2018 Section 5 - Rhinology and Allergic Disorders

Otolaryngology–Head and Neck Surgery 154(1)

evaluate BCD in specific patient populations, including those with prior ESS, nasal polyposis, and office-based pro- cedures; and (3) to identify differences in secondary outcome measures of recovery time, postoperative complications, deb- ridements, and revision surgery for patients with CRS under- going BCD versus ESS. Materials and Methods A systematic review with meta-analysis was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses statement (PRISMA guidelines) 17 to evaluate the outcomes of paranasal sinus BCD versus ESS in patients with CRS. Methods of the anal- ysis and inclusion criteria were specified in advance and documented in the Prospero database with registration number CRD42015015823 (http://www.crd.york.ac.uk/pros- pero/). The primary outcomes included changes in quality of life and radiographic opacification, as well as secondary outcome measures of recovery time, postoperative compli- cations, debridements, and revision surgery. Eligibility Criteria Experimental and cohort studies reporting original data in the evaluation of postoperative outcomes among patients undergoing transnasal paranasal sinus BCD for CRS were included. Trials included adult subjects ( . 18 years) with or without nasal polyposis. Only English articles were eligible. Case reports and pilot studies evaluating safety as the pri- mary outcome measure were excluded, as were studies eval- uating transantral approaches and ostial dilation techniques not involving a balloon catheter. Hybrid studies (BCD 1 ESS) were included only if data reporting included sub- groups of patients undergoing isolated BCD. Search Criteria A comprehensive review of the English-language literature was performed of the MEDLINE and EMBASE databases. Search criteria were designed to identify studies evaluating the use of BCD in the treatment of CRS in articles published between 1996 and December 2014. Search criteria encom- passed all occurrences in the title or abstract of the terms ‘‘chronic sinusitis’’ or ‘‘chronic rhinosinusitis’’ or ‘‘CRS’’ plus the term ‘‘balloon sinuplasty’’ or ‘‘sinuplasty’’ or ‘‘balloon dilation’’ or ‘‘dilatation.’’ Unpublished data were excluded. After searches were completed, duplicate records were discarded, and abstracts of all identified studies were inde- pendently reviewed by 2 authors (J.M.L., M.J.M.). Abstracts were screened to identify experimental and cohort studies reporting original outcomes in adult patients ( . 18 years) receiving BCD in the treatment of CRS. A full-text review of articles with eligible abstracts was then performed. Included in this review was a manual screening of the refer- ence lists to ensure identification of all relevant articles. Data Extraction A standardized data extraction template was used for each study. In the case of several manuscripts reporting the

outcomes of a single cohort at multiple durations of follow- up, the data from each time point was extracted as a separate entry. Primary outcomes included the impact of BCD on vali- dated measures of quality of life and sinonasal opacification. Additionally, secondary outcome measures of recovery time, global improvement, postoperative debridements, and compli- cations, when provided, were recorded and included in analy- sis. For each treatment arm, number of patients, sex, age, and preoperative 20-Item Sinonasal Outcome Test (SNOT-20) and Lund-Mackay scores were recorded when available. Additionally, the proportions of patients with nasal polyps and those undergoing office-based or revision procedures were recorded. The risk of bias was assessed at the study and outcome levels by examining each study for specific markers of validity—including randomization, concealment of subject allocation, blinding of subjects or investigators, proportion of subjects lost to follow-up, and specification of treatment for each subject. The level of evidence was determined to pro- vide an overall estimate of the strength of study design. 18 Additionally, conflict of interest was evaluated with reporting of corporate funding and financial disclosures. Statistical Analysis A narrative synthesis of findings in the systematic review of included studies describes demographic characteristics of the target population, intervention received, and reported outcome measures. RevMan 5.3 (Cochrane Group, London, UK) was utilized for pooling of data and completion of meta-analysis. A random-effects model was used to calcu- late standardized mean differences for continuous outcome variables; 95% confidence intervals (95% CIs) and 2-sided P values are reported. Heterogeneity among included studies was evaluated with the I 2 statistic, with value . 50% indicat- ing substantial heterogeneity. Descriptive statistics, where applicable, were calculated with SPSS Statistics 18 (IBM Corp, Armonk, New York). Subgroup analysis of weighted means for office versus operating room BCD and recovery time for BCD versus ESS was performed with a 2-sample independent t test of weighted means. Results The search strategy identified a total of 92 articles. Title and abstract review identified 22 that met eligibility criteria. Review of the references of these articles identified an addi- tional 17 articles. Full-text review was then completed of all 39 articles, yielding 17 that met eligibility criteria for sys- tematic review. Data extraction for meta-analysis was possi- ble in 11 of the identified studies ( Figure 1 ). Manuscripts identified for study inclusion are summarized in Table 1 . 11,13,19-33 There was homogeneity in primary out- come measures (SNOT-20 and Lund-Mackay), with varied duration of follow-up. Studies were subsequently pooled for quantitative analysis, with 2 additional subgroups reporting outcomes at early ( 6 months) and late ( 12 months) post- operative time points. Itemized assessment of risk of bias is reported in Table 2 . Potential conflicts of interest were

216