Methods
A series of 79 consecutive Caucasian patients undergoing rhinoplasty was studied. Nasal airway was evaluated before surgery in each patient of this group as part of the preoperative assessment.
This series consisted of 53 women and 26 men, with ages between 17 and 68 years old (mean 36.2 y/o). Sixty-five were primary cases and 14 were revision rhinoplasty cases. All the primary rhinoplasty cases were looking for an aesthetic improvement of the nose, with 29 also seeking a functional improvement of the nasal airway. For the 14 patients undergoing revision rhinoplasty, there were several different aesthetic indications for the revision surgery, with no functional complaints in none of the patients in this subgroup.
In each patient the NOSE score (Nasal Obstruction Symptom Evaluation) (24) was obtained and peak nasal inspiratory flow (PNIF) (25) was measured before undergoing surgery. The NOSE score was chosen as patient-reported assessment of nasal obstruction symptoms as it is a standardized and validated disease-specific quality of life instrument for measuring nasal obstruction (24). A version validated for the Portuguese language (26) was used in this study. The PNIF was obtained in each patient according to the standard rules for measuring PNIF (25). PNIF was chosen as objective measurement of the nasal airflow as its results not only measure nasal airflow but also correlate with nasal resistance (27). Furthermore, this method, together with rhinomanometry, has been found to strongly reflect both presence and severity of nasal airway obstruction (28).
Three groups of patients were created according to the symptoms of nasal obstruction reported by the patients: patients with NOSE score equal to or less than 35 (N=24), patients with NOSE score between 36 and 64 (N=26) and patients with a NOSE score equal to or higher than 65 (N=29).
Three other groups of patients were created according to the level of nasal airflow: patients with PNIF equal to or less than 70 l/min (N=20), patients with PNIF between 71 l/Min and 109 l/min (N=39) and patients with PNIF equal to or higher than 110 l/min (N=20).
Statistical analysis was made for the total population of this study and for each of the groups created based on NOSE score and on PNIF value. Mean and standard deviations (SD) were obtained for the PNIF and NOSE scores. A multivariable linear regression model was performed in order to investigate the association between the NOSE scores and the PNIF values after adjusting for age and gender of the individuals. The estimated correlation coefficient was obtained for the PNIF score and NOSE value to check the strength of the association between these measurements. The nonparametric Spearman’s rank correlation coefficient was employed. A significance level of α = 0.05 was used throughout the analysis. The statistical analysis was performed using the R statistical software (29), employing the RStudio environment (30). The graphs were obtained using the R package ggplot2 (31). The strength of the correlation (r ) was valued as very week forr <0.19, week for r between 0.20 and 0.39, moderate for r between 0.40 and 0.59, strong for r between 0.60 and 0.79 and very strong for r between 0.80 and 1 (32).