Background During the early COVID-19 pandemic travel in Uganda was tightly restricted which affected demand for and access to care for pregnant women and small and sick newborns. In this study we describe changes to neonatal outcomes in one rural central Ugandan newborn unit before and during the early phase of the COVID-19 pandemic. Methods We report outcomes from admissions captured in an electronic dataset of a well-established newborn unit before (September 2019 to March 2020) and during the early COVID-19 period (April–September 2020) as well as two seasonally matched periods one year prior. We report excess mortality as the percent change in mortality over what was expected based on seasonal trends. Findings The study included 2,494 patients, 567 of whom were admitted during the early COVID-19 period. During the pandemic admissions decreased by 14%. Patients born outside the facility were older on admission than previously (median 1 day of age vs. admission on the day of birth). There was an increase in admissions with birth asphyxia (22% vs. 15% of patients). Mortality was higher during COVID-19 than previously [16% vs. 11%, p = 0.017]. Patients born outside the facility had a relative increase of 55% above seasonal expected mortality (21% vs. 14%, p = 0.028). During this period patients had decreased antenatal care, restricted transport and difficulty with expenses and support. The hospital had difficulty with maternity staffing and supplies. There was significant community and staff fear of COVID-19. Interpretation Increased newborn mortality during the early COVID-19 pandemic at this facility was likely attributed to disruptions affecting maternal and newborn demand for, access to and quality of perinatal healthcare. Lockdown conditions and restrictions to public transit were significant barriers to maternal and newborn wellbeing, and require further focus by national and regional health officials.
Retrospective cohort study of a clinical database. Included in this study are patients admitted to the Kiwoko Hospital (KH) neonatal unit during the period of study. There were no exclusion criteria. Kiwoko Hospital is a rural, private, not-for-profit secondary level care hospital that acts as a referral center for three districts (total population 1,000,000) in central Uganda. During fiscal year 2019–2020, the perinatal mortality rate in the hospital’s district (Nakaseke) was in the range of 29–42 per 1,000 live births and 19–28 in the two other districts the hospital serves (Nakasongola and Luwero) [19, 20]. The neonatal mortality rate for Uganda was 27 per 1,000 live births [21]. The KH neonatal care unit was established in 2001 and is a regional leader in care of small and sick newborns [22]. The unit admits more than 1,000 patients annually, generally neonates with a gestational age greater than 24 weeks and up to 44 weeks. About half of admitted patients are born at KH and the remainder are “outborn,” or admitted after birth at home or another facility. Outborn patients are treated in the same unit and receive the same care as those born at KH. There are 5–7 nurses on duty each shift, as well as one assigned medical officer and one pediatric physician who round on the patients and are on call each day. Staff generally live on site. Electricity is continuously available with the help of a standby generator. The unit provides thermoregulatory support primarily from radiant warmers and incubators, infection control and treatment, nasogastric and cup feeding, intravenous hydration, phototherapy, blood transfusion, basic laboratory services, oxygen therapy, and pulse oximetry. Improvised bCPAP has been the standard of care for respiratory failure in the unit since 2012 and is assembled using donated RAM nasal cannulas [23]. The unit does not provide surfactant, mechanical ventilation, total parenteral nutrition nor therapeutic hypothermia for birth asphyxia. The national referral hospital is located two hours travel by car and referrals from the Kiwoko NICU are transported free of charge. Admissions via ambulance to KH are assisted by government primary health care subsidies. General care in the neonatal unit is subsidized and families pay for any additional lab tests or imaging. Mothers are offered free accommodation and basic meals during their baby’s stay. The first COVID-19 tests at Kiwoko hospital were conducted on 5 May 2020 and the first case of COVID-19 was confirmed on 2 August. Following a total of 752 tests of staff, patients and community members, 14 total cases were confirmed before October and transferred to isolation centers. No neonates were denied care and access to and provision of care continued unchanged in the neonatal unit. Mothers were screened for symptoms on entrance to the unit. Experience at KH was determined through interviews with neonatal and maternity unit doctors as well as supervising midwives and nurses (summarized in S1 Table). Nationally during the 2020 period of the COVID-19 pandemic in Uganda, restrictions included: banning mass gatherings (18 March), closing schools (20 March), suspending public transportation and requiring police presence for private transport (25 March), and a nationwide lockdown with curfew (30 March). All non-essential services and activities were closed in this period [24]. Until 19 April, approval from a Resident District Commissioner was required to move during a medical emergency, and each district had only one Commissioner [25, 26]. Transport restrictions began to ease on 26 May: public transport resumed at half capacity (and increased cost) on 4 June, and the most common form of transport (via motorcycle) resumed on 27 July. Overnight curfews remained in place through the end of 2020. The neonatal unit has maintained an electronic database of all neonatal admissions since late 2012. On admission, nurses record patient and maternal information on a designated bedside form. After discharge, a data entry team extracts additional information from the medical file including treatments and final diagnosis as assigned by the physician at discharge, death, or transfer. Data are manually entered into Epi Info version 7 [27]. The data was exported to STATA v.15 for analysis [28]. We utilized this database to explore outcomes from all infants admitted to the neonatal unit during the pre-COVID-19 period (October 2019 through March 2020) and the first six months of the pandemic in Uganda, which we will refer to as the early COVID-19 period, or the COVID-19 period (April through September 2020). These periods were selected to capture the first lockdown period in Uganda (April/May) as well as the subsequent phased relaxation of COVID-19 restrictions (June/July) and then two additional months (August/September) to allow for six month seasonal comparisons with previous years. We also show seasonal variation and pre-existing trends in mortality by reporting outcomes from reference periods one year prior to study periods: Oct 2018 –March 2019 and April–Sep 2019. Death before discharge was the primary outcome- defined as a death as an infant that was admitted to the neonatal unit and died before discharge home. The analysis was further stratified by inborn/outborn status. Patients were classified as ‘inborn’ if they were born at KH while others were classified as ‘outborn’ (born at another facility, at home or on the way to the hospital). Birthweight-specific and diagnosis-specific mortalities were also computed. When birthweight was unknown, admission weight was used if the patient was admitted within 3 days of birth. Gestational age is not reliably available from this dataset and is therefore not reported for this study. Primary diagnoses were assigned by doctor at patient discharge, transfer, or death. Additional data on availability of health workforce in the neonatal unit was retrospectively reported by the nurse in-charge. The mortality rate pre and during the early COVID-19 period was computed as a proportion of infants who died to the total number of infants admitted to the unit in the same period. To determine the potential impact of the COVID-19 pandemic on neonatal mortality in the unit, we computed the relative change in mortality rate between two time periods of the same months for the 2 sequential years. Proportion tests were used to compare the rates between two time periods of the same months. Human Subjects Approval was obtained from Makerere University School of Public Health Institutional Review Board (protocol number 917) and approved by the Uganda National Council for Science and Technology (registration number SS813ES). The data were fully anonymized before accessed by the research team and the ethics committee waived the requirement for informed consent. The University of Washington institutional review board designated this as an exempt study.