Menu
Menu
Menu
Menu
Background: A majority of women in the Democratic Republic of the Congo (DRC) give birth in a health facility, but maternal and newborn mortality remains high. In rural areas, the quality of facility-based delivery care is often low. This study examines clinical quality of intrapartum care in two provinces of the DRC. Methods: We observed process and input elements of delivery care provision at 29 facilities in Kwilu and Kwango provinces. Distinguishing non-performance attributable to provider behavior vs. input constraints, we compared both providers’ adherence to clinical standards (“competent care”) and non-adherence to processes for which required inputs were available (“deficient care”). Results: Observing a total of 69 deliveries, care was most competent for partograph use (75% cases) and hemorrhage prevention (73%), but least for postpartum monitoring (4%). Competent care was significantly associated with higher case volumes (p = ·03), skilled birth attendance (p = ·05), and nulliparous women (p = ·02). Care was most deficient for infection prevention (62%) and timely care (49%) and associated with cases observed at hospitals and lower delivery volume. Conclusions: Low quality was commonly not a result of missing equipment or supplies but related to providers’ non-adherence to standard protocols. Low case volumes and the absence of skilled attendants seemed to be main factors for sub-standard quality care. Birth assistance during labor stage 2 was the only intrapartum stage heavily affected by the unavailability of essential equipment. Future interventions should strengthen links between birth attendants’ practice to clinical protocols.
In the DRC, health facilities are organized in health zones (Zones de Santé), each zone comprising at least one hospital (Hôpital Général de Référence) and several health centers (Centres de Santé). Together, these health facilities are responsible for implementing the country’s primary health care strategy, including standards related to maternal, newborn and child health. This study reports observed clinical practices of birth attendants in two rural provinces, Kwilu and Kwango, with largely comparable maternal care statistics. In 2018, maternal mortality in Kwilu and Kwango was with 312 and 442 deaths per 100,000 live births, respectively, below national average [10]. Further, a vast majority of health facilities in both provinces offered essential obstetric care (96% in Kwilu, 95% in Kwango) [9], while hardly any of these facilities had the capacity to provide emergency obstetric care [11]. The two provinces however differed in the proportion of births attended by a skilled birth attendant, with 93% of births in Kwilu much higher compared to Kwango with 61% [8]. As outlined above, the purpose of this observational study was to assess the care provided to women giving birth in rural health facilities as well as whether gaps in clinical protocol adherence are driven by the availability of medical supplies. To do so, we combined two existing data streams collected by the Projet d’Appui des Services de Santé (PDSS), a World Bank Funded project that aims at improving maternal and child healthcare provision [12]. Although the project supports health zones in 11 provinces, five health zones in two provinces were purposively selected for this study. Three of the selected health zones (Gungu, Mosango, Pay Kongila) are in Kwilu and in two health zones (Boko, Kimbao) are in Kwango. Each of these five health zones comprises of one referral hospital and between 16 and 26 health centers. All referral hospitals were included in this study, as well health centers randomly selected for a wider PDSS evaluation. Ten health centers were selected in Gungu, six in Kimbao, and five in each of the other zones. The study sample therefore included a total of 36 facilities. Direct clinical observations of intrapartum care delivery were conducted at these 36 facilities. For this study, all non-complicated childbirth cases (i.e. normal vaginal births) presenting to each of these facilities during a five-day period were observed. Based on PDSS program data on childbirth frequencies in these facilities from 2017 (the year preceding the case observations) it was estimated that 5 days of observation at each facility would on average yield 2.6 deliveries in health centers located in both Kwango provinces. Given the logistic challenges for having research teams safely move between study facilities, a five-day observation stay at each facility was therefore the most feasible field strategy. In total, 69 case observations could be observed at 29 facilities (i.e. at seven health centers no births could be observed during the five-day stay given the low case volume). Content of care data was directly observed between August and September 2018 and contained information on the following intrapartum aspects: labor room set-up, patient history and physical assessment, labor monitoring, birth assistance, immediate newborn care, placenta management, and immediate postpartum monitoring. Information on birth attendants’ professional qualifications and patients’ parity status was also recorded. Case observation started once the laboring woman was admitted to the labor ward and ended 2 hours after birth. Observed information was recorded in a checklist. Checklist content reflected clinical standards taken from national treatment protocols for essential intrapartum and immediate newborn care [13]. Prior to conducting an observation, informed consent was obtained from both the involved birth attendants and the patient. Case observations were conducted by trained data collectors with a background in midwifery. Prior to data collection, data collectors participated in a 4-day training workshop, followed by a 2-day pilot to ensure observed activities were consistently observed and recorded across data collectors. During actual field data collection activities, data collectors were assigned to trained peer supervisors who ensured that recorded data met the completeness and quality standards. Further, all 36 PDSS facilities enrolled under the financing scheme report quarterly inventory data, including information on their’ capacity to provide childbirth services. For intrapartum care, this data included information on key input elements, such as the availability of essential equipment and supplies. To ensure timely overlap with the observed case data, we used quarterly facility data reported for the quarter July–September 2018. With a focus on technical quality only, available data points measured providers’ adherence to clinical intrapartum standards [14]. We assigned observed process measures to one of five clinical intrapartum stages: “labor monitoring”, “assistance in childbirth”, “assistance in delivery of placenta”, “immediate postpartum monitoring”, and “cross-cutting intrapartum care”. For each intrapartum stage, we then identified those indicators directly aligned with the concepts of “competent care” (i.e. evidence-based, effective care) and “capable systems “(i.e. environments that ensure patient safety, prevention and detection of negative health, continuity of care, and timely action) suggested in the Lancet Global Health Commission’s health systems assessment framework [15]. Our process selecting competent care indicators followed the rationale to focus on intrapartum care processes that could be directly linked to those clinical activities that prevent or control direct causes of maternal or newborn death (i.e. maternal and newborn sepsis, asphyxia, fetal distress, newborn distress, postpartum hemorrhage) and included a detailed review key documents related to intrapartum care provision at the community level [16–18], as well as by our previous work on routine intrapartum care assessment [19]. Resulting indicator definitions for each competent care domain are shown in Table 1 and include: “patient safety” (defined by infection prevention measures throughout case management), “early detection” (defined by the use of a partograph to identify signs of adverse labor progression during stage 1 labor), “timely access” (defined as accessibility to key birth equipment during stage 2 labor), “prevention” (defined as measures reducing the risk of postpartum hemorrhage during stage 3 labor), and “continuity of care” (defined as measures monitoring maternal and newborn well-being in the immediate postpartum). We identified two process indicators for each competent care domain, except for “early detection” with only one indicator. Outline of domains, indictors, definitions, and rationale applied to assessment of competent intrapartum care processes For clinical processes dependent on essential input elements, we further assessed “deficient care” defined as absent provider performance although the essential input element (e.g. equipment or supplies as outlined in Table Table1)1) would have been available at the facility. For instance, and as outlined in Table 5, infection prevention was considered deficient if any of its two processes (handwashing, glove use) was not observed in a case although the facility had the respective inputs (water source, sterile gloves) available. Selection of essential input elements followed a review of international recommendations and standards [16, 20]. Observed deficient care by domains and related indicator combinations Total n = 69 cases We assumed childbirth services provided at the hospital level to be of relative higher quality of care compared to health centers [21]. The variable “facility type” therefore classifies facilities as hospitals or health centers to reflect the stronger human and technical capacities (e.g. higher staff numbers including physicians or gynecologists, and more diagnostic and therapeutic options with respect to the management of pregnancy complications) accessible to hospitals. We further assumed that a higher volume of childbirth cases would also result in higher quality of intrapartum care as provided care at high volume facilities would be linked to more frequent exposure to different cases [22]. The variable “delivery volume” represents a facility’s annual volume of assisted non-complicated delivery cases (data taken from PDSS routine data for the year 2017) classifying facilities as “higher” or “lower” volume based on a cut-off at 183 noncomplicated births per year (i.e. equivalent to one delivery every other day). We also assumed that the number of skilled birth attendants (SBAs) at a facility will positively influence observed quality. The variable “presence of at least one SBA” indicates whether a SBA was present in the maternity during case observation, and if so, whether or not this SBA was directly or indirectly (i.e. closely supervising a non-skilled provider) in the care process. Our definition of SBAs is based on national provider qualification categories used in the DRC (i.e. midwives and nurse-midwives with at least 2 years undergraduate training, as well as gynecologists), assuming more competent care provided by skilled providers [13, 23]. Lastly, given the higher risk for negative birth outcomes for nulliparous women, we expected intrapartum care provided to these patients to more closely follow standard protocols [24]. The variable “client’s parity” groups cases into those involving a woman without previous childbirths (nulliparous) and those with one or more previous births (multiparous). Our analysis primarily focused on the 29 facilities for which case data could be observed to understand the overall frequency of both competent and deficient care as well as their association with observed facility and case characteristics. For each case, measures of competent care were computed based on the observed information. To compute deficient care measures, observed case data was matched with the respective facility’s capacity assessment. We then computed the frequencies of both identified competent and identified deficient care processes by indicator and care domains. To explore associations of frequency distributions with respect to the facility and case level characteristics described above we used. Fisher’s exact tests All analysis were computed using Stata.
N/A
