Validity of parentally reported febrile seizures: the generation R study

The objective of this study was to examine the validity of the diagnosis of "febrile seizure" as reported by parents. The study was embedded in a population-based prospective cohort study. Information on paroxysmal events was obtained by screening questions at ages 1, 2, and 3 years. One of these questions was the following: "Did your child have a febrile seizure?" If a screen-positive result was found, an additional questionnaire was sent and the medical record was consulted. Based on this information, paroxysmal events were classified by a pediatric neurologist as febrile seizure or other event. The validity of a positive reply to the screening question on febrile seizures was assessed, taking this classification as reference standard. Analyses were based on participants who reported at least one paroxysmal event (n = 610). The sensitivity of the positive reply to the question, "Did your child have a febrile seizure?", for the diagnosis of febrile seizures was 92%, the specificity 72%, the positive predictive value 41%, and the negative predictive value 98%. In conclusion, the sensitivity of the question, "Did your child have a febrile seizure?", is high. The positive predictive value is only 41%. Although this question may be appropriate as a screening instrument for febrile seizures, a second stage of evaluation is necessary to identify true cases.