval A, Tishelman C, Ahlberg BM. Obstetric care at the intersection of science and culture: Swedish doctors perspectives on obstetric care of women who have undergone female genital cutting. J Obstet Gynaecol. 2010;30:553 558. 8. Dieme Faye ME, Moreira P, Tamofo E, Diouf AA, Diouf A, Moreau JC. Monitoring pregnancy of women with a previous cesarean delivery: qualitative aspects and prognostic implications [in French]. Med Sante Trop. 2014;24:409 415. 9. Lewis L, Rowe J. Focus on the beat: current fetal monitoring practice in low risk labour. Aust J Midwifery. 2004;17:6 10. 10. Lewis L, Rowe J. Moving with the beat: changing fetal monitoring practices in low- risk labour. Aust J Midwifery. 2004;17:23 26. 11. Altaf S, Oppenheimer C, Shaw R, Waugh J, Dixon-Woods M. Practices and views on fetal heart monitoring: a structured observation and interview study. BJOG. 2006;113:409 418. 12. McKevitt S, Gillen P, Sinclair M. Midwives and doctors attitudes towards the use of the cardiotocograph machine. Midwifery. 2011;27: e279 e285. 13. Ogwang S, Karyabakabo Z, Rutebemberwa E. Assessment of partogram use during labour in Rujumbura Health Sub District, Rukungiri District, Uganda. Afr Health Sci. 2009;9(Suppl 1):S27 S34. 14. Silva RM, Barros NF, Jorge HM, Melo LP, Ferreira Junior AR. Qualitative evidence of monitoring by doulas during labor and childbirth [in Portuguese]. Cien Saude Colet. 2012;17:2783 2794. 15. Fawole A, Adekanle D, Hunyinbo K. Utilization of the partograph in primary health care facilities in southwestern Nigeria. Niger J Clin Pract. 2010;13:200 204. 16. Opiah MM, Ofi AB, Essien EJ, Monjok E. Knowledge and utilization of the partograph among midwives in the Niger Delta Region of Nigeria. Afr J Reprod Health. 2012;16:125 132. 17. Landry E, Pett C, Fiorentino R, Ruminjo J, Mattison C. Assessing the quality of record keeping for cesarean deliveries: results from a multicenter retrospective record review in five low- income countries. BMC Pregnancy Childbirth. 2014;14:139. 18. Oladapo OT, Souza JP, Bohren MA, et al. WHO Better Outcomes in Labour Difficulty (BOLD) project: innovating to improve quality of care around the time of childbirth. Reprod Health. 2015;12:48. 19. Souza JP, Oladapo OT, Bohren MA, et al. The development of a Simplified, Effective, Labour Monitoring- to- Action (SELMA) tool for Better Outcomes in Labour Difficulty (BOLD): study protocol. Reprod Health. 2015;12:49. 20. Bohren MA, Oladapo OT, Tuncalp O, et al. Formative research and development of innovative tools for Better Outcomes in Labour Difficulty (BOLD): study protocol. Reprod Health. 2015;12:50. 21. World Health Organization. WHO Country Cooperation Strategy: Nigeria. Brazzaville: World Health Organization Regional Office for Africa; 2001. 22. World Health Organization. Country Cooperation Strategy Uganda. Uganda: World Health Organization, Country Profile; 2009. 23. The Global Health Workforce Alliance. The Global Health Workforce Alliance 2012 Annual Report: Making Health Workers Count. Geneva: WHO; 2012. 24. Kyaddondo D, Mugerwa K, Byamugisha J, Oladapo OT, Bohren MA. Expectations and needs of Ugandan women for improved quality of childbirth care in health facilities: A qualitative study. Int J Gynecol Obstet. 2017;139:S1:38 46. 25. Ojelade OA, Titiloye MA, Bohren MA, et al. The communication and emotional support needs of childbearing women using health facilities in Southwest Nigeria: a qualitative study. Int J Gynecol Obstet. 2017;139:S1:27 37. 26. Patton MQ. Qualitative Research and Evaluation Methods. Saint Paul, Minnesota: SAGE Publishing; 2005. 27. Balde MD, Diallo BA, Bangoura A, et al. Perceptions and experiences of the mistreatment of women during childbirth in health facilities in Guinea: a qualitative study with women and service providers. Reprod Health. 2017;14:3. 28. Tibandebage P, Kida T, Mackintosh M, Ikingura J. Can managers empower nurse- midwives to improve maternal health care? A comparison of two resource- poor hospitals in Tanzania. Int J Health Plann Manage. 2016;31:379 395. 29. Klomp T, Witteveen AB, de Jonge A, Hutton EK, Lagro-Janssen AL. A qualitative interview study into experiences of management of labor pain among women in midwife- led care in the Netherlands. J Psychosom Obstet Gynaecol. 2017;38:94 102. 30. Karvande S, Sonawane D, Chavan S, Mistry N. What does quality of care mean for maternal health providers from two vulnerable states of India? Case study of Bihar and Jharkhand. J Health Popul Nutr. 2016; 35:6. 31. Thaddeus S, Maine D. Too far to walk: maternal mortality in context. Soc Sci Med. 1994;38:1091 1110. 32. Bohren MA, Titiloye MA, Kyaddondo D, et al. Defining quality of care during childbirth from the perspectives of Nigerian and Ugandan women: a qualitative study. Int J Gynecol Obstet. 2017;139:S1:4 16. SUPPORTING INFORMATION Additional Supporting Information may be found online in the supporting information tab for this article. Supporting Information S1: Full in-depth interview and focus group discussion guides.