Introduction

Olfactory loss is considered a defining symptom of chronic rhinosinusitis (CRS) [1, 2], closely related to mental health (MH) and quality of life (QoL) [3–6], although underestimated by patients and overlooked by doctors [7]. Endoscopic sinus surgery (ESS) has been described as the “gold” standard for the treatment of CRS [8–12] in several studies, because it is successful in obtaining symptom resolution, and furthermore it may improve olfactory function and QoL. However, the proportions of patients reporting clinically significant QoL improvement have not been adequately addressed in the literature [13]. In addition, although there are studies describing pre-operative predictors for QoL outcomes [13–22], there is no data available to the relevance of pre-operatively olfactory status to post-operative QoL recovery. The use of standardized olfactory tests and validated specific for olfaction and mental health psychometric instruments provides new opportunities for assessing olfaction as a possible determinant for QoL after ESS. The clinical importance is that this data may enable otolaryngologists to better counsel their patients about the expected benefit for their QoL after ESS and how it is correlated to olfactory status. This may also support clinical decisions on the optimal management of patients suffering from chronic rhinosinusitis, facilitating case selections.

Accordingly, this study was aimed to prospectively evaluate the proportion of patients suffering from CRS, who experience clinically significant QoL improvement after ESS, related to their pre-operative characteristics, and to identify pre-operative clinical phenotypes that best predict surgical outcomes for QoL, focusing mainly on the impact of patients’ olfactory status based on QoL results.

Materials and methods

The prospective study was carried out in the University Department of Otorhinolaryngology of Democritus University of Thrace in Alexandroupolis/Greece, during a 4-year period from March 2008 to June 2012. One hundred eleven patients suffering from chronic rhinosinusitis (either with nasal polyps (NP): CRSwNP or without NP: CRSsNP) were studied. Forty-eight healthy subjects who visited our outpatient department for a routine health check-up were recruited as controls. Diagnosis was established based on history, clinical examination, nasal endoscopy, sinus computed tomography scanning, skin prick testing (SPT) for atopy, test of pulmonary function and olfactory testing. None of the subjects used oral or nasal corticosteroids and antibiotics 4 weeks prior to inclusion and oral antihistamines 1 week prior to SPT, according to our protocol for CRS patients. Patients with a history of previous sinus surgery, cystic fibrosis or any malignancy were excluded. All patients fulfilled the criteria of CRS according to the EPOS [1] guidelines. Only patients with resistant CRS to maximal medical therapy (antibiotics, oral and nasal steroids) over a 6-month period were subjected to surgery. All subjects underwent a brief psychiatric interview to exclude those with pre-existing major psychiatric disorder. ESS was conducted by a single surgeon following Messerklinger [23] technique according to the extent of the disease, with the use of microdebrider device. Septoplasty and inferior nasal turbinates (INT) volume reduction—by submucosal radiofrequency tissue ablation—were implemented whenever significant septal deviation or INT enlargement were observed. The frequency of each treatment protocol performed is shown in Table 1. Post-operative care included multiple visits at the outpatient clinic where nasal endoscopy and toilette as needed were performed. All patients were asked to rinse the nose with sodium chloride solution and nasal steroids. The study protocol was approved by the local Institutional Review Board. All subjects signed informed consent. The study was performed in accordance with the Declaration of Helsinki/Hong Kong.

Table 1 Frequency of treatment procedures performed
Full size table

Olfactory function of all patients and healthy controls was quantitatively evaluated using “Sniffin’ sticks” test package (Burghart, Wedel, Germany) pre-treatment and 12 months after treatment, as it was an adequate time period for the healing process to be completed [24]. At these time points all subjects also filled in three validated, widely used questionnaires specific for olfaction-associated QoL (Questionnaire of Olfactory Deficits: QOD), for assessing psychology (Beck Depression Inventory: BDI) and a general-health survey (Short Form: SF-36).

“Sniffin’ Sticks” test battery for olfaction included specific tests for odor threshold (OT), odor discrimination (OD), and odor identification (OI) [25]. Results of each one of the three tests were combined to a “Threshold Discrimination Identification (TDI) score” [25]. TDI score ranges from 0 to 48 (values of 15 or less represent anosmia, values between 16 and 34.5 represent hyposmia and values over 34.5 represent normosmia for the mild climate conditions in Greece) [26].

QOD represents an olfaction-specific questionnaire for QoL, translated and validated for the Greek population [27]. It consists of 25 4-scale statements (17 “negative”, 2 “positive”, 6 “socially desired”) with a maximum score of 57 points [27]. High scores indicate a strong impairment in QoL. BDI was used for assessing QoL of patients detecting specifically their psychological profile [28]. BDI consists of 21 self-reporting items graded from 0 to 3, and a higher score indicates higher level of depression. Finally SF-36 is a widely used general-health related survey that assesses QoL in eight domains covering both physical and mental health from the patient’s point of view with chronic diseases [29, 30]. Scores range from 0 to 100, with a higher score representing better functioning [29, 30].

Statistical analysis of the data was performed using the Statistical Package for the Social Sciences (SPSS), version 19.0 (IBM). The normality of quantitative variables was ascertained with Kolmogorov–Smirnov test. The Chi-square test and Student’s t test were used to assess differences of demographic characteristics between patients and controls. The scores of olfactory function and all QoL questionnaires were expressed as the mean and standard deviation (SD). Differences in the scores of olfactory function and QoL questionnaires between (1) patients and controls and (2) pre- and post-treatment were assessed by Student’s t test for independent and related samples, respectively. Since the distribution of BDI score was skewed, the statistical analysis was performed on the log-transformed scores. The Chi-square test was used to evaluate any potential association between the incidence of clinically significant improvement for each QoL questionnaire with patients’ demographics and clinical characteristics. Multivariate stepwise logistic regression analysis was constructed to explore the independent effect of patient’s characteristics on clinically significant improvement. Odd ratios (OR) and 95 % confidence intervals (CI) were estimated as the measure of association between clinically significant improvement and all potential predictors. All tests were two tailed and statistical significance was considered for p values of less than 0.05.

Results

One hundred eleven patients [59 (53.2 %) males, mean age: 44.74 ± 16.21 years] and 48 healthy subjects [25 (52.1 %) males, mean age: 40.15 ± 15.50 years] participated. Demographics and disease characteristics of all patients and controls are shown in detail in Table 2. The table reveals no statistically significant differences related to gender (p = 0.901), age (p = 0.100), socio-economic status (p = 0.376) and smoking habits (p = 0.156) between patients and controls. Quantitative assessment of subjects’ olfactory function with the use of Sniffin’ sticks test revealed patients to present statistically significant lower TDI score as compared to controls [23.15 ± 12.75 (range 2–38) vs. 38.88 ± 1.38 (range 37–41), p < 0.001]. Based on this score, 84 (75.7 %) patients presented olfactory dysfunction; 36 (32.4 %) were anosmics and 48 (43.3 %) were hyposmics. The duration of olfactory deficit of these patients ranged from 2 to 30 years, with a median duration of 8 years (interquartile range 5–15 years). All the subjects of the control group were normosmics. Among patients, higher frequency of anosmia was associated with female gender (44.2 vs. 22.0 %, p = 0.013; OR = 2.81, 95 %CI = 1.23–6.40), ages older than 55 years (68.2 vs. 9.0 %, p < 0.001; OR = 21.79, 95 %CI = 7.61–62.35), lower socio-economic status (75.0 vs. 29.1 % in medium and high social status, p = 0.008; OR = 7.30, 95 %CI = 1.39–38.23), non-smoking habits (46.0 vs. 14.6 %, p < 0.001; OR = 5.00, 95 %CI = 1.95–12.82), asthma history (88.2 vs. 22.3 %, p < 0.001; OR = 26.07, 95 %CI = 5.52–123.23), nasal polyps (51.9 vs. 14.0 %, p < 0.001; OR = 6.60, 95 %CI = 2.63–16.53) and aspirin intolerance (75.0 vs. 27.3 %, p = 0.001; OR = 8.00, 95 %CI = 2.01–31.78).

Table 2 Demographics and disease characteristics of patients and healthy controls
Full size table

The scores of olfactory function and QoL questionnaires of patients and controls are presented in Table 3. Patients’ pre-treatment scores in all questionnaires were statistically significantly worse (all p < 0.05) compared to healthy controls. Throughout the 12-month follow-up time, it was observed a significant improvement of (1) all indices of olfactory function: odor threshold (OT) by 68.3 % (p < 0.001), odor discrimination (OD) by 33.2 % (p < 0.001), odor identification (OI) by 36.7 % (p < 0.001) and TDI score by 41.8 % (p < 0.001), and (2) the scores of all QoL questionnaires: QOD by −50.9 % (p < 0.001), QOD-NS by −85.2 % (p < 0.001), QOD-PS by 22.6 (p < 0.001), QOD-SD by −53.0 % (p < 0.001), SF-36 by 25.2 % (p < 0.001) and BDI by −31.5 % (p < 0.001) (Table 3). Clinically significant improvement was defined for each QoL questionnaire as a change of ≥1/2 SD of the pre-treatment score [31]. Using this criterion, improvement was defined as a decrease of 4.91 points for QOD and 4.33 for BDI and an increase of 7.65 points for SF-36. Among the entire cohort, clinically significant improvement was observed in 63 (56.8 %) patients for QOD, 72 (64.9 %) patients for SF-36 and 55 (49.5 %) patients for BDI. The incidence of clinically significant improvement of QOD, SF-36 and BDI in relation to patients’ and disease characteristics was studied next (Tables 4, 5, and 6). In univariate statistical analysis, it was found that the likelihood of clinically significant improvement of (1) QOD was higher for patients with female gender (67.3 vs. 47.5 %, p = 0.035), medium or high socio-economic status (68.6 % and 50.0 vs. 25.0 %, p = 0.028), nasal polyps (66.7 vs. 47.4 %, p = 0.040), hyposmia or anosmia (64.6 % and 75.0 vs. 18.5 %, p < 0.001) and lower for smokers (45.8 vs. 65.1 %, p = 0.043), (2) SF-36 was higher for patients with nasal polyps (85.2 vs. 45.6 %, p < 0.001) and hyposmia or anosmia (60.4 % and 86.1 vs 44.4 %, p = 0.002), and (3) BDI was higher for patients with nasal polyps (63.0 vs. 36.8 %, p = 0.006) and hyposmia or anosmia (54.2 % and 61.1 vs. 25.9 %, p = 0.015).

Table 3 Scores of olfactory function and quality of life (QoL) questionnaires pre-treatment and 12 months after treatment
Full size table
Table 4 Clinically significant improvement of QOD (Questionnaire of Olfactory Deficits) questionnaire in relation to patients’ and disease characteristics
Full size table
Table 5 Clinically significant improvement of SF-36 (Short Form-36) questionnaire in relation to patients’ and disease characteristics
Full size table
Table 6 Clinically significant improvement of BDI (Beck Depression Inventory) in relation to patients’ and disease characteristics
Full size table

Multivariate logistic regression analysis revealed that the following independent predictors were significantly associated with higher likelihood of clinically significant improvement: anosmia for all three questionnaires (QOD: aOR = 12.39, 95 %CI = 3.08–49.94, p < 0.001; SF-36: aOR = 4.92, 95 %CI = 1.11–21.80, p = 0.036; BDI: aOR = 7.68, 95 %CI = 2.16–27.40, p = 0.002), hyposmia for QOD and BDI (QOD: aOR = 8.55, 95 %CI = 2.65–27.52, p < 0.001; BDI: aOR = 3.97, 95 %CI = 1.31–12.02, p = 0.015), nasal polyps for all three questionnaires (QOD: aOR = 2.54, 95 %CI = 1.15–5.57, p = 0.021; SF-36: aOR = 6.14, 95 %CI = 2.03–18.56, p = 0.013; BDI: aOR = 3.31, 95 %CI = 1.52–7.25, p = 0.046) and medium or high social status for QOD (aOR = 8.92, 95 %CI = 1.43–55.70, p = 0.019); while smoking was an independent predictor significantly associated with lower likelihood of clinically significant improvement of BDI (aOR = 0.41, 95 % CI = 0.17–0.98, p = 0.046) (Table 7).

Table 7 Results of multivariate logistic regression analysis between predictor variables and clinically significant improvement of QOD, SF-36 and BDI questionnaires in patients suffering from chronic rhinosinusitis, expressed as adjusted odds ratios (aOR) with their 95 % confidence intervals (CI)
Full size table

Discussion

Chronic rhinosinusitis is already known as a common chronic sinonasal disease that adversely affects the quality of life of most involved patients [13]. Although the main concern of any medical treatment for these patients should be the improvement of QoL, there is lack of studies analyzing in their outcomes clearly defined proportions of patients who experienced clinically significant improvement on their daily lives and not just providing general conclusions for patients’ QoL. This study provides clinically relevant information for the individual patient and doctor, measuring the proportion of patients who experience a clinically significant improvement on QoL, related to their pre-operative clinical and demographic characteristics. In addition, although there are studies that highlight various demographic and clinical data and comorbidities that explain QoL outcomes after ESS [13–22], the importance of baseline olfactory status on QoL prediction is still worth pursuing and needs to be further explored, as olfactory loss is a major symptom in CRS directly affecting patients’ QoL [1–6]. To achieve this, additionally we used a specific for olfaction-associated QoL psychometric questionnaire, as well as quantitative smell tests for the assessment of olfactory performance, providing, thus, objective data for the olfactory function and not self-reporting symptoms.

In our study we observed that olfactory function in CRS patients after ESS recovers and this is expressed by a significant improvement in the TDI and all separate tests (OT, OD, OI) score. Olfactory rehabilitation is followed by patients’ QoL recovery as well, within 12 months post-operatively, that is expressed by a significant improvement in the results of all questionnaires used for the assessment of patients’ QoL. These results are in accordance with previous reports, in which positive effects of ESS on patients’ QoL [10–20] were found. A basic limitation of former studies was the fact that the conclusions were difficult to be expressed into meaningful terms that patients can understand and doctors can integrate into shared decision making. This is the reason why we explored clinically significant improvement for patients’ QoL, as defined by Norman et al. [31], adding new knowledge beyond what is already published on this topic in the literature. This definition is also clinically important as it permits us to build models to predict treatment outcomes. According to this criterion we observed that although all patients’ QoL improved significantly after ESS, a 49.5–64.9 % of patients with CRS experienced clinically significant improvement according to the psychometric scales for QoL assessment used (QOD: 56.8 %, BDI: 49.5 %, SF-36: 64.9 %).

Additionally, it is clinically important to mention that we for the first time identified pre-operative clinical phenotypes that best predict surgical outcomes for clinically significant QoL recovery, focusing mainly on the impact of patients’ baseline olfactory function on QoL results. After univariate screening of all possible clinical and demographic predictors we found that certain clinical phenotypes, such as CRS associated with anosmia or hyposmia, CRS associated with nasal polyps, CRS associated with non-smoking habits, and CRS in females and in medium or high socio-economic status patients were significant predictors of QoL outcomes; however, among these phenotypes, baseline olfactory dysfunction, irrespective of its duration, and the presence of nasal polyps as well, appears to be independent clinical predictors when multiple risk factors were accounted for in the predictive model. We observed that hyposmic patients were 1.54 up to 8.55 times more likely to experience clinically significant improvement on QoL according to the psychometric scales used (QOD: 8.55, SF-36:1.54, BDI: 3.97), whereas the results for anosmics were even higher (SF-36:4.92, BDI: 7.68, QOD: 12.39). Also CRSwNP patients were better candidates for presenting clinically significant improvement for QoL after ESS, presenting 2.5- up to 6-fold more increase on QoL results on each occasion compared to CRSsNP patients (QOD: 2.54, SF-36: 6.14, BDI: 3.31). In addition, in our study certain potential determinants of outcomes of ESS on patients’ QoL, including asthma [13], ASA intolerance [13, 15, 19], smoking [16, 21], nasal polyps [17], prior sinus surgery [13] as well as demographic factors including gender [13, 18, 20], socio-economic status [13], have been studied and described as important potential predictors of outcomes, although with occasionally conflicting results. However, to the best of our knowledge, this is the first study to address olfactory dysfunction as a significant predictor for QoL and psychological outcomes, using specific psychometric questionnaires as well as quantitative smell tests for the assessment of olfactory performance, providing, thus, objective data for olfactory function and not self-reporting symptoms.

In conclusion our results suggest that certain clinical phenotypes, such as CRS associated with anosmia and hyposmia, or with nasal polyps, or with non-smoking habits, as well as CRS in females and in medium or high socio-economic status patients, are significantly associated to better QoL outcomes, 12 months post-operatively, after ESS. However, multivariate logistic regression analysis revealed that olfactory dysfunction (anosmia or hyposmia) and the presence of nasal polyps were independent pre-operative predictors associated with higher likelihood of clinically significant improvement on patients’ QoL. This means that olfactory testing in CRS patients prior to surgery is of great clinical importance for appropriate case selection and consultation for treatment outcomes related to QoL results. This knowledge may further improve outcomes and patients’ QoL.