Nystagmus is an abnormal, rhythmic, repetitive oscillation of the eyes that can affect visual acuity due to the movement of images away from the fovea.1
It is a sign of cerebellar, vestibular or visual dysfunction. Based on the age of onset, nystagmus is generally classified as infantile nystagmus, which appears within the first 6 months of life, and acquired nystagmus, which appears later.2
In a single retrospective population-based study in children and adolescents, the annual incidence
The estimated rate of nystagmus was 6.72 per 100,000 inhabitants < 19 years of age.3 However, its incidence in the emergency care setting is unknown. Given the broad differential diagnosis of acute nystagmus (AN), its clinical evaluation can be challenging.4
When faced with a child with AN, the medical concern is to exclude that nystagmus is the sign of a significant neurological disorder that needs immediate intervention.1 In few previous studies, limited to the adult population, the authors investigated the clinical characteristics, underlying causes , and the management of AN in the emergency department setting.5,6 However, the spectrum of disorders that cause acute neurological dysfunction in children differs significantly from that in adults.7,8
To the authors’ knowledge, no studies have been used to investigate the presentation and management of AN in the pediatric and adolescent population in emergency departments or outpatient settings.
Through a retrospective analysis of a large multicenter cohort of children seen in pediatric emergency departments (PEDs), the authors’ objective was to describe the epidemiology, clinical characteristics, underlying causes, and management of AN in children.
The secondary objective was to identify clinical characteristics associated with an increased risk of significant underlying neurological abnormalities in children with AN to contribute to the improvement of their management in the emergency setting.
| Methods |
This multicenter and retrospective cohort study was carried out in the PEDs of 9 Italian hospitals (Padua, Genoa, Bologna, Florence, Siena, Perugia, L’Aquila, Caltagirone and Rome) after having obtained approval from the institutional ethics committee of the participating hospitals.
Patient identification was performed by searching keywords in the hospitals’ electronic database, and all patients < 18 years of age who attended the DEPs from January 2009 to December 2016 were included.
Medical records were selected by searching for the keyword “nystagmus” in the “history,” “clinical examination,” and “diagnosis” fields of the electronic notes. Potential cases were controlled manually by reviewing the medical history.
All patients referred to the DEP with a history of an eye movement abnormality of <30 days duration and in whom a diagnosis of nystagmus was confirmed were included.
The exclusion criteria were (1) abnormal eye movements other than nystagmus (such as ocular flutter, opsoclonus and/or supranuclear gaze disturbance), (2) patients treated in the PED for head injury or (3) epileptic seizures, and (4) patients affected by a known neurological condition that explained the nystagmus.
Patients who attended the PED for eye movement abnormalities and patients who attended for other symptoms whose nystagmus was detected during the clinical examination were included. In the latter case, nystagmus was considered new onset if it was reasonably linked to the same pathological process that caused the acute presenting symptoms (e.g., ataxia, vertigo, headache, altered mental status), if it had never been noticed or mentioned in the medical records, and if it could not be explained by any of the known pre-existing medical problems.
The following information was extracted from each medical record: demographic characteristics, medical history, clinical examination findings, investigations performed, hospital admission, and length of stay (if applicable).
In accordance with the Italian National Health Service guidelines, consultation priority upon admission to the DEP was based on a 4-color triage coding scale: • Code red: critical medical condition, altered vital signs requiring immediate life-saving intervention, high priority access to emergency care; • Code yellow: serious condition, risk of progression to critical conditions, intermediate priority access and re-evaluation necessary in 5 to 15 minutes; • Code green: stable condition and vital signs, medical consultation can be postponed without risk, low priority access and re-evaluation necessary in 30 to 60 minutes; and • White code: good condition, non-urgent query. |
The code is assigned by a triage-trained nurse at the entrance to the emergency department and is periodically re-evaluated during the wait time. This system was applied throughout the study period, after the Italian Ministry of Health reached an agreement with all Italian regions in 2001.9
The etiology of nystagmus was classified based on the diagnosis made at the end of the evaluation work. Causes of nystagmus that reflected a significant neurological abnormality, requiring further investigations and intervention (neoplastic, cerebrovascular, infectious, demyelinating, degenerative, or central nervous system [CNS] malformations), were considered urgent conditions (UC).
The clinical and demographic characteristics of the general cohort and the 2 subgroups (patients with and without UC) were described. Each variable was compared between the 2 subgroups to identify significant differences. χ2 and Student’s t tests were used for statistical comparison of categorical and continuous variables, respectively.
To detect predictive variables associated with a higher risk of UC in patients with AN, a logistic regression analysis model was applied. Clinical characteristics that revealed significant differences in the x2 and t tests were selected as independent variables.
Sex and age were included a priori to adjust the effect of each independent variable for the demographic characteristics of the cohort. Variables with a very unbalanced distribution between the 2 groups (frequency 0% in 1 group) were excluded.
Adjusted odds ratios (OR) and 95% confidence intervals (CI) were used as measures of effect. Statistical significance was set at a P value < 0.05 for all analyses.
| Results |
A total of 206 patients who met the inclusion criteria were identified (108 male patients; male-to-female ratio: 1.01). The average age at the time of consultation at the DEPs was 8 years 11 months. Thirty-seven patients (18%) were <2 years of age (25 of 37 were <6 months), 87 children (42%) were between 2 and 12 years, and 82 (40%) were >12 years.
In 77 cases (37%), patients presented to the PED complaining of abnormal eye movements. In the remaining cases, nystagmus was detected during clinical examination in patients with other complaints.
The median time from symptom onset to admission to the PED was 5 days, with 68% of the total cohort reporting symptom onset within the previous 3 days (median: 2 days).
The plane of nystagmus was horizontal in the vast majority of cases (71.4%). Less frequently, vertical (6.8%), rotary or torsional (1%), or combined nystagmus (3.4%) was reported. However, in a significant proportion of patients, the plane of oscillation was not reported in the clinical records (17.4%).
The most commonly reported symptoms during consultation with the DEP were headache (43.2%) and vertigo and/or dizziness (42.2%), followed by nausea and vomiting (25.7%) and visual disturbances (16.02%). ). Many patients presented with a constellation of associated symptoms.
The most commonly reported clinical findings during physical examination included ataxia (18.45%), strabismus (13.1%), or decreased level of consciousness (6.3%).
Sixteen patients (7.8%) had fever at the time of admission to the PED. In 54.9% of cases, nystagmus was the only neurological abnormality reported.
Interconsultations with specialists were requested for 83.5% of the patients, mainly neurological consultations (61.2%) or ophthalmological consultations (35.4%).
Approximately half of the patients underwent neuroimaging testing (53.9%); 60.4% of them took the test directly at the DEP. A total of 118 patients (57.3%) were hospitalized after consultation at the DEP.
Migraine was the most common cause of AN (representing 25.7% of all cases), followed by vestibular disorders (14.1%) . Transient vertigo, not otherwise identified, accounted for 12.6% of cases.
Idiopathic infantile nystagmus (IIP) was responsible for 6.8% of cases of AN, representing the first cause of consultation to the DEP for nystagmus in the first year of life. Other rarer causes of NA included toxic ingestion, post-infectious cerebellar ataxia, and periodic syndromes.
Thirty-nine patients were diagnosed with UC (18.9%). Brain tumors were the first UC causing AN (17 cases; 8.3% of the entire cohort).
Other causes included idiopathic intracranial hypertension, demyelinating disorders, degenerative conditions, and central nervous system (CNS) infections or malformations.
Patients with UC were found to be significantly younger than patients without UC (mean age: 6 years and 11 months vs. 12 years and 4 months), with the highest frequency of UC occurring in children between 1 and 6 years of age. age. The delay from symptom onset to PED presentation was significantly longer in patients with UC compared to patients without UC.
Diplopia, blurred vision, strabismus, cranial nerve palsies, ataxic gait, dysmetria, pyramidal weakness and papilledema, as well as the absence of accompanying symptoms, were significantly more frequent in patients with UC. In contrast, vertigo and the absence of any neurological sign were more commonly found in patients without UC, as well as the attribution of a non-urgent code (green or white).
On this basis, 14 variables were selected for the logistic regression model, including 199 patients (96.6%). According to this model, the presence of cranial nerve deficits, ataxia or strabismus was strongly associated with underlying UC, increasing its risk by 46.82, 9.29 and 9.17 times, respectively (P < 0.02). .
Although not reaching statistical significance, the presence of pyramidal weakness and abnormal head postures was also associated with a higher risk of UC (with an OR of 8.59 and 7.18, respectively).
A longer time from symptom onset to PED referral was found to increase the odds of underlying UC, with a 9% increased risk for each day from nystagmus onset (OR = 1.09; P < 0 ,01).
Despite the younger age at admission of the UC patients, this was not associated with an increased risk of UC when adjusted for other variables in the logistic regression model.
On the other hand, it was found that the occurrence of vertigo reduced the odds of an underlying UC (OR = 0.17; P < 0.01), as well as the attribution of a green or white triage code (OR = 0.30 ; P = 0.01).
| Discussion |
To the authors’ knowledge, this is the first study in which the epidemiology of AN in children was investigated. The most striking finding of this study is that UC represents almost 20% of AN cases in the DEP, confirming that nystagmus is an alarm signal.
It can be argued that the strict inclusion criteria may have led to an overestimation of CUs, determining the exclusion of several transient causes of gaze control abnormalities, such as postictal and posttraumatic states. However, the large size of the cohort and multicenter recruitment are in favor of good external validity of these findings.
The abrupt or subacute onset of abnormal eye movements represents the main cause of PED consultation, especially in infants and young children, while in older patients, AN is usually part of a more complex clinical picture and rarely represents the main complaint of the patient. patient.
Infantile nystagmus has been described as the main cause of nystagmus in children and adolescents, representing up to 87.3% of cases in a population-based study.3
In most cases, infantile nystagmus has a gradual onset and usually does not require urgent evaluation. As expected, in the emergency setting, infantile nystagmus explains a considerably lower proportion of cases (12.08% in the present cohort).
Despite the lower frequency in the emergency setting, this finding reveals that childhood nystagmus, including NII and ocular diseases, can have an abrupt presentation, raising concern of underlying brain injury.
In fact, underlying UC was diagnosed in 6 of 25 infants < 6 months (24%) of age, far exceeding the proportion previously reported in a community setting (2.8%)3 and suggesting that an underlying intracranial lesion is more likely. when infantile nystagmus has an abrupt onset that leads to consulting the DEP.
Even though it has been suggested that acquired nystagmus should be suspected when it begins after 4 months of age,10 in this series brain lesions were responsible for 3 of 15 cases of nystagmus in infants < 4 months of age, including leukodystrophies of early onset and complex brain malformations.
Taken together, the authors’ data reveal that, at least in the emergency setting, the proportion of young infants with congenital or early-onset brain injuries is not negligible, and a high level of suspicion is required to detect these cases.
Although not supported by strong evidence, direction and waveform have been widely reported to be useful indicators of the origin of nystagmus, with NII most frequently appearing as bilateral or conjugate, or occurring in the horizontal plane, with a shape of pendular or spasmodic wave.1,2
Unfortunately, most medical records lacked this information, preventing the inclusion of these variables in the analyses. However, when faced with a child with infantile nystagmus, it seems reasonable to consider these features as arguments in favor of NII, while a nystagmus with atypical features should prompt further investigation.
In older, otherwise healthy children, acutely or subacutely acquired nystagmus raises the obvious concern of an underlying intracranial injury.1
In the authors’ experience, AN is usually detected during visits to the PED for acute vertigo or dizziness, severe headache, or acute vision disturbances. However, these symptoms often occur at the same time, confusing the clinical picture.
Signs of ataxia (reported in 18.45% of cases, including gait ataxia and/or dysmetria) were the most useful clinical clues to localize the lesion in the cerebellum, brainstem, or vestibular pathways.
In contrast, other examination findings were detected more rarely. As a note, nystagmus was the only abnormal finding on neurological examination in 54.9% of patients admitted to the PED. Even in the UC group, in 12 of 39 patients, no other neurological abnormalities were observed on admission, and other signs only emerged in the following hours or days.
In adults, NA is primarily described in the context of acute dizziness or vestibular syndrome.11,12 In this population, vertebrobasilar strokes occur in 3.2% to 4.9% of all patients. with dizziness and nystagmus,5,6 representing the most common UC.13
In the authors’ cohort, despite the high frequency of dizziness and vertigo among the reported symptoms, the occurrence of a true vestibular syndrome (including dizziness, gait disturbance, and nausea or vomiting12) was rare, occurring in only 13 patients, with 10 additional patients presenting with ataxia and dizziness without nausea or vomiting.
Furthermore, CNS tumors represented more than a third of UC and no cases of stroke were identified. In fact, vertebrobasilar infarcts are a rare cause of stroke in children, their frequency varies from 6% to 13% of all pediatric strokes,14-18 and nystagmus is rarely cited among the signs presented.17
These findings suggest that posterior circulation infarcts are an exceptional cause of acute vestibular syndrome in children,19 while posterior fossa tumors are more frequently associated.
Among the benign etiologies of AN, migraine was the most common. As described previously, migraine and pediatric migraine variants are the most common cause of “brain attacks” in children and adolescents,7 with vestibular migraine and benign paroxysmal vertigo representing the most common causes of pediatric acute vertigo.20
In the present cohort, migraine frequency peaked between 8 and 14 years of age. As expected, this peak coincides with the peak of migraine diagnosis in children,21 with PED attendance less likely in older patients who have already received a migraine diagnosis. Among the differential diagnosis of acute headache and nystagmus, 4 cases of
idiopathic intracranial hypertension, confirming that AN is a possible, although atypical, feature of this condition.22
As expected, cranial nerve deficits, ataxia, and strabismus were found to be significantly associated with underlying UC, with their occurrence strongly suggesting a brainstem and/or cerebellar lesion.
It can be argued that ataxia and strabismus were among the most common findings in the overall cohort, sometimes occurring in the absence of any UC. In fact, it should be noted that many causes of cerebellar, vestibular, or ocular motility dysfunctions in children are transient and benign, including acute post-infectious cerebellar ataxias or viral vestibular neuritis.23,24
In some cases, nystagmus occurs as a transient finding during the recovery phase after peripheral gaze palsy, such as in abducens nerve neuritis.
These findings lead to the suggestion that ataxia and strabismus should be considered alarming findings in children with nystagmus, justifying the use of neuroimaging to exclude a central lesion, unless it is an obviously benign cause (e.g., chickenpox infection). that can be quickly identified.
Surprisingly, the association between pyramidal weakness or abnormal head postures and underlying UC did not reach statistical significance in the logistic model, even though these two characteristics were significantly more frequent among patients with UC in the univariate analysis and revealed a strong positive correlation with underlying UC.
Pyramidal weakness indicates damage to the corticospinal tract; however, its overall low frequency in this cohort was probably insufficient to reach significance in multiple regression analysis.
Abnormal head postures have been reported as possible signs present in pediatric brain tumors25 but they may also be postures adopted by the patient to reduce or block nystagmus to achieve better vision. Despite the lack of a significant association, the authors suggest that any child with these findings should be appropriately investigated and referred to a pediatric neurologist.
In contrast, vertigo was significantly associated with non-urgent conditions, reflecting the rare occurrence of central vestibular syndrome in children. Perhaps unexpectedly, patients with underlying UC attended the PED with a greater delay compared to patients without UC, resulting in an increased risk of UC in patients with persistent nystagmus.
In fact, with the notable exception of NII, most nonurgent conditions are paroxysmal disorders, and the persistence of nystagmus for several days should be considered an alarming finding.
In this series, UC affected all age groups, but its frequency reached a significant peak between 1 and 6 years of age (P = 0.02). This finding probably reflects the fall in frequency in this age group of NII and migraine, the main benign causes of AN respectively in younger and older children, with a relative increase in UC. Despite the lack of a significant correlation when adjusting for other variables in the logistic regression model, the higher frequency of UC in toddlers and preschool children should be considered in the evaluation of NA.
Additionally, other alarming findings not included in the logistic analysis should be considered. For example, papilledema was not included due to distribution imbalance between the 2 subgroups, but represents an obvious neuroimaging indication. Similarly, different abnormal eye movements (i.e., ocular flutter or opsoclonus) were considered as exclusion criteria, but their co-occurrence with nystagmus suggests a central origin of the gaze control abnormality.
Mainly due to its retrospective design, this study suffers from some limitations that may affect the conclusion.
First, during the study period, a routine protocol for NA was not applied, making the evaluation and management of the cohort highly heterogeneous. Furthermore, in a significant proportion of patients with transient symptoms, the definitive etiology was not identified. This is mainly attributable to the lack of detailed information on the diagnostic workup of non-hospitalized patients. Consequently, few patients received detailed vestibular function testing.
Finally, the logistic regression model was designed to identify patients with significant neurological abnormalities requiring urgent intervention, but a high risk of underlying UC is not informative per se of a specific disorder.
| Conclusions |
This cohort reveals that NA should be considered an alarming finding in children presenting to PEDs given the high prevalence of UC in this population.
Younger age or the absence of other neurological signs at presentation do not completely exclude the presence of intracranial lesions.
Cranial nerve palsies, ataxia, strabismus, upper motor neuron signs, and abnormal head postures should be considered red flags during the evaluation of a child with AN.
In contrast, acute dizziness and nystagmus, in the absence of other neurological signs, are rarely associated with intracranial lesions.
