243 - Identifying Risk Factors for Compromised Adolescent Confidentiality in the Emergency Department: A National Survey of Pediatric Emergency Medicine Clinicians

Publication Number: 3235.243

Daisy Ciener, Monroe Carell Jr. Children's Hospital at Vanderbilt, Nashville, TN, United States; Elizabeth Ebbens, University of Chicago, Oak Forest, IL, United States; Donald H. Arnold, Monroe Carell Jr. Children's Hospital at Vanderbilt, Nashville, TN, United States

Background: Adolescents may legally seek confidential care for sexual/reproductive health, substance use, and mental health concerns. Expansion of electronic health records (EHRs) and patient portals has complicated privacy protection following the 21st Century CURES Act and OpenNotes. Pediatric emergency medicine (PEM) clinicians frequently manage confidential issues but may lack clarity on appropriate disclosure and documentation.

Objective: Assess PEM clinicians' practices for confidential interviewing, knowledge of adolescent privacy rights, and understanding of EHR/portal capabilities for protecting confidential information.

Design/Methods: A 39-item cross-sectional survey was distributed via the Pediatric Emergency Medicine Collaborative Research Committee (PEMCRC). Part A evaluated confidential interview practices including 4 scored vignettes for handling sensitive information. Part B assessed EHR functionality, caregiver portal access, and documentation. States were grouped by adolescent-consent restrictiveness. Univariate associations were tested with chi-square or Wilcoxon-signed-rank, as appropriate. Multivariable/multivariate logistic/ordinal models generated adjusted odds ratios (aORs) with 95% CIs.

Results: Of 669 clinicians contacted, 248 responded (35%). Most were female attending physicians in urban academic centers within permissive states. Confidential interviews occurred more often with female patients (p < 0.001). Clinicians frequently provided confidential care for STIs (72%) and mental health (64%). 49% answered ≥1 vignette incorrectly. Although >90% reported portal availability, 40% were unsure if adolescent permission was required and 32% were unsure what caregivers could view. (Figure 1). EHR training was strongly associated with being sure of portal policies-permission aOR 2.90 (95% CI 1.49-5.62) and caregiver-access aOR 3.92 (1.98-7.76) (Table 1). Portal availability was higher at academic sites aOR 6.47 (1.72-24.32). Only 39% used appropriately segmented confidential notes; no covariates were significant though older clinician age trended toward higher odds aOR 1.37 (0.98-1.91, p = 0.07, Figure 2).

Conclusion(s): PEM clinicians frequently provide confidential care but demonstrate variability in knowledge and documentation practice. Uncertainty about caregiver portal access and permissions persists, particularly among clinicians without EHR privacy training. Despite modest response rate and self-report limitations, these findings identify EHR education and standardized portal privacy functions as modifiable risk factors to better safeguard adolescent information.

Figure 1
While most respondents reported that their institution offered a patient portal, many were unsure about caregiver access policies-40 % did not know whether caregiver permission was required and 32 % were unsure what caregivers could view.

Table 1
Multivariable logistic regression examining predictors of clinicians being knowledgeable about caregiver portal policies.
(A) Is adolescent permission required for caregivers to access the online portal?
(B) When caregivers access the portal, what level of information access do they have?
Models adjusted for clinician age, practice level, PEM board certification, restrictive-law state, academic medical center setting, vignette score, EHR training, and EHR system (Epic vs Cerner). Odds ratios (aOR) with 95 % CIs shown. EHR adolescent specific training was significantly associated with greater certainty about both portal permission (aOR 2.90 [1.49-5.62], p = 0.002) and access level (aOR 3.92 [1.98-7.76], p < 0.001).


Figure 2
Proportion of clinicians segmenting confidential information in separate note documentation versus documenting without protection or segmentation (n = 172). Binary logistic regression model (collapsed documentation outcome) adjusted for covariates including demographics, EHR training, and knowledge score. Older clinician age trended toward higher odds of using appropriately segmented notes (aOR 1.37 [0.98-1.91], p = 0.07); no other covariates were significant.