“Assessing Today for a Better Tomorrow”: An observational cohort study about quality of care, mortality and morbidity among newborn infants admitted to neonatal intensive care in Guinea

Background Neonatal mortality in Guinea accounts for about 30% of all fatalities in children younger than five years. Countrywide, specialized neonatal intensive care is provided in one single clinic with markedly limited resources. To implement targeted measures, prospective data on patient characteristics and factors of neonatal death are needed. Objective To determine the rates of morbidity and mortality, to describe clinical characteristics of admitted newborns requiring intensive care, to assess the quality of disease management, and to identify factors contributing to neonatal mortality. Methods Prospective observational cohort study of newborns admitted to the hospital between mid-February and mid-March 2019 after birth in other institutions. Data were collected on maternal/prenatal history, delivery, and in-hospital care via convenience sampling. Associations of patient characteristics with in-hospital death were assessed using cause-specific Cox proportional-hazards models. Results Half of the 168 admitted newborns underwent postnatal cardiopulmonary resuscitation. Reasons for admission included respiratory distress (49.4%), poor postnatal adaptation (45.8%), prematurity (46.2%), and infections (37.1%). 101 newborns (61.2%) arrived in serious/critical general condition; 90 children (53.9%) showed clinical signs of neurological damage. Quality of care was poor: Only 59.4% of the 64 newborns admitted with hypothermia were externally heated; likewise, 57.1% of 45 jaundiced infants did not receive phototherapy. Death occurred in 56 children (33.3%) due to birth asphyxia (42.9%), prematurity (33.9%), and sepsis (12.5%). Newborns in serious/critical general condition at admission had about a fivefold higher hazard to die than those admitted in good condition (HR 5.21 95%-CI 2.42–11.25, p = <0.0001). Hypothermia at admission was also associated with a higher hazard of death (HR 2.00, 95%-CI 1.10–3.65, p = 0.023). Conclusion Neonatal mortality was strikingly high. Birth asphyxia, prematurity, and infection accounted for 89.3% of death, aggravated by poor quality of in-hospital care. Children with serious general condition at admission had poor chances of survival. The whole concept of perinatal care in Guinea requires reconsideration. A prospective observational cohort study was conducted at the Institute of Nutrition and Child Health (Institut de Nutrition et de Santé de l’Enfant, INSE) of all admitted newborns and their follow-up until discharge or death during a 30-day period from February 15th to March 16th 2019. Ethical approval was granted from the Comité National d’Ethique pour la Recherche en Santé (CNERS) in Guinea (Ref. 035/CNERS/19). The written informed consent was signed by a parent or legal guardian prior to collecting and analyzing data. INSE is adjacent to the large public university hospital Donka and functions as the only specialized NICU in Guinea with a theoretical catchment area of the whole country with a population of 13.4 million inhabitants [16]. However, very few newborns outside the wider Conakry area, with a population of about 1.9 Mio inhabitants, are presented at INSE [17]. The unique shape of the capital city, a slim peninsula with all important facilities located at the coastal end and only two main roads connecting all suburbs with downtown Conakry, causes major traffic congestion and access difficulties to the medical care center. The Donka University Hospital has not offered obstetric services since 2015 due to renovation of the maternity ward. Therefore, all admitted neonates were born in other facilities or at home, and, consequently, had to be transported postnatally to INSE. On average, 4500 children are hospitalized annually at INSE, more than half of them in the neonatal department. Reasons for neonatal admissions include prematurity and low birth weight (LBW, <2500 g), poor postnatal adaptation, out-of-hospital reanimation, respiratory distress, bleeding, fever or other signs of infection, poor feeding, and congenital malformations. The architectural structure of the neonatal department offers only very limited space with overcrowded rooms and patients often sharing beds. The critical care unit accommodates 15–20 newborns. Further 25–30 newborn infants, once stabilized, are cared for in two more rooms. In each unit, one pediatrician is in charge supervising a general practitioner and an intern doctor from 9 am to 3 pm. After 3 pm, one general practitioner and two medical students assume the night shift for the whole neonatal department. Two specialized neonatologists work at the facility as consultants during daytime. The Kangaroo Mother Care (KMC) unit houses four premature newborns and their mothers, under the surveillance of a nurse teaching skin-to-skin-nursing. The other mothers stay in self-supply dormitories on the premises. The hospital is ill-equipped and underfunded. There had been no air conditioning at the time of the study and power outages were frequent. Solar panels power emergency lighting, oxygen extractors, and the one available phototherapy lamp. Equipment is of inferior quality and deteriorates quickly due to heat, dust, humidity, and power instability. Infants sleep in plastic cradles or makeshift wooden boxes. All medication and medical supplies have to be purchased by the parents after medical prescription, even in cases of emergency. Breastfeeding is encouraged according to the newborns’ general condition and gestational age; otherwise, the newborns are fed with expressed breastmilk or formula milk via nasogastric tube, cup, or spoon. Premature or hypothermic newborns are placed in plastic bags and externally heated under a shared radiant warmer, or with the help of warm water bottles. Managing admitted newborns, medical staff followed the guidelines of Advanced Neonatal Care by Médecins Sans Frontières (MSF) who had trained local personnel over several occasions until 2016. These guidelines are based on the World Health Organization (WHO) Essential Newborn Care guideline [18]. Vital parameters (axillary temperature, heart rate, respiratory rate, and peripheral oxygen saturation) were obtained using a digital thermometer and pulse oximeter, respectively. The body weight and findings of a full-body exam were recorded at admission. Respiratory distress was categorized as mild, moderate, or severe using the Silverman Score [19]. Interventions included airway suction for clearance, positive pressure ventilation for non-breathing infants with an Ambu® breathing bag, and oxygen therapy using oxygen concentrators with their output shared via Y-pieces of nasal cannula. Continuous positive airway pressure (CPAP), intubation, and mechanical ventilators were not available to assist newborns with respiratory distress. Nota bene, all pharmacological treatments were only administered if they were available, and had been purchased by the parents in the in-hospital or external pharmacies. Patients were routinely administered 1 mg of intramuscular Vitamin K if not yet injected in the obstetric clinic. Caffeine citrate (loading dose 10 mg/kg/day, maintenance dose 5 mg/kg/day) was given to premature newborns with gestational age <34 weeks or once apnea was noted. Cardiac arrest and bradycardia defined as heart rate below 60 beats/minute were treated with bag and mask ventilation and manual chest compressions. Anemia was defined as severe when the hemoglobin level was below 14 mg/dl. In case of severe anemia or active bleeding, blood transfusions with 20–25 ml/kg of concentrate red cells over 180–240 minutes were given. Neonatal jaundice was evaluated according to the Modified Kramer’s Scale as well as serum bilirubin levels, if available [20]. Phototherapy was used to treat jaundice with four hour-long sessions at a time; however, light therapy was often withheld due to hyperthermia. Intravenous antibiotic treatment was administered to newborns, for at least 48 hours, if either sepsis was being suspected (based on clinical signs and/or laboratory results), or if infection could not be excluded (unavailable prenatal history or laboratory results). No hemocultures were available. Ampicillin 50 mg/kg/day and gentamycin 5 mg/kg/day were administered to all patients without clinical signs of sepsis but suspected infection; ampicillin 100 mg/kg/day, gentamycin 5 mg/kg/day, and cefotaxime 100 mg/kg/day were given to children with clinical signs of sepsis or proven meningitis. Blood glucose levels were measured using a commercial glucometer. Hypoglycemia was defined as blood glucose below 2.5 mmol/l and treated with an intravenous bolus of 2 ml/kg of 5–10% dextrose followed by feeding or maintenance with 5–10% dextrose. In Guinea, 10% dextrose for intravenous application is not commercially available and had to be reconstituted from 30% dextrose and 5% dextrose. Children with birth asphyxia (BA) were categorized clinically as either mild, moderate, or severe hypoxic-ischemic encephalopathy (HIE) using the Sarnat staging [21]. Seizures were treated with 2mg of rectal diazepam, followed by oral maintenance with 5–10 mg/kg/d of phenobarbital. Radiological investigations, such as chest or abdominal X-rays, echocardiography, or ultrasound were not available on site and therefore rarely obtained due to lacking financial parental resources, restricted opening hours of the external imaging department, and difficulties related to transporting unstable newborns. Full blood counts and levels of C-reactive protein were available 24 hours per day; cerebrospinal fluid cell count with gram stain, blood films, bilirubin, and electrolyte measurements were only available during the daytime and were often not reliable due to equipment malfunction. Blood or liquor cultures were not available. Inclusion criteria were infants admitted to INSE during their newborn period (≤28 days of life) between February 15th and March 16th 2019, who were alive at the time of admission and whose parents had signed the informed consent form. Newborns who died before or at the time of their arrival at INSE and/or whose parents denied processing of their child’s data were excluded from the study. All infants were recruited in this study via convenience sampling. All information related to the hospitalized newborns was documented by physicians and nurses in the standard hospital paper charts. If available, data were collected on maternal history, prenatal and obstetric care, delivery, neonatal conditions, in-hospital management, and laboratory investigations. Provisional diagnoses and presumed causes of death were given by the treating doctor and discussed in the daily morning interdisciplinary staff meeting. Specific definitions used in this study are listed in Table 1. All data recorded for this study were entered in an excel sheet for further analysis. All analyses were performed using the R system for statistics and graphics version 4.0.4 [24]. Neonatal outcome was categorized into four different, mutually-exclusive states: patients who were cured during hospital stay (cure), deceased during hospital stay (death), were transferred to a different clinic (transfer), or were discharged early for whatever reason (early discharge). For simplicity, the latter two states were combined to a single one (other discharge). We descriptively analyzed how pre-, peri-, and postnatal characteristics may be related to neonatal mortality, stratifying characteristics by neonatal outcome. Frequency and percentage were tabulated for categorical variables, mean and standard deviation for continuous variables. In addition, we fitted a cause-specific Cox proportional-hazards model to assess the associations of patient characteristics at hospital admission with the outcome death (time to in-hospital death), censoring patients with other outcomes (cure or other discharge) at hospital discharge. Explanatory variables in the model were admission weight (in 100 g units), hypothermia (<36.5°C) at admission, neonatal CPR and a dichotomization of general condition and neurological condition into good/ fair vs. serious/critical. Neonatal time to death, cure or other discharge after hospital admission was visualized by cumulative incidence curves. A map of Conakry was produced using the R packages rnaturalearth and rnaturalearthhires, which provided the high-resolution polygon of Guinea, and ggplot2 for plotting the country together with the places of birth of the newborns [25–27].