In the initial search, PubMed yielded 542 articles, EMBASE yielded 1146, and Cochrane Central yielded 15. Articles were imported into Mendeley [23], a citation manager, and duplicates were automatically removed—yielding 1260 unique articles. Upon updating the literature search in 2023, PubMed yielded an additional 129 articles, Embase yielded an additional 84, and Cochrane an additional 3. Duplicates were removed resulting in 175 unique titles. In the first phase, two reviewers (GG, DH or PA) independently screened the articles by title to ensure that the articles were relevant to the inclusion criteria. The reviewers were blinded from each other’s decisions, and in the event of a disagreement, an ultrasound fellowship trained attending physician (AC) acted as the tie-breaking vote. Deviations were found to be minimal. The articles were re-screened by abstract, then by full-text using the same procedure. (GG, DH,AB, and PA).

A total of twenty nine studies published between 1994 and 2023 were selected for inclusion in this systematic review [12, 14, 26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52] (Table 1). Six out of the twenty nine studies were case series [28, 30, 37, 43, 45, 46]. Furthermore, seven studies were conducted in pediatric patients only [27, 28, 30, 31, 37, 48, 49], two had a mixed population of children and adults [14, 42], and the remainder were conducted in adult patients only. All studies utilized B scan ultrasonography with high-frequency transducers, ranging in frequency from 3 MHz (megahertz) to 20 MHz. One study also conducted A scan ultrasonographic measurements [42].

The included studies had a global distribution with eleven studies having been conducted in Europe, ten in North America, five in Asia, and one each in Oceania and Africa. There was also heterogeneity as to the setting for each study; seven studies were conducted in the emergency department, seven were conducted in the department of neurology, four were conducted in the department of ophthalmology or an eye center, and two were conducted in a neurotrauma or intensive care unit. Other settings included a perinatal center, clinical research facility, a mix of settings, or did not specify where enrollment was conducted.

The included studies examining optic nerve sheath diameter and optic disc elevation varied in their reference standard used for evaluating eICP. One study used ICP monitoring as the reference standard [26]. Other studies used elevated opening pressure on Lumbar Puncture (LP) as their reference standard [27, 32, 38, 47, 50,51,52]. Not all studies were evaluating elevated ICP, as some studies examined patients for optic disc abnormalities [12, 14, 27, 31, 36, 37, 42, 44, 45]. The included studies were a mix of observational studies [14, 26, 27, 29, 36, 42, 49], case control [12, 33, 35, 38,39,40,41, 44, 50, 51], cohort studies [29, 31, 32, 47, 48, 52] and case reports [28, 30, 37, 43, 45, 46] with their own inherent limitations. They ranged in sample sizes from a case report of two patients to a study of 137 patients. Most of the studies were on adult patients though seven studies involved only pediatric patients.

The QADAS-2 tool (Additional file 1) identified that in 23 out of the 29 studies, selection of patients may have introduced bias. However, all studies were low risk that the included patients did not match the review question. The conduct or interpretation of the index test had a high risk of introducing bias in nine studies, and unclear in three. All studies were low risk of concern that the index test, its conduct or interpretation differed from the review question. The reference standard had low risk of introducing bias in thirteen studies, while the remainder were high risk or did not use a reference standard. Of the studies which employed a reference study, all of them were low risk that the target condition as defined by the reference standard did not match the review question. Finally, the patient flow had a high risk of introducing bias in nine studies, low risk in thirteen studies, and was unclear in seven.

Mean optic disc height in patients with papilledema ranged from 0.6 mm to 1.2 mm, with some studies documenting a slight variation between the Optic Disc Height (ODH) in the left eye and the right eye. (Table 2) One study reported the ODH as a median of 0.95 mm in the right eye and 1.0 mm in the left eye [34]. Yu et al. 2023 reported a median of 0.81 mm [52].

Figures 3 and 4 show the sensitivity and specificity of ultrasound detected ODE in the identification of papilledema. The gold standards of the studies varied significantly, and included a diagnosis of Idiopathic intracranial hypertension (IIH), papilledema on fundoscopy, elevated Cerebrospinal fluid (CSF) opening pressure during lumbar puncture, evidence of elevated ICP on Computed tomography (CT) scan, amongst other endpoints. Some studies did not explicitly provide sensitivity and specificity values. These values were calculated by our study team based on the provided information or raw data. One study [41] did not provide ODE cut-off values, so the study team assigned values of 0.6 mm and 1 mm arbitrarily for data analysis based on the majority of studies having used these values. The sensitivity and specificity of two studies was extrapolated from reported area under the curve graphs for ODE cut off values. Confidence intervals could not be discerned [47, 52].

Reported and calculated sensitivity of ultrasound detected optic disc elevation in detecting papilledema

Reported and calculated specificity of ultrasound detected optic disc elevation in detecting papilledema

Two studies found that ODE and CSF opening pressure had no correlation (Spearman’s ρ =  − 0.016, P = 0.94) [34] and (r = 0.27, p = 0.186) [50]. Four studies found a positive correlation between ODE and CSF (r = 0.383, p = 0.025) [47] (r = 0.77, R2 = 0.59, p < 0.001) [51], (r = 0.613, p < 0.001) [