2017-18 HSC Section 4 Green Book

Otolaryngology–Head and Neck Surgery 156(5)

bias. MINORS is a validated instrument designed to assess the methodological quality of nonrandomized studies. No weight- ing of pooled means was performed due to MINORS score. Study results were pooled with a random effects model of meta-analysis, which assumes that the true effect could vary from study to study. 5 Change in NOSE score after surgery was assessed with both the mean difference between baseline and postoperative results and the standardized mean differ- ence (SMD), which is a measure of overall effect size (0.2 is a small effect, 0.5 moderate, and 0.8 large). Since the individ- ual studies did not report change scores consistently, we instead compared the mean and standard deviation of the baseline NOSE scores with the scores at a given postopera- tive time point (3.0-5.9, 6.0-11.9, or 12 months), adjusting the variance by the correlation of the patient sample at the 2 time points (1.00 if the measurements were on the same patients, lower if some of the original patients were lost to follow-up). 6 Heterogeneity was assessed and reported with the I 2 statistic, where 25% is low heterogeneity among stud- ies, 50% moderate, and 75% high. 7 Results The result of the database search was 665 articles after duplicates were removed. Screening of abstracts and full- text articles resulted in 16 articles included for final analysis ( Figure 1 ). Of the 16 included studies, 9 were from North America, 5 from Turkey, and the remaining from the United Kingdom and Brazil. The study samples ranged from 7 to 77. A summary of the characteristics and the techniques uti- lized in the included studies is seen in Table 1 . MINORS scores for the included studies are listed in Table 2 . MINORS criteria are based on a score of 16, with 16 being a perfect score for noncomparative studies. 4 In the 16 included studies, scores ranged from 8 to 16, with a median of 13 and a mean of 12.3. The major deficiencies were the inclusion of consecutive patients, the unbiased assessment of the study end point, and the prospective calcula- tion of the sample size. Many articles did not enroll patients consecutively and were, in fact, convenience samples. Also the majority of studies did not do any blinding with regard to col- lection of NOSE score. Finally, few studies made mention of prospective calculation of sample size; rather, many were pub- lished case series with whatever data were available to the authors. All or nearly all studies had clearly stated aims with appropriate follow-up periods. Given that only studies that used the NOSE score were included, nearly all of the studies had end points that were appropriate to the study aim. Patients in the included studies had moderate to severe nasal obstructive symptoms at baseline. The 16 studies included in final analysis had a pooled mean preoperative NOSE score of 67.4 (95% CI, 61-73.9) based on random effects meta-analysis. The range of scores was 34.8 to 86.5, with very high heterogeneity ( I 2 = 95). At 3-6 months, there was a substantial improvement in NOSE scores ( Figure 2 ). Six studies reported follow-up NOSE scores in this period. There was a mean improvement of 49.8 in NOSE scores (95% CI, 45.2-54.3), with high heterogeneity ( I 2 = 98). The

provide aggregate and quantitative data regarding the effect of functional rhinoplasty based on NOSE scores. We limited our search to the last 12 years, as the NOSE questionnaire was introduced in 2004. This is the first study to review the existing literature on functional rhinoplasty with attention paid to providing quantitative and aggregate data of the effects of this surgical intervention on patients’ QOL based on the validated NOSE instrument. Methods The investigators began this review with a PICO question— specifically, whether patients who had nasal obstruction due to nasal valve dysfunction ( population ) and who had functional rhinoplasty for the purposes of improving nasal obstruction ( intervention ) had improvement in their nasal obstruction, as measured by the NOSE scale ( outcome ); there was no compar- ison group as (1) the majority of studies that evaluated the effect of functional rhinoplasty did not have a control group or (2) comparison groups were studied among different tech- niques of functional rhinoplasty. A literature search was per- formed in similar fashion to the previous systematic review by Rhee et al. 3 A search for abstracts was performed in November 2015 with the PubMed, EMBASE, and Cochrane databases with the search terms ‘‘nasal obstruction’’ and ‘‘rhi- noplasty.’’ Given the robust results of this initial search, an information specialist was not utilized. The search was limited to English-only abstracts. Studies were included if they evalu- ated the effect of ‘‘functional rhinoplasty’’ on nasal obstruction with the NOSE score. Functional rhinoplasty was defined as follows: any surgical manipulation of the bony/cartilaginous nasal framework or soft tissues of the nose, specifically targeted toward improving static or dynamic collapse of the internal or external nasal valve or other elements of nasal vestibular stenosis. Case reports, editorials, narra- tives, studies with \ 3 patients, and articles that did not use the NOSE score were excluded. If a study included data on septoplasty alone or purely cosmetic rhinoplasty and those data could not be separated from functional rhinoplasty data, that study was also excluded. If nasal procedures such as turbinoplasty or sinus surgery were performed in addition to functional rhinoplasty, then the studies were excluded. We acknowledge, however, that additional proce- dures may have been inconsistently reported in the source articles; therefore, patients who received some of these proce- dures may have been included in our study sample. Septoplasty is generally considered to be a necessary step in functional rhinoplasty for the purpose of cartilage harvest, so with the exception of patients who had septoplasty alone (who were excluded), patients who had septoplasty in addi- tion to functional rhinoplasty were included, as this would entail nearly every functional rhinoplasty patient. Following the search with the above search terms, dual- investigator independent screening of abstracts and full-text articles was performed. Two independent investigators also extracted data—specifically, pre- and postoperative NOSE score and timing of follow-up. Methodological Index for Non- randomized Studies (MINORS) criteria 4 were used to measure

158