Sinonasal Quality of Life in Children With Cystic Fibrosis

Importance  Sinusitis is the most common otolaryngologic complaint in children with cystic fibrosis (CF). However, basic knowledge about the effect of sinusitis on these children is lacking.

Objective  To evaluate the incidence and quality-of-life impact of chronic rhinosinusitis (CRS) in an unbiased cohort of children with CF.

Design, Setting, and Participants  Survey study of consecutive pediatric patients with CF presenting for routine quarterly evaluation at a tertiary CF clinic at an academic pediatric hospital. Surveys were completed during the period from December 2012 to January 2013.

Main Outcomes and Measures  Surveys designed to assess major criteria for diagnosis of CRS and a validated pediatric sinonasal quality-of-life instrument, the Sinonasal-5 (SN-5). Statistical analysis was performed to evaluate association between demographic features and survey responses.

Results  Of the 102 consecutive eligible patients, 47 children (46%) aged 2 to 20 years (mean [SD] age, 12.9 [5.6] years; 24 [51%] female) completed the surveys. Depending on the exact diagnostic criteria used, 5 (11%) to 18 (38%) of children with CF had CRS. Mean domain (2.16; 95% CI, 2.02-2.30) and overall visual-analog scale (8.26; 95% CI, 8.01-8.51) scores on the SN-5 were consistent with minimal effect on quality of life and comparable to historical posttreatment scores. Mean scores on nasal obstruction (3.07; 95% CI, 2.80-3.34) and sinusitis (2.68;; 95% CI, 2.42-2.94) were the most affected domains, whereas allergy (1.83; 95% CI, 1.65-2.01), emotional disturbance (1.76; 95% CI, 1.56-1.96), and activity restriction (1.43; 95% CI, 1.31-1.57) were minimally affected. Children with a diagnosis of CRS had higher mean SN-5 scores (2.60; 95% CI, 2.31-2.89 vs 2.05; 95% CI, 1.90-2.20; difference of 0.55; 95% CI, 0.29-0.80). Twenty-five patients (53%) had undergone some treatment for sinusitis. There was no association between SN-5 score and CRS treatment history.

Conclusions and Relevance  In this study, the incidence of symptomatic CRS was high, but quality-of-life impact was relatively low among children with CF. Use of standardized assessment scales, including consensus diagnostic criteria and validated quality-of-life surveys, may be helpful to guide referral and management decisions.

Cystic fibrosis (CF) is well known to be associated with the physical findings of chronic sinusitis. Mucoperiosteal thickening is almost invariably seen on computed tomography, and nasal polyposis has been reported in up to 48% of children with CF.^1,2 Despite the ubiquity of these findings, the true incidence of symptomatic chronic rhinosinusitis (CRS) and its impact on quality of life (QOL) in children with CF are not well described. In 1 study, 10% of children studied reported sinonasal symptoms,³ whereas others have reported an incidence of symptoms in up to 62% of children.⁴ More recently, it was found that a cohort of children with CF had scores on the Sino-Nasal Outcome Test (SNOT-16), a validated age-nonspecific instrument of sinonasal QOL, that indicated slightly impaired QOL compared with asymptomatic adults but better QOL compared with adults with CRS.⁵ This instrument, however, does not assess nasal obstruction, which is the most common complaint among children with CF,⁶ and has not been validated in children.

The Sinonasal-5 (SN-5) is a validated pediatric sinonasal QOL instrument.⁷ Using parental report over 5 domains (sinus infections, nasal obstruction, allergy symptoms, emotional distress, and activity limitations), it exhibits good test-retest reliability and adequate correlation with external constructs for children as young as 2 years. It has been used to measure the treatment effect after both medical and surgical interventions for CRS in children.^7,8

A primary goal of medical and surgical treatment of CRS in children with CF is improvement in QOL. The lack of association suggested in the literature, however, between the physical findings and clinical symptoms of CRS makes determination of when to treat difficult. Recent clinical guidelines have provided some clarity and consistency in the definition of CRS,^9-11 although consensus on definitions for pediatric CRS remain elusive.

The 1999 clinical practice guidelines proposed at a single institution defined CRS as more than 3 months of 1 or more of the following symptoms: purulent nasal discharge, nasal obstruction, or head pain, together with a positive result on a computed tomographic scan.⁹ The 2007 American Rhinologic Society (ARS) guidelines, which did not differentiate between adults and children, require more than 12 weeks of 2 or more of the following symptoms: mucopurulent nasal discharge, nasal obstruction, facial pain/pressure, or decreased sense of smell.¹⁰ In addition, endoscopic or imaging confirmation of sinonasal mucosal inflammation is required. Finally, the 2012 European Task Force guidelines specifically define pediatric CRS as more than 12 weeks of 2 or more of the following symptoms: nasal obstruction, nasal discharge, facial pain/pressure, or cough (1 of which must be nasal obstruction or discharge).¹¹ Endoscopic or imaging confirmation of mucosal inflammation is not required.

In this study, we sought to determine the baseline rate of CRS, using a consensus set of these diagnostic guidelines, and sinonasal QOL, using the SN-5, in a nonbiased cohort of children with CF. Establishing a baseline for disease burden in the population is the first step in establishing clinical practice guidelines for this poorly understood and inconsistently managed disease entity.

Box Section Ref ID

Participants in this study receive ongoing comprehensive CF care through the Division of Pulmonary and Sleep Medicine at Seattle Children’s Hospital. Consecutive patients (N = 102) presenting for routine quarterly examination in the CF clinic were asked to complete a patient deidentified survey regarding their sinonasal health history and QOL. Patients were excluded if they had a nonroutine appointment in the clinic to work up a specific acute chief complaint. Parents of minors completed the score with input from adolescents. These deidentified surveys were collected as part of a quality improvement project. At the end of the quality improvement project, completed surveys were analyzed for purposes of this study. This project was approved by the institutional review board of Seattle Children’s Hospital. Written informed consent was not required; survey completion was taken to imply consent.

Sample size was chosen on the basis of clinical availability during the study period. We estimated that with the number of consecutive patients, accounting for incomplete survey responses, we would achieve a target of 50 respondents. This sample size should be adequate to detect a difference between 2 independent groups, with a power of 0.8, significance level of .05, and effect size of 1.0, which is equivalent to the posttreatment effect in previous studies of medical and surgical treatment for childhood CRS.^8,12-15

Surveys (Figure 1) were divided into 3 sections. The first section was used to establish a diagnosis of CRS based on consensus diagnostic criteria.^9-11 Based on current clinical practice guidelines, 4 of 5 symptoms were surveyed: nasal obstruction, nasal drainage, facial pain/pressure, and decreased sense of smell. Cough, being a highly common complaint in children with CF and thus a nonspecific symptom with respect to CRS, was not included. Imaging or endoscopic evidence of sinonasal inflammation was not included in the survey for the following reasons: (1) we did not have access to imaging or endoscopy results; (2) the European Task Force guidelines did not require it as part of the diagnosis of pediatric CRS; and (3) mucoperiosteal thickening is a ubiquitous finding on imaging studies of children with CF,² making it likely to be of very low specificity for diagnosis.

The second section of the survey was designed to assess the patient’s prior treatment for CRS, and the bottom section determined the child’s sinonasal QOL using the validated SN-5.⁷ The SN-5 includes 5 domains and an overall visual-analog scale (VAS) rating. Lower domain scores and higher VAS ratings are both associated with better QOL. The domain scores are constructed on a 7-point Likert scale, with a change in score of 0.5 suggested as a threshold for at least slight change. External validation against caregiver proxy assessment of clinical change suggested a higher threshold for “mild improvement” in the 0.57 to 1.55 range.⁷ Difference scores in this study are interpreted within this paradigm.

Unless otherwise indicated, categorical variables are reported as percentages and compared using the Fisher exact test; nonparametric values are reported as means and 95% confidence intervals and compared using the Mann-Whitney rank-sum test; and parametric values are reported as means and 95% confidence intervals and compared using the t test. Correlation coefficients and corresponding P values were calculated as a measure of linear association between variables. Analyses were exploratory; P < .05 was considered statistically significant, and results were not adjusted for multiple comparisons. All analyses were conducted using Microsoft Excel, version 14.3.9.

Of the 102 consecutive patients eligible for study participation, 47 (46%) returned the survey. Surveys were completed during the period from December 2012 to January 2013. Age ranged from 2 to 20 years, with a mean (SD) of 12.9 (5.6) years. Twenty-four patients (51%) were female.

A diagnosis of CRS was inferred from the responses to the following question: “Over the past 12 weeks, has your child complained of any of the following on a daily basis: (1) nasal congestion or obstruction; (2) facial pain, pressure, or fullness; (3) smelly drainage from the front of the nose, or back into the throat?; or (4) decreased sense of smell?” Seventeen of 47 (36%) complained of nasal congestion, 4 (9%) reported pain, 2 (4%) experienced nasal drainage, 6 (13%) had decreased smell, and 8 (17%) had multiple complaints.

Responses were scored as fulfilling criteria for diagnosis according to the 3 clinical practice guidelines described previously: Denver (1999),⁹ ARS (2007),¹⁰ and European (2012).¹¹ According to the Denver criteria, 18 of 47 children (38%) met diagnostic criteria for CRS; for ARS, 8 (17%) qualified, whereas 5 (11%) met the European criteria (when cough was excluded as a criterion for diagnosis). For further analysis, the ARS criteria were used to group patients into those with or without CRS. The ARS criteria were chosen because, compared with the Denver criteria, they represent more contemporary and, most importantly, multi-institutional consensus guidelines. Comparison of the 2 criteria was performed for analyses of age association, as well as sinonasal QOL. The European criteria were not used or assessed because of the major confounder of cough as a diagnostic criterion. Because children with CF invariably have cough, we did not believe that these criteria would be applicable for this population.

Association between age and CRS symptoms was assessed with binomial logistic regression. Significant associations were found between increased age and pain, decreased sense of smell, and presence of CRS by ARS criteria (Table). No association was seen between age and congestion, or CRS by Denver criteria, which include congestion as a sufficient single criterion for diagnosis of CRS. Analysis of the association between drainage and age was limited owing to only 2 individuals reporting this symptom.

Of the 47 participants, 21 (45%) had used medications that treat CRS in the past 3 months, and 18 (38%) had undergone surgery on the nose or sinuses at any time in the past. Overall, 25 of 47 (53%) had undergone some treatment for sinusitis. Twenty-three patients (49%) had seen an otolaryngologist. There was no significant difference between the participants with and without CRS in terms of past medical treatment (63% [5 of 8] vs 46% [15 of 33], respectively; difference of 17% [95% CI, −19% to 46%]), surgery (50% [4 of 8] vs 36% [12 of 33]; difference of 14% [95% CI, −20% to 45%]), and otolaryngologic evaluation (63% [5 of 8] vs 46% [15 of 33]; difference of 17% [95% CI, −19% to 46%]).

Overall, 41 of the 47 participants completed the SN-5 survey, with a mean domain score of 2.16 (95% CI, 2.02-2.30) and VAS score of 8.26 (95% CI, 8.01-8.51). Among all participants, higher mean domain score was strongly associated with lower VAS score (P < .001; R² = 0.38) (Figure 2A); both of these are associated with better QOL. Neither mean domain score nor VAS varied with age by linear regression (domain score: P = .22, R² = 0.04; VAS: P = .29, R² = 0.02) (Figure 2B). Mean scores on nasal obstruction (3.07 [95% CI, 2.80-3.34]) and sinusitis (2.68 [95% CI, 2.42-2.94]) were the most affected domains overall, whereas allergy (1.83 [95% CI, 1.65-2.01]), emotional disturbance (1.76 [95% CI, 1.56-1.96]), and activity restriction (1.43 [95% CI, 1.31-1.57]) were minimally affected. Children with CRS tended to have higher (more affected) scores on these QOL domains than those without CRS, with a statistically significantly higher mean domain score (2.60 [95% CI, 2.31-2.89] vs 2.05 [95% CI, 1.90-2.20]; difference of 0.55 [95% CI, 0.29-0.80]) (Figure 3). This difference is likely to be slightly clinically significant based on the design and validation of the SN-5 survey.⁷ The mean VAS scores were not statistically significantly different between the 2 groups (8.1 [95% CI, 7.4 to 8.6] for with CRS vs 8.3 [95% CI, 8.0 to 8.6] for without CRS; difference of 0.27 [95% CI, −0.29 to 0.84]). Three of 8 children (38%) had SN-5 overall domain scores that were clearly clinically significantly elevated (above the 1.06 mean change score associated with at least mild difference)⁷ relative to the mean SN-5 score for children without CRS.

These findings were consistent when the Denver criteria were used: SN-5 mean domain scores for children with vs without CRS by Denver criteria were 2.47 (95% CI, 2.30-2.64) vs 1.94 (95% CI, 1.75-2.13) (difference of 0.53 [95% CI, 0.37-0.69]), and VAS scores were 7.9 (95% CI, 7.5-8.3) for with CRS vs 8.4 (95% CI, 8.1-8.7) for without CRS (difference of 0.5 [95% CI, 0.1-0.9]).

There were no statistically or clinically significant differences seen in SN-5 mean domain or VAS scores between those children who had received prior treatment and those who had never been treated (SN-5: 2.27 [95% CI, 2.08 to 2.46] for treated vs 2.01 [95% CI, 1.81 to 2.21] for untreated, difference of 0.25 [95% CI, 0.10 to 0.41]; VAS: 8.1 [95% CI, 7.8 to 8.5] for treated vs 8.4 [95% CI, 8.0 to 8.7] for untreated, difference of 0.2 [−0.2 to 0.7]).

This study is a cross-sectional observational analysis of sinonasal QOL and CRS in an unbiased pediatric CF population. The instrument used here, the SN-5, is a validated sinonasal QOL instrument that has been used to evaluate treatment response to medical⁷ and surgical⁸ interventions. In contrast, to our knowledge, the only previously reported survey of sinonasal QOL in children with CF⁵ used an adult instrument (SNOT-16) that did not account for nasal obstruction, which is the most common complaint among these children.⁶ The primary limitation of this study is the response rate of 46%. The age range, mean age, and sex distribution of the respondents was comparable to the demographic profile of the CF clinic population as a whole and thus is unlikely to represent a biased subset. Respondent bias is a potential issue; however, the expectation would be that more affected individuals might be more likely to complete a study addressing sinonasal complaints. Therefore, the direction of the bias would tend to favor decreased sinonasal QOL, which strengthens the finding of this study. Although sample size was small, it should be adequate to detect a clinically meaningful difference based on previous studies.^8,12-15 A larger study, however, would be able to elicit more subtle differences and permit finer subgroup analyses.

A chief limitation of the SN-5 instrument is the lack of normative age- and season-matched data, which limits comparison. This was addressed by comparing the present results to historical cohorts. The SN-5 scores in the with and without CRS groups in the present cohort were compared with historical pediatric cohorts in 5 studies.^8,12-15 Mean domain values were reported before and after treatment including balloon sinuplasty, adenoidectomy, functional endoscopic sinus surgery, and/or nasal irrigation. The SN-5 scores for both the with and without CRS groups were comparable to the postoperative values in these studies and were significantly better than the preoperative values (Figure 4). The same was true for VAS scores, which were reported in 2 of these studies^8,14: the mean VAS scores of the with and without CRS groups in this CF cohort (8.00 and 8.32, respectively) were significantly higher than the pretreatment values before nasal irrigation (4.83) and functional endoscopic sinus surgery/adenoidectomy (4.3), and comparable to the values after treatment (7.6 and 8.0, respectively). Although the populations across these studies are not rigorously comparable, this comparison does provide some context in which to interpret the present data and suggests that sinonasal QOL in the CF cohort is better than in those who seek treatment for sinonasal complaints.

The validity of the SN-5 specifically in children with CF, however, has not been assessed; as a comorbid condition that significantly affects QOL, CF may mask some of the sinonasal QOL complaints, thus leading to an underestimate of the effect of sinonasal disease on overall QOL. Another limitation is lack of consensus guidelines for diagnosis of CRS in children. Indeed, the 3 more commonly used guidelines yielded different prevalence values for CRS in this cohort. However, the findings of the study were consistent regardless of the diagnostic criteria used. A final limitation is that the SN-5 is not validated for children older than 12 years. With the SNOT-16 not validated for children, this cohort of older children lacks a validated survey instrument for assessing sinonasal QOL. We found no statistically significant association between age and either SN-5 mean domain or VAS score by linear regression. Although the overall distribution and variance in scores between the 2 age groups was similar, suggesting that the SN-5 instrument is similarly valid, rigorous validation in the older cohort would be necessary for future studies.

In general, sinonasal QOL was high in this cohort, and the incidence of CRS according to consensus diagnostic criteria was low, especially considering the near-ubiquity of sinonasal mucosal inflammation reported previously. This may be due to distraction by other aspects of CF that focus attention away from sinonasal complaints. Indeed, we found that the SN-5 mean domain and VAS scores for the CF cohorts both with and without CRS were significantly better than those of general pediatric populations with CRS; instead, they were comparable to scores seen after effective medical and surgical treatment. This relative minimization of sinonasal QOL issues despite high prevalence of clinical and anatomic disease may be because these children have accommodated to the symptoms and notice them less, or it may be because other aspects of their disease are taking precedence and minimizing the QOL impact attributable to sinonasal disease.

Slightly more than half of the overall cohort had undergone some treatment for CRS, either medical or surgical; however, there was no association between either prior treatment or prior otolaryngologic evaluation and the presence of ongoing CRS symptoms, or impaired QOL. The lack of association between CRS, sinonasal QOL, and prior or ongoing treatment may be due to the small sample size in this cohort but may also reflect uncertainty as to when and how to treat children with CF and CRS. Indeed, guidelines for otolaryngologic referral and medical or surgical treatment of sinusitis in children with CF do not exist. The utility of imaging or endoscopy is already somewhat limited in the pediatric population for determination of need for surgical intervention for CRS; given the near-universal presence of sinonasal mucosal inflammation in children with CF, justification for surgery lies primarily in the improvement of QOL. Objective determination of sinonasal QOL using a validated instrument such as the SN-5 is thus a critical component in the evaluation of children with CF and suspected sinusitis. This study provides a baseline for future efforts to more sharply delineate the criteria for management of these frequently comorbid conditions.

Children with CF have a high incidence of CRS, but their baseline QOL indicates relatively low impairment of sinonasal quality of life. Use of a validated sinonasal QOL instrument can be valuable to guide management decisions for this challenging pediatric population.

Accepted for Publication: March 29, 2016.

Corresponding Author: Dylan K. Chan, MD, PhD, Division of Pediatric Otolaryngology, Department of Otolaryngology–Head and Neck Surgery, University of California, San Francisco, 2233 Post St, Third Floor, Box 1225, San Francisco, CA 94115 (dylan.chan@ucsf.edu).

Published Online: May 26, 2016. doi:10.1001/jamaoto.2016.0979.

Author Contributions: Drs Chan and Park had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Chan, Park, Parikh.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Chan, Park, Vajda.

Critical revision of the manuscript for important intellectual content: Chan, McNamara, Park, Gibson, Parikh.

Statistical analysis: Chan, Park, Vajda.

Administrative, technical, or material support: Vajda, Gibson, Parikh.

Study supervision: Chan, Gibson, Parikh.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.

Previous Presentation: Results of this study were presented at the American Society of Pediatric Otolaryngology Spring Meeting; April 22-26, 2015; Boston, Massachusetts.