Provision of essential evidence-based interventions during facility-based childbirth: Cross-sectional observations of births in northeast Nigeria

Objectives To measure the provision of evidence-based preventive and promotive interventions to women, and subsequently their newborns, during childbirth in a high-mortality setting. Design and participants Cross-sectional observations of care provided to women, and their newborns during the intrapartum and immediate postpartum period using a standardised checklist capturing healthcare worker behaviours regarding lifesaving and respectful care. Setting Ten primary healthcare facilities in Gombe state, northeast Nigeria. The northeast region of Nigeria has some of the highest maternal and newborn death rates globally. Main outcome measures Data on 50 measures of internationally recommended evidence-based interventions and good practice. Results 1875 women were admitted to a health facility during the observation period; of these, 1804 gave birth in the facility and did not experience an adverse event or death. Many clinical interventions around the time of birth were routinely implemented, including provision of uterotonic (96% (95% CI 93% to 98%)), whereas risk-assessment measures, such as history-taking or checking vital signs were rarely completed: just 2% (95% CI 2% to 7%) of women had their temperature taken and 12% (95% CI 9% to 16%) were asked about complications during the pregnancy. Conclusions The majority of women did not receive the recommended routine processes of childbirth care they and their newborns needed to benefit from their choice to deliver in a health facility. In particular, few benefited from even basic risk assessments, leading to missed opportunities to identify risks. To continue with the recommendation of childbirth care in primary healthcare facilities in high mortality settings like Gombe, it is crucial that birth attendant capacity, capability and prioritisation processes are addressed. We conducted direct observations of childbirth care in 10 primary healthcare facilities, in Gombe state, Nigeria, approximately every 6 months over a 2-year period between June 2016 and August 2018. Gombe state is one of six states in northeast Nigeria, it has an area of 20 265 km2 and a population of 2 857 042.20 Over 80% of the population live in rural areas and are reliant on subsistence farming as their primary source of income.21 The northeast region of Nigeria has some of the highest maternal and newborn death rates globally, estimated at 1549 per 100 000 live births in 2015 and 33 per 1000 live births in 2017, respectively.22 23 Access to maternal healthcare services is relatively low in Gombe state. In 2018, 46% of women in the state reported at least one antenatal care visit from a doctor, nurse, midwife or nurse/midwife and 28% delivered in a health facility.24 Over 70% of facility deliveries, in 2018, took place in a rural primary healthcare facility.25 Recent work in Gombe on the drivers of attending a facility for childbirth found that health system conditions including availability of staff, drugs and supply, and a clean environment had the biggest influence on respondents’ decision around where to give birth.26 Healthcare is predominantly delivered via a network of rural primary healthcare clinics run by the Gombe State Primary Healthcare Development Agency (GSPHCDA). In 2017, 460 primary healthcare clinics and 26 referral facilities provided childbirth services.27 In primary healthcare facilities care is typically delivered by lower cadres of healthcare workers, for example, community health extension workers (CHEWS), junior CHEWS and health officers.28 29 In response to the shortage and uneven distribution of healthcare workers, under its task-shifting and task-sharing policy for essential healthcare services, Nigeria classifies CHEWs as skilled birth attendants.30 Primary healthcare facilities in Gombe are poorly resourced, often lacking essential supplies and commodities to provide basic maternal and newborn healthcare.31–33 Led by the GSPHCDA, since 2016 intense non-governmental organization activity has been ongoing in 57 primary healthcare facilities across Gombe state, aimed at increasing the quality of care.34 35 Interventions include training of CHEWs in all aspects of skilled birth attendance and basic emergency obstetric care, and improving the supply of essential maternal and newborn health commodities.36 These facilities provide basic emergency obstetric and newborn care. Emergency care and complicated cases from these health facilities are referred to referral facilities. None of the 57 primary healthcare facilities have a medical doctor, 4% have at least one nurse and 19% have at least one midwife.37 Sampling methods have been described in detail elsewhere.19 32 Briefly, in November 2015, 10 primary healthcare facilities were selected from the 57 facilities for an in-depth assessment of quality of care. To achieve a sufficiently large number of observations and minimise the duration of data collection, the 10 primary healthcare facilities with the highest number of births in the preceding 6 months, as recorded in the maternity register, were purposively selected. The mean number of births per month in the 10 primary healthcare facilities was 15.7 (SD 12.0), compared with 4.3 (SD 6.3) births per facility per month across Gombe state as a whole.19 Five rounds of data collection took place over the 2-year study period. Each round lasted 3 weeks, during which observers aimed to collect data from a total of around 350 women. Two trained female observers (local midwives, not employed by the facility) and one clinical supervisor were assigned to each facility. Observers worked in 8 or 12 hours shifts to provide near continual data collection during the period. Depending on the observation team’s work schedule, the first point of contact for any observation may have been during initial assessment of a newly admitted pregnant woman or at a later stage of labour. Observers aimed to observe all women who were admitted irrespective of the cadre of the attending healthcare worker, but they prioritised observing women during the second and third stage of labour and immediately post partum rather than observing women earlier in the process. Observers stayed continuously with women from the first point of contact until the first hour after birth. The healthcare worker observed may have been different at different timepoints in the same facility. The clinical supervisor was always available onsite but not present in the delivery room. A structured clinical observation checklist, administered on a Lenovo A3300 tablet using CSPro V.7.0 (US Census Bureau and ICF Macro, Suitland, Maryland, USA), was used to record the processes of care and birth attendant–client interactions and client characteristics. The content of the checklist was developed from the United States Agency for International Development (USAID)-funded Maternal and Child Health Integrated Programme’s tool for observing vaginal births and the following complications: postpartum haemorrhage, pre-eclampsia/eclampsia and newborn asphyxia.38 The checklist was piloted and modified to the Gombe context. All women attending the facility in active labour or experiencing postpartum haemorrhage were invited to participate at the time of admission. All potential participants were provided with a study information sheet and a consent form in English and Hausa. Taking care to include any support persons accompanying potential participants, the observer read the information sheet, explained the purpose of the study, the risks and benefits of participating and answered questions before seeking written consent from the woman and verbal consent from the healthcare worker attending. Women who were not able to write their name were asked to provide a thumb print on the consent form. Participation was voluntary and participants were free to withdraw at any time. Before each round of data collection, observers underwent 4 days of training on how to conduct unobtrusive observations, the safety and confidentiality protocols and how to ensure consistency of rating between observers. Throughout the observation period, clinical supervisors conducted spot checks of observers and data to provide ongoing quality assurance. Observers were required to prioritise the safety of the mother and newborn; protocols were established on the actions to take during any life-threatening events. This included immediately stopping the observation activity and calling for the clinical supervisor who could advise the attending healthcare worker. A formal report detailing any actions and decisions made was made available to the Executive Secretary of the GSPHCDA. Where data collection was stopped, observations were excluded from the study. For this analysis, the content of the clinical observation checklist was mapped against current recommendations for high quality mother and newborn care.13 15 39–42 Fifty measures were identified (box 1), grouped into four organising categories based on the stage of childbirth: (1) initial assessment; (2) first stage of labour; (3) second and third stage of labour and (4) immediate newborn and postpartum care. Data from the five data collection periods were combined into a single dataset. Observations were excluded from the dataset if the woman’s outcome was not recorded. For all women observed, we mapped the different pathways from admission to the facility (childbirth or postpartum haemorrhage event) to their outcome. For women who experienced an uncomplicated labour the outcome of their baby was also mapped. An uncomplicated labour was defined as a woman who was sent to the ward for recuperation or discharged home after birth and who did not experience an adverse event to her own health (referral, postpartum haemorrhage or pre-eclampsia/eclampsia) or death. For the analysis of the provision of essential evidence based care, our population of interest was women with an uncomplicated labour and detailed information on their care and that of their newborn are included here. Women who were admitted but experienced an adverse event or death were excluded from the analysis because of their individual medical needs. For measures related to newborn care the analysis was further restricted to newborns recorded as being alive and who did not require resuscitation care or were not referred to another facility. For each measure, per cent frequencies and 95% CIs were calculated, adjusted for clustering by primary healthcare facility and stratified by time point using the svyset and svy commands in STATA V.15.1 (StataCorp). Results are presented graphically by time point to highlight any variability and the average across all five time points is presented in the text. Patients and the public were not involved in the design, conduct, reporting or dissemination plans of our research. Observations were recorded in English and pre-testing completed in health facilities by staff.