Predictors of High Costs of Care among Otolaryngology Patients

Objectives: Identify predictors of high-cost otolaryngology care. Study Design: Cross-sectional. Setting: Tertiary academic multispecialty hospital. Subjects/Methods: All patients undergoing ≥1 otolaryngologic procedures from 2011 to 2015. Encounter costs were standardized using previously described methods approximating Medicare reimbursement. Patients were stratified by adult/pediatric and inpatient/outpatient. “Outliers” were defined as total encounter costs ≥95th percentile. Logistic regression measured predictors of outlier status. Results: In total, 2433 adult inpatient encounters (95th percentile $57,611), 10,031 adult outpatient encounters ($10,772), 346 pediatric inpatient encounters ($84,639), and 3027 pediatric outpatient encounters ($8978) were included. For adult inpatient and outpatient, isolated head and neck oncologic procedures were the reference group. Among adult inpatients, laryngology and facial plastics procedures predicted higher odds of outlier status (odds ratio [OR] = 4.1 and 7.2). Involvement of multiple otolaryngology subspecialties increased the odds (OR = 4.7). Neck dissection and reconstructive procedures were the most common primary operations for adult inpatient outliers. For adult outpatients, several subspecialties had lower odds than head and neck (OR ≤0.44). Increased comorbidities predicted outliers for adult inpatient care (OR = 1.5); sex, age, race, and ethnicity did not. Cochlear implant was the most common primary operation among adult and pediatric outpatient outliers. Greater subspecialty involvement and increasing age predicted pediatric outpatient outliers (OR = 8.0 and 1.1); younger age and female sex predicted pediatric inpatient outliers (OR = 0.8 and 3.5). Airway procedures dominated pediatric inpatient outliers. Conclusion: This is the first large-scale study of high-cost otolaryngology care across multiple subspecialties. Specific procedures and subspecialties and increased comorbidities predicted high-cost care. Contrary to previous studies, patient sex, race, and ethnicity did not.