Adolescents and young adults (AYA) frequently use emergency departments (EDs) for care; utilization is even higher for transgender and nonbinary patients.1  Failure to accurately document demographics may result in delivery of non–patient-centered care and render certain minoritized populations invisible, making it more challenging to uncover and address health inequities.2  We compared concordance between self-reported confidential survey and electronic health record (EHR)-recorded demographic data in AYA who completed a tablet-based sexual health survey (SHS) in the ED.

Medically stable, neurologically intact, English-speaking 15- to 21-year-old patients presenting to 6 pediatric EDs in the Pediatric Emergency Care Applied Research Network between March 2021 and May 2022 were offered a tablet-based SHS.3  Adolescents being assessed for sexual assault were excluded. In addition to questions related to sexual health, adolescents were asked about their current gender identity (female, male, trans-feminine, trans-masculine, queer/gender-nonconforming, other), sex assigned at birth, and patient-identified race/ethnicity. Sex, race, and ethnicity obtained from the EHR at the time of the ED visit were abstracted. Patients were considered transgender if identified as transgender in the SHS or EHR or if EHR documentation listed a different gender identity or sex than the sex at birth. Percent agreement or odds ratios (ORs) using logistic regression were calculated for gender identity and race/ethnicity between the SHS and EHR documentation; patients in whom data were missing from the SHS or in the EHR were excluded from these analyses. One center routinely collected gender identity data in the EHR. Institutional review board approval and survey consent were obtained.

The SHS was completed by 5290 AYA (Table 1); paired gender identity, race, and ethnicity data were available for 4896 (92.6%), 4451 (84.1%), and 4689 (88.6%), respectively (Table 2). Discordant identification by self-report survey compared with EHR data were most common for patients who identified as Hispanic versus non-Hispanic ethnicity (OR, 21.29; 95% confidence interval [CI], 14.8–31.5); Asian (OR, 2.8; 95% CI, 1.8–4.3); or other race (OR, 10.4; 95% CI, 8.2–13.2) versus white race. Discordant gender identification was seen in 81% of trans-masculine and 57% of trans-feminine patients. Agreement in patients aged ≥18 years was higher than for younger adolescents for gender identity (97.1% vs 95.4%, P = .03) and ethnicity (96.7% vs 94.4%, P = .004), respectively; agreement regarding race was not significantly different by age (86.0% vs 84.1%, P = .21).

TABLE 1

Demographics of the Study Population who Completed the Sexual Health Survey (n = 5290)

VariableSubcategoryNo. (%)
Age, y Median (IQR) 16.8 (15.9–17.7) 
15 1478 (27.9%) 
16–17 2923 (55.3%) 
18–21 889 (16.8%) 
Sex (patient-reported) Female 3345 (63.2%) 
Male 1599 (30.2%) 
Missing 346 (6.6%) 
Gender identity (patient-reported) Female 3137 (59.3%) 
Male 1567 (29.6%) 
Trans-feminine 7 (0.1%) 
Trans-masculine 42 (0.8%) 
Gender queer/nonconforming 96 (1.8%) 
Other 49 (0.9%) 
Prefer not to answer 34 (0.6%) 
Missing 358 (6.8%) 
Race (patient-reported)a Black 2070 (39.1%) 
White 1934 (36.6%) 
Asian 130 (2.5%) 
Otherb 470 (8.9%) 
Prefer not to answer 227 (4.3%) 
Missing 459 (8.7%) 
Ethnicity (patient-reported)a Non-Hispanic 3499 (66.1%) 
Hispanic 1251 (23.6%) 
Missing 540 (10.2%) 
Insurance type Government 3044 (57.5%) 
Commercial 2118 (40.0%) 
Self-pay 111 (2.1%) 
Other 17 (0.3%) 
Triage category 1: resuscitation 2 (0.0%) 
2: emergent 1153 (21.8%) 
3: urgent 2691 (50.9%) 
4: less urgent 1305 (24.7%) 
5: nonurgent 139 (2.6%) 
ED sitec 1850 (35.0%) 
647 (12.2%) 
279 (5.3%) 
494 (9.3%) 
1815 (34.3%) 
205 (3.9%) 
VariableSubcategoryNo. (%)
Age, y Median (IQR) 16.8 (15.9–17.7) 
15 1478 (27.9%) 
16–17 2923 (55.3%) 
18–21 889 (16.8%) 
Sex (patient-reported) Female 3345 (63.2%) 
Male 1599 (30.2%) 
Missing 346 (6.6%) 
Gender identity (patient-reported) Female 3137 (59.3%) 
Male 1567 (29.6%) 
Trans-feminine 7 (0.1%) 
Trans-masculine 42 (0.8%) 
Gender queer/nonconforming 96 (1.8%) 
Other 49 (0.9%) 
Prefer not to answer 34 (0.6%) 
Missing 358 (6.8%) 
Race (patient-reported)a Black 2070 (39.1%) 
White 1934 (36.6%) 
Asian 130 (2.5%) 
Otherb 470 (8.9%) 
Prefer not to answer 227 (4.3%) 
Missing 459 (8.7%) 
Ethnicity (patient-reported)a Non-Hispanic 3499 (66.1%) 
Hispanic 1251 (23.6%) 
Missing 540 (10.2%) 
Insurance type Government 3044 (57.5%) 
Commercial 2118 (40.0%) 
Self-pay 111 (2.1%) 
Other 17 (0.3%) 
Triage category 1: resuscitation 2 (0.0%) 
2: emergent 1153 (21.8%) 
3: urgent 2691 (50.9%) 
4: less urgent 1305 (24.7%) 
5: nonurgent 139 (2.6%) 
ED sitec 1850 (35.0%) 
647 (12.2%) 
279 (5.3%) 
494 (9.3%) 
1815 (34.3%) 
205 (3.9%) 

ED, emergency department; IQR, interquartile range; N/A, not answered.

a

Race and ethnicity were asked as separate questions; for race, the question was phrased to indicate all that applied.

b

Included American Indian/Alaskan Native, multiracial, or patients who indicated “other.”

c

The sites were: Children’s Hospital of Colorado (Aurora [Denver], CO); Children’s Hospital of Philadelphia (Philadelphia, PA); Lurie Children’s Hospital (Chicago, IL); Medical College of Wisconsin (Milwaukee, WI); Nationwide Children’s Hospital (Columbus, OH); and Texas Children’s Hospital (Houston, TX). Only Chicago provided nonbinary classifications (transfeminine, transmasculine, and other).

TABLE 2

Agreement Between Self-Reported and EHR Documentation of Gender Identity, Race, and Ethnicity

Patient Self-ReportEHR DocumentationPercent Agreement
Gender IdentityFemaleMaleTrans-feminineTrans-masculineGender NonconformingOtherN/ATotal
Female 3127 (99.7%) 10 (0.3%) 3137 95.7 
Male 13 (0.8%) 1554 (99.2%) 1567 
Trans-feminine 3 (42.9%) 4 (57.1%) 
Trans-masculine 34 (81.0%) 7 (16.7%) 1 (2.4%) 42 
Gender non-conforming 79 (82.3%) 14 (14.6%) 3 (3.1%) 96 
Other 40 (81.6%) 5 (10.2%) 2 (4.1%) 2 (4.1%) 49 
Prefer not to answer 31 (91.2%) 3 (8.8%) 34 
Missing 215 (60.1%) 143 (39.9%) 358 
Race Black White Asian Other Multiracial N/A Total  Percent agreement 
 Black 1799 (86.9%) 87 (4.2%) 4 (0.2%) 122 (5.9%) 42 (2.0%) 16 (0.8%) 2070  84.4 
 White 9 (0.5%) 1684 (87.1%) 5 (0.3%) 197 (10.2%) 22 (1.1%) 17 (0.9%) 1934  
 Asian 1 (0.8%) 3 (2.3%) 88 (67.7%) 27 (20.8%) 6 (4.6%) 5 (3.8%) 130  
 Other 34 (7.2%) 199 (42.3%) 7 (1.5%) 185 (39.4%) 18 (3.8%) 27 (5.7%) 470  
 Prefer not to answer 29 (12.8%) 79 (34.8%) 2 (0.9%) 107 (47.1%) 5 (2.2%) 5 (2.2%) 227  
 Missing 142 (30.9%) 270 (58.8%) 6 (1.3%) 33 (7.2%) 2 (0.4%) 6 (1.3%) 459  
Ethnicity Hispanic Non-Hispanic N/A Total     Percent agreement 
 Hispanic 1030 (82.3%) 212 (16.9%) 9 (0.7%) 1251     94.8 
 Non-Hispanic 33 (0.9%) 3414 (97.6%) 52 (1.4%) 3499     
 Missing 93 (17.2%) 440 (81.5%) 7 (1.3%) 540     
Patient Self-ReportEHR DocumentationPercent Agreement
Gender IdentityFemaleMaleTrans-feminineTrans-masculineGender NonconformingOtherN/ATotal
Female 3127 (99.7%) 10 (0.3%) 3137 95.7 
Male 13 (0.8%) 1554 (99.2%) 1567 
Trans-feminine 3 (42.9%) 4 (57.1%) 
Trans-masculine 34 (81.0%) 7 (16.7%) 1 (2.4%) 42 
Gender non-conforming 79 (82.3%) 14 (14.6%) 3 (3.1%) 96 
Other 40 (81.6%) 5 (10.2%) 2 (4.1%) 2 (4.1%) 49 
Prefer not to answer 31 (91.2%) 3 (8.8%) 34 
Missing 215 (60.1%) 143 (39.9%) 358 
Race Black White Asian Other Multiracial N/A Total  Percent agreement 
 Black 1799 (86.9%) 87 (4.2%) 4 (0.2%) 122 (5.9%) 42 (2.0%) 16 (0.8%) 2070  84.4 
 White 9 (0.5%) 1684 (87.1%) 5 (0.3%) 197 (10.2%) 22 (1.1%) 17 (0.9%) 1934  
 Asian 1 (0.8%) 3 (2.3%) 88 (67.7%) 27 (20.8%) 6 (4.6%) 5 (3.8%) 130  
 Other 34 (7.2%) 199 (42.3%) 7 (1.5%) 185 (39.4%) 18 (3.8%) 27 (5.7%) 470  
 Prefer not to answer 29 (12.8%) 79 (34.8%) 2 (0.9%) 107 (47.1%) 5 (2.2%) 5 (2.2%) 227  
 Missing 142 (30.9%) 270 (58.8%) 6 (1.3%) 33 (7.2%) 2 (0.4%) 6 (1.3%) 459  
Ethnicity Hispanic Non-Hispanic N/A Total     Percent agreement 
 Hispanic 1030 (82.3%) 212 (16.9%) 9 (0.7%) 1251     94.8 
 Non-Hispanic 33 (0.9%) 3414 (97.6%) 52 (1.4%) 3499     
 Missing 93 (17.2%) 440 (81.5%) 7 (1.3%) 540     

% reflects within-row percentages and may not sum to 100% because of rounding. The percent agreement figures exclude adolescents for whom paired data from self-report and the EHR were unavailable (including self-report of “I prefer not to answer” and missing responses, as well as EHR documentation of unknown/missing). EHR, electronic health record; N/A, not available.

There are several potential reasons for discordance between AYA identification and EHR documentation. EHR categories may reflect parental identification for minors. Furthermore, staff entering data into the EHR may document gender, race, and ethnicity based on their perceptions of a patient’s appearance rather than self-report. Because race and ethnicity are social constructs, classifications used in the United States may not resonate with families from other countries.4  We surveyed adolescents who spoke English, but linguistic barriers to accurate documentation of demographic variables may have been present for families speaking languages other than English.5  Additionally, there is fluidity in how adolescents self-classify race/ethnicity and gender identity over time.6,7 

At least 2 unique barriers to accurate responses to the gender identity question may exist. First, only 1 of 6 children’s hospitals had the option of documenting gender identity at the time our study was conducted. Therefore, most transgender and nonbinary youth (∼2.5% of our cohort, in line with previous estimates8 ) would be misclassified in EHR data based only on their sex assigned at birth. Second, asking the gender identity question to parents will misclassify adolescents who have not disclosed their gender identity; in 1 survey, almost one-half of patients avoided disclosure to health care providers.9  The percentage agreement in our study may be higher for older AYA because the question is being asked directly to the patient, possibly without a parent or caregiver present.

There were limitations. Younger adolescents may not have understood demographic questions. There may have been variation in how these variables were documented in the EHR (eg, asking the family versus entered by staff) and whether variables were revisited after being initially entered.

Our study emphasizes the importance of obtaining demographic data by self-report at each visit, in a manner considering the developmental context of identity development and respecting adolescent confidentiality. Accurate collection of race/ethnicity and gender identity is essential to both provide the highest quality care and uncover and address health inequities for minoritized populations.

Dr Cruz conceptualized the study, drafted the initial manuscript, and reviewed and revised the manuscript; Dr Dowshen substantially contributed to study design and contributed substantially to critical review of the manuscript draft for important intellectual content; Drs Mollen, Pickett, Augustine, Stuckus, and Schmidt coordinated data acquisition from their hospitals and contributed substantially to critical review of the manuscript draft for important intellectual content; Ms Elsholz coordinated and supervised data acquisition across all sites and contributed substantially to critical review of the manuscript draft for important intellectual content; Dr Casper and Ms Palmer coordinated and supervised data analysis and interpretation and contributed substantially to critical review of the manuscript draft for important intellectual content; Drs Goyal and Reed conceptualized and designed the parent study, coordinated and designed the data collection instruments, and reviewed and revised the manuscript for important intellectual content; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

FUNDING: This study was funded by National Institutes of Health (NIH) grant R01HD094213 (Goyal/Reed). Pediatric Emergency Care Applied Research Network is supported by the Health Resources and Services Administration (HRSA) of the US Department of Health and Human Services (HHS), in the Maternal and Child Health Bureau (MCHB), under the Emergency Medical Services for Children program through the following cooperative agreements: DCC: University of Utah; GLEMSCRN: Nationwide Children’s Hospital; HOMERUN: Cincinnati Children’s Hospital Medical Center; PEMNEWS: Columbia University Medical Center; PRIME: University of California at Davis Medical Center; CHaMP node: State University of New York at Buffalo; WPEMR: Seattle Children's Hospital; and SPARC: Rhode Island Hospital/Hasbro Children's Hospital.

CONFLICT OF INTEREST DISCLOSURES: Drs Cruz and Goyal are on the editorial board for Pediatrics. The other authors have indicated they have no potential conflicts of interest to disclose. This information or content and conclusions are those of the author and should not be construed as the official position or policy of, nor should any endorsements be inferred by HRSA, HHS, or the US government.

AYA

adolescents and young adults

CI

confidence interval

ED

emergency department

EHR

electronic health record

OR

odds ratio

SHS

sexual health screen

1
Goyal
MK
,
Richardson
T
,
Masonbrink
A
, et al
.
Reliance on acute care settings for health care utilization: a comparison of adolescents with younger children
.
Pediatr Emerg Care
.
2021
;
37
(
12
):
e1128
e1132
2
Guss
CE
,
Eiduson
R
,
Khan
A
,
Dumont
O
,
Forman
SF
,
Gordon
AR
.
“It’d be great to have the options there:” a mixed-method study of gender identity questions on clinic forms in a primary care setting
.
J Adolesc Health
.
2020
;
67
(
4
):
590
596
3
Howe
JL
,
Reed
JL
,
Adams
KT
, et al
.
Workflow analysis driven recommendations for integration of electronically-enhanced sexually transmitted infection screening tools in pediatric emergency departments
.
J Med Syst
.
2020
;
44
(
12
):
206
4
Noe-Bustamante
L
,
Gonzalez-Barrera
A
,
Edwards
K
, et al
.
Pew Research Center.
Measuring the racial identity of Latinos. Available at: https://www.pewresearch.org/hispanic/2021/11/04/measuring-the-racial-identity-of-latinos/. Accessed May 15, 2023
5
Gore
A
,
Truche
P
,
Iskerskiy
A
,
Ortega
G
,
Peck
G
.
Inaccurate ethnicity and race classification of Hispanics following trauma admission
.
J Surg Res
.
2021
;
268
:
687
695
6
Booth
MZ
,
Gerard
JM
,
Deom
GM
,
Frey
CJ
.
Deconstructing adolescent ethnicity: a longitudinal analysis of ethnic-racial identity development and self-categorization
.
J Adolesc
.
2022
;
94
(
3
):
366
379
7
Katz-Wise
SL
,
Perry
NS
,
Nelson
KM
,
Gordon
AR
,
Ybarra
ML
.
Sexual fluidity in identity and behavior among cisgender youth: findings from the longitudinal Growing up with Media study
.
J Pediatr
.
2023
;
257
:
113355
8
Herman
JL
,
Flores
AR
,
O’Neill
KK
.
How many adults and youth identify as transgender in the United States? Available at: https://williamsinstitute.law.ucla.edu/wp-content/uploads/Trans-Pop-Update-Jun-2022.pdf. Accessed June 8, 2023
9
Sequeira
GM
,
Ray
KN
,
Miller
E
,
Coulter
RWS
.
Transgender youth’s disclosure of gender identity to providers outside of specialized gender center
.
J Adolesc Health
.
2020
;
66
(
6
):
691
698