Abstract
Background Accurate and timely diagnosis and management of common neonatal conditions is crucial for reducing preventable neonatal deaths. In low- and middle-income countries, there is sparse information on how neonatal diagnoses are made including use of clinical guidelines and access to appropriate laboratory and radiology services. An improved understanding of the diagnostic pathways for neonates in low-resource settings will help to inform future interventions and the development of best practice guidelines.
Objectives To identify challenges in neonatal diagnosis in resource-limited settings by describing diagnostic pathways for common neonatal conditions.
Methods Between August 2018 and April 2019, standard forms developed by the Neonatal Nutrition Network (NeoNuNet)* collaborators were completed by clinicians for all infants admitted to seven neonatal units in Nigeria and Kenya and diagnosed with birth asphyxia, respiratory disorders, abdominal conditions and suspected sepsis. Data was captured in REDCap** and the frequency of criteria used for each diagnosis in each NNU was analysed using SPSS.
Results 2852 neonates were included. Mean gestational age was 36 weeks (SD 4.27) and birthweight 2.42 kg (SD 0.94). 473 (16.6%) newborns died during admission and mortality was highest amongst very-low birth weight (<1500 g) infants (46.1%) and very preterm (<32 weeks’ gestation) infants (41.3%). 1230 (43.1%) newborns were diagnosed with suspected sepsis, 874 (30.6%) with respiratory conditions, 587 (20.6%) with birth asphyxia and 71 (2.5%) with abdominal conditions. For all diagnoses, the most frequently used diagnostic criteria were clinical, whilst laboratory and radiological criteria were rarely used. In addition, there was marked variation between the NNUs, including within each country, in the use of many of the criteria for each diagnosis. Using suspected neonatal sepsis as an example, temperature instability was the most used clinical criteria (1036/1230 (84.2%) infants; varied from 72% to 93% in each NNU) whereas hypotension was rarely used (1.8% infants; 0% to 9%). For laboratory criteria, abnormal white cell count was the most commonly used (22.7% infants; 8% to 37%) but all were used infrequently: raised C-reactive (9.4% infants; 0% to 27%) and presence of a pathogen in blood (16.3% infants; 1% to 74%) and cerebrospinal fluid (1.9% infants; 0% to 13%).
Conclusions Clinicians adopted a syndromic approach when making diagnoses but with marked variation in use of clinical criteria between NNUs. Laboratory and radiology technologies were mostly either unavailable or unaffordable. As neonatal conditions often have non-specific and overlapping clinical features, this has implications for management such as overuse of broad-spectrum antibiotics for suspected sepsis. There is an urgent need to facilitate standardisation of diagnostic pathways based on World Health Organisation and national diagnostic guidelines to optimise clinical care. In addition, investment in affordable, sustainable diagnostics suitable for low-resource settings, including point-of-care tests, is a priority. Standardisation of diagnostic pathways would also facilitate comparing disease burdens and outcomes between NNUs as a basis for research to improve neonatal outcomes.
* https://www.project-redcap.org/
** https://www.lstmed.ac.uk/nnu
