A Prediction Model to Determine Childhood Epilepsy After 1 Or More Paroxysmal Events Eric Van Diessen, Herm J
Total Page:16
File Type:pdf, Size:1020Kb
A Prediction Model to Determine Eric van Diessen, MD, PhD, a Herm J. Lamberink, MD, a Willem M. Otte, PhD, a, b Nynke Doornebal, MD, c ChildhoodOebele F. Brouwer, MD, PhD,c, d Floor Epilepsy E. Jansen, MD, PhD, a Kees AfterP.J. Braun, MD, PhD 1a or More Paroxysmal Events OBJECTIVES: abstract The clinical profile of children who had possible seizures is heterogeneous, and accuracy of diagnostic testing is limited. We aimed to develop and validate a prediction model that determines the risk of childhood epilepsy by combining available information at METHODS: first consultation. We retrospectively collected data of 451 children who visited our outpatient department for diagnostic workup related to 1 or more paroxysmal event(s). At least 1 year of follow-up was available for all children who were diagnosed with epilepsy or in whom diagnosis remained inconclusive. Clinical characteristics (sex, age of first seizure, event description,n medical history) and EEG report were used as predictor variables for building a multivariate logistic regression model. Performance was validated in an external cohort RESULTS: ( = 187). – Model discrimination was excellent, with an area under the receiver operating – – characteristic curve of 0.86 (95% confidence interval [CI]; 0.80 0.92), a positive predictive value of 0.93 (95% CI 0.83 0.97) and a negative predictive value of 0.76 (95% CI 0.70 0.80). Model discrimination in a selective subpopulation of children with uncertain – diagnosis after initial clinical workup was good, with an area under the receiver operating CONCLUSIONS: characteristic curve of 0.73 (95% CI 0.58 0.87). This model may prove to be valuable because predictor variables together with a first interictal EEG can be available at first consultation. A Web application is provided (http:// epilepsypredictio ntools. info/ first- consultation) to facilitate the diagnostic process for clinicians who are confronted with children with paroxysmal events, suspected of having an epileptic origin. WHAT’S KNOWN ON THIS SUBJECT: In clinical practice, epileptic seizures in children often go aDepartment of Pediatric Neurology, Brain Center Rudolf Magnus and bBiomedical MR Imaging and Spectroscopy Group, Image Sciences Institute, University Medical Center Utrecht, Utrecht, Netherlands; underrecognized because of the heterogeneous cDepartment of Pediatrics, Martini Hospital, Groningen, Netherlands; and dDepartment of Neurology, University clinical symptoms and limited sensitivity of the initial Medical Center Groningen, University of Groningen, Groningen, Netherlands interictal EEG. As a result, diagnosis of epilepsy may be delayed. Dr van Diessen designed the study, collected the patient data, performed statistical analysis, and wrote the manuscript; Dr Lamberink collected the patient data, performed statistical analysis, WHAT THIS STUDY ADDS: A decision model was built and wrote the manuscript; Dr Otte designed the study, performed statistical analysis, and wrote with the clinical information and the initial interictal the manuscript; Drs Doornebal and Brouwer collected the patient data and commented on the EEG available at first consultation of 638 children. statistical analysis and the manuscript; Dr Jansen designed the study, collected the patient data, This model, available to the general readership as and commented on the statistical analysis and the manuscript; Dr Braun designed the study and a Web application, may facilitate the diagnostic commented on the statistical analysis and the manuscript; and all authors approved the final process for clinicians. manuscript as submitted and agree to be accountable for all aspects of the work. DOI: https:// doi. org/ 10. 1542/ peds. 2018- 0931 To cite: van Diessen E, Lamberink HJ, Otte WM, et al. A Accepted for publication Sep 5, 2018 Prediction Model to Determine Childhood Epilepsy After 1 or More Paroxysmal Events. Pediatrics. 2018;142(6):e20180931 Downloaded from www.aappublications.org/news by guest on October 2, 2021 PEDIATRICS Volume 142, number 6, December 2018:e20180931 ARTICLE A paroxysmal event can be consultation after 1 or more possible department. The institutional frightful for1 both children and seizures. Model performance was ethical committees of both hospitals caregivers. When an epileptic validated with an external cohort. approved the study and concluded origin is considered, the associated To improve the implementation in that the Dutch Medical Research ’ psychosocial aspects may negatively clinical practice, we constructed Involving Human Subjects Act did not influence the child2, 3 s health-related a Web application that may apply, and written informed consent quality of life. Recurrent epileptic assist clinicians to estimate the Outcomewas not required. Definition seizures may be harmful– to the individualized probability of epilepsy developing brain, leading4 6 to cognitive in children who present after 1 or and behavioral deficits. This more possible seizures. The outcome measure was presence emphasizes the need for an early METHODS or absence of epilepsy, defined as: (1) and accurate diagnosis. However, at least 2 unprovoked seizures within establishing the epileptic origin of Subject Selection 1 year, judged by an experienced a paroxysmal event in children is child neurologist to be epileptic in often challenging in clinical practice, origin; or (2) 1 unprovoked seizure due to a limited sensitivity of We retrospectively collected clinical – and epileptiform EEG abnormalities heterogeneous clinical symptoms data of children referred after 1 or 7 9 at first consultation or at later sleep- and interictal EEG recordings. In more paroxysmal events, suspected deprived or prolonged EEG that children with uncertain diagnosis, of having an epileptic origin. Data confirm the diagnosis of an epilepsy additional investigations including from 451 children who visited the syndrome, or the demonstration of prolonged or sleep-deprived EEG and outpatient department of pediatric a typical epileptogenic brain lesion MRI are required, with subsequent neurology, University Medical Center 7, 10, 11 on MRI. If diagnosis of epilepsy was delay in making a final diagnosis. Utrecht, between January 2008 discarded on the basis of clinical Otherwise, diagnostic inaccuracy can and May 2013 were used to build a history, EEG results, other ancillary lead to a false diagnosis of epilepsy, clinical decision model (development investigations, or at least 1 year of resulting in unnecessary antiepileptic cohort). Data of 187 children who uneventful follow-up, we attempted drug treatment or outpatient visited the outpatient department to reappoint the diagnosis. Children admissions and delay of other of pediatrics, Martini Hospital in whom a definitive diagnosis was important ancillary investigations. Groningen, between January 2013 inconclusive at the end of follow-up, To this end, a clinical tool that and July 2016 were collected as an or who were lost to follow-up, were facilitates accurate diagnosis of first external cohort to validate the model classified as having no epilepsy. The paroxysmal events in children at (validation cohort). All children were consulting pediatric neurologist initial evaluation would be highly referred by a general practitioner, always made the definitive diagnosis. valuable. pediatrician, neurologist, or via the Predictor Variables emergency department. In addition Previous efforts to identify to clinical history and physical prognostic clinical variables after – examination, the report of a standard Potential predictor variables were first consultation, focused on the 8, 10, 12 14 EEG recording performed at time selected from previous cohort risk of seizure recurrence – of first consultation was used. At studies in children with newly in patients with first indisputable 16 20 least 1 year of clinical follow-up was diagnosed epilepsies. Predictor epileptic seizures and patients available for all children who were variables were required to be with ambiguous events, were eventually diagnosed with epilepsy routinely available at the time of first excluded from further analysis. and for those whose diagnosis consultation. Information on age at These restrictions may hamper remained inconclusive after initial consultation, sex, age at first event, its value in daily practice because evaluation. Children in whom medical history, relevant family diagnostic inaccuracy of first seizures 15 diagnosis of epilepsy was discarded history, psychomotor development, is common. Also, children are not directly after first consultation number of events and detailed event always referred directly after a first were referred back to their general description, and the EEG report was possible seizure because of delayed 4, 7 practitioner with explicit instructions available and extracted into separate recognition. to report new possible seizures. variables in a database. Age at Here, we determined the value Excluded were children in whom the consultation and at first event were of clinical and EEG data routinely diagnosis of epilepsy had already truncated at month level. Medical available at first consultation and been confirmed by the referring history of a child was categorized “ ” constructed a prediction model to health care professional before into 3 predictor variables, namely determine the risk of epilepsy at first consultation at the outpatient neurologic history in the presence Downloaded from www.aappublications.org/news by guest on October 2, 2021 2 VAN DIESSEN et al of a history