This work aims to describe the translation, cross-cultural adaptation, and validation process of the Polish version of SNOT-22. Reliability, validity, and responsiveness are specific characteristics of each context, so an instrument that has demonstrated satisfactory psychometric properties in a specific population is not necessarily appropriate for others. Thus, the validation of the Polish SNOT-22 will allow more specific treatment for CRS patients, as well as its generalization in the scientific community, and comparison between different countries. Among numerous disease-specific sinonasal outcome questionnaires the SNOT-22 questionnaire showed reliability, validity, responsiveness, and ease of use [33]. This assumes particular importance in the era of the introduction of biological treatment in CRSwNP patients with or without asthma (omalizumab, benralizumab, dupilumab). At present, the SNOT22 is widely used to assess the quality of life of these patients and it will be possible to compare results between patients of different nationalities, provided validated tools are used.

The number of participants in international SNOT-22 validation studies varied greatly. The largest group was that in the original validation study by Hopkins et al. [14] a prospective cohort study collecting data on 3128 adult patients undergoing sinonasal surgery in 87 NHS hospitals in England and Wales. In some studies on validation of the SNOT-22 the total number of subjects was under one hundred [34, 35]. In other studies, this number exceeded two hundred participants, for instance, 206 in Brazilian Portuguese validation [19], 341 in Italian [22], and 422 in French [20] studies. A total number of 148 enrolled subjects places our study at average in terms of sample size.

The participants in our study were all diagnosed with chronic rhinosinusitis with nasal polyps (CRSwNP). In other studies, the study sample included patients with CRS with and without nasal polyps.

Validation studies determine the accuracy, dependability, and consistency of a tool. The Polish SNOT-22 exhibited satisfactory psychometric properties. The results demonstrate that the Polish version of the SNOT-22 is a reliable outcome measure according to 2 reliability tests: internal consistency and test-retest reproducibility. The Cronbach’s alpha coefficient α = 0.89 showed similar values in the Polish SNOT-22 as compared with the original English version α = 0.91 [14] and compared with translations of the questionnaire to other languages. [15, 23,24,25,26,27, 36] Test-retest reproducibility of the Polish SNOT-22 was determined by the Intraclass Correlation Coefficient (ICC). The ICC value obtained for the test-retest resulting in 0.977 (95% confidence interval, lower band: 0.963; upper band: 0.985) for the Polish version of SNOT-22 indicates excellent reliability and stability over time, given the ICC can take a value from 0 to 1, with 0 indicating no agreement and 1 indicating perfect agreement. This result is in line with those obtained in other international SNOT-22 validation studies [17, 18].

The ability of the questionnaire to distinguish the disease-affected group was tested by comparison with asymptomatic subjects. Mean total SNOT-22 scores reported in different SNOT-22 translations range from 25.6 ± 13.3 (mean ± SD) in Persian translation [37], 29.7 (range 7–67) in Danish translation [38] to 62.4 ± 7.9 (mean ± SD) in Arabic [35] and 62.4 ± 25.3 (mean ± SD) in Brazilian Portuguese translation [23]. Our clinical sample of CRSwNP patients had a mean SNOT-22 score of 32.08 ± 8.34 (mean ± SD) which is in the lower range of the results reported in other studies. Several factors result in the discrepancy between the mean total scores reported in the literature, such as the demographic characteristics of the groups, or the patient recruitment methods [25]. It should also be taken into account that more than half of SNOT-22 domains are non-rhinology specific, therefore the scores can be influenced by the presence of various related conditions such, but not exclusively, as asthma, allergic rhinitis, chronic throat symptoms, Eustachian tube dysfunction, ear pathologies, sleep disturbances, depression, and also socioeconomic factors affecting QoL [39].

In 2020 Gallo et al. conducted a study aimed at verifying in an Italian CRS population whether SNOT-22 could assist physicians in predicting surgical outcomes, improving the shared decision-making process, and ameliorating patients’ understanding of their QoL expectations after treatment [40]. Based on the baseline SNOT-22 score, the cohort of patients was divided into 10 groups. The primary outcomes included measurement of the percentage of patients receiving a minimal clinically important difference (MCID) and the percentage of relative improvement (RI) after surgical treatment. Based on the results of the study, the mean percentage of achieving an MCID in groups 3–10 (patients with mean SNOT-22 scores of 30–110 points) is 91.6% with an average of 56.8% of RI. Contrarily, the mean percentage of achieving an MCID in groups 1–2 (patients SNOT-22 mean scores of 10–29) is 44.2% with an average of 38.9% of RI. A similar conclusion was drawn by Farhood et al. who conducted a cross-sectional study and systematic review on SNOT-22 in a control population. The authors underlined that in CRS patients with preoperative scores < 20, FESS is unlikely to yield clinical improvement [41]. In light of these data, the classification of patients with CRSwNP for FESS in our study was justified, as they were likely to benefit from surgery, with the SNOT-22 median score of 32 points standing a greater than 75% chance of achieving an MCID and on average obtain a 45% relative improvement in their QoL after ESS.

When validating a tool for use in the clinical setting It is necessary to establish its ‘normal’ value within the general population to set a reference point to identify those who may benefit from treatment. Gillet et al. [42] conducted a pilot study of the SNOT-22 score in adults with no sinonasal disease to determine a normal SNOT-22 score. Based on the results of a total of 116 subjects they concluded that in a clinical situation, a SNOT 22 score of 7 should be used as a guide for ‘‘normal’’, and that care should be taken when suggesting treatment on patients with a score below this level. The group of healthy controls in our study had a median total SNOT-22 score of 5 (range 0–20), and although slightly lower, it is in accordance with the result of the aforementioned study. In other studies validating the SNOT-22 the results in the control group were much higher. We found only one study with the controls’ result lower than the suggested reference point of 7 and it was the study by de los Santos et al. where the median SNOT-22 value for controls was 2 points [26]. In the remaining studies, the mean results in controls ranged from 8.7 ± 8.1 [mean ± SD] to 19.5 ± 13.1 [mean ± SD] [17, 18, 21,22,23,24, 27, 36].

It should be stressed at this point that despite the seemingly promising prognostic value of the SNOT-22, PROM questionnaires should not be used as a replacement for nasal endoscopy and/or CT sinus scan for diagnosis and assessment of CRSwNP [39]. Therefore, although the baseline SNOT-22 score and the chance of achieving the MCID are not intended to be used as an absolute threshold for eligibility for surgery, results reported in international studies suggest that a patient with a low preoperative score might be less likely to benefit from surgery and caution should be paid when operating on patients with a score < 10 [40].

When conducting the literature research on the SNOT-22 adaptation and validation in different languages, we found only a few studies that undertook the subject of diagnostic reliability of the SNOT-22 as determined by the Receiver Operating Characteristic (ROC) curve with Area Under Curve (AUC). In the validation study of the Turkish SNOT-22 Cakir et al. reported that with the cut-off value of 33.5, the sensitivity and specificity of the Turkish version of the SNOT‐22 were 54.5% and 75.9%, respectively (95% CI, (AUC): 0.69, range 0.624–0.756, p = 0.000) [36]. The Arabic validation study by Alanazy et al. in the ROC curve analysis of discriminant validity with the AUC of 0.87 at a cut-off threshold of 18.5, sensitivity was 77%, and specificity was 84% [35]. In the study of Vaitkus et al. the ROC test indicated that in the Lithuanian version of the SNOT-22, the total score of 29 was the optimal score distinguishing between patients and healthy controls The sensitivity of the Lithuanian version of SNOT-22 was 91.7%, and specificity 82.6% [25]. Adnane et al. conducted psychometric validation of a Moroccan version of the SNOT-22 in which they assessed the discriminant validity using a receiver operating characteristic (ROC) curve [17]. The determined area under the curve (AUC) was 0.994. The cut-off point was not reported in this study. In our study, the AUC equaled 0.997 with p < 0.001 and this result was the highest among the aforementioned studies. The cut-off threshold of 16 (sensitivity = 0.981, specificity = 0.995) was the closest to that in the study by Alanazy et al. [35].

Psychometric properties such as sensitivity and specificity are common measures used to evaluate the quality of a screening test. These psychometric properties are unbiased if the screening test results are compared with a gold standard measure. In our studies two objective methods of chronic rhinosinusitis were applied: endoscopic examination of the nasal cavities graded according to the Lund-Kennedy scale and computed tomography (CT) scans graded according to the Lund-Mackey scale.

Nasal endoscopy plays a main role in recognizing anatomical structural variations and mucosal changes of the middle meatus and osteomeatal complex. Adnane et al. reported a good statistically significant correlation between the SNOT-22 scores and the Lund-Kennedy endoscopic scores (r = 0.71) [17]. In our study, the correlation was also statistically significant, but moderate (r = 0.334). Other studies we found did not report a correlation between these two measures.

A computed tomography scan of the paranasal sinuses is regarded as the gold standard diagnostic radiological tool for CRS. Although it is widely used in the assessment of chronic rhinosinusitis [29] disagreement exists about the relationship between Lund-Mackay CT scores and quality-of-life outcome measures. In 2018 Brooks et al. [43]. reported that the preoperative Lund-Mackay scale scores were significantly associated with preoperative SNOT-22 scores (p < 0.01). Our study confirms this finding with a statistically significant, strong positive correlation between the Polish SNOT-22 and the Lund Mackey score (r = 0.469; p < 0.01). This result should, however, be interpreted with caution, as our patient sample was limited to those diagnosed with CRSwNP. In a study conducted by Bradley and Kountakis [44] in 2005, the results showed that the severity of rhinosinusitis based on a CT scan before surgery was not related to the severity of symptoms based on the SNOT-22 questionnaire after ESS surgery. Also, a CT scan could not predict the improvement of symptoms after FESS. Similarly, most of the available studies report no significant correlation between the SNOT-22 and the Lund-Mackay scores [18, 20].

The following study limitations are worth mentioning. First, the study population is limited to a sample selected for FESS. CRS patients without planned surgery were not included. However, these limitations could mostly affect the scores rather than the psychometric properties cc of the adapted version. It has been well documented that the SNOT-22 questionnaire is capable of being used as a before-after questionnaire in patients undergoing sinonasal surgeries. Our validation study is based on preoperative SNOT-22 results only. Therefore, our study does not provide information on SNOT-22 responsiveness that is the ability of the questionnaire to detect clinical changes. Future studies should focus on documenting the Polish SNOT-22’s ability to detect changes over time by comparing the final scores before and after the surgical intervention and other forms of CRS treatment. Another notion for future exploration might be an interaction of the SNOT-22 score with allergy, asthma, or smoking We did not show this interaction in the presented research, which may have resulted in a certain degree of a response bias.