Community-onset Sepsis Research Articles

Performance of Risk Models for Antimicrobial Resistance in Adult Patients With Sepsis

The results of prediction models that stratify patients with sepsis and risk of resistant gram-negative bacilli (GNB) infections inform treatment guidelines. However, these models do not extrapolate well across hospitals. To assess whether patient case mix and local prevalence rates of resistance contributed to the variable performance of a general risk stratification GNB sepsis model for community-onset and hospital-onset sepsis across hospitals. This was a retrospective cohort study conducted from January 2016 and October 2021. Adult patients with sepsis at 10 acute-care hospitals in rural and urban areas across Missouri and Illinois were included. Inclusion criteria were blood cultures indicating sepsis, having received 4 days of antibiotic treatment, and having organ dysfunction (vasopressor use, mechanical ventilation, increased creatinine or bilirubin levels, and thrombocytopenia). Analyses were completed in April 2024. The model included demographic characteristics, comorbidities, vital signs, laboratory values, procedures, and medications administered. Culture results were stratified for ceftriaxone-susceptible GNB (SS), ceftriaxone-resistant but cefepime-susceptible GNB (RS), and ceftriaxone- and cefepime-resistant GNB (RR). Negative cultures and other pathogens were labeled SS. Deep learning models were developed separately for community-onset (patient presented with sepsis) and hospital-onset (sepsis developed ≥48 hours after admission) sepsis. The models were tested across hospitals and patient subgroups. Models were assessed using area under the receiver operating characteristic curve (AUROC) and area under precision recall curve (AUPRC). A total of 39 893 patients with 85 238 sepsis episodes (43 207 [50.7%] community onset; 42 031 [48.3%] hospital onset) were included. Median (IQR) age was 65 (54-74) years, 21 241 patients (53.2%) were male, and 18 830 (47.2%) had a previous episode of sepsis. RS contributed to 3.9% (1667 episodes) and 5.7% (2389 episodes) of community-onset and hospital-onset sepsis episodes, respectively, and RR contributed to 1.8% (796 episodes) and 3.9% (1626 episodes), respectively. Previous infections and exposure to antibiotics were associated with the risk of resistant GNB. For example, in community-onset sepsis, 375 RR episodes (47.1%), 420 RS episodes (25.2%) and 3483 of 40 744 (8.5%) SS episodes were among patients with resistance to antimicrobial drugs (P < .001). The AUROC and AUPRC results varied across hospitals and patient subgroups for both community-onset and hospital-onset sepsis. AUPRC values correlated with the prevalence rates of resistant GNB (R = 0.79; P = .001). In this cohort study of 39 893 patients with sepsis, variable model performance was associated with prevalence rates of antimicrobial resistance rather than patient case mix. This variability suggests caution is needed when using generalized models for predicting resistant GNB etiologies in sepsis.

Development and Validation of the Hospital Medicine Safety Sepsis Initiative Mortality Model

When comparing outcomes after sepsis, it is essential to account for patient case mix to make fair comparisons. We developed a model to assess risk-adjusted 30-day mortality in the Michigan Hospital Medicine Safety sepsis initiative (HMS-Sepsis). Can HMS-Sepsis registry data adequately predict risk of 30-day mortality? Do performance assessments using adjusted vsunadjusted data differ? Retrospective cohort of community-onset sepsis hospitalizations in the HMS-Sepsis registry (April 2022-September 2023), with split-derivation (70%) and validation (30%) cohorts. We fit a risk-adjustment model (HMS-Sepsis mortality model) incorporating acute physiologic, demographic, and baseline health data and assessed model performance using concordance (C) statistics, Brier's scores, and comparisons of predicted vsobserved mortality by deciles of risk. We compared hospital performance (first quintile, middle quintiles, fifth quintile) using observed vsadjusted mortality to understand the extent to which risk adjustment impacted hospital performance assessment. Among 17,514 hospitalizations from 66 hospitals during the study period, 12,260 hospitalizations (70%) were used for model derivation and 5,254 hospitalizations (30%) were used for model validation. Thirty-day mortality for the total cohort was 19.4%. The final model included 13 physiologic variables, two physiologic interactions, and 16 demographic and chronic health variables. The most significant variables were age, metastatic solid tumor, temperature, altered mental status, and platelet count. The model C statistic was 0.82 for the derivation cohort, 0.81 for the validation cohort, and≥ 0.78 for all subgroups assessed. Overall calibration error was 0.0%, and mean calibration error across deciles of risk was 1.5%. Standardized mortality ratios yielded different assessments than observed mortality for 33.9%of hospitals. The HMS-Sepsis mortality model showed strong discrimination and adequate calibration and reclassified one-third of hospitals to a different performance category from unadjusted mortality. Based on its strong performance, the HMS-Sepsis mortality model can aid in fair hospital benchmarking, assessment of temporal changes, and observational causal inference analysis.

Trends in Empiric Broad-Spectrum Antibiotic Use for Suspected Community-Onset Sepsis in US Hospitals

Little is known about the degree to which suspected sepsis drives broad-spectrum antibiotic use in hospitals, what proportion of antibiotic courses are unnecessarily broad in retrospect, and whether these patterns are changing over time. To describe trends in empiric broad-spectrum antibiotic use for suspected community-onset sepsis. This cross-sectional study used clinical data from adults admitted to 241 US hospitals in the PINC AI Healthcare Database. Eligible participants were aged 18 years or more and were admitted between 2017 and 2021 with suspected community-onset sepsis, defined by a blood culture draw, lactate measurement, and intravenous antibiotic administration on admission. Empiric anti-methicillin-resistant Staphylococcus aureus (MRSA) and/or antipseudomonal β-lactam agent use. Annual rates of empiric anti-MRSA and/or antipseudomonal β-lactam agent use and the proportion that were likely unnecessary in retrospect based on the absence of β-lactam resistant gram-positive or ceftriaxone-resistant gram-negative pathogens from clinical cultures obtained through hospital day 4. Annual trends were calculated using mixed-effects logistic regression models, adjusting for patient and hospital characteristics. Among 6 272 538 hospitalizations (median [IQR] age, 66 [53-78] years; 443 465 male [49.6%]; 106 095 Black [11.9%], 65 763 Hispanic [7.4%], 653 907 White [73.1%]), 894 724 (14.3%) had suspected community-onset sepsis, of whom 582 585 (65.1%) received either empiric anti-MRSA (379 987 [42.5%]) or antipseudomonal β-lactam therapy (513 811 [57.4%]); 311 213 (34.8%) received both. Patients with suspected community-onset sepsis accounted for 1 573 673 of 3 141 300 (50.1%) of total inpatient anti-MRSA antibiotic days and 2 569 518 of 5 211 745 (49.3%) of total antipseudomonal β-lactam days. Between 2017 and 2021, the proportion of patients with suspected sepsis administered anti-MRSA or antipseudomonal therapy increased from 63.0% (82 731 of 131 275 patients) to 66.7% (101 003 of 151 435 patients) (adjusted OR [aOR] per year, 1.03; 95% CI, 1.03-1.04). However, resistant organisms were isolated in only 65 434 cases (7.3%) (30 617 gram-positive [3.4%], 38 844 gram-negative [4.3%]) and the proportion of patients who had any resistant organism decreased from 9.6% to 7.3% (aOR per year, 0.87; 95% CI, 0.87-0.88). Most patients with suspected sepsis treated with empiric anti-MRSA and/or antipseudomonal therapy had no resistant organisms (527 356 of 582 585 patients [90.5%]); this proportion increased from 88.0% in 2017 to 91.6% in 2021 (aOR per year, 1.12; 95% CI, 1.11-1.13). In this cross-sectional study of adults admitted to 241 US hospitals, empiric broad-spectrum antibiotic use for suspected community-onset sepsis accounted for half of all anti-MRSA or antipseudomonal therapy; the use of these types of antibiotics increased between 2017 and 2021 despite resistant organisms being isolated in less than 10% of patients treated with broad-spectrum agents.

The Impact of Common Variations in Sequential Organ Failure Assessment Score Calculation on Sepsis Measurement Using Sepsis-3 Criteria: A Retrospective Analysis Using Electronic Health Record Data.

To assess the impact of different methods of calculating Sequential Organ Failure Assessment (SOFA) scores using electronic health record data on the incidence, outcomes, agreement, and predictive validity of Sepsis-3 criteria. Retrospective observational study. Five Massachusetts hospitals. Hospitalized adults, 2015 to 2022. None. We defined sepsis as a suspected infection (culture obtained and antibiotic administered) with a concurrent increase in SOFA score by greater than or equal to 2 points (Sepsis-3 criteria). Our reference SOFA implementation strategy imputed normal values for missing data, used Pa o2 /F io2 ratios for respiratory scores, and assumed normal baseline SOFA scores for community-onset sepsis. We then implemented SOFA scores using different missing data imputation strategies (averaging worst values from preceding and following days vs. carrying forward nonmissing values), imputing respiratory scores using Sp o2 /F io2 ratios, and incorporating comorbidities and prehospital laboratory data into baseline SOFA scores. Among 1,064,459 hospitalizations, 297,512 (27.9%) had suspected infection and 141,052 (13.3%) had sepsis with an in-hospital mortality rate of 10.3% using the reference SOFA method. The percentage of patients missing SOFA components for at least 1 day in the infection window was highest for Pa o2 /F io2 ratios (98.6%), followed by Sp o2 /F io2 ratios (73.5%), bilirubin (68.5%), and Glasgow Coma Scale scores (57.2%). Different missing data imputation strategies yielded near-perfect agreement in identifying sepsis (kappa 0.99). However, using Sp o2 /F io2 imputations yielded higher sepsis incidence (18.3%), lower mortality (8.1%), and slightly lower predictive validity for mortality (area under the receiver operating curves [AUROC] 0.76 vs. 0.78). For community-onset sepsis, incorporating comorbidities and historical laboratory data into baseline SOFA score estimates yielded lower sepsis incidence (6.9% vs. 11.6%), higher mortality (13.4% vs. 9.6%), and higher predictive validity (AUROC 0.79 vs. 0.75) relative to the reference SOFA implementation. Common variations in calculating respiratory and baseline SOFA scores, but not in handling missing data, lead to substantial differences in observed incidence, mortality, agreement, and predictive validity of Sepsis-3 criteria.

Predictive value of laboratory indicators for in-hospital death in children with community-onset sepsis: a prospective observational study of 266 patients

Background and objectivesThis study aimed to identify predictors of sepsis-associated in-hospital mortality from readily available laboratory biomarkers at onset of illness that include haematological, coagulation, liver and kidney function, blood...

Epidemiology and patterns of empiric antimicrobial therapy practice in patients with community-onset sepsis using data from a Japanese nationwide medical claims database—the Japan Sepsis Alliance (JaSA) study group

ObjectivesWe aimed to describe empiric antimicrobial options for patients with community-onset sepsis using nationwide real-world data from Japan. MethodsThis retrospective cohort study used nationwide Japanese data from a medical reimbursement system database. Patients aged ≥20 years with both presumed infections and acute organ dysfunction who were admitted to hospitals from the outpatient department or emergency department between 2010 and 2017 were enrolled. We described the initial choices of antimicrobials for patients with sepsis stratified by intensive care unit (ICU) or ward. ResultsThere were 1,195,741 patients with community-onset sepsis; of these, 1,068,719 and 127,022 patients were admitted to the wards and ICU, respectively. Third-generation cephalosporins and carbapenem were most commonly used for patients with community-onset sepsis. We found that 1.7% and 6.0% of patients initially used antimicrobials for methicillin-resistant Staphylococcus aureus coverage in the wards and ICU, respectively. Although half of the patients initially used antipseudomonal agents, only a few patients used a combination of antipseudomonal agents. Moreover, few patients initially used a combination of antimicrobials to treat methicillin-resistant Staphylococcus aureus and Pseudomonas sp. ConclusionThird-generation cephalosporins and carbapenem were most frequently used for patients with sepsis. A combination therapy of antimicrobials for drug-resistant bacteria coverage was rarely provided to these patients.

Increase in Incidence Rates and Risk Factors for Multidrug Resistant Bacteria in Septic Children: A Nationwide Spanish Cohort Study (2013-2019).

The emergence of multidrug-resistant (MDR) bacteria in children is a growing concern, particularly among septic patients, given the need for first-right dosing. Our aim was to determine the incidence rates and factors associated with MDR-sepsis in the pediatric intensive care unit (PICU), using data from the Spanish ENVIN-HELICS PICU registry between 2013 and 2019. The rate of MDR bacteria among septic children ranged between 5.8 and 16.2% throughout this study period, with a significant increase since 2015 (p = 0.013). MDR-gram-negative bacteria (92%), particularly EBL-Enterobacterales (63.7%), were the most frequent causative microorganisms of MDR-sepsis. During this study period, sixteen MDR-sepsis (32.6%) corresponded to intrahospital infections, and 33 (67.4%) had community-onset sepsis, accounting for 10.5% of the overall community-onset sepsis. Independent risk factors associated with MDR-sepsis were antibiotics 48 h prior to PICU admission (OR 2.38) and PICU onset of sepsis (OR 2.58) in >1 year-old children, and previous malnourishment (OR 4.99) in <1 year-old children. Conclusions: There was an alarming increase in MDR among septic children in Spain, mainly by gram-negative (ESBL-Enterobacterales), mostly coming from the community setting. Malnourished infants and children on antibiotics 48 h prior to PICU are at increased risk and therefore require closer surveillance.

Improving Sepsis Outcomes in the Era of Pay-for-Performance and Electronic Quality Measures: A Joint IDSA/ACEP/PIDS/SHEA/SHM/SIDP Position Paper.

The Centers for Medicare & Medicaid Services (CMS) introduced the Severe Sepsis/Septic Shock Management Bundle (SEP-1) as a pay-for-reporting measure in 2015 and is now planning to make it a pay-for-performance measure by incorporating it into the Hospital Value-Based Purchasing Program. This joint IDSA/ACEP/PIDS/SHEA/SHM/SIPD position paper highlights concerns with this change. Multiple studies indicate that SEP-1 implementation was associated with increased broad-spectrum antibiotic use, lactate measurements, and aggressive fluid resuscitation for patients with suspected sepsis but not with decreased mortality rates. Increased focus on SEP-1 risks further diverting attention and resources from more effective measures and comprehensive sepsis care. We recommend retiring SEP-1 rather than using it in a payment model and shifting instead to new sepsis metrics that focus on patient outcomes. CMS is developing a community-onset sepsis 30-day mortality electronic clinical quality measure (eCQM) that is an important step in this direction. The eCQM preliminarily identifies sepsis using systemic inflammatory response syndrome (SIRS) criteria, antibiotic administrations or diagnosis codes for infection or sepsis, and clinical indicators of acute organ dysfunction. We support the eCQM but recommend removing SIRS criteria and diagnosis codes to streamline implementation, decrease variability between hospitals, maintain vigilance for patients with sepsis but without SIRS, and avoid promoting antibiotic use in uninfected patients with SIRS. We further advocate for CMS to harmonize the eCQM with the Centers for Disease Control and Prevention's (CDC) Adult Sepsis Event surveillance metric to promote unity in federal measures, decrease reporting burden for hospitals, and facilitate shared prevention initiatives. These steps will result in a more robust measure that will encourage hospitals to pay more attention to the full breadth of sepsis care, stimulate new innovations in diagnosis and treatment, and ultimately bring us closer to our shared goal of improving outcomes for patients.

Previously healthy adults among septic patients: Population-level epidemiology and outcomes

Purpose: Previously healthy adults with community-onset sepsis were recently reported to have, counterintuitively, higher short-term mortality than those with comorbid conditions. However, the population-level generalizability of this finding and its applicability to all hospitalized septic patients are unclear.Methods: We used a statewide dataset to identify hospitalizations aged ≥18 years with a diagnosis of sepsis in Texas during 2018–2019. Comorbidities were defined as those included in the Charlson Comorbidity Index and other prevalent conditions associated with mortality. Hierarchical models were used to estimate the association of comorbid state with short-term mortality (defined as in-hospital mortality or discharge to hospice), overall and in community-onset and hospital-onset sepsis.Results: Among 120,371 sepsis hospitalizations, 6611 (5.5%) were previously healthy and 105,455 (87.6%) had community-onset sepsis. Short-term mortality among the previously healthy and those with comorbidities was 11.7% vs 28.2% overall, 11.0% vs 25.2% in community-onset sepsis, and 22.0% vs 48.7% in hospital-onset sepsis, respectively. On adjusted analysis, being previously healthy remained associated with lower short-term mortality overall (adjusted odds ratio 0.62 [95% CI 0.57–0.69]), with findings consistent with the primary analysis in community-onset sepsis, hospital-onset sepsis.Conclusions: Previously healthy septic patients had lower short-term mortality compared to those with comorbid conditions.

The impact of comorbidities and COVID-19 on the evolution of community onset sepsis

Sepsis is a disease with high mortality and morbidity despite advances in diagnostic procedures and therapeutic strategies. The aim of this study was to evaluate the profile and outcomes of community-onset sepsis. This retrospective, multicenter study included five 24-h health care units and was conducted from January 2018 to December 2021. Patients were diagnosed with sepsis or septic shock according to the Sepsis 3.0 criterion. A total of 2630 patients diagnosed as having sepsis (68.4%, 1800) or septic shock (31.6%, 830) in the 24-h health care unit were included; 43.76% of the patients were admitted to the intensive care unit, 12.2% died, 4.1% had sepsis and 30% had septic shock. The comorbidities that were independent predictors of septic shock were chronic kidney disease on dialysis (CKD-d), bone marrow transplantation and neoplasia. CKD and neoplasia were also independent predictors of mortality, with ORs of 2.00 (CI 1.10–3.68) p = 0.023 and 1.74 (CI 1.319–2.298) p = < 0.0001, respectively. Mortality according to the focus of primary infection was as follows: pulmonary 40.1%; COVID-19 35.7%; abdominal 8.1% and urinary 6.2%. Mortality due to the COVID-19 outbreak had an OR of 4.94 (CI 3.08–8.13) p ≤ 0.0001. Even though community-onset sepsis can be potentially fatal, this study revealed that some comorbidities lead to an increased risk of septic shock (d-CKD and neoplasia) and mortality. COVID-19 infection as the primary focus was an independent predictor of mortality in patients with sepsis when compared to other foci.

Comparison of Administrative versus Electronic Health Record-based Methods for Identifying Sepsis Hospitalizations.

Rationale: Despite the importance of sepsis surveillance, no optimal approach for identifying sepsis hospitalizations exists. The Centers for Disease Control and Prevention Adult Sepsis Event Definition (CDC-ASE) is an electronic medical record-based algorithm that yields more stable estimates over time than diagnostic coding-based approaches but may still result in misclassification. Objectives: We sought to assess three approaches to identifying sepsis hospitalizations, including a modified CDC-ASE. Methods: This cross-sectional study included patients in the Veterans Affairs Ann Arbor Healthcare System admitted via the emergency department (February 2021 to February 2022) with at least one episode of acute organ dysfunction within 48 hours of emergency department presentation. Patients were assessed for community-onset sepsis using three methods: 1) explicit diagnosis codes, 2) the CDC-ASE, and 3) a modified CDC-ASE. The modified CDC-ASE required at least two systemic inflammatory response syndrome criteria instead of blood culture collection and had a more sensitive definition of respiratory dysfunction. Each method was compared with a reference standard of physician adjudication via medical record review. Patients were considered to have sepsis if they had at least one episode of acute organ dysfunction graded as "definitely" or "probably" infection related on physician review. Results: Of 821 eligible hospitalizations, 449 were selected for physician review. Of these, 98 (21.8%) were classified as sepsis by medical record review, 103 (22.9%) by the CDC-ASE, 132 (29.4%) by the modified CDC-ASE, and 37 (8.2%) by diagnostic codes. Accuracy was similar across the three methods of interest (80.6% for the CDC-ASE, 79.6% for the modified CDC-ADE, and 84.2% for diagnostic codes), but sensitivity and specificity varied. The CDC-ASE algorithm had sensitivity of 58.2% (95% confidence interval [CI], 47.2-68.1%) and specificity of 86.9% (95% CI, 82.9-90.2%). The modified CDC-ASE algorithm had greater sensitivity (69.4% [95% CI, 59.3-78.3%]) but lower specificity (81.8% [95% CI, 77.3-85.7%]). Diagnostic codes had lower sensitivity (32.7% [95% CI, 23.5-42.9%]) but greater specificity (98.6% [95% CI, 96.7-99.55%]). Conclusions: There are several approaches to identifying sepsis hospitalizations for surveillance that have acceptable accuracy. These approaches yield varying sensitivity and specificity, so investigators should carefully consider the test characteristics of each method before determining an appropriate method for their intended use.

Epidemiology of Sepsis in US Children and Young Adults.

Most multicenter studies of US pediatric sepsis epidemiology use administrative data or focus on pediatric intensive care units. We conducted a detailed medical record review to describe sepsis epidemiology in children and young adults. In a convenience sample of hospitals in 10 states, patients aged 30 days-21 years, discharged during 1 October 2014-30 September 2015, with explicit diagnosis codes for severe sepsis or septic shock, were included. Medical records were reviewed for patients with documentation of sepsis, septic shock, or similar terms. We analyzed overall and age group-specific patient characteristics. Of 736 patients in 26 hospitals, 442 (60.1%) had underlying conditions. Most patients (613 [83.3%]) had community-onset sepsis, although most community-onset sepsis was healthcare associated (344 [56.1%]). Two hundred forty-one patients (32.7%) had outpatient visits 1-7 days before sepsis hospitalization, of whom 125 (51.9%) received antimicrobials ≤30 days before sepsis hospitalization. Age group-related differences included common underlying conditions (<5 years: prematurity vs 5-12 years: chronic pulmonary disease vs 13-21 years: chronic immunocompromise); medical device presence ≤30 days before sepsis hospitalization (1-4 years: 46.9% vs 30 days-11 months: 23.3%); percentage with hospital-onset sepsis (<5 years: 19.6% vs ≥5 years: 12.0%); and percentage with sepsis-associated pathogens (30 days-11 months: 65.6% vs 13-21 years: 49.3%). Our data suggest potential opportunities to raise sepsis awareness among outpatient providers to facilitate prevention, early recognition, and intervention in some patients. Consideration of age-specific differences may be important as approaches are developed to improve sepsis prevention, risk prediction, recognition, and management.

Epidemiology, Resistance Profiles, and Outcomes of Bloodstream Infections in Community-Onset Sepsis in the United States.

To describe frequency of positive blood cultures, patterns of pathogens' characteristics and their resistance profile in patients with blood cultures drawn due to a presumed diagnosis of community-onset sepsis, and to examine the association between blood culture-positive pathogens and hospital mortality. Retrospective cohort study. Two hundred one U.S. hospitals from 2016 to 2020 using the Premier Healthcare Database. Adult patients presenting with community-onset sepsis who had blood cultures collected within 2 days of hospital admission. We defined sepsis using the U.S. Centers for Disease Control Adult Sepsis Event Surveillance criteria. None. We identified 147,061 patients with community-onset sepsis. The number of blood culture-positive sepsis episodes was 21,167 (14%) and the number of nonblood culture-positive sepsis episodes was 20,326 (14%). Among patients with blood culture-positive sepsis, Gram-negative rods were isolated in 55% of patients, Gram-positive cocci were isolated in 47%. Of those, methicillin-resistant Staphylococcus aureus (MRSA) was 11%, ceftriaxone-resistant Enterobacterales /extended-spectrum β-lactamase was 7%, and carbapenem-resistant Enterobacterales was 1.3%. The crude in-hospital mortality was 17% for culture-negative sepsis, 13% for nonblood culture-positive sepsis, and 17% for blood culture-positive sepsis. In multilevel logistic regression models, compared with culture-negative sepsis, blood culture-positive sepsis (adjusted odds ratio [aOR], 0.89; 95% CI, 0.85-0.94) and nonblood culture-positive sepsis (aOR, 0.82; 95% CI, 0.78-0.87) were associated with lower odds of in-hospital mortality. Acinetobacter species, Pseudomonas aeruginosa , methicillin-sensitive Staphylococcus aureus , and MRSA were associated with higher in-hospital mortality, whereas Escherichia coli , Klebsiella species, Proteus species, and Streptococcus species were associated with lower in-hospital mortality. In patients hospitalized with community-onset sepsis, the prevalence of blood culture-positive sepsis was 14%. Among positive blood culture sepsis resistant organisms were infrequent. Compared with culture-negative sepsis, blood culture-positive sepsis and nonblood culture-positive sepsis were associated with lower in-hospital mortality.

Benchmarking of two bioinformatic workflows for the analysis of whole-genome sequenced Staphylococcus aureus collected from patients with suspected sepsis

BackgroundThe rapidly growing area of sequencing technologies, and more specifically bacterial whole-genome sequencing, could offer applications in clinical microbiology, including species identification of bacteria, prediction of genetic antibiotic susceptibility and virulence genes simultaneously. To accomplish the aforementioned points, the commercial cloud-based platform, 1928 platform (1928 Diagnostics, Gothenburg, Sweden) was benchmarked against an in-house developed bioinformatic pipeline as well as to reference methods in the clinical laboratory.MethodsWhole-genome sequencing data retrieved from 264 Staphylococcus aureus isolates using the Illumina HiSeq X next-generation sequencing technology was used. The S. aureus isolates were collected during a prospective observational study of community-onset severe sepsis and septic shock in adults at Skaraborg Hospital, in the western region of Sweden. The collected isolates were characterized according to accredited laboratory methods i.e., species identification by MALDI-TOF MS analysis and phenotypic antibiotic susceptibility testing (AST) by following the EUCAST guidelines. Concordance between laboratory methods and bioinformatic tools, as well as concordance between the bioinformatic tools was assessed by calculating the percent of agreement.ResultsThere was an overall high agreement between predicted genotypic AST and phenotypic AST results, 98.0% (989/1006, 95% CI 97.3–99.0). Nevertheless, the 1928 platform delivered predicted genotypic AST results with lower very major error rates but somewhat higher major error rates compared to the in-house pipeline. There were differences in processing times i.e., minutes versus hours, where the 1928 platform delivered the results faster. Furthermore, the bioinformatic workflows showed overall 99.4% (1267/1275, 95% CI 98.7–99.7) agreement in genetic prediction of the virulence gene characteristics and overall 97.9% (231/236, 95% CI 95.0–99.2%) agreement in predicting the sequence types (ST) of the S. aureus isolates.ConclusionsAltogether, the benchmarking disclosed that both bioinformatic workflows are able to deliver results with high accuracy aiding diagnostics of severe infections caused by S. aureus. It also illustrates the need of international agreement on quality control and metrics to facilitate standardization of analytical approaches for whole-genome sequencing based predictions.

527. Antimicrobial stewardship for empirical treatment of bloodstream infection using machine learning clinical decision support

Abstract Background Undertreatment of sepsis is associated with increased mortality, but only 12 to 18% of people require broad-spectrum Gram negative or positive coverage. Unnecessary use of broad-spectrum treatments represents a modifiable driver for antimicrobial resistance (AMR). Here, we test the hypothesis that a recommendation algorithm built on ML models trained on electronic health record data can safely reduce the spectrum of activity for empiric antibiotic treatments relative to clinicians for patients with community-onset sepsis. Methods We included 29,508 people with a community-onset Adult Sepsis Event in the Mass General Brigham health system. We split the data into a training and test set by time, June 2016 to June 2020 and July 2020 to February 2021, respectively. We extracted 1,200 predictors and windowed time-varying features into non-overlapping windows. We used logistic regression, random forest and XGBoost to predict pathogen, and then predict AMR to treatments relevant to the most likely pathogen. We used dynamic thresholding to convert probabilities into AMR phenotypes and then chose the antibiotic of the narrowest spectrum among those susceptible. Outcomes were the percentage of decisions that were inappropriately narrow versus broad spectrum. We compared model performance to that of clinicians. Figure 1:Analytic pipeline Schematic of machine learning architecture, thresholding algorithm and calculation of outcomes. Results The highest performing models for pathogen prediction were XGBoost, with area under the receiver operator curve (AUROC) statistics ranging from 0.60 for Staphylococcus spp. to 0.68 for Proteus spp. Random forests performed best for AMR prediction, with AUROCs ranging from 0.59 for ampicillin-sulbactam to 0.74 for penicillin. The model utilized broad-spectrum therapy in 92.4% (92.2% - 92.7%) of patients compared to 89.3% (88.9% - 89.6%) for clinicians, a relative increase of 3.5%, but chose inappropriately narrow treatments in 7.3% (7.1% - 7.6%) of patients, compared to 9.4% (9.1% - 9.6%) for providers, a 29% reduction. Figure 2:Model performance for pathogen prediction AUROCs for prediction of pathogen groupings for patients with community-onset sepsis. All models are XGBoost. Figure 3:Model performance for AMR prediction AUROCs for prediction of antimicrobial resistance based on most likely predicted pathogens for patients with community-onset sepsis. All models are random forests. Figure 4:Post-hoc outcome analysis Comparison of rates of inappropriately narrow antibiotic treatment (IAT) and use of broad spectrum therapy for the machine learning decision algorithm versus clinicians for patients with community onset sepsis. Conclusion An ML decision support tool for antibiotic prescription in community-onset sepsis did not result in decreased broad spectrum use but was able to reduce inappropriately narrow treatment. Work is underway to refine feature sets, improve the thresholding algorithm and better account for time-varying confounders. Disclosures Sanjat Kanjilal, MD, MPH, GlaxoSmithKline: Advisor/Consultant|Roche Diagnostics: Honoraria David Sontag, PhD, Adobe: Grant/Research Support|ASAPP: Advisor/Consultant|Cureai Health: Stocks/Bonds|Facebook: Grant/Research Support|Google: Grant/Research Support|IBM: Grant/Research Support|SAP: Grant/Research Support|Takeda: Grant/Research Support.

1183: USE OF PERIPHERAL VASOPRESSORS IN EARLY SEPSIS-INDUCED HYPOTENSION ACROSS MICHIGAN HOSPITALS

Introduction: Recent data suggest it may be safe to administer vasopressors via peripheral IV (PIV), challenging convention that vasopressors must be delivered centrally. Surviving Sepsis Campaign 2021 guidelines suggest using peripheral vasopressors as a bridge to central access. However, little is known about vasopressor initiation in practice. Methods: Cohort study of patients hospitalized with community-onset sepsis at 12 hospitals in the Hospital Medicine Safety Consortium (HMS) sepsis initiative. HMS is a Collaborative Quality Initiative sponsored by Blue Cross Blue Shield of Michigan. A random sample of adult sepsis hospitalizations between 11/2020-1/2022 were included. Data were abstracted by trained abstractors. We sought to determine how commonly vasopressors were initiated via PIV vs central access across hospitals. HMS-Sepsis is expanding to 69 hospitals. Here we present pilot data; full cohort analysis is in process. Results: Of 1,901 patients in the HMS-Sepsis registry at the time of pilot data analysis, 440 (23.1%) had hypotension (defined by mean arterial pressure< 65mmHg, systolic blood pressure< 90mmHg, and/or vasopressor initiation) within 3 hours of hospital arrival. Of these, 160 (36.4%) received vasopressors within 6 hours of hospital arrival. Route of initial vasopressor was PIV in 122 (76.3%), central access in 30 (18.8%), midline catheter in 1 (0.6%), oral (ie, midodrine) in 5 (3.1%), and unknown in 2 (1.3%). Across all hospitals, 50.0% to 91.7% of vasopressor initiation was via PIV (median 83.3%). Among 122 patients with vasopressor initiation via PIV, 66 (54.1%) received a 2nd vasopressor, after a median of 2.8 hrs [IQR 1, 8] from 1st vasopressor. Route of 2nd vasopressor was PIV in 27 (40.9%) and central access in 30 (45.4%). Time from hypotension to vasopressor initiation did not differ between patients receiving initial vasopressor via PIV vs central access (median 1.9 vs 2.1 hrs, p=0.79). Likewise, IV fluids within 6 hrs (median 2.0 vs 2.1L, p=0.78), hospitalization length (median 7 vs 6 days, p=0.31), and in-hospital mortality (33.6% vs 40.0%, p=0.51) were similar. Conclusions: In this 12-hospital cohort, vasopressors were most frequently initiated peripherally. Outcomes were similar between patients in whom vasopressors were initiated via peripheral vs central access.

Short- versus long-course antibiotic therapy for sepsis: a post hoc analysis of the nationwide cohort study

BackgroundThe appropriate duration of antibiotic treatment in patients with bacterial sepsis remains unclear. The purpose of this study was to evaluate the association of a shorter course of antibiotics on 28-day mortality in comparison with a longer course using a national database in Japan.MethodsWe conducted a post hoc analysis from the retrospective observational study of patients with sepsis using a Japanese claims database from 2010 to 2017. The patient dataset was divided into short-course (≤ 7 days) and long-course (≥ 8 days) groups according to the duration of initial antibiotic administration. Subsequently, propensity score matching was performed to adjust the baseline imbalance between the two groups. The primary outcome was 28-day mortality. The secondary outcomes were re-initiated antibiotics at 3 and 7 days, during hospitalization, administration period, antibiotic-free days, and medical cost.ResultsAfter propensity score matching, 448,146 pairs were analyzed. The 28-day mortality was significantly lower in the short-course group (hazard ratio, 0.94; 95% CI, 0.92–0.95; P < 0.001), while the occurrence of re-initiated antibiotics at 3 and 7 days and during hospitalization were significantly higher in the short-course group (P < 0.001). Antibiotic-free days (median [IQR]) were significantly shorter in the long-course group (21 days [17 days, 23 days] vs. 17 days [14 days, 19 days], P < 0.001), and short-course administration contributed to a decrease in medical costs (coefficient $-212, 95% CI; − 223 to − 201, P < 0.001). Subgroup analyses showed a significant decrease in the 28-day mortality of the patients in the short-course group in patients of male sex (hazard ratio: 0.91, 95% CI; 0.89–0.93), community-onset sepsis (hazard ratio; 0.95, 95% CI; 0.93–0.98), abdominal infection (hazard ratio; 0.92, 95% CI; 0.88–0.97) and heart infection (hazard ratio; 0.74, 95% CI; 0.61–0.90), while a significant increase was observed in patients with non-community-onset sepsis (hazard ratio; 1.09, 95% CI; 1.06–1.12).ConclusionsThe 28-day mortality was significantly lower in the short-course group, even though there was a higher rate of re-initiated antibiotics in the short course.

Sequence-based genotyping of extra-intestinal pathogenic Escherichia coli isolates from patients with suspected community-onset sepsis, Sweden

Extra-intestinal pathogenic Escherichia coli (ExPEC) strains are responsible for a large number of human infections globally. The management of infections caused by ExPEC has been complicated by the emergence of antimicrobial resistance, most importantly the increasing recognition of isolates producing extended-spectrum β-lactamases (ESBL). Herein, we used whole-genome sequencing (WGS) on ExPEC isolates for a comprehensive genotypic characterization. Twenty-one ExPEC isolates, nine with and 12 without ESBL-production, from 16 patients with suspected sepsis were sequenced on an Illumina MiSeq platform. Analysis of WGS data was performed with widely used bioinformatics software and tools for genotypic characterization of the isolates. A higher number of plasmids, virulence and resistance genes were observed in the ESBL-producing isolates than the non-ESBL-producing, although not statistically significant due to the low sample size. All nine ESBL-producing ExPEC isolates presented with at least one bla gene, as did three of the 12 without ESBL-production. Multi-locus sequence typing analysis revealed a diversity of sequence types whereas phylogroup A prevailed among isolates both with and without ESBL-production. In conclusion, this limited study shows that analysis of WGS data can be used for genotypic characterization of ExPEC isolates to obtain in-depth information of clinical relevance.

Mortality of patients with hospital-onset sepsis in hospitals with all-day and non-all-day rapid response teams: a prospective nationwide multicenter cohort study

BackgroundHospital-onset sepsis is associated with a higher in-hospital mortality rate than community-onset sepsis. Many hospitals have implemented rapid response teams (RRTs) for early detection and timely management of at-risk hospitalized patients. However, the effectiveness of an all-day RRT over a non-all-day RRT in reducing the risk of in-hospital mortality in patient with hospital-onset sepsis is unclear. We aimed to determine the effect of the RRT’s operating hours on in-hospital mortality in inpatient patients with sepsis.MethodsWe conducted a nationwide cohort study of adult patients with hospital-onset sepsis prospectively collected from the Korean Sepsis Alliance (KSA) Database from 16 tertiary referral or university-affiliated hospitals in South Korea between September of 2019 and February of 2020. RRT was implemented in 11 hospitals, of which 5 (45.5%) operated 24-h RRT (all-day RRT) and the remaining 6 (54.5%) had part-day RRT (non-all-day RRT). The primary outcome was in-hospital mortality between the two groups.ResultsOf the 405 patients with hospital-onset sepsis, 206 (50.9%) were admitted to hospitals operating all-day RRT, whereas 199 (49.1%) were hospitalized in hospitals with non-all-day RRT. A total of 73 of the 206 patients in the all-day group (35.4%) and 85 of the 199 patients in the non-all-day group (42.7%) died in the hospital (P = 0.133). After adjustments for co-variables, the implementation of all-day RRT was associated with a significant reduction in in-hospital mortality (adjusted odds ratio 0.57; 95% confidence interval 0.35–0.93; P = 0.024).ConclusionsIn comparison with non-all-day RRTs, the availability of all-day RRTs was associated with reduced in-hospital mortality among patients with hospital-onset sepsis.

Performance of hospital administrative data for detection of sepsis in Australia: The sepsis coding and documentation (SECOND) study.

Sepsis is the world's leading cause of death and its detection from a range of data and coding sources, consistent with consensus clinical definition, is desirable. To evaluate the performance of three coding definitions (explicit, implicit, and newly proposed synchronous method) for sepsis derived from administrative data compared to a clinical reference standard. Extraction of administrative coded data from Australian metropolitan teaching hospital with 25,000 annual overnight admissions compared to clinical review of medical records; 313 (27.9%) randomly selected adult multi-day stay hospital separations from 1,123 separations with acute infection during July 2019. Estimated prevalence and performance metrics, including positive (PPV) and negative predictive values (NPV), and area under the receiver operator characteristic curve (ROC). Clinical prevalence of sepsis was estimated at 10.7 (95% CI = 10.3-11.3) per 100 separations, and mortality rate of 11.6 (95% CI = 10.3-13.0) per 100 sepsis separations. Explicit method for case detection had high PPV (93.2%) but low NPV (55.8%) compared to the standard implicit method (74.1 and 66.3%, respectively) and proposed synchronous method (80.4% and 80.0%) compared to a standard clinical case definition. ROC for each method: 0.618 (95% CI = 0.538-0.654), 0.698 (95% CI = 0.648-0.748), and 0.802 (95% CI = 0.757-0.846), respectively. In hospitalised Australian patients with community-onset sepsis, the explicit method for sepsis case detection underestimated prevalence. Implicit methods were consistent with consensus definition for sepsis, and proposed synchronous method had better performance.