Prediction models for postoperative uncontrolled chronic rhinosinusitis in daily practice.

Abstract

The European Position Paper on Rhinosinusitis and Nasal Polyps proposes an assessment of clinical control of chronic rhinosinusitis (CRS). However, there are limited data about the percentage of postoperative control, and no prediction models for uncontrolled CRS have been reported. The aim of the study was to develop prediction models for postoperative uncontrolled CRS. Retrospective case series. Patients (n = 136) who had undergone endoscopic sinus surgery at least 1 year prior to the study were recruited to assess the clinical control. Risk factors were determined by logistic models and presented as odds ratio (OR) with a 95% confidence interval. Receiver operating characteristics curves were constructed to set the cutoff points and create predictive models. Approximately 47.8% of patients had controlled, 22.1% partially controlled, and 30.1% uncontrolled CRS. Univariate regression models revealed the risk factors for uncontrolled CRS: tissue eosinophilia, blood eosinophilia, high computed tomography (CT) score, bilateral disease, asthma, and allergic rhinitis. Multiple regression models found tissue eosinophil ratio >0.206 (OR: 12.96, P = .001) or blood eosinophil ratio >0.025 (OR: 4.56, P = .003), Lund-Mackay (LM) score ≥ 15 (OR: 15.50, P < .001) and CT ethmoid (E) score ≥ maxillary (M) score (OR: 3.51, P = .037) were independent risk factors. We generated a pathological model (tissue eosinophil ratio and LM score) and a clinical model (blood eosinophil ratio, LM score and E ≥ M score) to categorize CRS into mild, moderate, and severe. This research provides simplified and efficient prediction models for uncontrolled CRS. It may help otolaryngologists to predict the prognosis before surgery in daily practice. 2b Laryngoscope, 128:2673-2680, 2018.