ABSTRACT
OBJECTIVES: Persons 65 years and older (older persons), particularly residents of nursing homes (NHs), disproportionately access the emergency department (ED) and utilize more medical resources. The goal of this study is to provide a contemporary description of healthcare utilization patterns and disposition decisions for United States (US) NH residents presenting to EDs. METHODS: Older persons presenting to EDs in the US were identified in the National Hospital Ambulatory Medical Care Survey (NHAMCS) 2017, 2018 and 2019 datasets. We examined demographic, clinical, and resource use characteristics and outcomes. After survey weighting, we compared the frequency of different imaging, medications, clinical interventions, and outcomes in the ED between NH residents and those residing outside NHs. RESULTS: From 2017 to 2019, older persons made 24,441,285 annual visits to the ED, comprising 17.5% of all visits. Among these, 1,579,916 visits (6.5%) were by NH residents. Compared with non-NH residents, NH residents were older (mean age: 81.2 [95%CI 81.5-82.9] vs 76.1 [95%CI 75.8-76.4]), underwent more imaging (82.8% [95%CI 79.5-86.1] vs 71.6% [95%CI 69.9-73.3]), were administered fewer potentially inappropriate medications (PIMs) in the ED or upon discharge (9.5% [95%CI 6.2-2.7] vs 17.1% [95%CI 15.8-18.4]), and had a higher proportion of visits resulting in hospital admission (44.1% [95%CI 38.2-49.9] vs 26.0% [95%CI 23.3, 28.7]). CONCLUSIONS: Older NH residents presenting to the ED use more resources and are more likely to be hospitalized compared to older persons residing outside NHs. The resource-intensive nature of these visits highlights the importance of targeted, multi-disciplinary interventions that optimize ED care for this population.
Subject(s)
Hospitalization , Nursing Homes , Humans , United States , Aged , Aged, 80 and over , Cross-Sectional Studies , Patient Discharge , Emergency Service, HospitalABSTRACT
Rationale: Shorter time-to-antibiotics is lifesaving in sepsis, but programs to hasten antibiotic delivery may increase unnecessary antibiotic use and adverse events. Objectives: We sought to estimate both the benefits and harms of shortening time-to-antibiotics for sepsis. Methods: We conducted a simulation study using a cohort of 1,559,523 hospitalized patients admitted through the emergency department with meeting two or more systemic inflammatory response syndrome criteria (2013-2018). Reasons for hospitalization were classified as septic shock, sepsis, infection, antibiotics stopped early, and never treated (no antibiotics within 48 h). We simulated the impact of a 50% reduction in time-to-antibiotics for sepsis across 12 hospital scenarios defined by sepsis prevalence (low, medium, or high) and magnitude of "spillover" antibiotic prescribing to patients without infection (low, medium, high, or very high). Outcomes included mortality and adverse events potentially attributable to antibiotics (e.g., allergy, organ dysfunction, Clostridiodes difficile infection, and culture with multidrug-resistant organism). Results: A total of 933,458 (59.9%) hospitalized patients received antimicrobial therapy within 48 hours of presentation, including 38,572 (2.5%) with septic shock, 276,082 (17.7%) with sepsis, 370,705 (23.8%) with infection, and 248,099 (15.9%) with antibiotics stopped early. A total of 199,937 (12.8%) hospitalized patients experienced an adverse event; most commonly, acute liver injury (5.6%), new MDRO (3.5%), and Clostridiodes difficile infection (1.7%). Across the scenarios, a 50% reduction in time-to-antibiotics for sepsis was associated with a median of 1 to 180 additional antibiotic-treated patients and zero to seven additional adverse events per death averted from sepsis. Conclusions: The impacts of faster time-to-antibiotics for sepsis vary markedly across simulated hospital types. However, even in the worst-case scenario, new antibiotic-associated adverse events were rare.
Subject(s)
Sepsis , Shock, Septic , Humans , Anti-Bacterial Agents/adverse effects , Shock, Septic/drug therapy , Retrospective Studies , Sepsis/drug therapy , Hospitalization , Emergency Service, Hospital , Hospital MortalityABSTRACT
Rationale: Predicting long-term outcomes in sepsis survivors remains a difficult task. Persistent inflammation post-sepsis is associated with increased risk for rehospitalization and death. As surrogate markers of inflammation, complete blood count parameters measured at hospital discharge may have prognostic value for sepsis survivors. Objective: To determine the incremental value of complete blood count parameters over clinical characteristics for predicting 90-day outcomes in sepsis survivors. Methods: Electronic health record data was used to identify sepsis hospitalizations at United States Veterans Affairs hospitals with live discharge and relevant laboratory data (2013 to 2018). We measured the association of eight complete blood count parameters with 90-day outcomes (mortality, rehospitalization, cause-specific rehospitalizations) using multivariable logistic regression models. Measurements and main results: We identified 155,988 eligible hospitalizations for sepsis. Anemia (93.6%, N=142,162) and lymphopenia (28.1%, N=29,365) were the most common blood count abnormalities at discharge. In multivariable models, all parameters were associated with the primary outcome of 90-day mortality or rehospitalization and improved model discrimination above clinical characteristics alone (likelihood ratio test, p<0.02 for all). A model including all eight parameters significantly improved discrimination (AUROC, 0.6929 v. 0.6756) and reduced calibration error for the primary outcome. Hemoglobin had the greatest prognostic separation with a 1.5 fold increased incidence of the primary outcome in the lowest quintile (7.2-8.9 g/dL) versus highest quintile (12.70-15.80 g/dL). Hemoglobin and neutrophil lymphocyte ratio provided the most added value in predicting the primary outcome and 90-day mortality alone, respectively. Absolute lymphocyte count added little value in predicting 90-day outcomes. Conclusions: The incorporation of discharge complete blood count parameters into prognostic scoring systems could improve prediction of 90-day outcomes. Hemoglobin had the greatest prognostic value for the primary composite outcome of 90-day rehospitalization or mortality. Absolute lymphocyte count provided little added value in multivariable model comparisons, including for infection- or sepsis-related rehospitalization.
Subject(s)
Lymphocytes , Sepsis , Humans , United States/epidemiology , Retrospective Studies , Inflammation , HospitalsABSTRACT
STUDY OBJECTIVE: Bystander cardiopulmonary resuscitation increases the likelihood of out-of-hospital cardiac arrest survival by more than two-fold. A common barrier to the prompt initiation of compressions is moving victims to the floor, but compression quality on a "floor" versus a "mattress" has not been tested among lay bystanders. METHODS: We conducted a prospective, randomized, cross-over trial comparing lay bystander compression quality using a manikin on a bed versus the floor. Participants included adults without professional health care training. We randomized participants to the order of manikin placement, either on a mattress or on the floor. For both, participants were instructed to perform 2 minutes of chest compressions on a cardiopulmonary resuscitation Simon manikin Gaumard (Gaumard Scientific, Miami, FL). The primary outcome was mean compression depth (cm) over 2 minutes. We fit a linear regression model adjusted for scenario order, age, sex, and body mass index with robust standard errors to account for repeated measures and reported mean differences with 95% confidence intervals (CIs). RESULTS: Our sample of 80 adults was 66% female with a mean age of 50.5 years (SD 18.2). The mean compression depth on the mattress was 2.9 cm (SD 2.3) and 3.5 cm (SD 2.2) on the floor, a mean difference of 0.58 cm (95% CI 0.18, 0.98). Compression depth fell below the 5 to 6 cm depth recommended by the American Heart Association on both surfaces. In the adjusted model, the mean depth was greater when the manikin was on the floor than the mattress (adjusted mean difference 0.62 cm; 95% CI 0.23 to 1.01), and mean depth was less for females than males (adjusted mean difference -1.42 cm, 95% CI -2.59, -0.25). In addition, the difference in compression depth was larger for female participants (mean difference 0.94 cm; 95% CI 0.54, 1.34) than for male participants (mean difference -0.01 cm; 95% CI -0.80, 0.78), and the interaction was statistically significant (P = .04). CONCLUSION: The mean compression depth was significantly smaller on the mattress and with female bystanders. Further research is needed to understand the benefit of moving out-of-hospital cardiac arrest victims to the floor relative to the detrimental effect of delaying chest compressions.
Subject(s)
Cardiopulmonary Resuscitation , Out-of-Hospital Cardiac Arrest , Adult , Humans , Male , Female , Middle Aged , Out-of-Hospital Cardiac Arrest/therapy , Cross-Over Studies , Prospective Studies , Cardiopulmonary Resuscitation/education , Hand , ManikinsABSTRACT
Importance: Some experts have cautioned that national and health system emphasis on rapid administration of antimicrobials for sepsis may increase overall antimicrobial use even among patients without sepsis. Objective: To assess whether temporal changes in antimicrobial timing for sepsis are associated with increasing antimicrobial use, days of therapy, or broadness of antimicrobial coverage among all hospitalized patients at risk for sepsis. Design, Setting, and Participants: This is an observational cohort study of hospitalized patients at 152 hospitals in 2 health care systems during 2013 to 2018, admitted via the emergency department with 2 or more systemic inflammatory response syndrome (SIRS) criteria. Data analysis was performed from June 10, 2021, to March 22, 2022. Exposures: Hospital-level temporal trends in time to first antimicrobial administration. Outcomes: Antimicrobial outcomes included antimicrobial use, days of therapy, and broadness of antibacterial coverage. Clinical outcomes included in-hospital mortality, 30-day mortality, length of hospitalization, and new multidrug-resistant (MDR) organism culture positivity. Results: Among 1â¯559â¯523 patients admitted to the hospital via the emergency department with 2 or more SIRS criteria (1â¯269â¯998 male patients [81.4%]; median [IQR] age, 67 [59-77] years), 273â¯255 (17.5%) met objective criteria for sepsis. In multivariable models adjusted for patient characteristics, the adjusted median (IQR) time to first antimicrobial administration to patients with sepsis decreased by 37 minutes, from 4.7 (4.1-5.3) hours in 2013 to 3.9 (3.6-4.4) hours in 2018, although the slope of decrease varied across hospitals. During the same period, antimicrobial use within 48 hours, days of antimicrobial therapy, and receipt of broad-spectrum coverage decreased among the broader cohort of patients with SIRS. In-hospital mortality, 30-day mortality, length of hospitalization, new MDR culture positivity, and new MDR blood culture positivity decreased over the study period among both patients with sepsis and those with SIRS. When examining hospital-specific trends, decreases in antimicrobial use, days of therapy, and broadness of antibacterial coverage for patients with SIRS did not differ by hospital antimicrobial timing trend for sepsis. Overall, there was no evidence that accelerating antimicrobial timing for sepsis was associated with increasing antimicrobial use or impaired antimicrobial stewardship. Conclusions and Relevance: In this multihospital cohort study, the time to first antimicrobial for sepsis decreased over time, but this trend was not associated with increasing antimicrobial use, days of therapy, or broadness of antimicrobial coverage among the broader population at-risk for sepsis, which suggests that shortening the time to antibiotics for sepsis is feasible without leading to indiscriminate antimicrobial use.
Subject(s)
Sepsis , Aged , Anti-Bacterial Agents/therapeutic use , Cohort Studies , Emergency Service, Hospital , Hospital Mortality , Hospitalization , Humans , Male , Retrospective Studies , Sepsis/drug therapy , Systemic Inflammatory Response Syndrome/epidemiologyABSTRACT
OBJECTIVE: To create and validate a simple and transferable machine learning model from electronic health record data to accurately predict clinical deterioration in patients with covid-19 across institutions, through use of a novel paradigm for model development and code sharing. DESIGN: Retrospective cohort study. SETTING: One US hospital during 2015-21 was used for model training and internal validation. External validation was conducted on patients admitted to hospital with covid-19 at 12 other US medical centers during 2020-21. PARTICIPANTS: 33 119 adults (≥18 years) admitted to hospital with respiratory distress or covid-19. MAIN OUTCOME MEASURES: An ensemble of linear models was trained on the development cohort to predict a composite outcome of clinical deterioration within the first five days of hospital admission, defined as in-hospital mortality or any of three treatments indicating severe illness: mechanical ventilation, heated high flow nasal cannula, or intravenous vasopressors. The model was based on nine clinical and personal characteristic variables selected from 2686 variables available in the electronic health record. Internal and external validation performance was measured using the area under the receiver operating characteristic curve (AUROC) and the expected calibration error-the difference between predicted risk and actual risk. Potential bed day savings were estimated by calculating how many bed days hospitals could save per patient if low risk patients identified by the model were discharged early. RESULTS: 9291 covid-19 related hospital admissions at 13 medical centers were used for model validation, of which 1510 (16.3%) were related to the primary outcome. When the model was applied to the internal validation cohort, it achieved an AUROC of 0.80 (95% confidence interval 0.77 to 0.84) and an expected calibration error of 0.01 (95% confidence interval 0.00 to 0.02). Performance was consistent when validated in the 12 external medical centers (AUROC range 0.77-0.84), across subgroups of sex, age, race, and ethnicity (AUROC range 0.78-0.84), and across quarters (AUROC range 0.73-0.83). Using the model to triage low risk patients could potentially save up to 7.8 bed days per patient resulting from early discharge. CONCLUSION: A model to predict clinical deterioration was developed rapidly in response to the covid-19 pandemic at a single hospital, was applied externally without the sharing of data, and performed well across multiple medical centers, patient subgroups, and time periods, showing its potential as a tool for use in optimizing healthcare resources.
Subject(s)
COVID-19/diagnosis , Clinical Decision Rules , Hospitalization/statistics & numerical data , Machine Learning , Risk Assessment/methods , Adolescent , Adult , Aged , Aged, 80 and over , Area Under Curve , Clinical Deterioration , Electronic Health Records , Female , Hospitals , Humans , Linear Models , Male , Middle Aged , Predictive Value of Tests , Prognosis , ROC Curve , Retrospective Studies , SARS-CoV-2 , Young AdultABSTRACT
BACKGROUND/PURPOSE: Extreme events such as hurricanes adversely impact healthcare systems and the communities they serve. The degree to which hurricanes affect healthcare use among high need groups such as older adults with chronic conditions has not been well examined, nor has the impact of hurricane severity on health outcomes. We characterized hospitalizations among older adults by chronic condition after eight large-scale hurricanes in the United States. METHODS: Using a combination of administrative healthcare data and the Federal Emergency Management Agency's Disaster Declaration database we conducted a self-controlled case series analysis. We identified Medicare beneficiaries who were exposed to one of eight hurricanes and compared hospitalizations in the 30-days after a hurricane to hospitalizations in the rest of the calendar year of the hurricane. We examined hospitalizations (1) in total, (2) separately for diabetes, congestive heart failure (CHF) and chronic obstructive pulmonary disease (COPD) admissions, and (3) by hurricane damage category. RESULTS: Among all older adults exposed, hospitalizations in the 30-day period after each disaster increased for all three chronic conditions; diabetes (incidence rate ratio [IRR] = 1.06, 95% confidence interval [CI] 1.03, 1.10), COPD (IRR = 1.06, 95% CI 1.04, 1.08), and CHF (IRR = 1.19, 95% CI 1.17, 1.21. In the 30-to-60-day period hospitalizations also increased for each chronic condition; diabetes (IRR = 1.06, 95% CI 1.03, 1.10), COPD (IRR = 1.12, 95% CI 1.10, 1.15), and CHF (IRR = 1.32, 95% CI 1.30, 1.34). Substantial differences in hospitalizations were observed according to individual hurricane and by the chronic disease examined. CONCLUSION: Exposure to hurricanes is associated with an increase in hospitalizations for chronic conditions across all hurricane damage categories. As disasters are expected to increase in strength and frequency, our results underscore the need for response strategies and health policy planning for healthcare systems designed to address the health needs of older Americans with chronic conditions.
Subject(s)
Cyclonic Storms , Heart Failure , Pulmonary Disease, Chronic Obstructive , Aged , Chronic Disease , Hospitalization , Humans , Medicare , United States/epidemiologyABSTRACT
Importance: Patients who survive pediatric critical illness and their caregivers commonly experience physical, emotional, and cognitive sequelae. However, the rate and duration of school absence among patients and work absence among their caregivers are unknown. Objective: To determine the rates and duration of school absence among children who survived hospitalization with acute respiratory failure and work absence among their caregivers. Design, Setting, and Participants: The Randomized Evaluation of Sedation Titration for Respiratory Failure (RESTORE) cluster randomized trial included 2449 children from 31 sites to protocolized sedation (intervention) vs usual care (control) from June 6, 2009, to December 2, 2013. In total, 1360 children survived hospitalization and were selected for follow-up at 6 months after pediatric intensive care unit (PICU) discharge, which was completed from January 12, 2010, to April 13, 2015. This secondary analysis was conducted from July 1, 2020, to September 30, 2021. Exposures: PICU hospitalization for acute respiratory failure, including invasive mechanical ventilation. Main Outcomes and Measures: Postdischarge assessments with caregivers of eligible participants at 6 months after PICU discharge, including questions about school and work absence. Risk factors associated with longer absence from school and work were identified. Results: Postdischarge assessments were completed for 960 children who survived treatment for acute respiratory failure, of whom 443 (46.1%) were girls and 517 (53.9%) were boys; 509 of 957 (53.2%) were non-Hispanic White. Median age was 1.8 years (IQR, 0.4-7.9 years). In total, 399 children (41.6%) were enrolled in school, of whom 279 (69.9%) missed school after discharge. Median duration of postdischarge absence was 9.1 days (IQR, 0-27.9 days) among all children enrolled in school and 16.9 days (IQR, 7.9-43.9 days) among the 279 children with postdischarge absence. Among 960 primary caregivers, 506 (52.7%) were employed outside the home, of whom 277 (54.7%) missed work. Median duration of postdischarge work absence was 2 days (IQR, 0-10 days) among all employed primary caregivers, and 8 days (IQR, 4-20 days) among the 277 caregivers who missed work after discharge. The odds of postdischarge school absence and greater duration of absence increased for children 5 years or older (compared with 0-4 years, odds ratios [ORs] for 5-8 years, 3.20 [95% CI, 1.69-6.05] and 2.09 [95% CI, 1.30-3.37], respectively; ORs for 9-12 years, 2.49 [95% CI, 1.17-5.27] and 2.32 [95% CI, 1.30-4.14], respectively; and ORs for 13-18 years, 2.37 [95% CI, 1.20-4.66] and 1.89 [95% CI, 1.11-3.24], respectively) and those with a preexisting comorbidity (ORs, 1.90 [95% CI, 1.10-3.29] and 1.76 [95% CI, 1.14-2.69], respectively). Conclusions and Relevance: In this secondary analysis of a cluster randomized trial, 2 in 3 children hospitalized for acute respiratory failure missed school after discharge, for a median duration of nearly 2 weeks. In addition, more than half of primary caregivers missed work after discharge. The magnitude of school absenteeism suggests that children may be at increased risk for lower educational achievement, economic hardship, and poor health outcomes in adulthood.
Subject(s)
Absenteeism , Caregivers/statistics & numerical data , Child, Hospitalized , Critical Care , Critical Illness/therapy , Respiratory Distress Syndrome/therapy , Female , Humans , Infant , Male , Risk FactorsABSTRACT
Importance: It is unclear whether antimicrobial timing for sepsis has changed outside of performance incentive initiatives. Objective: To examine temporal trends and variation in time-to-antibiotics for sepsis in the US Department of Veterans Affairs (VA) health care system. Design, Setting, and Participants: This observational cohort study included 130 VA hospitals from 2013 to 2018. Participants included all patients admitted to the hospital via the emergency department with sepsis from 2013 to 2018, using a definition adapted from the Centers for Disease Control and Prevention Adult Sepsis Event definition, which requires evidence of suspected infection, acute organ dysfunction, and systemic antimicrobial therapy within 12 hours of presentation. Data were analyzed from October 6, 2020, to July 1, 2021. Exposures: Time from presentation to antibiotic administration. Main Outcomes and Measures: The main outcome was differences in time-to-antibiotics across study periods, hospitals, and patient subgroups defined by presenting temperature and blood pressure. Temporal trends in time-to-antibiotics were measured overall and by subgroups. Hospital-level variation in time-to-antibiotics was quantified after adjusting for differences in patient characteristics using multilevel linear regression models. Results: A total of 111â¯385 hospitalizations for sepsis were identified, including 107â¯547 men (96.6%) men and 3838 women (3.4%) with a median (interquartile range [IQR]) age of 68 (62-77) years. A total of 7574 patients (6.8%) died in the hospital, and 13â¯855 patients (12.4%) died within 30 days. Median (IQR) time-to-antibiotics was 3.9 (2.4-6.5) hours but differed by presenting characteristics. Unadjusted median (IQR) time-to-antibiotics decreased over time, from 4.5 (2.7-7.1) hours during 2013 to 2014 to 3.5 (2.2-5.9) hours during 2017 to 2018 (P < .001). In multilevel models adjusted for patient characteristics, median time-to-antibiotics declined by 9.0 (95% CI, 8.8-9.2) minutes per calendar year. Temporal trends in time-to-antibiotics were similar across patient subgroups, but hospitals with faster baseline time-to-antibiotics had less change over time, with hospitals in the slowest tertile decreasing time-to-antibiotics by 16.6 minutes (23.1%) per year, while hospitals in the fastest tertile decreased time-to-antibiotics by 7.2 minutes (13.1%) per year. In the most recent years (2017-2018), median time-to-antibiotics ranged from 3.1 to 6.7 hours across hospitals (after adjustment for patient characteristics), 6.8% of variation in time-to-antibiotics was explained at the hospital level, and odds of receiving antibiotics within 3 hours increased by 65% (95% CI, 56%-77%) for the median patient if moving to a hospital with faster time-to-antibiotics. Conclusions and Relevance: This cohort study across nationwide VA hospitals found that time-to-antibiotics for sepsis has declined over time. However, there remains significant variability in time-to-antibiotics not explained by patient characteristics, suggesting potential unwarranted practice variation in sepsis treatment. Efforts to further accelerate time-to-antibiotics must be weighed against risks of overtreatment.
Subject(s)
Anti-Bacterial Agents/administration & dosage , Hospitalization , Sepsis/drug therapy , Time-to-Treatment/trends , Veterans , Aged , Cohort Studies , Drug Administration Schedule , Emergency Service, Hospital , Female , Humans , Male , Middle Aged , United States , United States Department of Veterans AffairsABSTRACT
Importance: The Epic Sepsis Model (ESM), a proprietary sepsis prediction model, is implemented at hundreds of US hospitals. The ESM's ability to identify patients with sepsis has not been adequately evaluated despite widespread use. Objective: To externally validate the ESM in the prediction of sepsis and evaluate its potential clinical value compared with usual care. Design, Setting, and Participants: This retrospective cohort study was conducted among 27â¯697 patients aged 18 years or older admitted to Michigan Medicine, the academic health system of the University of Michigan, Ann Arbor, with 38â¯455 hospitalizations between December 6, 2018, and October 20, 2019. Exposure: The ESM score, calculated every 15 minutes. Main Outcomes and Measures: Sepsis, as defined by a composite of (1) the Centers for Disease Control and Prevention surveillance criteria and (2) International Statistical Classification of Diseases and Related Health Problems, Tenth Revision diagnostic codes accompanied by 2 systemic inflammatory response syndrome criteria and 1 organ dysfunction criterion within 6 hours of one another. Model discrimination was assessed using the area under the receiver operating characteristic curve at the hospitalization level and with prediction horizons of 4, 8, 12, and 24 hours. Model calibration was evaluated with calibration plots. The potential clinical benefit associated with the ESM was assessed by evaluating the added benefit of the ESM score compared with contemporary clinical practice (based on timely administration of antibiotics). Alert fatigue was evaluated by comparing the clinical value of different alerting strategies. Results: We identified 27â¯697 patients who had 38â¯455 hospitalizations (21â¯904 women [57%]; median age, 56 years [interquartile range, 35-69 years]) meeting inclusion criteria, of whom sepsis occurred in 2552 (7%). The ESM had a hospitalization-level area under the receiver operating characteristic curve of 0.63 (95% CI, 0.62-0.64). The ESM identified 183 of 2552 patients with sepsis (7%) who did not receive timely administration of antibiotics, highlighting the low sensitivity of the ESM in comparison with contemporary clinical practice. The ESM also did not identify 1709 patients with sepsis (67%) despite generating alerts for an ESM score of 6 or higher for 6971 of all 38â¯455 hospitalized patients (18%), thus creating a large burden of alert fatigue. Conclusions and Relevance: This external validation cohort study suggests that the ESM has poor discrimination and calibration in predicting the onset of sepsis. The widespread adoption of the ESM despite its poor performance raises fundamental concerns about sepsis management on a national level.
Subject(s)
Anti-Bacterial Agents/therapeutic use , Hospitalization/statistics & numerical data , Intensive Care Units/statistics & numerical data , Sepsis , Decision Support Systems, Clinical/standards , Female , Hospital Mortality , Humans , Male , Michigan/epidemiology , Middle Aged , Organ Dysfunction Scores , Predictive Value of Tests , Prognosis , Reproducibility of Results , Retrospective Studies , Sepsis/diagnosis , Sepsis/epidemiology , Sepsis/prevention & controlABSTRACT
Causal graphs provide a key tool for optimizing the validity of causal effect estimates. Although a large literature exists on the mathematical theory underlying the use of causal graphs, less literature exists to aid applied researchers in understanding how best to develop and use causal graphs in their research projects. We sought to understand why researchers do or do not regularly use DAGs by surveying practicing epidemiologists and medical researchers on their knowledge, level of interest, attitudes, and practices towards the use of causal graphs in applied epidemiology and health research. We used Twitter and the Society for Epidemiologic Research to disseminate the survey. Overall, a majority of participants reported being comfortable with using causal graphs and reported using them 'sometimes', 'often', or 'always' in their research. Having received training appeared to improve comprehension of the assumptions displayed in causal graphs. Many of the respondents who did not use causal graphs reported lack of knowledge as a barrier to using DAGs in their research. Causal graphs are of interest to epidemiologists and medical researchers, but there are several barriers to their uptake. Additional training and clearer guidance are needed. In addition, methodological developments regarding visualization of effect measure modification and interaction on causal graphs is needed.
Subject(s)
Attitude of Health Personnel , Causality , Computer Graphics , Data Interpretation, Statistical , Epidemiologic Research Design , Epidemiologists , Female , Humans , Male , Qualitative Research , Research Personnel , Surveys and QuestionnairesABSTRACT
Importance: Therapies that improve survival in critically ill patients with coronavirus disease 2019 (COVID-19) are needed. Tocilizumab, a monoclonal antibody against the interleukin 6 receptor, may counteract the inflammatory cytokine release syndrome in patients with severe COVID-19 illness. Objective: To test whether tocilizumab decreases mortality in this population. Design, Setting, and Participants: The data for this study were derived from a multicenter cohort study of 4485 adults with COVID-19 admitted to participating intensive care units (ICUs) at 68 hospitals across the US from March 4 to May 10, 2020. Critically ill adults with COVID-19 were categorized according to whether they received or did not receive tocilizumab in the first 2 days of admission to the ICU. Data were collected retrospectively until June 12, 2020. A Cox regression model with inverse probability weighting was used to adjust for confounding. Exposures: Treatment with tocilizumab in the first 2 days of ICU admission. Main Outcomes and Measures: Time to death, compared via hazard ratios (HRs), and 30-day mortality, compared via risk differences. Results: Among the 3924 patients included in the analysis (2464 male [62.8%]; median age, 62 [interquartile range {IQR}, 52-71] years), 433 (11.0%) received tocilizumab in the first 2 days of ICU admission. Patients treated with tocilizumab were younger (median age, 58 [IQR, 48-65] vs 63 [IQR, 52-72] years) and had a higher prevalence of hypoxemia on ICU admission (205 of 433 [47.3%] vs 1322 of 3491 [37.9%] with mechanical ventilation and a ratio of partial pressure of arterial oxygen to fraction of inspired oxygen of <200 mm Hg) than patients not treated with tocilizumab. After applying inverse probability weighting, baseline and severity-of-illness characteristics were well balanced between groups. A total of 1544 patients (39.3%) died, including 125 (28.9%) treated with tocilizumab and 1419 (40.6%) not treated with tocilizumab. In the primary analysis, during a median follow-up of 27 (IQR, 14-37) days, patients treated with tocilizumab had a lower risk of death compared with those not treated with tocilizumab (HR, 0.71; 95% CI, 0.56-0.92). The estimated 30-day mortality was 27.5% (95% CI, 21.2%-33.8%) in the tocilizumab-treated patients and 37.1% (95% CI, 35.5%-38.7%) in the non-tocilizumab-treated patients (risk difference, 9.6%; 95% CI, 3.1%-16.0%). Conclusions and Relevance: Among critically ill patients with COVID-19 in this cohort study, the risk of in-hospital mortality in this study was lower in patients treated with tocilizumab in the first 2 days of ICU admission compared with patients whose treatment did not include early use of tocilizumab. However, the findings may be susceptible to unmeasured confounding, and further research from randomized clinical trials is needed.
Subject(s)
Antibodies, Monoclonal, Humanized/therapeutic use , COVID-19 Drug Treatment , Hospital Mortality , Respiratory Insufficiency/therapy , Adolescent , Adrenal Cortex Hormones/therapeutic use , Adult , Aged , Anticoagulants/therapeutic use , COVID-19/physiopathology , Cohort Studies , Critical Illness , Early Medical Intervention , Female , Hospitalization , Humans , Intensive Care Units , Male , Middle Aged , Mortality , Organ Dysfunction Scores , Patient Positioning , Prone Position , Proportional Hazards Models , Receptors, Interleukin-6/antagonists & inhibitors , Respiration, Artificial , Respiratory Insufficiency/physiopathology , SARS-CoV-2 , Young AdultSubject(s)
COVID-19/mortality , Patient Discharge/statistics & numerical data , Patient Readmission/statistics & numerical data , SARS-CoV-2 , Adult , Age Factors , Aged , Aged, 80 and over , COVID-19/epidemiology , COVID-19/therapy , Female , Heart Failure/epidemiology , Humans , Kaplan-Meier Estimate , Male , Middle Aged , Respiration, Artificial/statistics & numerical data , Time Factors , Vasoconstrictor Agents/therapeutic use , Veterans Health Services/statistics & numerical dataABSTRACT
Rationale: The Epic Deterioration Index (EDI) is a proprietary prediction model implemented in over 100 U.S. hospitals that was widely used to support medical decision-making during the coronavirus disease (COVID-19) pandemic. The EDI has not been independently evaluated, and other proprietary models have been shown to be biased against vulnerable populations. Objectives: To independently evaluate the EDI in hospitalized patients with COVID-19 overall and in disproportionately affected subgroups. Methods: We studied adult patients admitted with COVID-19 to units other than the intensive care unit at a large academic medical center from March 9 through May 20, 2020. We used the EDI, calculated at 15-minute intervals, to predict a composite outcome of intensive care unit-level care, mechanical ventilation, or in-hospital death. In a subset of patients hospitalized for at least 48 hours, we also evaluated the ability of the EDI to identify patients at low risk of experiencing this composite outcome during their remaining hospitalization. Results: Among 392 COVID-19 hospitalizations meeting inclusion criteria, 103 (26%) met the composite outcome. The median age of the cohort was 64 (interquartile range, 53-75) with 168 (43%) Black patients and 169 (43%) women. The area under the receiver-operating characteristic curve of the EDI was 0.79 (95% confidence interval, 0.74-0.84). EDI predictions did not differ by race or sex. When exploring clinically relevant thresholds of the EDI, we found patients who met or exceeded an EDI of 68.8 made up 14% of the study cohort and had a 74% probability of experiencing the composite outcome during their hospitalization with a sensitivity of 39% and a median lead time of 24 hours from when this threshold was first exceeded. Among the 286 patients hospitalized for at least 48 hours who had not experienced the composite outcome, 14 (13%) never exceeded an EDI of 37.9, with a negative predictive value of 90% and a sensitivity above this threshold of 91%. Conclusions: We found the EDI identifies small subsets of high-risk and low-risk patients with COVID-19 with good discrimination, although its clinical use as an early warning system is limited by low sensitivity. These findings highlight the importance of independent evaluation of proprietary models before widespread operational use among patients with COVID-19.
Subject(s)
COVID-19 , Adult , Aged , Female , Hospital Mortality , Hospitalization , Humans , Intensive Care Units , Male , Middle Aged , Pandemics , Retrospective Studies , SARS-CoV-2ABSTRACT
Importance: The US is currently an epicenter of the coronavirus disease 2019 (COVID-19) pandemic, yet few national data are available on patient characteristics, treatment, and outcomes of critical illness from COVID-19. Objectives: To assess factors associated with death and to examine interhospital variation in treatment and outcomes for patients with COVID-19. Design, Setting, and Participants: This multicenter cohort study assessed 2215 adults with laboratory-confirmed COVID-19 who were admitted to intensive care units (ICUs) at 65 hospitals across the US from March 4 to April 4, 2020. Exposures: Patient-level data, including demographics, comorbidities, and organ dysfunction, and hospital characteristics, including number of ICU beds. Main Outcomes and Measures: The primary outcome was 28-day in-hospital mortality. Multilevel logistic regression was used to evaluate factors associated with death and to examine interhospital variation in treatment and outcomes. Results: A total of 2215 patients (mean [SD] age, 60.5 [14.5] years; 1436 [64.8%] male; 1738 [78.5%] with at least 1 chronic comorbidity) were included in the study. At 28 days after ICU admission, 784 patients (35.4%) had died, 824 (37.2%) were discharged, and 607 (27.4%) remained hospitalized. At the end of study follow-up (median, 16 days; interquartile range, 8-28 days), 875 patients (39.5%) had died, 1203 (54.3%) were discharged, and 137 (6.2%) remained hospitalized. Factors independently associated with death included older age (≥80 vs <40 years of age: odds ratio [OR], 11.15; 95% CI, 6.19-20.06), male sex (OR, 1.50; 95% CI, 1.19-1.90), higher body mass index (≥40 vs <25: OR, 1.51; 95% CI, 1.01-2.25), coronary artery disease (OR, 1.47; 95% CI, 1.07-2.02), active cancer (OR, 2.15; 95% CI, 1.35-3.43), and the presence of hypoxemia (Pao2:Fio2<100 vs ≥300 mm Hg: OR, 2.94; 95% CI, 2.11-4.08), liver dysfunction (liver Sequential Organ Failure Assessment score of 2-4 vs 0: OR, 2.61; 95% CI, 1.30-5.25), and kidney dysfunction (renal Sequential Organ Failure Assessment score of 4 vs 0: OR, 2.43; 95% CI, 1.46-4.05) at ICU admission. Patients admitted to hospitals with fewer ICU beds had a higher risk of death (<50 vs ≥100 ICU beds: OR, 3.28; 95% CI, 2.16-4.99). Hospitals varied considerably in the risk-adjusted proportion of patients who died (range, 6.6%-80.8%) and in the percentage of patients who received hydroxychloroquine, tocilizumab, and other treatments and supportive therapies. Conclusions and Relevance: This study identified demographic, clinical, and hospital-level risk factors that may be associated with death in critically ill patients with COVID-19 and can facilitate the identification of medications and supportive therapies to improve outcomes.
Subject(s)
COVID-19/mortality , Critical Illness/mortality , Intensive Care Units , Adult , Age Factors , Aged , Aged, 80 and over , Critical Illness/therapy , Female , Hospital Mortality , Humans , Male , Middle Aged , Pandemics , Risk Factors , United StatesABSTRACT
INTRODUCTION: The Epic Deterioration Index (EDI) is a proprietary prediction model implemented in over 100 U.S. hospitals that was widely used to support medical decision-making during the COVID-19 pandemic. The EDI has not been independently evaluated, and other proprietary models have been shown to be biased against vulnerable populations. METHODS: We studied adult patients admitted with COVID-19 to non-ICU care at a large academic medical center from March 9 through May 20, 2020. We used the EDI, calculated at 15-minute intervals, to predict a composite outcome of ICU-level care, mechanical ventilation, or in-hospital death. In a subset of patients hospitalized for at least 48 hours, we also evaluated the ability of the EDI to identify patients at low risk of experiencing this composite outcome during their remaining hospitalization. RESULTS: Among 392 COVID-19 hospitalizations meeting inclusion criteria, 103 (26%) met the composite outcome. Median age of the cohort was 64 (IQR 53-75) with 168 (43%) African Americans and 169 (43%) women. Area under the receiver-operating-characteristic curve (AUC) of the EDI was 0.79 (95% CI 0.74-0.84). EDI predictions did not differ by race or sex. When exploring clinically-relevant thresholds of the EDI, we found patients who met or exceeded an EDI of 68.8 made up 14% of the study cohort and had a 74% probability of experiencing the composite outcome during their hospitalization with a median lead time of 24 hours from when this threshold was first exceeded. Among the 286 patients hospitalized for at least 48 hours who had not experienced the composite outcome, 14 (13%) never exceeded an EDI of 37.9, with a negative predictive value of 90% and a sensitivity above this threshold of 91%. CONCLUSION: We found the EDI identifies small subsets of high- and low-risk COVID-19 patients with fair discrimination. We did not find evidence of bias by race or sex. These findings highlight the importance of independent evaluation of proprietary models before widespread operational use among COVID-19 patients.
ABSTRACT
Identifying persons with hepatitis C virus (HCV) infection has become an urgent public health challenge because of increasing HCV-related morbidity and mortality, low rates of awareness among infected persons, and the advent of curative therapies (1). Since 2012, CDC has recommended testing of all persons born during 1945-1965 (baby boomers) for identification of chronic HCV infection (1); urban emergency departments (EDs) are well positioned venues for detecting HCV infection among these persons. The United States has witnessed an unprecedented opioid overdose epidemic since 2013 that derives primarily from commonly injected illicit opioids (e.g., heroin and fentanyl) (2). This injection drug use behavior has led to an increase in HCV infections among persons who inject drugs and heightened concern about increases in human immunodeficiency virus (HIV) and HCV infection within communities disproportionately affected by the opioid crisis (3,4). However, targeted strategies for identifying HCV infection among persons who inject drugs is challenging (5,6). During 2015-2016, EDs at the University of Alabama at Birmingham; Highland Hospital, Oakland, California; Johns Hopkins Hospital, Baltimore, Maryland; and Boston University Medical Center, Massachusetts, adopted opt-out (i.e., patients can implicitly accept or explicitly decline testing), universal hepatitis C screening for all adult patients. ED staff members offered HCV antibody (anti-HCV) screening to patients who were unaware of their status.* During similar observation periods at each site, ED staff members tested 14,252 patients and identified an overall 9.2% prevalence of positive results for anti-HCV among the adult patient population. Among the 1945-1965 birth cohort, prevalence of positive results for anti-HCV (13.9%) was significantly higher among non-Hispanic blacks (blacks) (16.0%) than among non-Hispanic whites (whites) (12.2%) (p<0.001). Among persons born after 1965, overall prevalence of positive results for anti-HCV was 6.7% and was significantly higher among whites (15.3%) than among blacks (3.2%) (p<0.001). These findings highlight age-associated differences in racial/ethnic prevalences and the potential for ED venues and opt-out, universal testing strategies to improve HCV infection awareness and surveillance for hard-to-reach populations. This opt-out, universal testing approach is supported by new recommendations for hepatitis C screening at least once in a lifetime for all adults aged ≥18 years, except in settings where the prevalence of positive results for HCV infection is <0.1% (7).
Subject(s)
Emergency Service, Hospital , Hepatitis C/epidemiology , Hospitals, Urban , Adult , Aged , Alabama/epidemiology , Baltimore/epidemiology , Boston/epidemiology , California/epidemiology , Female , Hepatitis C/diagnosis , Humans , Male , Mass Screening , Middle Aged , PrevalenceABSTRACT
BACKGROUND: Claims-based algorithms are commonly used to identify sepsis in health services research because the laboratory features required to define clinical criteria may not be available in administrative data. METHODS: We evaluated claims-based sepsis algorithms among adults in the US aged ≥65 years with Medicare health insurance enrolled in the REasons for Geographic And Racial Differences in Stroke (REGARDS) study. Suspected infections from baseline (2003-2007) through December 31, 2012 were analyzed. Two claims-based algorithms were evaluated: (1) infection plus organ dysfunction diagnoses or sepsis diagnoses (Medicare-Implicit/Explicit) and (2) Centers for Medicare and Medicaid Services Severe Sepsis/Septic Shock Measure diagnoses (Medicare-CMS). Three classifications based on clinical criteria were used as standards for comparison: (1) the sepsis-related organ failure assessment (SOFA) score (REGARDS-SOFA), (2) "quick" SOFA (REGARDS-qSOFA), and (3) Centers for Disease Control and Prevention electronic health record criteria (REGARDS-EHR). RESULTS: There were 2217 suspected infections among 9522 participants included in the current study. The total number of suspected infections classified as sepsis was 468 for Medicare-Implicit/Explicit, 249 for Medicare-CMS, 541 for REGARDS-SOFA, 185 for REGARDS-qSOFA, and 331 for REGARDS-EHR. The overall agreement between Medicare-Implicit/Explicit and REGARDS-SOFA, REGARDS-qSOFA, and REGARDS-EHR was 77, 79, and 81%, respectively, sensitivity was 46, 53, and 57%, and specificity was 87, 82, and 85%. Comparing Medicare-CMS and REGARDS-SOFA, REGARDS-qSOFA, and REGARDS-EHR, agreement was 77, 87, and 85%, respectively, sensitivity was 27, 41, and 36%, and specificity was 94, 92, and 93%. Events meeting the REGARDS-SOFA classification had a lower 90-day mortality rate (140.7 per 100 person-years) compared with the Medicare-CMS (296.1 per 100 person-years), REGARDS-qSOFA (238.6 per 100 person-years), Medicare-Implicit/Explicit (219.4 per 100 person-years), and REGARDS-EHR classifications (201.8 per 100 person-years). CONCLUSION: Claims-based sepsis algorithms have high agreement and specificity but low sensitivity when compared with clinical criteria. Both claims-based algorithms identified a patient population with similar 90-day mortality rates as compared with classifications based on qSOFA and EHR criteria but higher mortality relative to SOFA criteria.
Subject(s)
Algorithms , Medicare/statistics & numerical data , Sepsis/diagnosis , Stroke/diagnosis , Black or African American/statistics & numerical data , Aged , Aged, 80 and over , Cohort Studies , Female , Geography , Health Services Research/methods , Humans , Male , Reproducibility of Results , Sensitivity and Specificity , Sepsis/ethnology , Stroke/ethnology , United States , White People/statistics & numerical dataABSTRACT
The original version of this article unfortunately contained a mistake. There was an error in figure one. The correct figure can be found below. We apologize for the mistake.