Search

BACKGROUND: Early detection of sepsis in patients admitted to the emergency department (ED) is an important clinical objective to help reduce morbidity and mortality. We aimed to use data from Electronic Health Records (EHR) system to characterize the relative importance of a new biomarker called Monocyte Distribution Width (MDW) that has been recently approved by the US Food and Drug Administration (FDA) for sepsis screening in the presence of routinely available hematologic parameters and vital signs measures. METHODS: In this retrospective cohort study, we included ED patients admitted to the MetroHealth hospital (a large regional safety-net hospital in Cleveland, OH, USA) with suspected infection who later developed severe sepsis. All adult patients presenting to the ED were eligible for inclusion and encounters that did not have complete blood count with differential data or vital signs data were excluded. We developed seven data models and an ensemble of four high accuracy machine learning (ML) algorithms using the Sepsis-3 diagnostic criteria for validation. Using the results generated by the high accuracy ML models, we applied the Local Interpretable Model-Agnostic Explanation (LIME) and Shapley Additive Value (SHAP) post-hoc ML interpretability methods to characterize the contributions of individual hematologic parameters, including MDW, vital signs measures in screening for severe sepsis. FINDINGS: We evaluated 7071 adult patients from 303,339 adult ED visits occurring between May 1 (st) , 2020 and August 26 (th) , 2022. Implementation of the seven data models reflected the ED clinical workflow with incremental addition of standard complete blood count (CBC), CBC with differential, with MDW, and finally vital signs measures. Random forest and deep neural network model reported classification area under the receiver operating characteristic curve (AUC) value of up to 93% (CI 92 - 94) and 90% (CI 88 – 91) over data model with hematologic parameters and vital signs measures. We applied the LIME and SHAP ML interpretability methods on these high accuracy ML models. Both the interpretability methods were consistent in their findings that the value of MDW is grossly attenuated (low feature importance scores of 0.015 (SHAP) and 0.0004 (LIME)) in the presence of other routinely reported hematologic parameters and vital signs measures for severe sepsis detection. INTERPRETATION: Using ML interpretability methods applied to EHR data, we show that MDW can be replaced with routinely reported CBC with differential together with vital signs measures for severe sepsis screening. MDW requires specialized laboratory equipment and modification of existing care protocols; therefore, these results could guide decisions about allocation of limited resources in cost constrained care settings. Additionally, the analysis shows the practical application of ML interpretability methods in clinical decision making. FUNDING: National Institute of Biomedical Imaging and Bioengineering, National Institutes of Health/National Center for Advancing Translational Sciences, National Institute on Drug Abuse

Spirometry is necessary to diagnose chronic obstructive pulmonary disease (COPD), yet a large proportion of patients are diagnosed and treated without having received testing. This study explored whether the effects of interventions using the electronic health record (EHR) to target patients diagnosed with COPD without confirmatory spirometry impacted the incidence rates of spirometry referrals and completions. This retrospective before and after study assessed the impact of provider-facing clinical decision support that identified patients who had a diagnosis of COPD but had not received spirometry. Spirometry referrals, completions, and results were ascertained 1.5 years prior to and 1.5 years after the interventions were initiated. Inhaler prescriptions by class were also tallied. There were 10,949 unique patients with a diagnosis of COPD who were eligible for inclusion. 4,895 patients (44.7%) were excluded because they had completed spirometry prior to the cohort start dates. The pre-intervention cohort consisted of 2,622 patients, while the post-intervention cohort had 3,392. Spirometry referral rates pre-intervention were 20.2% compared to 31.6% post-intervention (p < 0.001). Spirometry completion rates rose from 13.2% pre-intervention to 19.3% afterwards (p < 0.001). 61.7% (585 of 948) had no evidence of airflow obstruction. After excluding patients with a diagnosis of asthma, 25.8% (126 of 488) patients who had no evidence of airflow obstruction had prescriptions for long-acting bronchodilators or inhaled steroids. A concerted EHR intervention modestly increased spirometry referral and completion rates in patients with a diagnosis of COPD without prior spirometry and decreased misclassification of disease.

Additional file 4: Supplementary Fig. 3. Age Distribution of Hospitalized Patients with SARS-CoV2 who Experienced AKI within First 7 days of Hospitalization-baseline creatinine estimator MDRD equation removing race. Main figure presents percentage per age bracket who developed acute kidney injury (AKI) among all hospitalized patients and further stratified by severity of illness status. Severe illness is defined as a composite indicator of invasive ventilation, use of vasopressor(s)/inotrope(s), and/or use of extracorporeal membrane oxygenation. Moderate illness is defined as admitted to an intensive care unit but without use of above organ support measures. Mild illness is defined as patient required hospitalization but not in an intensive care unit and without use of above organ support measures. Insert presents the adjusted odds ratio (OR) with 95% confidence intervals (CI) of developing AKI within the first week by age bracket compared to young adults (30���35-year-olds) as the referent category. Adjusted for sex, race/ethnicity, pre-existing hypertension, diabetes mellitus, cancer, chronic kidney disease, and severity of illness. AKI defined per KDIGO guidelines, but baseline creatinine estimator uses modified MDRD equation removing race component for adults (���18 years) and height-independent equation for children (

Additional file 3: Supplementary Fig. 2. Age Distribution of Hospitalized Patients with SARS-CoV2 who Experienced AKI within First 7 days of Hospitalization-baseline creatinine estimator FAS equation. Main figure presents percentage per age bracket who developed acute kidney injury (AKI) among all hospitalized patients and further stratified by severity of illness status. Severe illness is defined as a composite indicator of invasive ventilation, use of vasopressor(s)/inotrope(s), and/or use of extracorporeal membrane oxygenation. Moderate illness is defined as admitted to an intensive care unit but without use of above organ support measures. Mild illness is defined as patient required hospitalization but not in an intensive care unit and without use of above organ support measures. Insert presents the adjusted odds ratio (OR) with 95% confidence intervals (CI) of developing AKI within the first week by age bracket compared to young adults (30���35-year-olds) as the referent category. Adjusted for sex, race/ethnicity, pre-existing hypertension, diabetes mellitus, cancer, chronic kidney disease, and severity of illness. AKI defined per KDIGO guidelines, but baseline creatinine estimator uses full-age spectrum (FAS) equation for all participants.

Additional file 1: Supplementary Table 1. Initial Hospital-related Associations with SARS-CoV2 -related AKI. These therapies or complications occur within the first 7 days of hospitalization when SARS-CoV2-related AKI is defined. Data presented as number (column percentiles), except where specified. ACE-I = angiotensin-converting enzyme-inhibitors; AKI = acute kidney injury; ARB = angiotensin receptor blockers; IVIG = intravenous immunoglobulin; NSAID = non-steroidal anti-inflammatory drugs; PRISM = Pediatric Risk of Mortality Score; SOFA = Sequential Organ Failure Assessment. aInitial PRISM score missing for 497 pediatric patients. Baseline SOFA score missing for 2741 adult patients; maximum SOFA score missing for 2016 adult patients.

Additional file 5: Supplementary Fig. 4. Age Distribution of Hospitalized Patients with SARS-CoV2 who Experienced AKI within First 7 days of Hospitalization-baseline creatinine estimator as minimum serum creatinine. Main figure presents percentage per age bracket who developed acute kidney injury (AKI) among all hospitalized patients and further stratified by severity of illness status. Severe illness is defined as a composite indicator of invasive ventilation, use of vasopressor(s)/inotrope(s), and/or use of extracorporeal membrane oxygenation. Moderate illness is defined as admitted to an intensive care unit but without use of above organ support measures. Mild illness is defined as patient required hospitalization but not in an intensive care unit and without use of above organ support measures. Insert presents the adjusted odds ratio (OR) with 95% confidence intervals (CI) of developing AKI within the first week by age bracket compared to young adults (30���35-year-olds) as the referent category. Adjusted for sex, race/ethnicity, pre-existing hypertension, diabetes mellitus, cancer, chronic kidney disease, and severity of illness. AKI defined per KDIGO guidelines, but baseline creatinine estimator uses minimum serum creatinine value within first 7 days of hospitalization for all participants.

Additional file 2: Supplementary Fig. 1. Age Distribution of Hospitalized Patients with SARS-CoV2 who Experienced AKI within First 7 days of Hospitalization-different baseline creatinine estimators. Main figure presents percentage per age bracket who developed acute kidney injury (AKI) among all hospitalized patients. The original AKI definition (blue) assumes a baseline creatinine based on KDIGO guidelines for adults (eGFR 75 ml/min/1.73m2 and back calculates using MDRD equation) and common pediatric definitions assuming an eGFR of 120 ml/min/1.73m2 and back calculating using height-independent equation, except for patients with CKD when minimum serum creatinine during first 7 days of hospitalization is assumed to be their baseline creatinine value. Orange line assumes that the minimum creatinine during the first 7 days of hospitalization is the baseline creatinine for all participants. Gray line uses the KDIGO guidelines but back calculates the baseline creatinine for all participants using the FAS equation. Yellow line uses the original definition but uses the MDRD equation minus the race variable. Abbreviations: AKI = acute kidney injury, CKD = chronic kidney disease, eGFR = estimated glomerular filtration rate, FAS = full age spectrum, KDIGO=Kidney Disease Improving Global Outcomes, MDRD = modification of diet in renal disease.

Background Hospitalized patients with SARS-CoV2 develop acute kidney injury (AKI) frequently, yet gaps remain in understanding why adults seem to have higher rates compared to children. Our objectives were to evaluate the epidemiology of SARS-CoV2-related AKI across the age spectrum and determine if known risk factors such as illness severity contribute to its pattern. Methods Secondary analysis of ongoing prospective international cohort registry. AKI was defined by KDIGO-creatinine only criteria. Log-linear, logistic and generalized estimating equations assessed odds ratios (OR), risk differences (RD), and 95% confidence intervals (CIs) for AKI and mortality adjusting for sex, pre-existing comorbidities, race/ethnicity, illness severity, and clustering within centers. Sensitivity analyses assessed different baseline creatinine estimators. Results Overall, among 6874 hospitalized patients, 39.6% (n = 2719) developed AKI. There was a bimodal distribution of AKI by age with peaks in older age (≥60 years) and middle childhood (5–15 years), which persisted despite controlling for illness severity, pre-existing comorbidities, or different baseline creatinine estimators. For example, the adjusted OR of developing AKI among hospitalized patients with SARS-CoV2 was 2.74 (95% CI 1.66–4.56) for 10–15-year-olds compared to 30–35-year-olds and similarly was 2.31 (95% CI 1.71–3.12) for 70–75-year-olds, while adjusted OR dropped to 1.39 (95% CI 0.97–2.00) for 40–45-year-olds compared to 30–35-year-olds. Conclusions SARS-CoV2-related AKI is common with a bimodal age distribution that is not fully explained by known risk factors or confounders. As the pandemic turns to disproportionately impacting younger individuals, this deserves further investigation as the presence of AKI and SARS-CoV2 infection increases hospital mortality risk.

Additional file 6: Supplementary Fig. 5. Age Distribution of Hospitalized Patients with SARS-CoV2 who Experienced AKI within First 7 days of Hospitalization Stratified by U.S. versus non-U.S. Hospital. Presents percentage of hospitalized patients who developed acute kidney injury (AKI) among all hospitalized patients and further stratified by hospital center based in the United States versus not in the United States. AKI defined per KDIGO guidelines.

Viral infections such as influenza have been shown to predispose hosts to increased colonization of the respiratory tract by pathogenic bacteria and secondary bacterial pneumonia. To examine how viral infections and host antiviral immune responses alter the upper respiratory microbiome, we analyzed the nasal bacterial composition in healthy volunteers at baseline, 1-2 weeks, and 4-6 weeks after instillation of a live attenuated influenza vaccine or intranasal sterile saline. In parallel, we also assessed changes in the nasal microbiome over similar time points in patients presenting with flu-like illness in ambulatory settings. All samples were subjected to 16S rRNA gene sequencing for assessment of the microbiota, while a subset of samples from vaccinated subjects was submitted for microarray host gene expression profiling. We found that live attenuated influenza vaccination led to significant changes in microbial community structure, diversity and core taxonomic membership as well as increases in the relative abundances of Staphylococcus and Bacteroides (both p