Use of cranial ultrasonography to improve prompt and early diagnosis of meningitis at the Neonatal Centre of Excellence, University Teaching Hospitals, Lusaka, Zambia
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
Despite cerebrospinal fluid (CSF) analysis being the gold standard for definitively diagnosing meningitis, its practical application has presented considerable difficulties, especially in environments with limited resources. Cranial ultrasound (CUS), while not a replacement for CSF analysis, provides rapid imaging to identify meningeal irregularities. Nevertheless, achieving a timely diagnosis of meningitis, particularly in its nascent stages, remains problematic, and healthcare professionals exhibit a notably low awareness of CUS’s utility in diagnosing this condition.
Methods
This was a prospective cohort that recruited 273 term mother-neonate pairs. The neonates were initially divided into those with sepsis-meningitis (exposure group) and those with sepsis (non-exposure group) based on clinical assessment done by the admitting team. The research team further divided the participants into 4 diagnostic subgroups based on CUS findings: positive CUS for meningitis, positive clinical diagnosis for meningitis, positive CUS & clinical diagnosis for meningitis and those with negative CUS & clinical diagnosis for meningitis /sepsis only. Data on socio-demographics, clinical characteristics, blood works and CSF analysis reports, neurologic deficits and mortality outcomes were recorded for each neonate. Descriptive and inferential statistics were performed.
Results
The 4 diagnostic subgroups based on CUS findings were: positive CUS for meningitis (24.5%), positive clinical diagnosis for meningitis (4.4%), positive CUS & clinical diagnosis for meningitis (6.6%) and those with negative CUS & clinical diagnosis for meningitis /sepsis only (64.5%). Overall, duration of hospitalization was 11 days (range 2-45 days; interquartile range [IQR] 6,16) and the median chronological age of the neonates was 13 days (IQR 7, 21). Meningitis was suspected in 11% of neonates admitted with clinical sepsis. Uptake for lumbar puncture (LP) or ventricular tap (VT) was low at 4.4% (n = 12), underlining barriers to CSF-based diagnosis in this setting. Late-onset sepsis was associated with only 31.9% of NSM. The chronological age of the neonates at admission (p=0.002) and their duration of hospitalization (p=0.042) were significantly different across the 4 categories of neonates. Prominent sulci, hyperechoic brain lesions, ventriculitis, and lateral ventriculomegaly were the most common abnormal findings on CUS. Overall, neonates with meningitis presented later in age and stayed longer on the ward than those with sepsis: clinical meningitis diagnosis was likely to be made in older neonates and was associated with a shorter duration of hospitalization than CUS-diagnosed meningitis. Overall mortality rate was 2.2%. Mortality (16.7%) was highest amongst those with clinical & CUS diagnosed-NSM. Prevalence of Near-Miss cases of NSM was at least 27.6%. A neonate with clinical diagnosis of NSM was 3.94 times (95% CI 1.80, 8.62; p<0.001) more likely to have abnormal CUS at the time of admission. Percent agreement between the clinical and CUS diagnosis for NSM was 71.1%. Nine (75%) out of the 12 CSF reports was positive for meningitis; and the majority 55.6% (5) of the positive CSF report belonged to the category of neonates with sepsis only. Compared to CSF analysis, sensitivity of clinical diagnosis of NSM was 22.2%; specificity was 33.3%; positive predictive value (PPV) was 50%; negative predictive value (NPV) was 12.5% and likelihood ratio (LR) was 0.33. Whereas, sensitivity of CUS was 44.4%; specificity of CUS was 66.7%; PPV was 80%; NPV was 28.6% and LR was 1.33.
Conclusion
The CUS was only moderately effective at diagnosing meningitis whereas the presence of positive CSF among neonates with sepsis only within this study, further reaffirms the irreplaceability of CSF analysis in the diagnosis of NSM. Nevertheless, the integration of clinical assessment and CUS findings for diagnosing NSM emerged as possessing greater clinical significance in contexts characterized by limited resources. The systematic adoption of CUS for neonates exhibiting features suggestive of sepsis or meningitis (while exploring measures to improve on uptake of CSF analysis) could enhance the promptness of diagnosis, inform the selection of suitable therapeutic interventions, and potentially mitigate mortality rates and the occurrence of long-term neurological impairments, especially within environments facing resource limitations.
