Victor Certal1,2, Hélder Silva2, Carlos Carvalho2, Altamiro Costa-Pereira3, Inês Azevedo4, João Winck5, Robson Capasso6, Macario Camacho7. 1. Department of Otorhinolarygology/Sleep Medicine Centre-Hospital CUF, Porto. 2. Department of Otorhinolarygology, Hospital São Sebastião, Sta Maria da Feira, Portugal. 3. Centre for Research in Health Technologies and Information Systems. 4. Department of Pediatrics, Faculty of Medicine. 5. Department of Pulmonology, Faculty of Medicine, University of Porto. 6. Department of Otolaryngology-Head and Neck Surgery, Division of Sleep Surgery, Stanford Hospital and Clinics, Stanford, California. 7. Department of Otolaryngology, Division of Sleep Surgery and Medicine, Tripler Army Medical Center, Honolulu, Hawaii, U.S.A.
Abstract
OBJECTIVES/HYPOTHESIS: Obstructive sleep apnea (OSA) is a syndrome frequently diagnosed in children; however, it lacks optimal diagnostic methods. This study aimed to provide a clinical decision rule for predicting pediatric OSA using commonly available clinical information. STUDY DESIGN: A prospective cohort study. METHODS: Children between the ages of 3 to 6 years-old, referred for an otorhinolaryngology consultation due to clinical suspicion of OSA, were recruited from January to June 2014. At baseline age, weight, height, gender, body mass index, Pediatric Sleep Questionnaire (PSQ) scores, tonsil size, and oxygen desaturation index (ODI) were assessed. A logistic regression modeling was used with backward stepwise elimination to develop a prediction model. RESULTS: Sixty-seven children were included with a mean age of 4.51 years. Of the 67 children included in this study, 25 (37.3%) subjects were diagnosed with pediatric OSA. Significant predictors of pediatric OSA in the final model (odds ratio, 95% confidence interval) included PSQ score (5.12; 3.3-6.5), ODI (1.34; 1.0-1.79) and tonsil size (6.7; 3.22-9.75). The final decision rule had a sensitivity of 88% and a specificity of 86%. The area under the receiver operating characteristic curve was 0.897. CONCLUSION: The proposed clinical decision rule, based on three readily available variables, is a promising discriminating instrument for prediction of OSA among children between 3 and 6 years. LEVEL OF EVIDENCE: 2b.
OBJECTIVES/HYPOTHESIS: Obstructive sleep apnea (OSA) is a syndrome frequently diagnosed in children; however, it lacks optimal diagnostic methods. This study aimed to provide a clinical decision rule for predicting pediatric OSA using commonly available clinical information. STUDY DESIGN: A prospective cohort study. METHODS:Children between the ages of 3 to 6 years-old, referred for an otorhinolaryngology consultation due to clinical suspicion of OSA, were recruited from January to June 2014. At baseline age, weight, height, gender, body mass index, Pediatric Sleep Questionnaire (PSQ) scores, tonsil size, and oxygen desaturation index (ODI) were assessed. A logistic regression modeling was used with backward stepwise elimination to develop a prediction model. RESULTS: Sixty-seven children were included with a mean age of 4.51 years. Of the 67 children included in this study, 25 (37.3%) subjects were diagnosed with pediatric OSA. Significant predictors of pediatric OSA in the final model (odds ratio, 95% confidence interval) included PSQ score (5.12; 3.3-6.5), ODI (1.34; 1.0-1.79) and tonsil size (6.7; 3.22-9.75). The final decision rule had a sensitivity of 88% and a specificity of 86%. The area under the receiver operating characteristic curve was 0.897. CONCLUSION: The proposed clinical decision rule, based on three readily available variables, is a promising discriminating instrument for prediction of OSA among children between 3 and 6 years. LEVEL OF EVIDENCE: 2b.