Health facility readiness to care for high risk newborn babies for early childhood development in eastern Uganda

A cross sectional study was carried out in February 2020 in six hospitals and ten higher level health centers (HC IVs) in Busoga region in eastern Uganda. This was before the first Covid-19 case had been identified in the country. In Uganda Health Centres (HC) vary in their care provision across three levels – HC II, HC III, and HC IV. The higher level HC IVs offer comprehensive maternal and newborn care for HRBs. Busoga region has a population of about four million people, with a NMR estimated at 30/1000 live births in 2015[32]. The region is served by eleven hospitals [33].

The hospitals included in this study were 4 public and 2 ‘private not for profit’ health facilities (PNFP). All the HCIVs were government owned health facilities. One of the public hospitals was a regional referral hospital to which the rest of the general hospitals refer complicated cases. All the six hospitals were sites for the Preterm Birth Initiative (PTBi) study which was conducted between 2016 and 2019. The PTBi study aimed at reducing preterm morbidity and mortality through four intervention components: Data strengthening, use of the modified WHO Safe Childbirth Checklist, health provider training and mentorship, and use of collaborative quality improvement approach [34]. The PTBi study also provided some equipment and supplies at the start of implementation to address the critical gaps identified during the baseline study.

In the current study, the facilities were assessed to determine their readiness for care of HRBs: availability of inputs based on standards and clients’ experiences of the maternal and newborn care received. The health facilities were selected on the basis of being high volume facilities that are mandated to offer emergency obstetric and newborn care. We selected public and PNFP hospitals, and health centers IVs. This allowed us to assess the facility readiness based on the variation of the type of facilities in the region.

The WHO Services Availability and Readiness Assessment (SARA) tool, previously used by the PTBi study, was adapted based on the matrix developed by Moxon et al. 2018 [35], and used to assess the readiness of included facilities to care for HRBs. The adaptation of the SARA tool involved adding infrastructure for thermoregulation beyond KMC, and management of jaundice, use of a continuous positive airway pressure (CPAP) ventilation, and neurodevelopmental support. The inputs for neurodevelopmental support included: availability of cyclic lights, sound control measures, and guidelines for positive interactions with newborns and communication with carers. The infrastructure assessed included space for special care and resuscitation, stabilization and KMC. The staffing levels and availability of the skilled cadres were determined. An inventory was taken of equipment and commodities including nasal gastric tubes for feeding, availability of antibiotics for treatment of neonatal bacterial infection, intravenous fluids, oxygen, use of pulse oximetry, and use of a phototherapy machine for effective case management of pathological jaundice. The inventory data were collected by the first author, with the support of one research assistant, using a paper checklist written in English.

The experience of service provision for HRBs from the client perspective was assessed through client exit interviews in the hospitals. The HRBs were defined as: babies with APGAR score of less than 7 five minutes after birth, preterms with a gestation age of less than 37 weeks at birth, babies with a birth weight of less than 2500 g at birth, and infections characterized by either convulsions, failure or cessation of feeding, fast breathing of > 60 breaths per minute, severe chest in-drawing, temperature > 37.5 °C or < 35.5 °C, movement only when stimulated or no movement at all. Additionally, HRBs included those with pathological jaundice: a condition where a term newborn baby presents with jaundice within 24 h after birth, or the total serum bilirubin level is higher than 17 mg/dl in infants 25 to 48 h old, and the infant has signs and symptoms of serious illness. On average, 250 HRBs were admitted in the SNCU on a monthly basis from the six hospitals [36]. After excluding 10 runaway cases, 14 patients referred to other facilities and 25 deaths, 201 mothers with HRBs were included in the study. Proportionate to size sampling methods were used to distribute the sample size across the six hospitals.

Facility readiness to manage maternal and neonatal conditions that result in NDD was determined by the availability of infrastructure, medical commodities, skilled providers and client satisfaction with service provision. These were based on the list of evidence-based treatments (inclusive of items for diagnosis, treatment, and monitoring) for the common neonatal conditions developed by Morgan and team (essential routine newborn care; neonatal resuscitation; feeding and hypothermia; respiratory distress/apnea of prematurity; infection, convulsions and jaundice) [20], and on the WHO quality of care standards on developmental support for sick and small newborn babies [6]. Neonatal complications are known to be the main causes of neonatal deaths and also responsible for neonatal developmental disabilities among survivors. In addition, availability of resources for antenatal and emergency obstetric care (EmOC) were included given their critical role in neonatal survival and developmental potential during prenatal and intrapartum period.

Data from the SARA tool and exit interviews were entered in ACCESS with consistence checks and later exported to STATA version 15 for analysis. Descriptive statistics using frequencies, percentages, means, and standard deviations were used to summarize the data stratified by type of facility. Readiness scores were derived from proportions of facilities with availability of essential equipment, supplies and medicines for care of HRBs. The scores were then compared across the public hospitals, PNFP, and HCIV facilities.

The satisfaction of mothers of HRBs was assessed on several services received. We considered satisfaction to include those who mentioned ‘very satisfied’ and ‘satisfied’ for each variable. Chi square test statistic was computed to determine whether there was any difference in satisfaction between the Public and PNFP hospitals.

Ethical approval to conduct the study was obtained from the Higher Degrees and Research Ethical Committee (HDREC) of Makerere University School of Health Sciences (Ref. 2017- 011) and Uganda National Council of Science and Technology (#SS4600). Permission to access the health facilities was obtained from the district health authorities and the hospital administration prior to facility assessment and exit interviews. Written informed consent was obtained from all mothers of HRBs before data collection.

All the six hospitals were conducting deliveries 24 h a day, with capacity to do a caesarian section and with a neonatal special care unit including a dedicated unit for KMC. Although all the HCIVs were conducting deliveries 24 h a day only six were able to carry out Caesarian sections, two were unable due to inadequacies of theatre, while another two lacked anesthetists and relied on medical officers for giving anesthesia. All the health facilities had basic amenities such as electricity and running water. The average monthly deliveries were more at the public hospitals (450) than at the PNFP (79) and HCIVs (106) (Table 1). All the health facilities had medical officers attached to maternity and SNCU units, but only the regional referral hospital and the high volume district level hospital had obstetricians and pediatricians. The clinical officers at hospital level were only attached to the general outpatient department and did not work in the antenatal or maternity units, while at HCIV they could work in other inpatient departments rather than maternity. There was only one neonatal nurse based at the regional referral hospital. The rest of the health facilities had midwives that run the NSCU, and these were also the same midwives that conducted deliveries in the general hospitals and HCIVs. The average monthly number of HRBs admitted in the public hospitals was 61 compared to 15 in the PNFP, and 12 in HCIVs. There were no guidelines in any facility for positive interactions with the HRBs and for communication with the caregivers on how to ensure proper neurodevelopment of the vulnerable newborn babies.

Readiness for provision of ANC was high in HCIVs (83%) but generally low in hospitals with an average score of 58% (Table 2). The readiness was lower in public health facilities due to lack of access to ultrasonography at the facilities during pregnancy, lack of examination/flash light, and limited availability of diagnostic equipment particularly for hemoglobin and blood sugar level estimation. Two (50%) of the public health facilities did not have a single functioning blood pressure machine in the antenatal departments, only one out of the four public hospitals had stock of ferrous sulphate tablets, and one had run out of stock of folic acid tablets.

The readiness average scores for EmOC and blood transfusion services were 70% in hospitals and 74% in HCIVs (Table 2). However, there were limitations in availability of oxygen, found in only 2 public hospitals and pulse oximeters were available in only one public hospital. Similarly, there was lack of access to ultrasound scan, which was only available in one general high volume hospital and four of HCIVs. Although resuscitation tables for newborn babies existed in all the health facilities, they lacked overhead heaters. The resuscitation algorithm and referral protocols were available in most (70%) of the HCIVs but in only two (50%) of the public hospitals while both PNFP hospitals did not have referral protocols.

Readiness scores for essential routine newborn care were higher in HCIVs (85%) compared to hospitals at 78% (Table 3). However, the readiness for resuscitation was not so different in the hospitals (78%) and in HCIVs (77%). Only two hospitals (one PNFP and one public) had umbilical cord ties. Most health facilities were improvising by using the end parts of gloves as cord ties. The resuscitation area lacked a heat source in half of hospitals and HCIVs. Similarly, only half of the hospitals had the pulse oximeters with accompanying probes.

Overall, the average score for infection control, management + convulsions readiness was 60% in hospitals, public and PNFP facilities, and about 50% in HCIVs (Table 3). None of the health facilities assessed had chlorhexidine for cord care to prevent infection, and none of the facilities had C-reactive protein tests for assessing the possibility of infection. Similarly, all the facilities had no neonatal lancets for taking off blood for laboratory testing. They were using ordinary needles to prick the neonates. Gentamycin, an essential first line antibiotic for treatment of neonatal septicemia, was out of stock in all public hospitals. Only two high volume public hospitals had the multi-functional monitors to monitor the vital functions of sick newborn babies. None of the health facilities had an infusion pump, and only two public facilities had burettes for controlled provision of intravenous fluids, and dextrose (Table 3). None of the HCIVs had readily available 10% dextrose for treatment of hypoglycemia, and also lacked 50% dextrose to enable preparation locally.

Readiness of all health facilities was poorest regarding preterm and hypothermia care. The overall average score was 38% for hospitals and 18% for HCIVs (Table 3). The poor score was due to limited facilities for sound and light control, lack of surfactant, and KMC beds or chairs. Only the referral hospital had one KMC bed provided by the KMC study. The rest of the health facilities were using ordinary beds and plastic chairs for mothers to practice KMC. In addition, only about half (3 hospitals and 4 HCIVs) of the health facilities had phototherapy machines for care of jaundiced newborn babies.

The overall score for feeding of sick and small babies was about 50% in HCIVs and 67% in hospitals, and this was similar between public (67%) and PNFP (65%) hospitals (Table 3). The lancets, glucose test strips, breast pumps and referral guidelines were lacking in almost all health facilities.

A total of 201 mothers (166 from public and 35 from PNFP health facilities) were interviewed during exit interviews. There were significantly younger mothers (mean age 21.4 years SD ± 4.4) whose HRBs were admitted to the PNFP hospitals, compared to those in public facilities, p = 0.008) (Table 4). However, there was no difference in the period babies spent in the SNCU, with an average duration of about 5.3 days SD ± 4.3 in public hospitals compared to 4.9 days SD ± 2.3 in PNFP hospitals, p = 0.552. Similarly, there was no difference in the diagnosis captured on the discharge forms. Most babies had a diagnosis of asphyxia (37% in public facilities and 43% in PNFP hospitals) followed by preterm/LBW (28% and 26% respectively, p = 0.971).

Overall, client satisfaction with the services provided, and satisfaction with waiting time, health facility cleanness, and providers’ attitudes were over 90% in PNFP hospitals, significantly higher compared to public hospitals (p < 0.05) (Table 5). However, satisfaction with privacy was not significantly different in the two settings (p = 0.303).

The results further revealed that counselling on the prognosis of baby’s condition, and on newborn danger signs was done for all mothers in PNFP hospitals compared to 74% and 65.8% respectively in public facilities. However, only 29% of mothers in PNFP facilities were able to mention at least two newborn danger signs compared to 50% of those in public facilities. The proportion of mothers counseled on family planning before discharge (60%) was similar in both settings.

All babies were reportedly examined within one hour after birth in PNFP hospitals compared to only 71% in public health facilities. In fact, almost all babies were examined at least twice per day in both settings. However, there were missed opportunities for immunization of the babies before discharge in both public (19%) and PNFP (26%) settings.

There are some limitations to the study: We were unable to do knowledge and skills assessment of health workers, yet availability of equipment and supplies may not translate into quality service provision without skilled human resource. We did not also assess for availability of vaccines given within the first week after birth which is a main factor for reducing early infections. In addition, few hospitals were included in the study limiting the generalizability of the study findings. Furthermore, exit interviews have an inherent weakness of social desirability bias, and mothers may have had recall limitations after a stressful child birth experience. However, the study findings are still important in informing policy and programming for improved service delivery aimed at transforming care for improved neurodevelopment of the most vulnerable newborn babies.