On chronic rhinosinusitis and the prevalence of fungal sinus disease: problems of diagnostic accuracy and a proposed classification of chronic rhinosinusitis

Abstract

There is considerable controversy in the diagnosis and classification of the type of inflammation that is attributed to various forms of chronic rhinosinusitis (CRS). Specimens obtained during surgical treatment of CRS have been invaluable resources for identifying the underlying inflammatory process. The classification of sinus inflammation is based on histopathologic examination of these surgical specimens. Accurate identification of the pathology and standardized reporting are invaluable for postsurgical treatment options and our understanding of CRS. In a large multispecialty referral hospital where multiple surgeons and pathologists are involved in clinical practice, the lack of standardization in specimen collection, specimen processing, and reporting introduce several variables that make it extremely difficult for retrospective analysis. This report focuses on consecutive endoscopic sinus surgical procedures performed by 4 different sinus surgeons over a period of 4 years in a Central Texas multispecialty hospital. This is an analysis of the reality of clinical practice without intervention. At the core of this analysis are pathology reporting practices for fungal sinus disease and the undesirable variables introduced by nonstandardized reporting. A practical classification of CRS based on pathology is proposed.

Introduction

Classification of chronic rhinosinusitis (CRS) has been an unresolved issue. There are several factors for the poor definition. As of 2012, we still do not have validated criteria to assess the presence and severity of sinusitis with confidence. Most symptoms used to categorize and classify sinus disease are not specific to the sinuses, reflecting the presence of nasal disease or lower respiratory tract disease. Even headaches and sinus pressure, the most commonly presumed associates of sinus disease, have not achieved relevance in our classification efforts because the sinuses do not seem to have pain sensory nerves. Similar limitations apply to imaging studies because comparing sinusitis symptom scores with computed tomographic (CT) scans to arrive at clinical relevance is not much better than random in predicting the presence or severity of CRS. Adding to the mentioned imperfections in our science, most research, analysis, and classification is conducted in academic tertiary centers, but most patients are treated by otolaryngologists in community hospitals, and results remain unreported and are not analyzed systematically. Many classification parameters used at academic centers are not part of clinical practice because of impracticality or inaccessibility to practitioners. Therefore, reports from academic centers may overestimate or underestimate distribution of disease. For any meaningful system of classification and calculations of prevalence and incidence, our research efforts need large standardized and comparable data sets. As the move to electronic health records gains momentum, usable research database can expand only if data acquisition through participation of community otolaryngologist and pathologists can be improved. To achieve this objective, a simpler, more feasible classification system based on objective parameters acquired through standardized methods using established practice patterns would have the easiest applicability and, therefore, the most impact. Such a system would use minimal subjective interpretation and rely in large part on histopathology. To demonstrate problems associated with data quality, this study includes a 4-year retrospective analysis of operative notes and surgical pathology reports of CRS submitted by 4 different surgeons and analyzed by 10 different pathologists. The focal point is selected as fungal sinus disease because this disease has relevant importance and high prevalence in the Southern United States [1]. Accuracy in diagnostic analysis and reporting for CRS and fungal sinus disease is crucial in identifying nuances that may have implications for our poorly evidenced classification attempts.