A comparative study of three systems for quantifying the spike and wave index in patients with continuous spikes and waves during slow sleep
*Correspondencia: Dr. Iñaki García de Gurtubay. Servicio de Neurofisiología Clínica. Complejo Hospitalario de Navarra. Irunlarrea, 4. E-31008 Pamplona (Navarra).
E-mail: igarciag@cfnavarra.es
Introduction: Continuous spikes and waves during slow sleep (CSWS) is an epileptic encephalopathy of childhood with a pattern of epileptiform discharges during sleep, which, if prolonged over time, produce severe neuropsychological impairment. Quantification of the paroxysms by means of a spike and wave index (SWI) > 85% establishes a presumptive diagnosis and allows early therapy to be started to prevent such impairment.
Aims: To compare the results of the classic method for calculating the SWI against two proposals that optimise the relation between the analysis time employed and the diagnostic sensitivity.
Patients and methods: The nocturnal electroencephalographic registers of 17 patients with CSWS were studied. Two observers calculated the SWI with the classic method, as well as with two other methods, M2 and M3, first in the active phase and then in the remission phase. The time required by each method, the individual SWI values and the agreement between methods and observers were compared.
Results: With M3 two of the patients failed to reach the cut-off point of SWI > 85%. Agreement in the active phase of CSWS after M2 and M3 was 0.762 and 0.704, respectively, while in the remission phase it was 0.951 and 0.830. Inter-observer agreement exceeded 0.905 in all cases.
Conclusions: The two abbreviated methods can be used in both the active and the remission phases, with a substantial reduction in the analysis time that is needed. Our results support the current tendency to consider SWI > 60% as suggestive of CSWS. Method M2 yields results that are closer to those of the classic method than those of M3.
Objetivos Comparar los resultados del método clásico de cálculo del SWI con dos propuestas que optimicen la relación entre el tiempo de análisis empleado y la sensibilidad diagnóstica.
Pacientes y métodos Se estudiaron los registros electroencefalográficos nocturnos de 17 pacientes con POCS. Dos observadores calcularon el SWI con el método clásico, así como con otros dos métodos, M2 y M3, primero en la fase activa y posteriormente en la fase de remisión de la POCS. Se comparó el tiempo consumido por cada método, los valores individuales de SWI y la concordancia entre métodos y observadores.
Resultados Con el M3 dos pacientes no alcanzaron el corte del SWI > 85%. La concordancia en la fase activa de la POCS tras el M2 y el M3 fue de 0,762 y 0,704, respectivamente, mientras que en la fase de remisión fue de 0,951 y 0,830. La concordancia entre observadores superó el 0,905 en todos los casos.
Conclusiones Los dos métodos abreviados se pueden utilizar tanto en la fase activa de la POCS como en la fase de remisión, con una sustancial reducción del tiempo de análisis empleado. Nuestros resultados apoyan la tendencia actual de considerar el SWI > 60% como sugestivo de POCS. El método M2 arroja resultados más cercanos a los del método clásico que los de M3.