Unique Patients Research Articles

Establishing Medical Intelligence-Leveraging Fast Healthcare Interoperability Resources to Improve Clinical Management: Retrospective Cohort and Clinical Implementation Study.

FHIR (Fast Healthcare Interoperability Resources) has been proposed to enable health data interoperability. So far, its applicability has been demonstrated for selected research projects with limited data. This study aimed to design and implement a conceptual medical intelligence framework to leverage real-world care data for clinical decision-making. A Python package for the use of multimodal FHIR data (FHIRPACK [FHIR Python Analysis Conversion Kit]) was developed and pioneered in 5 real-world clinical use cases, that is, myocardial infarction, stroke, diabetes, sepsis, and prostate cancer. Patients were identified based on the ICD-10 (International Classification of Diseases, Tenth Revision) codes, and outcomes were derived from laboratory tests, prescriptions, procedures, and diagnostic reports. Results were provided as browser-based dashboards. For 2022, a total of 1,302,988 patient encounters were analyzed. (1) Myocardial infarction: in 72.7% (261/359) of cases, medication regimens fulfilled guideline recommendations. (2) Stroke: out of 1277 patients, 165 received thrombolysis and 108 thrombectomy. (3) Diabetes: in 443,866 serum glucose and 16,180 glycated hemoglobin A1c measurements from 35,494 unique patients, the prevalence of dysglycemic findings was 39% (13,887/35,494). Among those with dysglycemia, diagnosis was coded in 44.2% (6138/13,887) of the patients. (4) Sepsis: In 1803 patients, Staphylococcus epidermidis was the primarily isolated pathogen (773/2672, 28.9%) and piperacillin and tazobactam was the primarily prescribed antibiotic (593/1593, 37.2%). (5) PC: out of 54, three patients who received radical prostatectomy were identified as cases with prostate-specific antigen persistence or biochemical recurrence. Leveraging FHIR data through large-scale analytics can enhance health care quality and improve patient outcomes across 5 clinical specialties. We identified (1) patients with sepsis requiring less broad antibiotic therapy, (2) patients with myocardial infarction who could benefit from statin and antiplatelet therapy, (3) patients who had a stroke with longer than recommended times to intervention, (4) patients with hyperglycemia who could benefit from specialist referral, and (5) patients with PC with early increases in cancer markers.

Cohort Profile: VZNKUL–NMIBC Quality Indicators Program: A Flemish Prospective Cohort to Evaluate the Quality Indicators in the Treatment of Non-Muscle-Invasive Bladder Cancer

Purpose: Bladder cancer (BC) is a heterogeneous disease with varying outcomes, influenced by disease heterogeneity and variability in treatment and follow-up. Risk groups have been established for non–muscle-invasive BC (NMIBC) to standardize therapy, and several quality control indicators (QCIs) monitor adherence to these risk group-based guidelines. However, controversial results had been obtained regarding the oncological benefits of these QCIs until recent high-quality studies from large registries showed their usefulness. To improve adherence to the European Association of Urology (EAU) Guidelines and benchmark current care in Flemish hospitals within Vlaams Ziekenhuisnetwerk–KU Leuven (VZNKUL), a QCI program for NMIBC was initiated in 2013. This study aims to describe the demographic, clinical, and treatment data of patients enrolled in this program. Participants: The VZNKUL–NMIBC Quality Indicators Program Registry is a prospective cohort including patients treated and followed up with at seven academic and non-academic Flemish hospitals since June 2013. Data collection includes patient characteristics, tumor data, treatment, and oncological outcomes. Findings to date: From June 2013 to December 2020, 4744 transurethral resections of bladder tumors (TURBTs) from 2237 unique patients were analyzed. Most patients (80%) were men with a median age of 73. The median time from diagnosis to TURBT was 19 days. A single tumor was detected in 37% of TURBTs. Tumors larger than 3 cm were found in 20.8% of cases. In 46% of TURBTs, a reTURBT was scheduled according to guidelines. The complication rates were 7.5% and 2.4% for bladder perforation and bleeding, respectively. Postoperative single intravesical instillation of chemotherapy (SIVIC) was administered to 56.9% of 1533 indicated patients with a median time to administration of 4.7 h. Among the cohort, 60.4% had NMIBC, and 9.3% had muscle-invasive BC. Of 972 high-risk patients, 60.7% received adequate BCG induction, while 39.4% received adequate maintenance. After BCG induction ± maintenance, 39.7% were tumor-free, with 17.7% recurrence and 4% progression to muscle-invasive BC. BCG treatment was terminated early for 17% of patients due to intolerance. Early cystectomy was performed for 2.4% of the BCG-naïve patients, and 27.7% of patients with BCG failure underwent a BCG rechallenge. For intermediate-risk patients, 2.1% received adequate BCG, and 23% received intravesical chemotherapy. The median follow-up was 57 months. Five-year recurrence-free, progression-free, cancer-free, overall, and cancer-specific survival rates were 53%, 91.6%, 89%, 70.6%, and 95.6%, respectively, for the NMIBC patients. Of 400 non-metastatic MIBC patients, 217 (54.3%) underwent radical cystectomy (RC), of whom 46% received neoadjuvant chemotherapy, while 18 (4.5%) refused RC, and 74 (18.5%) were considered unfit for the surgery. Future plans: The VZNKUL–NMIBC Quality Indicators Program Registry will continue collecting data to evaluate QCIs and monitor treatment quality, enabling hospitals to benchmark their performance and improve patient care. Additionally, the registry’s real-world data can support research and international collaboration. Trial registration: The study was registered on ClinicalTrials.gov (NCT04167332).

Biomarkers of cardiohepatic syndrome in the cardiac intensive care unit

Abstract Introduction Non-cardiovascular end-organ dysfunction is an important determinant of outcomes in cardiac intensive care unit (CICU) patients. We examined the association between in-hospital mortality with the severity and extent of admission liver function test (LFT) abnormalities in a heterogeneous CICU population. Purpose We hypothesized that a greater severity or extent of LFT abnormalities would be associated with higher in-hospital mortality. Methods We included consecutive unique adult CICU patients with available data for one or more of the admission LFT values of interest. Admission laboratory values were defined as those closest to CICU admission. We used the NIH Common Terminology Criteria for Adverse Events (CTCAE) version 5.0 to categorize the severity of each LFT abnormality from Grade 0 (normal) to Grade 4 (life-threatening). We defined the extent of LFT abnormality as the number of individual LFTs meeting criteria for CTCAE Grade 1 or greater. The primary outcome of interest was all-cause in-hospital mortality. Odds ratio (OR) and 95% confidence interval (CI) values for in-hospital mortality were estimated using logistic regression, before and after multivariable adjustment. Results Of 12,428 unique CICU patients, 5,144 were excluded due to a lack of available data for any admission LFT values. The remaining 7,284 patients comprised the final study population. Among patients with available data for each individual LFT, abnormal values were present in: AST, 2,743/5,733 (47.8%); ALT 1,576/5,512 (28.6%); ALK 611/3,684 (16.6%); and TB 898/5,155 (17.4%). A total of 864 (11.9%) patients died during hospitalization. In-hospital mortality was higher for patients with one or more LFTs meeting criteria for CTCAE Grade 1 or greater (17.9% vs. 6.4%, adjusted OR 1.54 [1.27-1.86], p <0.001). A stepwise increase in in-hospital mortality was observed with increasing CTCAE grade, both overall (adjusted OR 1.25 per each higher CTCAE grade [1.16-1.34], p <0.001); (Figure 1A), and for each individual LFT (Figure 1B). Conclusion Cardiohepatic syndrome is an important predictor of prognosis in CICU patients, as the extent and severity of LFT abnormalities are strongly associated with in-hospital mortality. This is the first large-scale study to examine the association between the magnitude of hepatic biomarker derangement with in-hospital mortality in unselected CICU patients. Patients with markedly elevated LFT values exhibited the highest risk of in-hospital mortality. This analysis advocates for the inclusion of commonly obtained LFTs in future risk-prediction tools for enhanced prognostication.

Identifying heart failure dynamics using multi-point electrocardiograms and deep learning

Abstract Background Heart failure (HF) hospitalizations are related to poor survival outcomes, emphasizing the need for early detection and intervention. Recent studies have reported the potential of deep learning algorithms in identifying heart failure through electrocardiogram (ECG) with notable accuracy. However, evidence supporting their utility in the ongoing monitoring and follow-up of HF patients remains uncertain. Advanced diagnostic tools that can effectively track heart failure progression or improvement over time are needed. Purpose To address this challenge, we developed a deep learning model analyzing 12-lead ECG waveforms at two time points, aiming to detect HF deterioration or improvement, potentially revolutionizing HF management. Methods This study analyzed 32,045 ECGs from 6,892 unique adult patients (aged 18 years and older) visiting the Department of Cardiology at Kobe University Hospital. The participants were randomly assigned to training (4,284 patients), validation (1,034 patients), and test (1,034 patients) datasets. ElecTransformer was developed to classify HF status into deteriorated, improved, and no change classes based on ECG waveform signals at two different time points and to estimate brain natriuretic peptide (BNP) levels. Performance metrics, such as the area under the receiver operating characteristic curve (AUROC) and accuracy, were calculated, and attention mapping via gradient-weighted class activation mapping was utilized to interpret the model's decision-making ability. Results The patients had an average age of 65 years (±15.4 years) and BNP levels averaged 71.5 pg/mL (25.8 to 185.2 pg/mL). For HF status classification, ElecTransformer achieved an AUROC of 0.845 and an accuracy of 0.821, improving to an AUROC of 0.885 and an accuracy of 0.833 with baseline BNP integration. The ElecTransformer accurately estimated BNP levels from ECG data, evidenced by a correlation coefficient of 0.798. Averaged attention mapping showed that the QRS complex was the most important feature for determining model outputs with lesser effects seen for P waves and T waves variance Conclusions Harnessing a Transformer-based deep learning model to analyze multiple ECGs markedly improves the accuracy in detecting changes in heart failure status. This method offers a valuable non-invasive option for monitoring heart failure progression, potentially leading to better patient management and earlier intervention in heart failure treatment.Architecture and Performance of ModelsAveraged attention mapping

Echocardiographic predictors of mortality in cardiac intensive care unit patients with pulmonary hypertension

Abstract Background Pulmonary Hypertension (PH) is a diverse condition associated with elevated mortality. We sought to determine the association of Transthoracic Echocardiography (TTE) parameters with in-hospital mortality in patients with PH admitted to Cardiac Intensive Care Unit (CICU). Methods We included 1 Clinic CICU admissions from 2007 to 2018, with available TTE parameters within the first day of CICU admission who had PH defined as an estimated right ventricular systolic pressure (RVSP) ≥36 mmHg. The primary outcome was in-hospital mortality, evaluated using logistic regression. Results We included 3085 unique CICU patients with PH, predominantly WHO group 2 and 3. Median age was 73.7 (63.8, 82.4) years, and 1343 (43.5%) were females. Heart failure (65.6%) and respiratory failure (34.0%) were the most common admission diagnoses. The median RVSP was 47 (41, 56) mmHg, and 1314 (42.6%) had RVSP ≥50 mmHg. A total of 337 (10.9%) patients died during hospitalization. In-hospital deaths had higher RVSP (51 versus 47 mmHg, p <0.001), reflecting higher right atrial (RA) pressure (14 versus 10 mmHg, p <0.001) (Figure). In-hospital mortality increased with worse TTE parameters reflecting Right Ventricle-Pulmonary Artery (RV-PA) coupling, such as Tricuspid Annular Plane Systolic Excursion/ Right Ventricular Systolic Pressure (RVSP), Tricuspid Annular Systolic Velocity (TASV)/RVSP and TASV/ Tricuspid Regurgitation velocity ratios. Conclusion In CICU patients with PH, we have identified the specific Doppler TTE parameters with strong associations with outcomes. Specifically, RA pressure and parameters of RV-PA coupling had the highest discrimination for in-hospital mortality. Comprehensive Doppler TTE is valuable for identification of high-risk PH patients to facilitate further investigation plans, management, and prognostication. Figure: In-hospital mortality in association with right ventricular (RV) function and right atrial pressure (RAP).In-hospital mortality in association wit

Lipoprotein(a) and atherosclerotic cardiovascular disease in a Brazilian tertiary hospital population

Abstract Background/Introduction Elevated levels of lipoprotein(a) [Lp(a)] are a strong risk marker of atherosclerotic cardiovascular disease (ASCVD), including myocardial infarction (MI), ischemic stroke, and peripheral arterial disease, as well as of aortic valve calcification and stenosis. However, few studies have assessed the prevalence of high Lp(a) levels in patients followed at a tertiary hospital specialized in cardiovascular diseases (CVD). In addition, there is scant information about Lp(a) concentrations in Brazilian individuals. Purpose The aim of the study was to describe the demographic characteristics and comorbidities of patients with at least one measure of Lp(a) at a tertiary treatment center for CVD in our city – Brazil. Methods Patients treated at a heart institute had their data extracted between the years 1999 and 2021. Data were obtained through direct electronic extraction of patients' medical records. The patients were categorized according to different Lp(a) cutoff levels (> 50 and < 70, > 70 and < 90, and > 90 mg/dL) and by quartiles. Results Of 4,970 unique patients included in the study, the mean age at baseline was 59.9 years (SD ± 14.5), 2,634 (53%) were male, and the median Lp(a) was 23.2 mg/dL (interquartile range 8.9 - 56.8 mg/dL). In the whole cohort, 460 (9.3%) had Lp(a) > 50 and < 70, 344 (6.9%) > 70 and < 90, 612 (12.3%) > 90 mg/dl and 138 (2.8%) > 150 mg/dL. Established ASCVD was reported in 1,223 (24%) of the individuals, with a progressively higher prevalence across the Lp(a) categories above (31.2%, 33.8%, 34.2%, respectively) (table 1). Moreover, the Lp(a) median (interquartile range) in mg/dL was significantly higher in those with baseline history of ASCVD (31.4 [10.8-71.2] vs 21.2 [8.3-52]), MI (27.8 [10.6-65.9] vs 22.3 [8.5-54.7]) and stroke (24.9 [10-69.5] vs 23.1 [8.9-56.5]), all p < 0.05 (table 2). Lp(a) was also significantly higher among those with LDL-C ≥ 100 mg/dL relative to LDL-C < 100 mg/dl (25.5[9.5-63.2] vs 21.4[8.4 vs 52.2] p < 0.05). Conclusions In this cohort of patients followed at a tertiary center specialized in CVD, approximately a third had Lp(a) levels above 50 mg/dL, a cutoff robustly associated with ASCVD. This finding likely represents an important source of cardiovascular residual risk, particularly in those with established ASCVD, exposing the need of further therapies.

Aortic valve replacement and mortality with left ventricular hypertrophy in setting of normal LVEF

Abstract Background Without symptoms, indications for aortic valve replacement (AVR) in severe aortic stenosis (AS) differ for individuals with normal vs. low/decreasing left ventricular ejection fraction (LVEF). Conceptually, the development of left ventricular hypertrophy (LVH) could also indicate ventricular injury and potential benefit with earlier AVR. However, little is known about the comparative effectiveness of AVR in individuals with LVH. As such, unlike low or dropping LVEF, development of LVH is not in itself an indication for AVR in clinical guidelines. Methods Initially, 1,232,492 echocardiography reports from 12 hospitals were identified from 1 January 2000 to 22 November 2022, and previously validated natural language processing modules were then applied to identify aortic gradients, LVEF, and presence of LVH. These reports were linked to mortality data, key comorbidities with a 2-year lookback covariate window, and date of any AVR. We then isolated echocardiograms demonstrating normal flow severe AS, LVEF ≥ 55% and LVH. Chart reviews were conducted to reduce missing data and outcomes were censored on the date of the chart review (October 4, 2023). To assess the association between AVR and mortality, Cox proportional hazards models considering treatment status (ie, AVR or not) were used with the index echocardiogram as time zero. Covariates used in risk-adjustment were chosen a priori based on conditions and laboratory values used in cardiac surgery mortality risk-adjustment, since those variables could plausibly confound the treatment decision. To address immortal time bias, AVR was considered a time-varying covariate. Multiple additional approaches with propensity scores were performed as sensitivity analyses. Results After application of inclusion and exclusion criteria, 4856 unique patients remained for analysis. Of those, 2043 (42.1%) received AVR with 841/2043 (41.2%) receiving surgical AVR and 1202/2043 (58.8%) receiving transcatheter AVR. Patients who received AVR had lower unadjusted mortality than those who did not receive AVR (33.6% vs. 47.1%, p < 0.001). After adjustment, and accounting for time-dependence of AVR, AVR was associated with reduced mortality (HR 0.81, 95% CI 0.73-0.89, p < 0.0001). The sensitivity analyses were all consistent with the primary analysis. Conclusions AVR was associated with reduced mortality for patients with severe AS, preserved LVEF, and LVH. This association persisted even after accounting for time-dependence, which tends to reduce the observed effectiveness of procedures in observational data. With over a million echocardiograms over more than 20 years, this is the largest known analysis of AVR outcomes in individuals with severe AS and LVH. Our findings suggest LVH may be a high-risk phenotype of individuals with AS, similar to dropping LVEF. As such, these results raise the potential of more proactive AVR clinical guidelines in individuals with LVH, even when LVEF is normal.Survival in AVR Vs No AVR

Diagnosis of hypertrophic cardiomyopathy by artificial intelligence using standard transthoracic echocardiography

Abstract Introduction Misdiagnosis of hypertrophic cardiomyopathy (HCM) is not uncommon among patients with left ventricular hypertrophy (LVH). Purpose We hypothesized that a convolutional neural network (CNN) could be trained to identify HCM patients among individuals with LVH using standard transthoracic echocardiography (TTE). Methods Echocardiographic reports of 121,767 unique patients linked to clinical data from the largest tertiary medical centre in the country (2007-2022) were reviewed. Patients with no septal hypertrophy were excluded. All HCM cases were retrieved from electronic medical records and were adjudicated by the authors. Data was then divided into 3 mutually exclusive cohorts: training (N=7,368), testing (N=2,456) and validation (2,456) cohorts. Deep learning models were used in order to train a model that identifies HCM patients. The model was based solely on parasternal long axis and apical four chamber views. Results Final study population included 12,281 patients who underwent a total of 27,242 echocardiography studies. There were 1535 (12%) patients with HCM. Mean age of study population was 72±14 and 7,943 (65%) were men. Overall, the model obtained 98.1% classification accuracy with an area under the curve of 0.851 in the independent test data (FIGURE) which could translate to sensitivity and specificity of 68% and 99%, respectively with positive and negative predictive values of 50% and 99%, respectively. Conclusions Applying artificial intelligence to the standard TTE can identify HCM patients with a reasonable accuracy. With further evaluation, this tool can be used to assist in HCM diagnosis.

Healthcare utilization following acute myocardial infarction findings from RECORD MI registry

Abstract Background As therapeutic options and epidemiology have evolved rapidly, the contemporary healthcare utilization following acute myocardial infarction (MI) is not well known. Methods All patients admitted with MI between 01/2015 and 12/2021 in the Baylor Scott & White health system were studied. Patient characteristics and outcomes including hospital readmissions, cardiovascular (CV)-related hospital readmissions, emergency department (ED) encounters, and outpatient visits were studied. Results Of 6804 patients, 6556 were discharged alive. Median age was 69 years, 60% were male, and 77% had non-ST elevation MI (NSTEMI); 1172 (22%) had an LVEF <40%, 493 (7.5%) were revascularized within 3 months of discharge, and 1090 (17%) had multivessel disease. The number of unique patients with first hospital readmission within 30-days, 3-months, and 12-months were 844 (13%), 1372 (21%) and 2306 (35%), respectively, with a higher risk in patients with NSTEMI, prior heart failure (HF), new HF, and LVEF <40%. ED visits at 12 months for any cause were 2401 (37%) and for CV cause were 1321 (20%), with a higher risk in patients with prior HF. Patients with at least 1 primary care visit within 12 months of discharge were captured for 4102 (63%), CV specialty visits for 5009 (76%), and non-CV visits for 3860 patients (59%), with similar distribution across subgroups (Table). Conclusions Post-MI patients had a high risk of readmissions, and ED and outpatient visits, especially amongst those with a prior or new HF diagnosis.

Transitional Care Program in Reducing Acute Hospital Utilization in Singapore

(1) Background: The evidence to support transitional care in reducing acute hospital utilization is variable. Despite changes in the healthcare landscape with a rapidly aging population, there is a lack of local and regional studies to evaluate the effectiveness of transitional home care programs. This study investigates whether a transitional home care program delivered by an acute tertiary hospital can reduce acute hospital utilization. (2) Methods: A pre-post design was used to evaluate the effectiveness of the program. A total of 2004 enrolments from 1679 unique patients that fulfilled the criteria of enrolment were included. The transitional care program is delivered through telephone follow-up and home visits. The Wilcoxon Signed-Rank Test was used to assess the differences between the three periods of baseline, enrolment, and post-enrolment. (3) Results: All 2004 enrolments were analyzed. The re-attendances at the emergency department reduced significantly by 31.2% and 71.9% during enrolment and post-enrolment (p < 0.001), respectively. Similarly, patients had a 38.7% and 76.2% reduction in hospital admissions during enrolment and post-enrolment (p < 0.001), respectively. For patients who were admitted, there was no significant difference in the length of stay between these groups (p = 0.23). (4) Conclusions: The transitional home care program can effectively reduce emergency department re-attendances and inpatient admissions. Not only was the total number of emergency department re-attendances reduced significantly, but the number of frequent re-attendances also dropped significantly. The outcomes were consistent during COVID-19 and post-pandemic phases. These findings can be used as a guide in program planning and future scalability.

Artificial intelligence to detect pulmonary congestion on chest computed tomography in consecutive dyspnoeic patients

Abstract Background Pulmonary congestion is a hallmark of acute heart failure (AHF), and essential for the diagnosis, treatment, and risk stratification. We have previously shown that a low-dose chest CT has four times greater odds than chest radiographs to detect pulmonary congestion. However, implementation is limited by shortage of radiologists. We hypothesized that artificial intelligence (AI) can detect pulmonary congestion on chest CT and aimed to develop and validate an AI-algorithm to detect pulmonary congestion on chest CT. Methods We extracted all chest CT-scans ordered acutely (N= 4840 unique patients) by medical departments at X Hospital from 2016-2021. A random subset of these CT-scans (N= 2187) was used to develop an AI-algorithm to detect pulmonary congestion, using the radiology reports as reference. The remaining subset (N= 1094) was used for internal validation. We externally validated the AI-algorithm in a different cohort (X), comprising of 236 consecutive dyspnoeic patients admitted to the emergency department, all examined with a low-dose, non-contrast chest CT (LDCT). The primary outcome was pulmonary congestion adjudicated by clinical on-call radiologists with access to the medical record, and two independent, blinded research thoracic radiologists. The default AHF thresholds for the optimal cut-off, rule-in threshold and rule-out threshold was established from the internal validation. Results Of 236 patients (median age 74 years, 57% male), 62 (26%), 64 (27%) and 60 (26%) had pulmonary congestion on LDCT evaluated by the on-call radiologists and the two thoracic research radiologists, respectively. The AI-algorithm included ten discriminating features of pulmonary congestion, and the results demonstrate an AUC of 0.95 [95%CI: 0.93-0.98] to predict pulmonary congestion by the on-call radiologists. The results were confirmed by the two expert research radiologists, with an AUC of 0.91 [95%CI: 0.88-0.95] and 0.94 [95%CI: 0.90-0.97], respectively. The sensitivity and specificity of the AI-default threshold was 89% [95%CI: 78-95%] and 89% [95%CI: 83-93%], respectively. The rule-in threshold demonstrated a sensitivity of 97% [95%CI: 86-100%] and the rule-out threshold had a specificity of 96% [95%CI: 92-98%], Figure 1. Conclusion We developed and validated a novel AI-algorithm with promising potential for the rapid automatic detection of pulmonary congestion as the underlying cause of dyspnoea in consecutive acute patients.

Acute kidney injury stage 2/3 during acute heart failure admission - clinical contributors and outcomes

Abstract Introduction Acute Heart failure (AHF) may be complicated by acute kidney injury (AKI) as a result of cardiorenal syndrome.(1) However other factors may contribute to AKI. AKI stage 2 or 3 predicts negative inpatient outcomes in the general population.(2) We have reported data in a contemporary cohort on AKI in AHF. This study looks at a larger cohort in 5 previous years to test whether those findings were consistent. Better understanding of this clinical scenario may lead to improvements in care. Aim To assess the clinical factors associated with AKI 2/3 in AHF inpatients. Methods This study used retrospective data from a single teaching hospital collected for inclusion in the UK National Heart Failure National Audit. All AHF admissions over a 5-year period, from April 2017 – April 2022 inclusive, were analysed. Renal function on admission and discharge/death were assessed. AKI stage was defined according to KDIGO guidelines based on serum creatinine levels. Electronic patient records were reviewed and the clinical course leading up to and following AKI 2/3 was reviewed. Reported mode of death was also collected. Results 3219 AHF admissions (2360 unique patients) were assessed. 257 patients died during admission (8% admission mortality, 10.9% patients). 157 (4.9%) admissions had AKI 2/3 on admission/discharge. 85 (54.1%) AKI 2/3 died during their hospital admission vs 6.1% mortality in admissions without AKI 2/3. Baseline population characteristics are summarised in Table 1. Admissions with AKI 2/3 who died were older (77.6±11.5 vs 71.5 ±15.8) and had higher rates of atrial fibrillation (48.2% vs 39.9%), diabetes (40% vs 36%), ischaemic cardiomyopathy (56.5% vs 41.7%) and LVEF <35% (38.8% vs 27.7%). The most frequent contributors to AKI 2/3 were sepsis (55, 64.7%) and congestive cardiac failure/cardiorenal syndrome (48, 56.6%). Hypotension (25, 29.4%), over-diuresis (17, 20%), radioiodine contrast (10, 11.8%) and aminoglycosides (8, 9.4%) were less common factors in development of AKI 2/3. ≥2 contributors were present in 49, 57.6%. The dominant contributor for AKI 2/3 was adjudicated to be sepsis in 33 (38.8%) of admissions, followed by cardiorenal syndrome (congestive heart failure with AKI) in 29 (34.1%), and cardiogenic shock in 7 (8.2%). AKI 2/3 contributors are summarised in Table 2. In AKI 2/3 cases who died, 35 (41.2%) involved renal specialist input; 60 (70%) involved heart failure specialist review. Conclusion AKI in AHF is associated with high inpatient mortality. Despite this, specialist renal input was uncommon. AKI 2/3 was usually multifactorial. Cardiorenal syndrome accounted for one third of cases and may not be preventable. In keeping with our findings from a smaller contemporary cohort, sepsis was the most common dominant cause of severe AKI. Prevention, early identification, and treatment of infection could reduce this risk. Awareness of the high mortality of AKI in AHF and specialist review may improve outcomes in AHF.

Prospective and External Validation of Machine Learning Models for Short- and Long-Term Mortality in Acutely Admitted Patients Using Blood Tests

Background: Predicting mortality in emergency departments (EDs) using machine learning models presents challenges, particularly in balancing simplicity with performance. This study aims to develop models that are both simple and effective for predicting short- and long-term mortality in ED patients. Our approach uses a minimal set of variables derived from one single blood sample obtained at admission. Methods: Data from three cohorts at two large Danish university hospitals were analyzed, including one retrospective and two prospective cohorts where prognostic models were applied to predict individual mortality risk, spanning the years 2013–2022. Routine biochemistry analyzed in blood samples collected at admission was the primary data source for the prediction models. The outcomes were mortality at 10, 30, 90, and 365 days after admission to the ED. The models were developed using Light Gradient Boosting Machines. The evaluation of mortality predictions involved metrics such as Area Under the Receiver Operating Characteristic Curve (AUC), sensitivity, specificity, negative predictive values, positive predictive values, and Matthews correlation coefficient (MCC). Results: A total of 43,648 unique patients with 65,484 admissions were analyzed. The models showed high accuracy, with very good to excellent AUC values between 0.87 and 0.93 across different time intervals. Conclusions: This study demonstrates that a single assessment of routine clinical biochemistry upon admission can serve as a powerful predictor for both short-term and long-term mortality in ED admissions.

Clinical utility of EBV DNA testing of blood in immunocompetent and immunocompromised patients: A single-center experience.

Epstein-Barr virus (EBV) causes different clinical presentations in immunocompetent and immunocompromised persons, and thus indications for testing vary between these populations. We reviewed our institution's EBV DNA testing across these populations to understand its clinical utility and appropriateness. We conducted a retrospective chart review of adult patients with positive EBV nucleic acid amplification (NAAT) testing from November 2022 to 2023. We recorded demographics, indications for testing, EBV-related diagnosis, laboratory results, and whether testing influenced patient treatment. Of 3560 EBV NAAT tests, 187 (5.3%) were positive, representing 124 unique adult patients (51 immunocompetent and 73 immunocompromised). Reactivation of EBV was the most common diagnosis in both populations, followed by acute EBV infection in the immunocompetent and non-posttransplant lymphoproliferative disorder malignancies in the immunocompromised. In immunocompromised patients, positive tests led to treatment changes more frequently than in immunocompetent patients (27.4% vs 11.7%, P=.06). Testing affected clinical management in 0% to 2% of cases when sent for sepsis/shock and nonspecific viral syndromes compared to 37% to 49% of cases when sent for posttransplant and malignancy screening/monitoring. EBV NAAT testinginfrequently influenced clinical management in immunocompetent individuals but had a notable impact in immunocompromised patients, particularly those undergoing posttransplant or malignancy screening. This underscores the need for targeted testing to optimize resource utilization and cost-effectiveness.

Using Technology to Support Measurement-Based Care: Design and Implementation of the Measurement-Assisted Care Program.

The authors aimed to increase measurement-based care in an outpatient academic psychiatry service. The Measurement-Assisted Care program (MAC) was implemented as a clinical quality-improvement effort that included 4,665 unique patients and 176 clinicians over 12 months. The Patient Health Questionnaire-9 (PHQ-9), Generalized Anxiety Disorder scale (GAD-7), PTSD Checklist for DSM-5, and Brief Inventory of Thriving were automatically included in the patient electronic check-in process for new visits; the PHQ-9 and GAD-7 were included for return visits. Patient responses were automatically routed to clinicians' electronic health record inboxes and documentation templates. MAC was rolled out in two phases, and clinicians were surveyed about their use of MAC data 3-6 months after the start of the program. After implementation of MAC, PHQ-9 completion rates increased from 5% to 66% of visits for phase 1 and from 5% to 60% for phase 2. Post-MAC completion rates were higher for telehealth (70%) than for in-person (40%) visits. More than 90% of clinicians reported that MAC was useful with at least one of their three most recent patients, and 51% reported that it was useful with all three of their three most recent patients. Clinician adoption was high, with 78% reporting that they use MAC data to guide care. MAC increased the completion rate of patient-reported outcome measures, especially for telehealth visits. Clinicians reported that they frequently use and discuss MAC data with patients, implying that technology-supported workflows can help systems with high telehealth utilization capture and use patient-reported outcomes.

Poor correlation of venous lactate with systemic oxygen saturation in the paediatric cardiac ICU: a pilot study.

Cardiac intensive care providers require a comprehensive understanding of cardiac output and oxygen delivery. The estimation of cardiac output in clinical practice often relies on thermodilution and the Fick principle. Central venous saturation and lactate levels are commonly used indicators for cardiac output assessment. However, the relationship between venous lactate levels and venous oxygen saturation in paediatric cardiac intensive care patients remains unclear. This is a single-centre retrospective pilot study aimed to investigate the correlation between venous lactate and venous oxygen saturation in paediatric patients. Data collected included venous saturation, heart rate, mean arterial blood pressure, arterial saturation by pulse oximetry, cerebral and renal near-infra-red spectroscopy values, and the presence of a functionally univentricular heart. Statistical analyses included Bayesian Pearson correlation and regression analyses. A total of 203 data points from 37 unique patients were included in the analysis. There was no significant correlation between serum lactate and venous saturation (correlation coefficient = -0.01; Bayes factor 10 = 0.06). Serum lactate also did not correlate with other haemodynamic metrics. Venous saturation showed correlations with arterial saturation and cerebral and renal near-infra-red spectroscopy. Regression analysis revealed that parallel circulation, arterial saturation, and cerebral near-infra-red spectroscopy were predictive of venous saturation. The following equation resulted from the regression analysis: 68.0 - (12.7 x parallel circulation) - (0.8 x arterial saturation) + (0.3 x cerebral near-infra-red spectroscopy). This model had a Bayes factor 10 of 0.03 and adjusted R-squared was 0.29. In paediatric cardiac intensive care patients, there is no significant correlation between venous lactate and venous saturation, suggesting that lactate may not be a reliable marker for assessing the adequacy of oxygen delivery in this population. Only a weak correlation could be identified once the venous saturation was 70% or lower. Additional research is needed to explore alternative markers for monitoring oxygen delivery in critically ill paediatric patients.

Matched oligoclonal bands: Diagnostic utility and clinical characteristics.

To describe patient clinical characteristics associated with matched oligoclonal bands (OCB). A retrospective review at the University of Utah examined patients with matched OCB from 2015 to 2020. Clinical data, diagnosis, and outcomes were collected. Patients were classified with either multiple sclerosis (MS), other inflammatory neurologic disorder (other-IND), or noninflammatory neurologic disorder (NIND). Of 539 identified patients, 436 (53.4% female) were matched-only, while 103 (43.7% female) were matched + unique. Patients with matched-only bands were older (57.4 ± 16 vs. 52 ± 14.2, p < 0.001) and more likely to have a history of autoimmune disease (40.1% vs. 28.2%, p = 0.024) and/or cancer (28.7% vs. 16.5%, p = 0.012). Patients with matched + unique bands were more likely to have CSF pleocytosis (52.4% vs. 25.9%, p < 0.001), high IgG index (52.2% vs. 7.6%, p < 0.001), and an abnormal MRI (86.9% vs. 63.1%, p < 0.001). More than two-thirds of matched-only patients had NIND, while 33% and 41.7% of matched + unique patients had MS and other-IND, respectively. Patients exhibiting matched-only bands and a high IgG index demonstrated a significantly higher incidence of other-IND compared to those with matched-only bands and a normal IgG index (55.6% vs. 30.4%, p = 0.013). While Kaplan-Meier survival curves demonstrated higher mortality in the matched-only cohort compared to the matched + unique cohort (p = 0.02), multivariable Cox regression analysis showed this difference was not statistically significant when adjusting for various factors. A history of cancer was the significant predictor of increased mortality risk (Hazard ratio = 3.147, 95% CI [2.196, 4.51]). Patients with matched only versus matched + unique OCB have distinct clinical profiles.

Onychomycosis in the US Pediatric Population-AnEmphasis on Fusarium Onychomycosis.

Onychomycosis is a common nail disease that is often difficult to treat with a high risk of recurrence. To update our current understanding of the etiologic profile in pediatric patients with onychomycosis utilizing molecular diagnosis by polymerase chain reaction (PCR) combined with histopathologic examination. Records of 19,770 unique pediatric patients were retrieved from a single diagnostic laboratory in the United States spanning over a 9-year period (March 2015 to April 2024). This cohort represents patients clinically suspected of onychomycosis seen by dermatologists and podiatrists. Dermatophytes, nondermatophyte molds (NDMs), and yeasts were identified by multiplex real-time PCR corroborated by the demonstration of fungal invasion on histopathology. An average of 37.0% of all patients sampled were mycology-confirmed to have onychomycosis. Most patients were between ages 11 and 16 years, and the rate of mycologically confirmed onychomycosis was significantly higher among the 6- to 8-year (47.2%) and 9- to 11-year (42.7%) age groups compared to the 0- to 5-year (33.1%), 12- to 14-year (33.2%), and 15- to 17-year (36.7%) age groups. The majority of infections were caused dermatophytes (74.7%) followed by NDMs (17.4%). The Trichophyton rubrum complex represents the dominant pathogen with higher detection rates in the 6- to 11-year-olds. Fusarium was the most commonly isolated NDM with an increasing prevalence with age. Elementary school-aged children have a higher risk of contracting onychomycosis which may be attributed to the onset of hyperhidrosis at puberty, use of occlusive footwear, nail unit trauma, and walking barefoot. Fusarium onychomycosis may be more prevalent than expected, and this may merit consideration of management strategies.

Improving the Diagnostic Accuracy of RECAM in North American Patients with Suspected Idiosyncratic Drug-Induced Liver Injury: Improving RECAM accuracy.

The RECAM is an updated, electronic version of the RUCAM to diagnose DILI. The primary aim of this study was to compare RECAM versus RUCAM in patients with suspected DILI. Patient encounters from 10/1/2015 to 9/30/2019 were searched for suspected DILI using ICD-10 K71 codes for toxic liver disease. DILIN expert opinion scores were assigned to each case (1/2/3= probable DILI, 4/5= non-DILI). RECAM and RUCAM scores were compared with DILIN expert opinion scores. Among 766,930 encounters searched, 120 unique patients met inclusion criteria with 72 (60%) adjudicated as probable-DILI. The most frequent suspect drugs were antimicrobials (38.3%), antineoplastics (8.3%), and antirheumatic drugs (8.3%). Mean age was 49.2 + 15.6 years and 50% were female with 45.8% having hepatocellular injury. RUCAM had better agreement with DILIN expert opinion for probable-DILI versus RECAM (66.7% vs. 44.4%, p=0.018). Both had 100% agreement with DILIN expert opinion for non-DILI. Frequently missing laboratory data included HCV RNA (64.3%) and anti-HEV IgM testing (70%), leading to loss of up to 6 points in RECAM scoring but not impacting RUCAM scores. A modified RECAM that made HCV RNA and anti-HEV IgM optional had better agreement with DILIN expert opinion compared to RUCAM (79.2% vs. 66.7%, p=0.09). Among 120 suspected DILI cases, RUCAM had better agreement with DILIN expert opinion scores versus RECAM. Making HCV RNA and anti-HEV IgM testing optional significantly improved agreement between RECAM and DILIN expert opinion. Future modifications to RECAM are needed to improve causality assessment in North American patients with suspected DILI.

Pregnancy Outcomes from a Multidisciplinary Obstetric-Medicine/Rheumatology Clinic in the United States: A Five-Year Retrospective Analysis.

At Women & Infants Hospital in Providence, Rhode Island, the Specialty Care in Pregnancy clinic combines obstetric-medicine internists with rheumatologists to care for pregnant patients with rheumatologic conditions. These clinics are scarce, with only three known similar clinics in the United States. This study aims to characterize the population cared for in this clinic, identify interventions, and analyze pregnancy outcomes for the birthing parents and newborns. A five-year retrospective chart review was performed from January 1st, 2016, through December 31st, 2021. Of 81 patients, 62% had a clinically diagnosed rheumatic disorder. Of 87 patient visits, which included preconception, prenatal, and postpartum encounters, 54% of patients were taking conventional synthetic disease modifying antirheumatic drugs, and 17% were taking biologic disease modifying antirheumatic drugs. New medications were started in 52% of patients. A total of 52% of pregnancies resulted in live births, with 2% resulting in miscarriages. Prematurity occurred in 19% of newborns, and 9% had intrauterine growth restriction. Our study illustrates the benefits of multidisciplinary care in patients with rheumatologic disorders during their prenatal and perinatal periods. The expertise from both the obstetric-medicine internists and rheumatologists was critical in making complex decisions that weighed the benefits of therapy against potential risks for the fetus. Our multidisciplinary approach resulted in doubling of the number of patients initiating disease modifying therapy and increased prophylaxis with hydroxychloroquine and/or aspirin therapy, as recommended by current guidelines. Additional multidisciplinary clinics of this type would help coordinate care among physicians who frequently treat these high-risk, unique patients and open the door for more research of this understudied population.