There are over 1 million transgender people living in the United States, and 33% report negative experiences with a healthcare provider, many of which are connected to data representation in electronic health records (EHRs). We present recommendations and common pitfalls involving sex- and gender-related data collection in EHRs. Our recommendations leverage the needs of patients, medical providers, and researchers to optimize both individual patient experiences and the efficacy and reproducibility of EHR population-based studies. We also briefly discuss adequate additions to the EHR considering name and pronoun usage. We add the disclaimer that these questions are more complex than commonly assumed. We conclude that collaborations between local transgender and gender-diverse persons and medical providers as well as open inclusion of transgender and gender-diverse individuals on terminology and standards boards is crucial to shifting the paradigm in transgender and gender-diverse health.

Training GPs with the reattribution training package appears to be extremely cost-effective.

Objectives To assess fairness and bias of a previously validated machine learning opioid misuse classifier. Materials & Methods Two experiments were conducted with the classifier’s original (n = 1000) and external validation (n = 53 974) datasets from 2 health systems. Bias was assessed via testing for differences in type II error rates across racial/ethnic subgroups (Black, Hispanic/Latinx, White, Other) using bootstrapped 95% confidence intervals. A local surrogate model was estimated to interpret the classifier’s predictions by race and averaged globally from the datasets. Subgroup analyses and post-hoc recalibrations were conducted to attempt to mitigate biased metrics. Results We identified bias in the false negative rate (FNR = 0.32) of the Black subgroup compared to the FNR (0.17) of the White subgroup. Top features included “heroin” and “substance abuse” across subgroups. Post-hoc recalibrations eliminated bias in FNR with minimal changes in other subgroup error metrics. The Black FNR subgroup had higher risk scores for readmission and mortality than the White FNR subgroup, and a higher mortality risk score than the Black true positive subgroup (P < .05). Discussion The Black FNR subgroup had the greatest severity of disease and risk for poor outcomes. Similar features were present between subgroups for predicting opioid misuse, but inequities were present. Post-hoc mitigation techniques mitigated bias in type II error rate without creating substantial type I error rates. From model design through deployment, bias and data disadvantages should be systematically addressed. Conclusion Standardized, transparent bias assessments are needed to improve trustworthiness in clinical machine learning models.

No abstract

Objective: In 2017 an academic health center in Chicago launched the multidisciplinary Substance Use Intervention Team (SUIT) to address opioid misuse across 18 inpatient units and in a new outpatient addiction medicine clinic. This report assesses the first five months of implementation and associations with patient health and healthcare utilization. Methods: Patient demographic and screening data were extracted from the administrative data warehouse of the electronic health record infrastructure. Distribution of sample characteristics for positive initial screens for opioid misuse was tested against those of all patients screened using a two-tailed test of proportions (p < 0.05). A second analysis compared length of stay and 30-day readmissions within a cohort of patients with a secondary diagnosis of substance use disorder. Results: Between November 2017-March 2018, 76% of 15,054 unique patients were screened, 578 had positive scores on the AUDIT and DAST, 131 had positive scores for opioid misuse, and 52 patients initiated medication treatment. Patients with a secondary diagnosis of substance use disorder who received a SUIT consult (n = 161), compared to those who did not (n = 612), had a shorter average length of stay (5.91 v. 6.73 days) and lower 30-day readmission rate (13.6% v. 15.7%). Conclusion: Leveraging the electronic health record to conduct standardized screenings and treatment has helped identify an at-risk population, disproportionately younger, black, and male, and treat new cases of opioid and substance misuse. The intervention indicates trends toward a shortened length of stay, reduced 30-day readmissions, and has linked patients to outpatient care.

Purpose: In 2015, the Centers for Medicare and Medicaid Services ruled that health organizations comply with additional requirements for electronic health records (EHRs), known as “Meaningful Use,” and develop the capacity to collect gender identity data. Research has established effectiveness of a two-step gender identity question to collect these data. This study examines transgender patient perspectives on the use of a two-step question and experiences with privacy and sensitive disclosures in EHRs and healthcare settings.Methods: Four focus groups (N=30) were conducted in Chicago, Illinois in 2014–2015. Participants were asked to compare two intake forms—one with a two-step question and one with a single question—and discuss experiences with gender identity disclosure, privacy, and access to care. Narratives were transcribed verbatim to identify patterns and themes; the extended case method was used and grounded the data analysis process in the concept of intersectionality.Results: Participants expressed appreciation for improved reliability and competencies that the two-part question may afford. Narratives reveal concerns related to patient privacy, safety, and access because of the contexts in which these data are collected and transmitted. Virtually all participants described situations whereby sensitive gender identity information had been involuntarily disclosed, misinterpreted, or abused, and safety and care were compromised.Conclusion: Participants recognized the potential of the two-part question as a measurement and competency tool, but anticipated new privacy violations and involuntary disclosures. Narratives indicate that effects of sensitive disclosures may vary intersectionally, whereby white participants experienced lesser harms than their immigrant, HIV-positive, and black trans feminine counterparts. Discrimination and privacy violations may occur regardless of a two-part or one-part gender identity question, but increasing these sensitive disclosures within expanding EHR infrastructures may require a range of mechanisms that have flexibility across contexts to safeguard sensitive information and access to care.

Objective: This study analyzes effectiveness of screening, referrals, and treatment uptake of a collaborative care for depression intervention across 10 primary care clinics in Chicago.Methods: Patients were screened with the PHQ-2/9 based on an eligibility algorithm. Electronic health record data were analyzed for sample characteristics, screening rates, referrals, and treatment pathways. To identify disparities, a test of proportions was conducted between eligible and screened patients and referred and treated patients.Results: From November 2016 -December 2017, 25,369 patients were eligible for screening, and rates rose to 79%, versus 7% in the prior year. Screenings, referrals, and uptake occurred proportionately across subgroups except for patients ages 12-17. Adolescent age was associated with disproportionate PHQ-9 screenings and with treatment disengagement. Conclusion:The intervention shows promise in expanding access to care and reducing disparities. Greater access to psychotherapies and innovative treatment modalities, particularly for adolescents, may improve overall treatment uptake.

Purpose In response to the opioid crisis, public health advocates urge hospitals to perform substance use disorder (SUD) screening, brief intervention, discharge planning with referral to treatment, and naloxone education. Universal screening makes specialized treatment available to all patients and decreases stigma around SUDs, allowing patients and providers to address SUDs during their hospitalization. Additionally, hospital and emergency department–initiated medications to treat SUD improve patient engagement with treatment and decrease opioid use, and use of medications for opioid use disorder after nonfatal overdoses decreases mortality. Summary A substance use intervention team (SUIT) service was established to offer universal screening and consultation by an interdisciplinary team at our urban academic medical center. The SUIT program provides inpatient consultation services as well as medical and behavioral clinic visits to transition patients to long-term treatment and is comprised of physicians, nurse practitioners, a clinical pharmacist, social workers, and a nurse. Successes attributed to enhanced medication use as a function of having a designated pharmacist as an integral member of the team are highlighted. Our medical center initiated screening efforts in tandem with its interdisciplinary team and clinic. The team attempts to start appropriately selected patients with SUD on medications for SUD while hospitalized. From January through December 2018, 87.2% of patients admitted to the hospital received initial SUD screening. Of the patients who screened positive, 1,400 received a brief intervention by a unit social worker; the SUIT service was consulted on 880 patients, and multiple medications for SUD were started during inpatient care. Conclusion A screening, brief intervention, and referral to treatment service was successfully implemented in our hospital, with the SUIT program in place to provide interdisciplinary addiction care and initiate medications for SUD in appropriate patients.