2018 Section 5 - Rhinology and Allergic Disorders

Otolaryngology–Head and Neck Surgery 153(1)

hypersensitivity, nasal polyposis, characteristic bony expan- sion on radiographic imaging, eosinophilic mucin, and pres- ence of noninvasive fungi. 8 The literature has failed to show consistent evidence linking disease severity and markers of socioeconomic status in CRS patients. Kilty et al 9 found that socioeconomic status was a negative modifying factor for CRS disease severity measured by sinonasal assessment questionnaire (SNAQ-11). Different racial subgroups of all CRS patients have been found to have different insurance status, access to health care, delay of care due to financial restrictions, and perceptions of disease sever- ity, mostly with disadvantages noted among the African American and Hispanic populations. 10 Among the pediatric CRS patient population, Smith et al 11 found that within an urban academic setting, children with CRS are more likely to be white and privately insured compared with the general pop- ulation. AFRS, in contrast, has demonstrated an association between disease severity and factors of lower socioeconomic status. In past studies, AFRS has been noted to be more common with a later presentation in young African American males. 12 Patients with AFRS tend to present at a younger age, with more severe disease often indicated by osseous erosion, and are more likely to be uninsured or Medicaid recipients. 13,14 Miller et al 15 found that in a North Carolina tertiary academic center, AFRS severity was higher among patients living in poorer conditions, with lower income and less health care access, and in rural communities. The present investigation was designed to systematically compare the differences in disease severity, demographic factors, and indicators of socioeconomic status between patients with CRS (further stratified into CRSwNP and CRSsNP) and those with AFRS. We hypothesized that patients with AFRS would demonstrate higher severity of disease and would be associated with lower socioeconomic status and less access to primary medical care than those with other CRS subtypes. Methods Study Design and Inclusion and Exclusion Criteria Approval was obtained from the University of North Carolina Institutional Review Board. A retrospective chart review was conducted to extract CRS (both CRSwNP and CRSsNP) patients who had undergone functional endo- scopic sinus surgery (FESS) between January 2000 and April 2013, from the same source population as used by Miller et al, 15 in a tertiary academic rhinology practice. Both CRS and AFRS patients were extracted by simple random sampling using Nth-Name Select. Ultimately, this review yielded 93 CRS patients, who were combined with the previously extracted AFRS cohort of 93 patients for a total of 186 patients. Patients were diagnosed with AFRS by the previously documented Bent and Kuhn criteria. 8 It should be noted that a definitive diagnosis of AFRS was used when 3 of 5 Bent and Kuhn criteria were met, as the work of Melroy et al 16 revealed a significant delay between the time of initial presentation and the diagnosis of AFRS.

Similar to that study, many of the current study’s AFRS cohort presented at various stages of evolution of their dis- ease process, and for this reason patients with at least 3 of 5 Bent and Kuhn criteria were included as the diagnosis of AFRS was satisfied. 8 All preoperative computed tomogra- phy (CT) scans of patients with suspected AFRS were reviewed by the senior authors (C.S.E and S.M.M), and those patients with definitive CT characteristics of AFRS were retained in the study. 17 Patients with CRS were diagnosed based on conventional criteria mentioned previously. 2 Preoperative radiographic imaging was graded by the senior authors using the Lund- MacKay scoring system. 18 For both the CRS and AFRS groups, patients with a history or current diagnosis of head or neck malignancy, cystic fibrosis, primary ciliary dyskinesia, or immune compromise and those with residence outside of North Carolina were excluded from the study. Data Extraction Self-reported demographic data, such as age, gender, race, ethnicity, insurance status, and county of residence, were collected from medical records. Insurance status was further categorized into uninsured or subsidized insurance (eg, Medicaid, self-pay), Medicare, military insurance, and pri- vate insurance. County-specific demographic information, indicators of socioeconomic status, and measures of health care access were obtained from the North Carolina Database, provided by the North Carolina Data Center. The National Center for Health Statistics further divided the 100 counties in North Carolina into 6 Health Services Areas (HSAs) based on population, health resources, and geographic features ( Figure 1 ). HAS-based data were obtained in a similar fash- ion. These variables included income, rural population, occu- pied housing units that were more than 30 years old, occupied housing units designated overcrowded at last inspection, and access to primary care physician (PCP). Markers of disease severity were extracted from medical records. Variables associated with disease severity included comorbid conditions, such as asthma and allergic rhinitis, as well as diagnostic and affiliated tests (number of CT scans, skin or in vitro allergen testing, and total quantitative serum immunoglobulin E [IgE] level). Parameters of treatment modalities were also included, such as treatment with subcu- taneous immunotherapy and/or budesonide-impregnated sinonasal irrigations as well as the number of FESS proce- dures before and after a visit with a tertiary center rhinolo- gist. Other parameters of disease severity characteristics included the presence of polyps, age at diagnosis, and pres- ence of unilateral disease. Last, preoperative CT images and corresponding radiology reports by the University of North Carolina Department of Radiology were reviewed and graded by 2 fellowship-trained rhinologists (C.S.E. and S.M.M) based on the Lund-Mackay CT scoring system. 18 Statistical Analysis Univariate analysis was used to assess baseline characteris- tics of the CRS patients. Bivariate analysis was used to

88