Process for validating the LINNE scoring.

BackgroundAnkyloglossia may restrict newborn infant’s tongue movements, complicating breastfeeding. However, due to lacking evidence-based guidelines, patient selection for early frenotomy has remained a challenge. The present prospective, observation study validated a scoring aimed to detect infants who required early tongue-tie treatment. It is a part of the Lingual Frenulum in Newborn Infants (LINNE) project, investigating tongue-tie epidemiology along with a randomized treatment trial.MethodsHealthy mother-infant-dyads were assessed by independent and mutually blinded study physicians and midwives. They used the LINNE scoring, including: 1. Infant examination using a picture assessment tool for tongue-tie in breastfed babies (TABBY); 2. Maternal breastfeeding experience scoring (MBES), a psychometric testing on acute breastfeeding symptoms; 3. Three anamnestic factors associated with tongue-ties and breastfeeding problems. Validation tests were conducted using the Cronbach alpha for the whole LINNE scoring, and for the subscorings: 1. TABBY: areas under the curve (AUC), multi-rater Fleiss kappa; 2. MBES: content and construct validity and responsiveness tests; 3. Anamnesis: accuracy studies for diagnostic tests.ResultsOf the studied 556 mother-infant-dyads, 72 (12.9%) infants fulfilled the early tongue-tie treatment criteria. The overall LINNE scoring’s internal consistency was acceptable (alpha = 0.723). The pooled TABBY scoring (n = 1094) accurately detected the need for treatment (AUC = 0.914, 95%CI 0.891–0.938) with moderate inter-rater agreement (kappa = 0.441, 95%CI 0.406–0.477). MBES responsiveness, content, and construct validities reached the required levels. Anamnestic factors had high specificity but fair sensibility for the tongue-tie causing breastfeeding symptoms.ConclusionThe LINNE scoring demonstrated moderate inter-rater agreement; however, the accuracy of the TABBY tool was very good with AUC 0.9 for ankyloglossia needing early treatment.