2017-18 HSC Section 4 Green Book

Original Investigation Research

Body Dysmorphic Disorder in Facial Plastic Surgery Clinics

Figure 1. Surgeon BDD Screening Form

Table 1. Demographics of Participants by Study Site

No. (%)

1. Does this patient have BDD? (circle one)

Patient Characteristic

Site 1 (n = 342)

Site 2 (n = 158)

Site 3 (n = 97)

P Value

Yes

No

Age, mean (SD), y

44.2 (16.5) 46.0 (16.2) 56.3 (15.5) .56

Sex

<.001

2. Rate your certainty of your response for question 1 (please put an X on the line)

Male

128 (37.4)

31 (19.6) 38 (39.2)

Female

214 (62.6) 126 (79.8) 58 (59.8)

0%

50%

100%

Unknown

0 (0)

1 (0.6)

1 (1.0)

Race

.004

BDD indicates body dysmorphic disorder.

White

223 (65.2) 125 (79.1) 81 (83.5)

African American

54 (15.8)

11 (7)

6 (6.2) 6 (6.2) 2 (2.1)

short series of questions to determine whether they believed each patient met diagnostic criteria for BDD and their cer- tainty in the diagnosis ( Figure 1 ). Surgeons were blinded to the results of the BDDQ instrument throughout the study. Statistical Analysis Descriptive statistics were used to assess baseline demo- graphicvariables. For categorical variables, theFischer exact test or χ 2 test was used to test for differences across study sites. Analysis of variance (ANOVA) methods were used to compare continuous demographic variables (age) across study sites. Simple logistic regression was used to assess for differences in baseline patient-rated satisfaction scores based on BDDQ sta- tus. Surgeon diagnostic accuracy was compared with the vali- datedBDDQscreening instrument. Sensitivity, specificity, posi- tive predictive value, negative predictive value, prevalence, receiver operating curve (ROC) area, and positive and negative likelihood ratioswere calculated fromfrequency tables. Our pri- mary performance metric—our surgeon’s ability to classify BDD—were the positive and negative likelihood ratios. We generated the pretest vs posttest probability curves using Mathematica 10.0 (Wolfram). These curves illustratehowmuch a surgeon’s diagnostic ability changes the likelihood that a pa- tient is diagnosed or not diagnosed as a BDD case. Missing data were assumed to bemissing at randomand a sensitivity analy- sis of our results tomissing datawas performed usingmultiple imputation; 200 imputations were used for this analysis. All tests were 2-sided and P values less than .05 were considered statistically significant. Besides what was otherwise noted, all remaining statistical analyseswere performedusing StataData Analysis and Statistical Software, version 14.0 (StataCorp LP). A total of 597 patients completed the surveys and were in- cluded in the study population. Baseline demographic vari- ables are shown in Table 1 . Patients presenting to each study site had similar age. Across study sites, there was a difference in the distribution of gender (χ 2 [2, N = 595] = 17.36; P < .001), education (χ 2 [10, N = 595] = 21.13; P = .049), self-reported race (χ 2 [10, N = 595] = 25.81; P = .004), and marital status (χ 2 [4, N = 594] = 34.08; P < .001). Furthermore, theproportionof pa- tients presenting primarily for cosmetic concerns differed across sites (χ 2 [2, N = 597] = 58.21; P < .001). Results Study Population

Asian

24 (7)

8 (5.1) 8 (5.1) 2 (1.3) 3 (1.9)

Hispanic

15 (4.4)

Other

17 (5)

0 (0)

Prefer not to answer Unknown

8 (2.3)

2 (2.1)

1 (0.3)

1 (0.6)

0 (0)

Marital status

<.001

Single

140 (40.9)

68 (43)

12 (12.4)

Married

164 (48)

69 (43.7) 60 (61.9)

Separated/ divorced/widow

38 (11.1)

19 (12)

24 (24.7)

Unknown

0 (0)

2 (1.3)

1 (1)

Education

.049

Less than high school

4 (1.2)

4 (2.5)

1 (1)

High school/GED 36 (10.5)

9 (5.7)

15 (15.5)

Some college

66 (19.3) 26 (7.6) 108 (31.6) 68 (19.9)

26 (16.5) 17 (17.5)

2 y college 4 y college

10 (6.3)

14 (14.4)

54 (34.2) 27 (27.8) 44 (27.8) 17 (17.5)

Masters degree Doctoral degree

34 (9.9)

10 (6.3) 1 (0.6)

5 (5.2)

Unknown

0 (0)

1 (1)

Visit type

<.001

Reconstructive

207 (59.5)

43 (27.2) 69 (69.0)

Cosmetic

141 (40.5) 115 (72.8) 31 (31.0)

Abbreviation: SD, standard deviation.

Table 2. Participants Screening Positive for Body Dysmorphic Disorder as Determined by the Body Dysmorphic Disorder Questionnaire by Study Site

No. (%)

Site 1 (n = 342)

Site 2 (n = 158)

Site 3 (n = 9)

P Value a

Response

.57

Body Dysmorphic Disorder Questionnaire

Positive Negative

37 (10.8)

13 (8.2)

8 (8.2)

305 (89.2) 145 (91.8) 89 (91.8)

a P value for heterogeneity in prevalence across sites.

BDDQ Screening A total of 58 patients (9.7%) across study sites screened posi- tive for BDD based on the BDDQ instrument. The proportion of patients screening positive at each study site is shown in Table 2 . There was no significant difference in the proportion

(Reprinted) JAMA Facial Plastic Surgery Published online December 8, 2016

jamafacialplasticsurgery.com

Copyright 2016 American Medical Association. All rights reserved.

153