Study design
Paediatric patients were included in this prospective observational study (2018–2021) according to the following criteria: All patients presented with clinical symptoms suggestive for raised ICP (headache, vomiting, irritability, visual disturbance, e.g. blurred/double vision). Some of them appeared for the first time, some were already known and pre-treated, and all required further investigation due to the duration or severity of symptoms. All patients underwent fundoscopy by independent ophthalmologists (inpatient/outpatient). Additionally, all patients underwent transorbital ultrasound of ONSD and ODE by an investigator blinded to fundoscopy. US findings were not included in decision making regarding further diagnostics and treatment. Only patients tolerating US investigations without sedation were included. The maximum time-interval between fundoscopy and transorbital US was 24 h, with no ICP-affecting therapy in between. Depending on kind, duration and severity of symptoms and fundoscopic findings, further imaging diagnostics were performed.
In 28 patients, an additional invasive ICP measurement (by intraparenchymal ICP probe) was performed, either because the diagnostic work-up required it, e.g. in PTC [
22], or the preceding diagnostics did not yield clear findings (e.g. arachnoid cyst or clinical suspicion of craniostenosis in absence of papilledema).
According to findings and treatment decisions, patients were stratified into 2 groups: treatment group (TG) and non-treatment group (NTG).
Inclusion criteria for TG included the following in addition to the existing symptoms as described above: (1) imaging evidence of disease/disease-recurrence associated with increased ICP (hydrocephalus, space-occupying arachnoid cyst, brain tumour, craniosynostosis with copper-beaten skull or involvement of multiple sutures) and/or (2) evidence of raised intracranial pressure ≥ 18.5 mmHg (≥ 20.5 mmHg if child was sedated and/or obese) and/or (3) diagnosis of PTC according to the Friedman criteria [
22].
Inclusion criteria for NTG encompassed the following: (1) absence of fundoscopic papilledema and absence of radiological evidence of a disease associated with increased ICP or (2) absence of fundoscopic papilledema, absence of imaging findings and intracranial pressure < 18.5 mmHg.
Informed consent was obtained from parents and children old enough to understand the study prior to US-investigation. According to the written ophthalmologic reports, fundoscopic findings were classified into (A) papilledema (in any manifestation) and (B) no papilledema.
All procedures were in accordance with the Code of Ethics of the World Medical Association (Declaration of Helsinki). The study protocol was approved by the institutional ethics committee (180/2018BO2). The study meets the STROBE guidelines for reporting observational studies.
Study population
Seventy-two patients, aged 9 months to 18 years, (mean age 9.3 ± 4.7 years), were included; 69 children were > 1 year. Forty-six children were male (64%), and 26 were female (36%).
The diagnoses included: PTC (n = 21), brain tumor (n = 16), arachnoid cyst (n = 8), craniosynostosis (n = 9), other intracranial pathologies (sinus vein thrombosis, intracranial haemorrhage, hydrocephalus; n = 14) and patients without proven disease (n = 4).
In 40/50 patients of the TG, fundoscopy, US-ONSD and US-ODE were acquired before and after ICP decreasing therapy. In 24 of those 40 patients, at least one further follow-up investigation was performed (maximum follow-up interval 50 weeks).
PoC ultrasound
One examiner (SRK), highly experienced in transorbital ultrasound, performed all US-investigations. US was performed in supine position, head straight and not elevated. A 12-MHz linear transducer (Philips, Epiq 5G) was placed to the closed eyelid. US-ONSD determination was performed 3 mm behind the bulb, 90° to the optic nerve in axial plane (Fig.
4e). From three measurements per eye, mean ONSD and resulting mean binocular US-ONSD were calculated as previously described [
23]. Additionally, the optic disc elevation was measured three times per side (Fig.
4e) and mean binocular US-ODE was calculated.
Sample size calculation
The sample size estimation for comparison of US-ONSD means was calculated at 16 patients per group to achieve 99% power with an alpha of 0.1 (p < 0.001). The sample size estimate for comparison of US-ODE means was calculated at 22 patients per group to achieve 95% power with an alpha of 1 (p < 0.01). With an estimated loss of 10% due to exclusion criteria and data loss, the study aimed to recruit a minimum of 50 patients.
Statistical analysis
SPSS (PASW Statistics 27, IBM) statistical software was used. Data were tested for normal distribution using the Kolmogorov-Smirnov or Shapiro-Wilk test. Parametric data were reported as means \(\pm\) standard deviation (sd). Depending on normality of distribution, correlation of parametric variables was tested using Pearson’s or Spearman’s correlation; point-biserial correlation r was used for comparing parametric to nominal variables. Receiver operating characteristic (ROC) curves were created to define US-ONSD and US-ODE cut-off values according to Youden-Index calculation. Analysis of variance (ANOVA) was used to determine intra-observer variability of US-ODE measurement. Cronbach’s alpha (α), a model of internal consistency, was used to test repeatability and observer reliability of US-ODE determination. Repeatability and reliability are considered reasonable when Cronbach’s alpha is > 0.7, strong if > 0.8 and excellent if > 0.9. The Mann-Whitney U test and dependent t test were used for comparing mean values or paired samples, respectively. Chi-Quadrat-test was applied for comparing qualitative parameters. Bonferroni’s correction was used to correct multiple comparisons. Statistical significance was set at p < 0.05: p < 0.05*, p < 0.01**, p < 0.001***.