Determining the Electronic Signature of Infection in Electronic Health Record Data.

OBJECTIVES: Recent sepsis studies have defined patients as "infected" using a combination of culture and antibiotic orders rather than billing data. However, the accuracy of these definitions is unclear. We aimed to compare the accuracy of different established criteria for identifying infected patients using detailed chart review. DESIGN: Retrospective observational study. SETTING: Six hospitals from three health systems in Illinois. PATIENTS: Adult admissions with blood culture or antibiotic orders, or Angus International Classification of Diseases infection codes and death were eligible for study inclusion as potentially infected patients. Nine-hundred to 1,000 of these admissions were randomly selected from each health system for chart review, and a proportional number of patients who did not meet chart review eligibility criteria were also included and deemed not infected. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The accuracy of published billing code criteria by Angus et al and electronic health record criteria by Rhee et al and Seymour et al (Sepsis-3) was determined using the manual chart review results as the gold standard. A total of 5,215 patients were included, with 2,874 encounters analyzed via chart review and a proportional 2,341 added who did not meet chart review eligibility criteria. In the study cohort, 27.5% of admissions had at least one infection. This was most similar to the percentage of admissions with blood culture orders (26.8%), Angus infection criteria (28.7%), and the Sepsis-3 criteria (30.4%). Sepsis-3 criteria was the most sensitive (81%), followed by Angus (77%) and Rhee (52%), while Rhee (97%) and Angus (90%) were more specific than the Sepsis-3 criteria (89%). Results were similar for patients with organ dysfunction during their admission. CONCLUSIONS: Published criteria have a wide range of accuracy for identifying infected patients, with the Sepsis-3 criteria being the most sensitive and Rhee criteria being the most specific. These findings have important implications for studies investigating the burden of sepsis on a local and national level.

Distinguishing pulmonary hypertension in interstitial lung disease by ventilation and perfusion defects measured by cardiopulmonary exercise testing.

BACKGROUND: Pulmonary hypertension (PH) is common in interstitial lung disease (ILD). Since cardiopulmonary exercise testing (CPET) is useful in understanding the pathophysiology of respiratory disorders and can distinguish between ventilation and perfusion (V/Q) defects, it may have a role in the detection of PH in ILD. We evaluated whether CPET can detect PH through analysis of V/Q defects in ILD. OBJECTIVES: We aimed to use CPET to determine if there are changes in the ventilation and the activity pattern of mixed-expired carbon dioxide pressure (PECO2) and end-tidal carbon dioxide pressure (PetCO2) in ILD patients with and without PH. METHODS: A retrospective chart review was done of all patients who received lung transplants at the Columbia University Medical Center between 2000 and 2011 with the diagnosis of ILD. CPETs were performed during the 2 years prior to transplantation; right heart catheterizations and pulmonary function tests were performed within 4 months of CPET. RESULTS: The ILD patients with PH demonstrated significantly lower PetCO2 and PECO2 during certain levels of exercise with a distinctive activity pattern for PECO2/PetCO2. CONCLUSIONS: Evaluation of V/Q defects through the PECO2 and PetCO2 patterns on CPET in ILD patients can distinguish between patients with and without PH.

Impact of Sexual Assault Survivor Identity on Patient Care in the Emergency Department.

Interactions between emergency department (ED) staff and sexual assault (SA) survivors can be a source of retraumatization for survivors, increasing their risk of posttraumatic stress and decreasing utilization of longitudinal medical care. Little is known about nationwide trends in ED staff attitudes and behaviors toward survivors, including the impact of survivor identity. We conducted a survey to determine if survivor identity influenced ED staff behaviors. A nationwide survey of SA patient advocates was conducted between June and August 2021. Advocates are volunteers or staff dispatched from rape crisis centers to support survivors during ED care. Advocates participated in an online survey to determine the frequency of observing six potentially retraumatizing provider attitudes and behaviors. Adaptive questioning was used to explore the impact of survivor identity on each attitude or behavior and which identity groups were more likely to be affected. Three hundred fifteen advocates responded to the survey. Less than 10% indicated that ED staff often or always convey disbelief or blame to survivors. Almost 75% of advocates indicated that they often or always observe ED staff showing empathy to survivors. Disparities were found in provider attitudes. Over 75% of advocates observed that survivors' mental health status or substance use impacted conveyed belief from providers. Patients who were intoxicated when assaulted, had psychiatric disorders, were Black, Hispanic/LatinX, or indigenous, or were not cis-females were more likely to experience disbelief. Patients who were white and/or cis-gender females were more likely to be pressured by ED staff to complete the forensic exam and/or report to the police. Our study documents disparities in ED staff behavior towards SA survivors according to survivor identity. Given that post-assault ED interactions are critical turning points in survivors' future medical processes, disparities in ED care may be linked to larger disparities in healing from trauma.

Patient Advocates' Perspectives on the Care of Sexual Assault Survivors in Chicago-Area Emergency Departments.

Emergency departments (EDs) providing care and forensic examinations for sexual assault (SA) survivors are often supported by SA patient advocates. This study explored advocates' perspectives regarding problems and potential solutions in SA patient care through a focus group with 12 advocates. Thematic analysis identified two major themes: provider-patient interactions and ED-hospital systems. Challenging aspects of provider-patient interactions included (a) provider attitudes and (b) disempowering behaviors. Within ED-hospital systems, themes included time constraints, efficiencies, and hospital preparation. Advocates surveyed were optimistic about an increased presence of SA nurse examiners and enhanced protocols and provider training to improve survivors' experiences.

Emergency Department Preparedness to Care for Sexual Assault Survivors: A Nationwide Study.

INTRODUCTION: Emergency departments (ED) provide trauma-informed care to sexual assault (SA) survivors and connect them with comprehensive services. Through surveying SA survivor advocates, we aimed to 1) document updated trends in the quality of care and resources offered to SA survivors and 2) identify potential disparities according to geographic regions in the US, urban vs rural clinic locations, and the availability of sexual assault nurse examiners (SANE). METHODS: We conducted a cross-sectional study between June-August 2021, surveying SA advocates who were dispatched from rape crisis centers to support survivors during ED care. Survey questions addressed two major themes in quality of care: staff preparedness to provide trauma-response care; and available resources. Staff preparedness to provide trauma-informed care was assessed through observations of staff behaviors. We used Wilcoxon rank-sum and Kruskal-Wallis tests to analyze differences in responses according to geographic regions and SANE presence. RESULTS: A total of 315 advocates from 99 crisis centers completed the survey. The survey had a participation rate of 88.7% and a completion rate of 87.9%. Advocates who indicated that a higher proportion of their cases were attended by SANEs were more likely to report higher rates of trauma-informed staff behaviors. For example, the recalled rate of staff asking patients for consent at every step of the exam was significantly associated with SANE presence (P < 0.001). With respect to access to resources, 66.7% of advocates reported that hospitals often or always have evidence collection kits available; 30.6% reported that resources such as transportation and housing are often or always available, and 55.3% reported that SANEs are often or always part of the care team. The SANEs were reported to be more frequently available in the Southwest than in other US regions (P < 0.001) and in urban as opposed to rural areas (P < 0.001). CONCLUSION: Our study indicates that support from sexual assault nurse examiners is highly associated with trauma-informed staff behaviors and comprehensive resources. Urban-rural and regional disparities exist regarding access to SANEs, suggesting that elevating nationwide quality and equity in care of survivors of sexual assault requires increased investments in SANE training and coverage.

Internal Medicine Resident Barriers to Advance Care Planning in the Primary Care Continuity Clinic.

Background: While primary care providers regularly engage in Advance Care Planning (ACP) conversations, it is not well known what challenges resident physicians face to achieving this core competency. Objectives: We aimed to assess resident perceptions of barriers and potential interventions to outpatient ACP. Methods: We distributed an electronic survey to Internal Medicine and Medicine-Psychiatry residents at our institution in 2022. Questions addressed outpatient ACP barriers and potential interventions in several domains: structural issues, personal knowledge, and communication skills. We reported results using descriptive statistics and Wilcoxon rank-sum tests, comparing responses by residency year (interns vs upperyears). Likert-scale responses were dichotomized to a "not at all or slightly" vs "moderate or extreme" barrier or helpful intervention. Results: Of 149 residents, 71 completed the survey (48%). Highest scoring barriers were structural, including 1) lack of clinic time (99%), 2) need to prioritize other medical problems (94%), and 3) lack of patient continuity (62%). Highest scoring interventions included the ability to schedule dedicated ACP visits with themselves (96%) or another clinician (82%). Interns were statistically significantly less confident in their ability to conduct ACP, and more likely to report lack of knowledge (i.e., not understanding ACP, patient prognosis, or how to complete paperwork, P < .05). Conclusions: Residents report significant structural barriers to outpatient ACP, including limitations in time, continuity, and competing medical priorities, that may warrant greater program attention to interventions such as clinic schedules and work-flow. Additional trainings may be most beneficial if targeted to the beginning of intern year.

Assessing the accuracy of the lung allocation score.

BACKGROUND: The United States (US) Lung Allocation Score (LAS) relies on the performance of 2 survival models that estimate waitlist and post-transplant survival. These models were developed using data from 2005 to 2008, and it is unknown if they remain accurate. METHODS: We performed an observational cohort study of US lung transplantation candidates and recipients greater than 12 years of age between February 19, 2015 and February 19, 2019. We evaluated the LAS waitlist and post-transplant models with the concordance probability estimate and by comparing predicted vs observed 1-year restricted mean survival times by risk decile. We then compared a nonparametric estimate of the observed LAS with the predicted LAS for each percentile of recipients. RESULTS: The waitlist model ranked candidates (N = 11,539) in the correct risk order 72% of the time (95% CI 71%-73%), and underestimated candidate one-year survival by 136 days for the highest risk decile (p < 0.001). The post-transplant model ranked recipients (N = 9,377) in the correct risk order 57% of the time (95% CI 55-58%), and underestimated recipient one-year survival by 70 days for the highest risk decile (p < 0.001). Overall, the LAS at transplant explained only 56% of the variation in observed outcomes, and was increasingly inaccurate at higher predicted values. CONCLUSIONS: The waitlist and the post-transplant models that constitute the LAS are inaccurate, limiting the ability of the system to rank candidates on the waitlist in the correct order. The LAS should therefore be updated and the underlying models should be modernized.

Identifying infected patients using semi-supervised and transfer learning.

OBJECTIVES: Early identification of infection improves outcomes, but developing models for early identification requires determining infection status with manual chart review, limiting sample size. Therefore, we aimed to compare semi-supervised and transfer learning algorithms with algorithms based solely on manual chart review for identifying infection in hospitalized patients. MATERIALS AND METHODS: This multicenter retrospective study of admissions to 6 hospitals included "gold-standard" labels of infection from manual chart review and "silver-standard" labels from nonchart-reviewed patients using the Sepsis-3 infection criteria based on antibiotic and culture orders. "Gold-standard" labeled admissions were randomly allocated to training (70%) and testing (30%) datasets. Using patient characteristics, vital signs, and laboratory data from the first 24 hours of admission, we derived deep learning and non-deep learning models using transfer learning and semi-supervised methods. Performance was compared in the gold-standard test set using discrimination and calibration metrics. RESULTS: The study comprised 432 965 admissions, of which 2724 underwent chart review. In the test set, deep learning and non-deep learning approaches had similar discrimination (area under the receiver operating characteristic curve of 0.82). Semi-supervised and transfer learning approaches did not improve discrimination over models fit using only silver- or gold-standard data. Transfer learning had the best calibration (unreliability index P value: .997, Brier score: 0.173), followed by self-learning gradient boosted machine (P value: .67, Brier score: 0.170). DISCUSSION: Deep learning and non-deep learning models performed similarly for identifying infection, as did models developed using Sepsis-3 and manual chart review labels. CONCLUSION: In a multicenter study of almost 3000 chart-reviewed patients, semi-supervised and transfer learning models showed similar performance for model discrimination as baseline XGBoost, while transfer learning improved calibration.

Pediatric Cancer Data Commons: Federating and Democratizing Data for Childhood Cancer Research.

The international pediatric oncology community has a long history of research collaboration. In the United States, the 2019 launch of the Children's Cancer Data Initiative puts the focus on developing a rich and robust data ecosystem for pediatric oncology. In this spirit, we present here our experience in constructing the Pediatric Cancer Data Commons (PCDC) to highlight the significance of this effort in fighting pediatric cancer and improving outcomes and to provide essential information to those creating resources in other disease areas. The University of Chicago's PCDC team has worked with the international research community since 2015 to build data commons for children's cancers. We identified six critical features of successful data commons design and implementation: (1) establish the need for a data commons, (2) develop and deploy the technical infrastructure, (3) establish and implement governance, (4) make the data commons platform easy and intuitive for researchers, (5) socialize the data commons and create working knowledge and expertise in the research community, and (6) plan for longevity and sustainability. Data commons are critical to conducting research on large patient cohorts that will ultimately lead to improved outcomes for children with cancer. There is value in connecting high-quality clinical and phenotype data to external sources of data such as genomic, proteomics, and imaging data. Next steps for the PCDC include creating an informed and invested data-sharing culture, developing sustainable methods of data collection and sharing, standardizing genetic biomarker reporting, incorporating radiologic and molecular analysis data, and building models for electronic patient consent. The methods and processes described here can be extended to any clinical area and provide a blueprint for others wishing to develop similar resources.

A Needs Assessment and Educational Intervention Addressing the Care of Sexual Assault Patients in the Emergency Department.

Emergency department (ED) providers serve as the primary point-of-contact for many survivors of sexual assault but are often ill-prepared to address their unique treatment needs. Sexual assault nurse examiners (SANEs) are therefore an important resource for training other ED providers. The objective of this project was to create a SANE-led educational intervention addressing this training gap. We achieved this objective by (a) conducting a needs assessment of ED providers' self-reported knowledge of, and comfort with, sexual assault patient care at an urban academic adult ED and, (b) using these results to create and implement a SANE-led educational intervention to improve emergency medicine residents' ability to provide sexual assault patient care. From the needs assessment survey, ED providers reported confidence in medical management but not in providing trauma-informed care, conducting forensic examinations, or understanding hospital policies or state laws. Less than half of the respondents felt confident in their ability to avoid retraumatizing sexual assault patients, and only 29% felt comfortable conducting a forensic examination. On the basis of these results, a SANE-led educational intervention was developed for emergency medicine residents, consisting of a didactic lecture, two standardized patient cases, and a forensic pelvic examination simulation. Preintervention and postintervention surveys showed an increase in respondents' self-perceived ability to avoid retraumatizing patients, comfort with conducting forensic examinations, and understanding of laws and policies. These results show the value of an interprofessional collaboration between physicians and SANEs to train ED providers on sexual assault patient care.

Ventilatory efficiency before and after lung volume reduction surgery.

BACKGROUND: Lung volume reduction surgery (LVRS) has been shown to improve the clearance of carbon dioxide and minute ventilation (V̇E) in select patients with COPD. One variable often assessed in COPD is ventilatory efficiency (V̇E/V̇CO2 ). METHODS: We compared 55 LVRS subjects with 25 controls from the National Emphysema Treatment Trial. V̇E/V̇CO2 was calculated from cardiopulmonary exercise testing at baseline and 6-months. We sought to assess V̇E/V̇CO2 changes with LVRS compared with controls who only received standard medical care. RESULTS: At 6 months, the LVRS group significantly increased peak V̇O2 , work load, V̇E, V̇CO2 , and tidal volume while lowering peak and lowest V̇E/V̇CO2 (improved ventilatory efficiency) and end-tidal carbon dioxide pressure. The control group did not display these changes. The changes were greatest in the LVRS subjects who improved their exercise capacity after surgery (> 10 W). CONCLUSIONS: The changes were greatest in the LVRS subjects who showed the most functional improvement, indicating an association of improved ventilation with improved functional outcome.