Background: A strong evidence base exists regarding routine and emergency services that can effectively prevent or reduce maternal and new-born mortality. However, even when skilled providers care for women in labour, many of the recommended services are not provided, despite being available. Barriers to the provision of appropriate childbirth services may include lack of availability of supplies, limited health worker knowledge and confidence, or inadequate time. The WHO Safe Childbirth Checklist (SCC) includes reminders for evidenced-based practices at specific points in the childbirth process. Zambia is currently considering nation-wide adoption of the SCC, but there is a need for context-specific evidence. Beginning in September 2017, a program is being implemented in Nchelenge District to pilot use of the SCC, along with coaching that focuses on strengthening the systems that allow the essential practices in childbirth to be performed. Methods: This study will use a pre-post study design to measure health worker adherence to the essential practices for delivery care outlined in the SCC. Data will be collected through observations of health workers as they care for mothers during childbirth at four facilities. Data collection will take place before the start of the intervention, at 3 months, and at 6 months post-intervention. The primary outcome interest is the change in the average proportion of essential childbirth practices completed. A health worker questionnaire will be administered at the time that the SCC is introduced and 6 months later to gather their perspectives on incorporating the SCC into clinical practice in Zambia. Discussion: Findings are expected to inform plans for introducing the SCC in Zambia. This evaluation will aim to understand uptake and impact of the SCC and associated coaching in the context of a basic level of mentorship that the government could feasibly provide at a national scale. Trial registration: Clinical Trials.gov (NCT03263182) Registered August 28, 2017.
This is a pre-post study design that will include observing consecutive deliveries carried out in four selected health facilities before and after the introduction of the SCC. The study protocol was registered in the Clinical Trials.gov, a resource provided by the U.S. National Library of Medicine ({“type”:”clinical-trial”,”attrs”:{“text”:”NCT03263182″,”term_id”:”NCT03263182″}}NCT03263182). The WHO SCC will be introduced in eight purposively selected health facilities in Nchelenge District, Luapula Province. Nchelenge District was chosen by the MOH to be the location for this pilot based on need. The eight facilities were selected based on criteria including: high demand for services, perceived need for support for quality improvement, and presence of a SBA to mentor. However, data collection for this evaluation will take place in only four facilities with high volumes for the observations to occur. The four sites include three health centers and one general hospital. For facility observations, all consecutive deliveries will be observed. If a pregnant woman or health worker declines to participate in the consent process, no observations will be recorded. The SCC will be piloted as part of a broader program to improve delivery of routine care for all women through a systems improvement approach focusing on health worker knowledge and skills; availability of supplies and equipment; ensuring good documentation and data use; and creating an enabling environment at health facilities in Zambia. This broader program is taking place in 14 districts in Zambia and aims to increase the availability, accessibility, and quality of maternal and newborn health services. While this program will be operating in 14 districts (and is being evaluated separately), the introduction of the SCC as a part of this package will be piloted in one district. The SCC intervention will involve several key steps and components: The primary outcome variable of interest is the change in the average proportion of essential childbirth practices completed per birth out of those observed from the list of selected tasks. We have developed a list of selected tasks that represent a sub-set of items in the SCC which are tasks that can be visually observed and that are non-conditional (in other words they should be performed on all mother-infant pairs regardless of health status). The list of selected tasks includes the following 21 items: We calculated the expected power that this observation period will give us to detect changes in practices based on data about delivery volumes at the four sites, using the following formula [22]: Due to the small number of facilities the equation utilizes the difference in the main proportion assuming a t-test will be utilized [23]. Where: c = number of facilities needed, Za/2 and ZB = type I (1.96) and type II error (0.84) respectively, s0 and s1 are the standard error at baseline (17%), ρ = interclass correlation (0.25). We estimate that at baseline 34% (m0) of the 15 SCC items will be observed per birth, based on previous work in India indicating that an average of 9.8 out of 29 items were observed per birth at baseline [12]. According to the Health Management Information System data from January to October 2016, there was an average of 390 deliveries per month across the four study facilities. This means that if we observed all deliveries for approximately 48 h in each site, we could expect to observe an average of 6.5 (n) deliveries per site. With such a sample size, the calculations indicate that we could detect an absolute increase of 24% in the proportion of adherence to SCC essential best practices (m1 = 58%). For the secondary outcomes related to health workers characteristics and feedback, all health workers in the four study sites will be invited to participate as a convenient sample. Data will be collected through observation of health worker practices at baseline, midline and end line. Observations will be conducted by a team of four observers with clinical expertise so that they will be knowledgeable about clinical issues and capable of providing accurate assessments of whether SCC practices are completed or not. Observations will be conducted for a minimum of 3 days in all facilities, or as long as required to observe at least seven births. It may not be possible for the observer to document every delivery process at all times, particularly if multiple deliveries are taking place at one time, so observations will relate to activities at four specific periods (pause points). As a part of the training process, the observing data collectors will observe births in parallel with a trainer or supervisor to ensure accuracy and consistency in recording the necessary information. Data collectors will be trained to recognize each of the practices on the checklist according to the definitions adapted from previous work [8]. Observation data will be recorded on standardized data sheets over the course of the delivery process. Data will be collected on paper forms and later entered electronically on portable touch-screen tablets that each observer will be equipped with. A self-administered questionnaire [14] will be administered to health workers at the end of the facility-based SCC training and after all end line observations are completed to understand the overall success and acceptability of the tool and supportive activities. The questionnaires will collect basic information about the participant, their feedback on the process of using the SCC, and the appropriateness of the SCC for Zambia. For the primary outcome, data from each delivery will be considered as a percentage of the 21 SCC items that were practiced, or in other words each pause point observation will be assigned a score from 0 to 100%. The differences over time will be examined using linear regression and accounting for clustering at the facility level as well as transformation to account for the bounded data. We will also consider clustering at the level of the primary provider responsible for overseeing the delivery. Sub-analysis will be conducted to account for the impact of availability of supplies on conduct of tasks. Within each facility in the sample, we will also compare the average score for all deliveries at baseline, midline and end line using a paired, weighted t-test to account for differences in the standard errors. Finally, to better explore meaningful change due to the program, we will examine if there is an increase in the proportion of births where the majority (75%) of tasks are completed; specifically, we will examine if the difference overtime in the average proportion of births where 75% of tasks were completed increases due to the intervention. To this end we will run individual-level logistic regression, accounting for clustering at the facility level. For the health worker questionnaires, we will report descriptive statistics only, with baseline and end line comparisons. This evaluation is a partnership between the MOH of Zambia and the Clinton Health Access Initiative (CHAI) under the Demand-Driven Evaluations for Decisions (3DE) program. Since 2012, 3DE has been working with the MOH of Zambia to generate evidence in response to some of its most critical needs for evidence-based decision-making [23–27]. By using policy-relevant and rapid designs, 3DE evaluations have been able to improve the implementation of programs and policies. The idea for this evaluation emerged out of requests from the MOH for evidence on practical ways to reduce maternal and newborn mortality. CHAI and the MOH were both aware of the SCC and reached out to Ariadne Labs for support in considering how the SCC could be brought to Zambia. The SCC introduction and coaching activities are being carried out by the Human Resources for Health program at CHAI, which aims to increase access to skilled attendance at birth, while the evaluation described in this paper is being carried out by 3DE.
N/A
DIMA AI Care
