Objective: In the Netherlands, many children are surgically treated for OM. Recent publications question the need for surgical treatment in common uncomplicated OM, although there is certainly a subgroup of infants that do need further assessment and possible treatment. The present study explores the possibility of using known and presumed risk factors for OM as an instrument for selecting and routing an infant with OM to further care. Methods: Two questionnaires were used. A questionnaire embracing a wide range of OM-related factors was sent to 6531 children aged nine months that were routinely invited for the hearing screen at nine months. In a second stage, a structured history questionnaire regarding ear and/or hearing problems, subsequent referral and/or treatment, was sent to all parents of children at age 21 months, responding to the first questionnaire. Univariate analysis was performed for identification of potential predictors for surgical treatment of OM for the whole sample as well as for 4 different subsets. Multivariable regression analysis with stepwise backward deletion was applied to arrive at a model for optimal prediction of tube insertion. A ROC (receiver operating characteristic) curve and the accompanying sensitivity and specificity values were analyzed to determine cut off values. Results: Univariate analysis found 10 items predicting surgical treatment for OM. Multivariable regression analysis resulted in a model with a ROC curve having an area of 0.801 and estimated coefficients for risk factors which were used to calculate a OM-score for each case. Conclusion: The developed scoring sheet, e.g., to be used in combination with physical examinations and/or tympanometry looks promising as a predictor for those children that might benefit from further assessment and eventually surgically treatment for OM.
|Journal||International Journal of Pediatric Otorhinolaryngology|
|Publication status||Published - 1 Jan 2011|