Disparities in Utilization of Jaw Surgery for Treatment of Sleep Apnea: A Nationwide Analysis

2016 
BACKGROUND: Maxillomandibular advancement has been shown to be one of the most effective operations for management of severe obstructive sleep apnea, yet pharyngeal surgery is more commonly performed. The goal of this study was to identify socioeconomic factors associated with this phenomenon. METHODS: Patients aged 14 or older with a primary hospital diagnosis of sleep apnea were identified using the National Inpatient Sample from 2005 to 2012. ICD9 codes were used to determine whether a pharyngeal or jaw procedure was performed. Patient demographics, comorbidities, and complications were compared. RESULTS: Among 6316 sleep surgeries, 5964 (94.4%) were pharyngeal and 352 (5.6%) were jaw procedures. Women were significantly more likely to receive jaw surgery than men (odds ratio [OR] = 1.68, P = 0.0007). African Americans (OR = 0.19, P < 0.0001), Hispanics (OR = 0.42, P = 0.0009), Asians (OR = 0.41, P = 0.0009), and other non-Caucasians (OR = 0.19, P = 0.0008) had a significantly lower odds of receiving jaw surgery than Caucasians. Patients falling into lower-income brackets (OR = 0.39 and 0.57, P = 0.02 and 0.04) and patients with Medicare compared with private or Health Maintenance Organization insurance (OR = 0.46, P = 0.008) also had significantly decreased odds of undergoing jaw surgery. Comorbidities were similar between surgical groups, and there were no significant differences in bleeding, infection, or cardiopulmonary complications. CONCLUSIONS: We identified no significant difference in complication rates between pharyngeal and jaw procedures. Nonetheless, African American, Hispanic, and Asian patients, in addition to lower-income patients and patients with Medicare, had a significantly lower odds of receiving jaw surgery. Awareness of these disparities may help guide efforts to improve patients' surgical options for sleep apnea.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    31
    References
    0
    Citations
    NaN
    KQI
    []