Correlation of the Standardized Cosmesis and Health Nasal Outcomes Survey With Psychiatric Screening Tools

Abstract

Background

Identifying mental health disorders, including body dysmorphic disorder (BDD), is important prior to rhinoplasty surgery; however, these disorders are underdiagnosed, and screening tools are underutilized in clinical settings.

Objectives

The authors sought to evaluate the correlation of a rhinoplasty outcomes tool (Standardized Cosmesis and Health Nasal Outcomes Survey [SCHNOS]) with psychiatric screening tools.

Methods

Patients presenting for rhinoplasty consultation were prospectively enrolled and administered mental health instruments to assess depression (Patient Health Questionnaire-9), anxiety (Generalized Anxiety Disorder-7), and BDD (BDD Questionnaire-Aesthetic Surgery [BDDQ-AS]) as well as the SCHNOS rhinoplasty outcomes scale. Convergent validity of SCHNOS scores with these mental health instruments was assessed as well as calculation of an optimal SCHNOS-C score to screen for BDD.

Results

A total 76 patients were enrolled in the study. The average SCHNOS-O score (standard deviation) was 46.1 (34.0) and the average SCHNOS-C score was 61.1 (27.0). Five (7%) patients screened positive for depression, and 24 (32%) patients screened positive for mild, 5 (7%) for moderate, and 4 (5%) for severe anxiety. Twenty-four (32%) patients screened positive for BDD by BDDQ-AS scores. SCHNOS-O and SCHNOS-C did not correlate with Patient Health Questionnaire-9 or Generalized Anxiety Disorder-7 scores; SCHNOS-C did correlate with BDDQ-AS. A score of 73 or greater on SCHNOS-C maximized the sensitivity and specificity of also screening positive for BDD with BDDQ-AS. This score correlated with a sensitivity of 62.5%, specificity of 80.8%, and number needed to diagnose of 2.3, meaning for every 2 patients with a score of ≥73 on SCHNOS-C, 1 will have a positive BDDQ-AS score.

Conclusions

SCHNOS-C correlates with BDDQ-AS and may help screen rhinoplasty patients at higher risk for BDD.

Level of Evidence: 2

Mental health disorders such as depression, anxiety, and body dysmorphic disorder (BDD) occur at higher rates in the cosmetic surgery patient population compared with the general population, with estimated rates as high as 20%.^1-9 Rates of BDD are particularly high in the cosmetic population, with an estimated prevalence between 5% and 20% compared with 0.7% to 2.4% in the general population.^1-7,10-20 Prevalence of BDD in the rhinoplasty patient population specifically has been reported as high as 43%.⁵ Additionally, up to 76% of patients with BDD seek consultation for cosmetic procedures, and up to 60% actually undergo these procedures.^4,21-25

The DSM-5 criteria for diagnosis of BDD include appearance fixation, repetitive or compulsive behaviors in response to the appearance fixation, impairment of function as a result, differentiation from an eating disorder, and assessment for the 2 BDD subtypes: muscle dysmorphia and insight level.²⁶ Commonly, patients with BDD seek out rhinoplasty consultation, because the nose is a frequent source of body image fixation.^5,7,27,28 Moreover, these patients often have worse outcomes after undergoing surgery.^{1,4,6,7,21,23-25,29-38} In a review of the literature, the worst surgical and psychological outcomes were found in BDD patients who underwent cosmetic rhinoplasty or revision surgery.³⁰ Surveys show that between 9% and 29% of cosmetic surgeons were threatened legally by patients with BDD, and 2% were threatened with physical harm.³² Treatment for patients suspected of having BDD includes psychiatric care in the form of cognitive behavioral therapy or serotonin-reuptake inhibitors rather than surgical procedures.^{4,21,32,39-48}

Despite the overall consensus that it is important to recognize patients with mental health disorders before surgery, a 2002 survey of the American Society for Aesthetic Plastic Surgery found that more than 80% of surveyed physicians realize patients may have BDD only after surgery was performed.³² Yet these same surgeons believed the incidence of BDD to be only 2% in their practice, and only 30% believed BDD was a contraindication to cosmetic surgery.³² A 2015 survey of the American Society for Dermatologic Surgery also showed that 60% of surveyed physicians realized patients had BDD after treatment was performed, but unlike the earlier survey, 63% believed BDD was a contraindication for cosmetic treatment.⁴⁹

Due to infrequent screening, the rate of BDD is underestimated and underdetected, and there are no guidelines for proper identification of BDD in patients seeking aesthetic surgery.^4,32,50-52 The ability of surgeons to detect BDD in their patients is very poor, as evidenced by Joseph et al, who found that of 597 patients surveyed, 58 (9.7%) screened positive on a BDD questionnaire; however, only 2 of these patients were correctly identified by cosmetic surgeons.⁵³ Thus, it is recommended that validated screening tools for BDD be used to assess for this diagnosis at the time of initial consultation.

Despite the relatively short nature of validated BDD screening tools such as body dysmorphic disorder questionnaire (BDDQ) or BDDQ-Aesthetic Surgery (BDDQ-AS) available for utilization in the rhinoplasty patient population, many surgeons do not employ these, because added patient questionnaires become time-consuming to manage for both the patient and the physician.^4,32,50,51 Therefore, it would be useful if a patient questionnaire already being utilized to assess rhinoplasty outcomes could also serve as an initial screening tool for the important diagnosis of BDD or other mental health disorders. One such screening tool recently validated for both cosmetic and functional rhinoplasty is the Standardized Cosmesis and Health Nasal Outcomes Survey (SCHNOS). This survey is a 10-item patient-reported outcome measure that reports 2 outcomes: 1 for nasal obstruction (SCHNOS-O, items 1-4) and 1 for nasal cosmesis (SCHNOS-C, items 5-10).⁵⁴ Convergent validity is a statistical method to test the correlation of scores on 1 test (such as the SCHNOS) with other tests designed to assess the same topic (in this case mental health diagnoses). Thus, the objective of this study was to assess the convergent validity between tools to identify depression severity (Patient Health Questionnaire-9 [PHQ-9]), anxiety severity (Generalized Anxiety Disorder-7 [GAD-7]), and BDD (BDDQ-AS) with SCHNOS scores, and to determine a SCHNOS score at which BDD diagnosis may be more likely, prompting further workup.

METHODS

Participants

The institutional review board of Stanford University approved this study. Informed consent for study participation was obtained for each patient. The study location represents an academic facial plastic and reconstructive surgery practice. Patients presenting for rhinoplasty consultation were sequentially enrolled between November 15, 2017 and January 15, 2019. Other inclusion criteria included age over 18 years and new patient status. Questionnaires were distributed to and completed by study participants on paper, and although not completed anonymously, participation was only known to the study coordinator. Participation in the study and results to screening questions were blinded to the surgeon (S.P.M.) and had no impact on clinical decision-making. Patients completed each screening instrument (PHQ-9, GAD-7, and BDDQ-AS) as well as a SCHNOS. Other data collected included patient age, sex, marital status, history of prior rhinoplasty surgery, whether they underwent rhinoplasty surgery after consultation, and if the procedure was cosmetic, functional, or both.

Clinical Instruments

Validated tools used to detect mental health disorders included the PHQ-9 for depression severity, the GAD-7 for anxiety severity, and the BDDQ-AS for BDD screening. The PHQ-9 is a self-administered diagnostic instrument for depression severity, which scores each of the 9 DSM criteria for depression as “0” (not at all) to “3” (nearly every day). A PHQ-9 score of ≥10 has a sensitivity of 88% and specificity of 88% for major depression.⁵⁵

The GAD-7 is a self-administered diagnostic instrument for anxiety severity based on DSM symptoms criteria for generalized anxiety disorder. For each item, patients score from “0” (not at all) to “3” (nearly every day). A GAD-7 score of ≥10 has a sensitivity of 89% and specificity of 82% for generalized anxiety disorder. Scores of 0 to 4, 5 to 9, 10 to 14, and 5 to 21, represent minimal, mild, moderate, and severe anxiety, respectively.⁵⁶

The BDDQ is a self-administered diagnostic instrument for BDD based on the DSM definition of BDD, with binary outcomes (positive or negative for BDD) and demonstrated to have 100% sensitivity and 89% specificity in a psychiatric setting, and was also validated in the facial plastic and reconstructive surgery setting.^57,58 The BDDQ-AS is a 7-item questionnaire developed from the BDDQ as an instrument to screen for BDD in the aesthetic rhinoplasty population. It has a binary outcome (positive or negative for BDD) and was found to have a sensitivity of 89.6% and specificity of 81.4% for rhinoplasty patients.⁵⁷

The SCHNOS scale is a validated 10-item clinical instrument utilized to assess rhinoplasty outcomes. The scale has 2 scores, SCHNOS-O (for obstruction, items 1-4) and SCHNOS-C (for cosmesis, items 5-10), scored out of a possible 100 points each. The greater the score, the greater the disability.^54,59-61

Statistical Analysis

Power analysis was performed employing G*Power 3.1.7 (Franz Faul Universität, Kiel, Germany). To achieve a power of 0.95, with alpha of 0.05 utilizing a point biserial correlation and 2-tailed P value, an effect size >0.4 was chosen, which would assess a medium difference between scores of the clinical instruments being evaluated with convergent validity statistics. With these parameters, a sample size >70 was calculated. The corresponding number for the bivariate normal model was >45. Thus, a sample of at least 45 patients was required to sufficiently power the study.

The baseline characteristics of the sample were described in proportions and percentage or in means, standard deviations (SDs), and ranges when appropriate. The BDDQ-AS was considered as gold standard for BDD presence or absence. The baseline characteristics of patients who were BDD-positive according to the BDDQ-AS were compared with those who were BDD-negative. Statistical methods employed were independent t test for continuous variables and chi-squared for binary variables. P values were reported, with P ≤ 0.05 considered significant.

Next, convergent validity of the mental health tools with the SCHNOS was undertaken. Correlations between PHQ-9, GAD-7, BDDQ-AS, individual items of SCHNOS, and SCHNOS-C and SCHNOS-O total scores were evaluated by utilizing a Spearman rank correlation and Somer’s D test for binary and ordinal data. All correlation estimates were accompanied by 95% confidence intervals. Correlation coefficients were interpreted as follows: 0.00 to 0.19, very weak; 0.20 to 0.39, weak; 0.40 to 0.59, moderate; 0.60 to 0.79, strong; and 0.80 to 1.00, very strong correlation. All analyses were conducted utilizing Stata/IC Statistical Software: Release 15 (StataCorp LP, College Station, TX).

An area under the curve model utilizing the sensitivity and specificity of the SCHNOS-C scale evaluated against BDDQ-AS was constructed. The Youden index (J) was employed to define a score on SCHNOS-C that maximized the difference between the true positive rate and false positive rate of the SCHNOS-C with BDDQ-AS over all possible values of SCHNOS-C, thus maximizing the positive predictive value.⁶² The receiver operator characteristic analysis was conducted utilizing MedCalc 19.0.7 (StataCorp LP, Ostend, Belgium).

RESULTS

Of 76 study participants, the average age (SD) was 32.3 (12.2) years, with a range of 18 to 77 years; 26 (34%) participants were male, and 50 (66%) were female. Fourteen (18%) patients had a history of prior rhinoplasty surgery, and 40 (53%) underwent rhinoplasty after the consultation. The average (SD) SCHNOS-O score was 46.1 (34.0) with a range of 0 to 100, and the average (SD) SCHNOS-C score was 61.1 (27.0) with a range of 0 to 100. Five (7%) respondents screened positive for depression according to a PHQ-9 score of 10 or greater. Based on GAD-7 scores, 43 (57%) respondents did not have significant anxiety, 24 (32%) had mild, 5 (7%) had moderate, and 4 (5%) had severe anxiety.

Twenty-four (32%) patients had BDD according to the BDDQ-AS scores. Comparison of the cohort of patients with BDD to patients without BDD according to BDDQ-AS scores showed no difference in age, sex, marital status, history of prior rhinoplasty, type of rhinoplasty (functional or cosmetic), or undergoing rhinoplasty after consultation between the 2 groups (Table 1).

Convergent validity between SCHNOS items and total scores with BDDQ-AS, PHQ-9, and GAD-7 scores are shown in Table 2. Neither SCHNOS items nor total scores correlated with tests of general depression and anxiety (PHQ-9 and GAD-7). Neither SCHNOS-O nor its items correlated with BDDQ-AS. The SCHNOS-C total score did correlate with BDDQ-AS, especially in items number 5 (decreased mood or self-esteem due to my nose), number 6 (the shape of my nasal tip), and number 8 (the shape of my nose from the side), with the strongest correlation in item number 5.

To identify an optimal score at which SCHNOS-C may correspond to a positive BDDQ-AS test, the area under the curve for sensitivity and specificity of SCHNOS-C with BDDQ-AS was conducted and shown in Figure 1. From this graph and using receiver operator characteristic analysis, the Youden index identified an optimal score of 73 or greater on SCHNOS-C, which maximized the difference between the true positive rate and false positive rate over all possible SCHNOS-C values. A score of 73 on SCHNOS-C corresponded to a Youden index of 0.4327, number needed to diagnosis of 2.3, sensitivity of 62.5%, and specificity of 80.8%. Thus, for every 2 patients with a SCHNOS-C >73, 1 will have a positive BDDQ-AS score. Sensitivity and specificity values for various SCHNOS-C values can be found in Supplemental Table 1.

DISCUSSION

This study supports prior literature reporting a much higher prevalence of BDD and other psychiatric conditions such as anxiety and depression in the cosmetic surgery, and specifically rhinoplasty patient population compared with the general population.^1-20 Moreover, this study showed that the SCHNOS-C questionnaire correlated with the BDDQ-AS questionnaire and may be useful as an initial screening method to identify patents needing further workup for BDD. By conducting the convergent validity of these 2 instruments for BDD, the SCHNOS-C could be employed for preliminary BDD screening in addition to assessing rhinoplasty outcomes in a busy clinical practice.

The prevalence of BDD in this study was 32% based on the BDDQ-AS scores. The prevalence of BDD has been reported at rates between 5% and 20% in the general cosmetic surgery population and as high as 43% in the rhinoplasty patient population.^1-20 Although patients in this study presented for rhinoplasty for both cosmetic and functional reasons, a majority (61 patients, or 80% of the study population) presented for either cosmetic or combined functional and cosmetic surgery. Therefore, the high rate of positive BDD screening in these patients is in line with the high rate of BDD found in other studies of cosmetic rhinoplasty patients.^5,7,10 Picavent et al found a rate of 33%, Veale et al found a rate of 20.7%, and Alavi et al found a rate of 24.5%.^5,7,10 Prevalence of other psychiatric disorders, particularly anxiety and depression, has also been shown to be higher in the cosmetic surgery patient population compared with the general population, and indeed was higher in this study as well.^1-9 Close to 40% of patients participating in this study demonstrated at least mild anxiety by the GAD-7 screening tool.

Assessment of the SCHNOS scale shows no correlation with tests of general depression and anxiety (PHQ-9 and GAD-7). SCHNOS-O (measuring nasal obstruction symptoms) or its items do not correlate with BDDQ-AS, PHQ-9, or GAD-7. This means that patients’ nasal obstruction symptoms do not seem to be worse in those screening positive for psychiatric diagnoses. SCHNOS-C does moderately correlate with BDDQ-AS, most strongly in items number 5, which is the most body image-directed item on the SCHNOS-C. As evidenced by the high rate of rhinoplasty patients with BDD compared with the general population and that the nose is a frequent source of body image fixation in BDD patients, it is understandable that questions of the SCHNOS-C assessing nasal appearance correlate with the BDDQ-AS.^5,7,27,28

It was determined that a score of 73 or greater on the SCHNOS-C optimized this instrument’s ability to screen for BDD. A score of 73 was determined to have the greatest difference between the true positive and false positive rate compared with the BDDQ-AS, correlating to a sensitivity of 62.5%, specificity of 80.8%, and number needed to diagnose of 2.3. Thus, for every 2 patients with a score of >73 on the SCHNOS-C, 1 will have a positive BDDQ-AS score. The BDDQ-AS, which was also validated in the rhinoplasty patient population, has a sensitivity of 89.6% and specificity of 81.4%.⁵⁷ The BDDQ, which does not have questions specific to rhinoplasty, was also validated in the rhinoplasty patient population, with a sensitivity of 100% and specificity of 90.3%.⁵⁸ Thus, although SCHNOS-C is not as sensitive or specific as these other tools, a high score can alert rhinoplasty surgeons that further work-up for BDD may be warranted.

Limitations of this study include that participants were enrolled from a single center and single surgeon practice, which may not represent the general rhinoplasty patient population. Additionally, patients who screened positive on the BDDQ-AS questionnaire did not undergo further workup for BDD with more rigorous or specific assessment because the results were blinded to the clinical care team.

Future areas of study in this patient population would include assessing surgical outcomes based on the SCHNOS for patients who screened positive for psychiatric diagnoses, and in particular BDD, compared with patients who did not. Additionally, rates of revision surgery or other complications may also be assessed.

CONCLUSIONS

SCHNOS-C does correlate with BDDQ-AS, and a SCHNOS-C score of ≥73 may help identify patients presenting for rhinoplasty who would benefit from further work-up for BDD.

Disclosures

The authors declared no potential conflicts of interest with respect to the research, authorship, and publication of this article.

Funding

The authors received no financial support for the research, authorship, and publication of this article.

REFERENCES