A Common Longitudinal Intensive Care Unit data Format (CLIF) to enable multi-institutional federated critical illness research.

Background: Critical illness, or acute organ failure requiring life support, threatens over five million American lives annually. Electronic health record (EHR) data are a source of granular information that could generate crucial insights into the nature and optimal treatment of critical illness. However, data management, security, and standardization are barriers to large-scale critical illness EHR studies. Methods: A consortium of critical care physicians and data scientists from eight US healthcare systems developed the Common Longitudinal Intensive Care Unit (ICU) data Format (CLIF), an open-source database format that harmonizes a minimum set of ICU Data Elements for use in critical illness research. We created a pipeline to process adult ICU EHR data at each site. After development and iteration, we conducted two proof-of-concept studies with a federated research architecture: 1) an external validation of an in-hospital mortality prediction model for critically ill patients and 2) an assessment of 72-hour temperature trajectories and their association with mechanical ventilation and in-hospital mortality using group-based trajectory models. Results: We converted longitudinal data from 94,356 critically ill patients treated in 2020-2021 (mean age 60.6 years [standard deviation 17.2], 30% Black, 7% Hispanic, 45% female) across 8 health systems and 33 hospitals into the CLIF format, The in-hospital mortality prediction model performed well in the health system where it was derived (0.81 AUC, 0.06 Brier score). Performance across CLIF consortium sites varied (AUCs: 0.74-0.83, Brier scores: 0.06-0.01), and demonstrated some degradation in predictive capability. Temperature trajectories were similar across health systems. Hypothermic and hyperthermic-slow-resolver patients consistently had the highest mortality. Conclusions: CLIF facilitates efficient, rigorous, and reproducible critical care research. Our federated case studies showcase CLIF's potential for disease sub-phenotyping and clinical decision-support evaluation. Future applications include pragmatic EHR-based trials, target trial emulations, foundational multi-modal AI models of critical illness, and real-time critical care quality dashboards.

Association between Dexmedetomidine Use and Mortality in Patients with COVID-19 Receiving Invasive Mechanical Ventilation: A U.S. National COVID Cohort Collaborative (N3C) Study.

(1) Background/Objectives: Dexmedetomidine is a sedative for patients receiving invasive mechanical ventilation (IMV) that previous single-site studies have found to be associated with improved survival in patients with COVID-19. The reported clinical benefits include dampened inflammatory response, reduced respiratory depression, reduced agitation and delirium, improved preservation of responsiveness and arousability, and improved hypoxic pulmonary vasoconstriction and ventilation-perfusion ratio. Whether improved mortality is evident in large, multi-site COVID-19 data is understudied. (2) Methods: The association between dexmedetomidine use and mortality in patients with COVID-19 receiving IMV was assessed. This retrospective multi-center cohort study utilized patient data in the United States from health systems participating in the National COVID Cohort Collaborative (N3C) from 1 January 2020 to 3 November 2022. The primary outcome was 28-day mortality rate from the initiation of IMV. Propensity score matching adjusted for differences between the group with and without dexmedetomidine use. Adjusted hazard ratios (aHRs) for 28-day mortality were calculated using multivariable Cox proportional hazards models with dexmedetomidine use as a time-varying covariate. (3) Results: Among the 16,357,749 patients screened, 3806 patients across 17 health systems met the study criteria. Mortality was lower with dexmedetomidine use (aHR, 0.81; 95% CI, 0.73-0.90; p < 0.001). On subgroup analysis, mortality was lower with earlier dexmedetomidine use-initiated within the median of 3.5 days from the start of IMV-(aHR, 0.67; 95% CI, 0.60-0.76; p < 0.001) as well as use prior to standard, widespread use of dexamethasone for patients on respiratory support (prior to 30 July 2020) (aHR, 0.54; 95% CI, 0.42-0.69; p < 0.001). In a secondary model that was restricted to 576 patients across six health system sites with available PaO2/FiO2 data, mortality was not lower with dexmedetomidine use (aHR 0.95, 95% CI, 0.72-1.25; p = 0.73); however, on subgroup analysis, mortality was lower with dexmedetomidine use initiated earlier than the median dexmedetomidine start time after IMV (aHR, 0.72; 95% CI, 0.53-0.98; p = 0.04) and use prior to 30 July 2020 (aHR, 0.22; 95% CI, 0.06-0.78; p = 0.02). (4) Conclusions: Dexmedetomidine use was associated with reduced mortality in patients with COVID-19 receiving IMV, particularly when initiated earlier, rather than later, during the course of IMV as well as use prior to the standard, widespread usage of dexamethasone during respiratory support. These particular findings might suggest that the associated mortality benefit with dexmedetomidine use is tied to immunomodulation. However, further research including a large randomized controlled trial is warranted to evaluate the potential mortality benefit of DEX use in COVID-19 and evaluate the physiologic changes influenced by DEX that may enhance survival.

Studying Outcomes after Steroid-Sparing Immunosuppressive Agent vs. Steroid-Only Treatment for Immune-Related Adverse Events in Non-Small-Cell Lung Cancer (NSCLC) and Melanoma: A Retrospective Case-Control Study.

BACKGROUND: The effects of steroid-sparing immunosuppressive agents (SSIAs), used for the treatment of immune-related adverse events (irAEs), on immune checkpoint inhibitor (ICI) antitumor activity is not well known. We compared tumor outcomes of patients who received corticosteroid monotherapy (CS) versus a corticosteroid plus SSIA (CS-SSIA) for irAE treatment, using statistical methods to address immortal time bias. METHODS: We conducted a retrospective case-control study on patients ≥ 18 years with melanoma or non-small-cell lung cancer (NSCLC) treated with ≥1 ICI at a quaternary care center between 1 January 2016 and 11 January 2021. Patients were divided into two cohorts: CS or CS-SSIA. We used propensity score nearest-neighbor matching to match on tumor type, stage, and prior lines of therapy. Primary outcomes were progression-free survival (PFS) and overall survival (OS). Secondary outcomes included the time from the start of the irAE treatment to the irAE resolution. Hazard ratios (HRs) for PFS and OS were calculated using the Cox proportional hazard regression method with both (1) the time to the steroid and SSIA as time-varying covariates and (2) a binary exposure classification not accounting for the time to the treatment. RESULTS: A total of 167 patients were included after matching (132 in the CS cohort and 35 in the CS-SSIA cohort). Sixty-six percent of all the patients had melanoma. The most common irAEs requiring treatment were gastroenterocolitis and hepatitis. In an adjusted analysis not accounting for immortal time bias, there were no significant differences in PFS (HR 0.75, 95% CI [0.46-1.23]) or OS (HR 0.82, 95% CI [0.46-1.47]). In analyses using a time-varying treatment indicator, there was a trend toward improved PFS in patients treated with SSIAs (HR 0.54, CI 0.26-1.10). There was no difference in OS (HR 1.11, CI 0.55-2.23). Patients with melanoma who specifically received infliximab had improved PFS compared to patients with CS only, after adjusting for immortal time bias (HR 0.32, CI 0.24-0.43). CONCLUSIONS: The use of SSIAs with CS did not have worse outcomes than CS monotherapy. In melanoma, our findings showed improved PFS for the use of infliximab versus steroid monotherapy for irAEs. Large, prospective, randomized controlled trials are needed to confirm these findings and guide the optimal treatment of irAEs.

When is vancomycin prophylaxis necessary? Risk factors for MRSA surgical site infection.

Background: The 2022 SHEA/IDSA/APIC guidance for surgical site infection (SSI) prevention recommends reserving vancomycin prophylaxis to patients who are methicillin-resistant Staphylococcus aureus (MRSA) colonized. Unfortunately, vancomycin prophylaxis remains common due to the overestimation of MRSA risk and the desire to cover MRSA in patients with certain healthcare-associated characteristics. To optimize vancomycin prophylaxis, we sought to identify risk factors for MRSA SSI. Methods: This was a single-center, case-control study of patients with a postoperative SSI after undergoing a National Healthcare Safety Network operative procedure over eight years. MRSA SSI cases were compared to non-MRSA SSI controls. Forty-two demographic, medical, and surgical characteristics were evaluated. Results: Of the 441 patients included, 23 developed MRSA SSIs (rate = 5.2 per 100 SSIs). In the multivariable model, we identified two independent risk factors for MRSA SSI: a history of MRSA colonization or infection (OR, 9.0 [95% CI, 1.9-29.6]) and hip or knee replacement surgery (OR, 3.8 [95% CI, 1.3-9.9]). Hemodialysis, previous hospitalization, and prolonged hospitalization prior to the procedure had no measurable association with odds of MRSA SSI. Conclusions: Patients with prior MRSA colonization or infection had 9-10 times greater odds of MRSA SSI and patients undergoing hip and knee replacement had 3-4 times greater odds of MRSA SSI. Healthcare-associated characteristics, such as previous hospitalization or hemodialysis, were not associated with MRSA SSI. Our findings support national recommendations to reserve vancomycin prophylaxis for patients who are MRSA colonized, as well as those undergoing hip and knee replacement, in the absence of routine MRSA colonization surveillance.

Suicide Methods and Trends Across Race/Ethnicity, Age, and Sex Groups in Chicago, Illinois, 2015-2021.

Objectives. To measure differences in suicide rates across race/ethnicity, age, and sex groups in Chicago, Illinois, from 2015 to 2021. Methods. We calculated the incidence rate and annual percentage change in suicides among Asian, Black, Latino/a, and White persons in Chicago. We also analyzed patterns in suicide method across race/ethnicity, age, and sex groups. Results. Suicides increased significantly among Black males (incidence rate ratio [IRR] = 1.10; 95% confidence interval [CI] = 1.01, 1.20), Black females (IRR = 1.18; 95% CI = 1.04, 1.33), and Latino males (IRR = 1.23; 95% CI = 1.11, 1.38) between 2015 and 2021. Suicides decreased overall among White Chicagoans during this period. A significantly greater proportion of Black males than Black females died by suicide using a firearm (55.79% vs 24.05%; P < .001). Similar results were detected for Latino males and females (32.99% vs 9.09%; P = .001) and White males and females (30.10% vs 11.73%; P < .001). Conclusions. Black persons in Chicago were the only group to experience significant increases in suicide among both males and females from 2015 to 2021, although specific methods used varied by race/ethnicity and sex group. (Am J Public Health. 2024;114(3):319-328. https://doi.org/10.2105/AJPH.2023.307511).

Reducing Physical Therapy Consults for Patients With High Functional Mobility in the Acute Medical Inpatient Setting: A Difference-in-Difference Analysis.

OBJECTIVE: To evaluate the effectiveness of clinical decision support for reducing misallocation of physical therapy (PT) consults. DESIGN: A prospective quasi-experimental study. Between October 2018 and November 2021, routinely documented data on functional status and physical therapy referrals were collected from electronic medical records. SETTING: Hospital Medicine and General Internal Medicine service lines at a large quaternary academic medical center. PARTICIPANTS: 20,810 adult patients hospitalized on any of the included treatment (hospital medicine) or control (general internal medicine) service lines. MAIN OUTCOME MEASURE: The primary outcome was "change in proportion of misallocated PT consults" measured as likelihood of PT consults for patients admitted with high functional mobility scores. Changes in the primary outcome from the pre-intervention to post-intervention period were compared in the control and treatment groups using propensity score-weighted difference-in-differences multivariable logit regression adjusting for clinically relevant covariates. INTERVENTION: The intervention period was measured for 20 months and consisted of a clinical decision support tool embedded in the daily note templates for hospital medicine providers. The tool provided education on patient mobility scores and their relation to need for PT consult. The tool was rolled out without any further announcements or education. RESULTS: Our cohort included 20,810 unique admissions (mean age 58.9, 55% women, 83% Black). Post-intervention, the likelihood of PT referrals for patients with high baseline mobility (AM-PAC >18) decreased by 7.3% (P<.001) for the treatment group compared with control, adjusted for age, sex, race, ethnicity, length-of-stay, and mobility change. CONCLUSION: Mobility score-based clinical decision support can decrease unneeded PT consults in the inpatient setting. This could help allocate therapy time for at-risk patients while also having a positive effect on health care systems.

Pharmacotherapy and pulmonary fibrosis risk after SARS-CoV-2 infection: a prospective nationwide cohort study in the United States.

Background: Pulmonary fibrosis is characterized by lung parenchymal destruction and can increase morbidity and mortality. Pulmonary fibrosis commonly occurs following hospitalization for SARS-CoV-2 infection. As there are medications that modify pulmonary fibrosis risk, we investigated whether distinct pharmacotherapies (amiodarone, cancer chemotherapy, corticosteroids, and rituximab) are associated with differences in post-COVID-19 pulmonary fibrosis incidence. Methods: We used the National COVID-19 Cohort Collaboration (N3C) Data Enclave, which aggregates and harmonizes COVID-19 data across the United States, to assess pulmonary fibrosis incidence documented at least 60 days after COVID-19 diagnosis among adults hospitalized between January 1st, 2020 and July 6th, 2022 without pre-existing pulmonary fibrosis. We used propensity scores to match pre-COVID-19 drug-exposed and unexposed cohorts (1:1) based on covariates with known influence on pulmonary fibrosis incidence, and estimated the association of drug exposure with risk for post-COVID-19 pulmonary fibrosis. Sensitivity analyses considered pulmonary fibrosis incidence documented at least 30- or 90-days post-hospitalization and pulmonary fibrosis incidence in the COVID-19-negative N3C population. Findings: Among 5,923,394 patients with COVID-19, we analyzed 452,951 hospitalized adults, among whom pulmonary fibrosis incidence was 1.1 per 100-person-years. 277,984 hospitalized adults with COVID-19 were included in our primary analysis, among whom all drug exposed cohorts were well-matched to unexposed cohorts (standardized mean differences <0.1). The post-COVID-19 pulmonary fibrosis incidence rate ratio (IRR) was 2.5 (95% CI 1.2-5.1, P = 0.01) for rituximab, 1.6 (95% CI 1.3-2.0, P < 0.0001) for chemotherapy, and 1.2 (95% CI 1.0-1.3, P = 0.02) for corticosteroids. Amiodarone exposure had no significant association with post-COVID-19 pulmonary fibrosis (IRR = 0.8, 95% CI 0.6-1.1, P = 0.24). In sensitivity analyses, pre-COVID-19 corticosteroid use was not consistently associated with post-COVID-19 pulmonary fibrosis. In the COVID-19 negative hospitalized population (n = 1,240,461), pulmonary fibrosis incidence was lower overall (0.6 per 100-person-years) and for patients exposed to all four drugs. Interpretation: Recent rituximab or cancer chemotherapy before COVID-19 infection in hospitalized patients is associated with increased risk for post-COVID-19 pulmonary fibrosis. Funding: The analyses described in this publication were conducted with data or tools accessed through the NCATS N3C Data Enclave https://covid.cd2h.org and N3C Attribution & Publication Policy v1.2-2020-08-25b supported by NIHK23HL146942, NIHK08HL150291, NIHK23HL148387, NIHUL1TR002389, NCATSU24 TR002306, and a SECURED grant from the Walder Foundation/Center for Healthcare Delivery Science and Innovation, University of Chicago. WFP received a grant from the Greenwall Foundation. This research was possible because of the patients whose information is included within the data and the organizations (https://ncats.nih.gov/n3c/resources/data-contribution/data-transfer-agreement-signatories) and scientists who have contributed to the on-going development of this community resource (https://doi.org/10.1093/jamia/ocaa196).