Development and Validation of a Machine Learning Model for Early Detection of Untreated Infection.

BACKGROUND: Early diagnostic uncertainty for infection causes delays in antibiotic administration in infected patients and unnecessary antibiotic administration in noninfected patients. OBJECTIVE: To develop a machine learning model for the early detection of untreated infection (eDENTIFI), with the presence of infection determined by clinician chart review. DERIVATION COHORT: Three thousand three hundred fifty-seven adult patients hospitalized between 2006 and 2018 at two health systems in Illinois, United States. VALIDATION COHORT: We validated in 1632 patients in a third Illinois health system using area under the receiver operating characteristic curve (AUC). PREDICTION MODEL: Using a longitudinal discrete-time format, we trained a gradient boosted machine model to predict untreated infection in the next 6 hours using routinely available patient demographics, vital signs, and laboratory results. RESULTS: eDENTIFI had an AUC of 0.80 (95% CI, 0.79-0.81) in the validation cohort and outperformed the systemic inflammatory response syndrome criteria with an AUC of 0.64 (95% CI, 0.64-0.65; p < 0.001). The most important features were body mass index, age, temperature, and heart rate. Using a threshold with a 47.6% sensitivity, eDENTIFI detected infection a median 2.0 hours (interquartile range, 0.9-5.2 hr) before antimicrobial administration, with a negative predictive value of 93.6%. Antibiotic administration guided by eDENTIFI could have decreased unnecessary IV antibiotic administration in noninfected patients by 10.8% absolute or 46.4% relative percentage points compared with clinicians. CONCLUSION: eDENTIFI could both decrease the time to antimicrobial administration in infected patients and unnecessary antibiotic administration in noninfected patients. Further prospective validation is needed.

Causes, Diagnostic Testing, and Treatments Related to Clinical Deterioration Events Among High-Risk Ward Patients.

IMPORTANCE: Timely intervention for clinically deteriorating ward patients requires that care teams accurately diagnose and treat their underlying medical conditions. However, the most common diagnoses leading to deterioration and the relevant therapies provided are poorly characterized. OBJECTIVES: We aimed to determine the diagnoses responsible for clinical deterioration, the relevant diagnostic tests ordered, and the treatments administered among high-risk ward patients using manual chart review. DESIGN, SETTING, AND PARTICIPANTS: This was a multicenter retrospective observational study in inpatient medical-surgical wards at four health systems from 2006 to 2020. Randomly selected patients (1000 from each health system) with clinical deterioration, defined by reaching the 95th percentile of a validated early warning score, electronic Cardiac Arrest Risk Triage, were included. MAIN OUTCOMES AND MEASURES: Clinical deterioration was confirmed by a trained reviewer or marked as a false alarm if no deterioration occurred for each patient. For true deterioration events, the condition causing deterioration, relevant diagnostic tests ordered, and treatments provided were collected. RESULTS: Of the 4000 included patients, 2484 (62%) had clinical deterioration confirmed by chart review. Sepsis was the most common cause of deterioration (41%; n = 1021), followed by arrhythmia (19%; n = 473), while liver failure had the highest in-hospital mortality (41%). The most common diagnostic tests ordered were complete blood counts (47% of events), followed by chest radiographs (42%) and cultures (40%), while the most common medication orders were antimicrobials (46%), followed by fluid boluses (34%) and antiarrhythmics (19%). CONCLUSIONS AND RELEVANCE: We found that sepsis was the most common cause of deterioration, while liver failure had the highest mortality. Complete blood counts and chest radiographs were the most common diagnostic tests ordered, and antimicrobials and fluid boluses were the most common medication interventions. These results provide important insights for clinical decision-making at the bedside, training of rapid response teams, and the development of institutional treatment pathways for clinical deterioration.

Rheumatoid Arthritis-Associated Interstitial Lung Disease (RA-ILD): Update on Prevalence, Risk Factors, Pathogenesis, and Therapy.

PURPOSE OF REVIEW: Rheumatoid arthritis is frequently complicated by interstitial lung disease (RA-ILD), an underappreciated contributor to excess morbidity and mortality. The true prevalence of RA-ILD is difficult to define given the variability in diagnostic criteria used. The lack of standardized screening methods, an incomplete understanding of disease pathogenesis, and dearth of validated biomarkers have limited the development of controlled clinical trials for this disease. RECENT FINDINGS: Numerous studies have focused on clinical, radiographic, genetic, molecular, and/or serologic markers of disease severity as well as risk of disease progression. In addition to defining valuable clinical biomarkers, these studies have provided insights regarding the pathogenesis of RA-ILD and potential therapeutic targets. Additional studies involving immunomodulatory and/or anti-fibrotic agents have assessed new therapeutic options for different stages of RA-ILD. RA-ILD continues to be a major contributor to the increased morbidity and mortality associated with RA. Advancements in our understanding of disease pathogenesis at a molecular level are necessary to drive the development of more targeted therapy.

State Mandates To Cap Out-Of-Pocket Insulin Costs Are No Longer Necessary.

During the past five years, many states have imposed out-of-pocket spending caps on insulin. In most cases, these reforms have had limited impact, in part because of the limits of state authority. Meanwhile, changes at the federal level and actions by manufacturers and commercial plans have made some of the caps nonbinding. It is not surprising that efforts to measure the impact of these caps yield conflicting results.

Comparative B cell epitope profiling in Japanese and North American cohorts of MDA5+ dermatomyositis reveals a direct association between immune repertoire and pulmonary mortality.

OBJECTIVES: Anti-melanoma differentiation-associated gene 5 antibody-positive (MDA5+) dermatomyositis patients exhibit clinical features that vary by geographical and ethnic/genetic distribution. We therefore investigated whether B cell epitope profiles and corresponding clinical features distinguished two independent cohorts of MDA5+ dermatomyositis. METHODS: We used ELISA-based methods to determine the relationship between antibody recognition of overlapping 155 amino acid MDA5 subfragments and clinical features of 17 MDA5+ dermatomyositis patients from Japan. Associations between clinical features and standardized anti-MDA5 subfragment antibody titers were assessed via Brunner Munzel testing and compared with clinical/serological profiles of an independent North American cohort. ROC analyses and Kaplan-Meier curves were used to further assess the relationship between anti-MDA5 fragment antibody levels and specific clinical features/outcomes. RESULTS: Clinical characterization of a Japanese cohort of 17 MDA5+ dermatomyositis patients revealed a high prevalence of arthritis (47%) and interstitial lung disease (ILD) (100%). Serological profiling demonstrated predominant antibody recognition of MDA5 fragments A (aa 1-155), B (aa 130-284), and E (aa 517-671) in a pattern that was distinct from North American MDA5+ patients (n = 24) whose sera preferentially recognized fragment H (aa 905-1026). Statistical analysis revealed a striking association between anti-fragment A antibody levels and rapidly progressive ILD (RP-ILD) among Japanese patients (p< 0.01). ROC and Kaplan Meier curves also demonstrated a strong relationship between anti-fragment A antibody levels, RP-ILD, and pulmonary death in combined cohort analyses. CONCLUSIONS: Japanese and North American MDA5+ dermatomyositis patients manifest markedly different B cell epitope profiles that are associated with higher prevalence of RP-ILD and worse clinical outcome among Japanese patients.

Clinical Outcomes of a Novel Hybrid Transtibial Technique for Femoral Tunnel Drilling in Anterior Cruciate Ligament Reconstruction: A Large Single-Center Case Series With a Minimum 2-Year Follow-up.

Background: A novel hybrid transtibial (HTT) approach to femoral tunnel drilling in anterior cruciate ligament reconstruction (ACLR) has been developed that circumvents the need for knee hyperflexion and orients the graft in the most anatomic position without sacrificing the tunnel length or aperture. Hypothesis: Patients who underwent ACLR utilizing the HTT technique would achieve excellent patient-reported outcome scores and experience low rates of graft failure and reoperations. Study Design: Case series; Level of evidence, 4. Methods: Patients who underwent primary ACLR at a single institution between 2005 and 2020 were retrospectively reviewed. Patients treated with the HTT, anteromedial portal (AMP), and transtibial (TT) approaches were matched based on age, sex, and body mass index ±3 kg/m2. Demographic and surgical data as well as femoral tunnel angle measurements on anteroposterior and lateral radiographs were collected for the 3 groups. However, clinical outcomes were only reported for the HTT group because of concerns of graft heterogeneity. Results: A total of 170 patients (median age, 26.5 years [interquartile range (IQR), 18.0-35.0 years]) who underwent ACLR using the HTT approach were included. The median coronal- and sagittal-plane femoral tunnel angles were 47° (IQR, 42°-53°) and 40° (IQR, 34°-46°), respectively. The sagittal-plane femoral tunnel angles in the HTT group were significantly more horizontal compared with those in the TT group (P < .0001), whereas the coronal-plane femoral tunnel angles in the HTT group were found to be significantly more vertical compared with those in the AMP group (P = .001) and more horizontal compared with those in the TT group (P < .0001). The graft failure and reoperation rates in the HTT group at a minimum 2-year follow-up were 1.8% (3/170) and 4.7% (8/170), respectively. The complication rate was 6.5% (11/170), with the most common complication being subjective stiffness in 7 patients. The median Lysholm score was 89.5 (IQR, 79.0-98.0); the median International Knee Documentation Committee score was 83.9 (IQR, 65.5-90.8); and the median Veterans RAND 12-Item Health Survey physical and mental component summary scores were 55.0 (IQR, 52.6-55.9) and 56.2 (IQR, 49.1-59.3), respectively. Conclusion: ACLR using the HTT technique was associated with low graft retear and revision surgery rates and good patient-reported outcome scores at medium-term follow-up and demonstrated femoral tunnel obliquity on postoperative radiographs that correlated with optimal parameters previously reported in cadaveric and biomechanical studies.

Patient Preferences for Graft Selection in Anterior Cruciate Ligament Reconstruction.

Background: Selecting an appropriate graft for anterior cruciate ligament (ACL) reconstruction requires consideration of a patient's preferences, goals, age, and physical demands alongside the risks and benefits of each graft choice. Purpose: To determine the most popular ACL reconstruction grafts among patients and the most important factors influencing their decisions. Study Design: Cross-sectional study; Level of evidence, 3. Methods: Patients undergoing ACL reconstruction between October 2022 and April 2023 completed a survey either before (nonconsult group) or after (consult group) speaking with their surgeon, who provided an evidence-based description of the pros and cons of an allograft and the following autografts: bone-patellar tendon-bone (BPTB), hamstring tendon (HT), and quadriceps tendon (QT). Patient characteristics, graft choice, information influencing their graft choice, and surgeon recommendation were collected and compared between the groups. Results: Among the 100 included patients, 59.0% were male, and the mean age was 28.3 ± 10.4 years. The most popular grafts were the BPTB (56.0%), followed by the QT (29.0%), HT (8.0%), and allograft (7.0%). No significant difference was observed in the graft selection between the consult group (n = 60; BPTB, 46.7%; QT, 38.3%; HT, 8.3%; allograft, 6.7%) and nonconsult group (n = 40; BPTB, 70.0%; QT, 15.0%; HT, 7.5%; allograft, 7.5%) (P = .0757). In the consult group, 81.7% of patients selected the graft recommended to them by their surgeon. The top 2 graft selection reasons were usage in professional athletes and failure rates, while the top 2 ACL surgery concerns were returning to their desired level of athletics and graft failure risk. Among the 93 patients who researched their ACL graft options before their visit, the most popular information source was some form of media (72.0% [67/93]). Conclusion: The study findings underscore the importance of patient preference and surgeon recommendation in a patient's graft selection and highlight the need to be cognizant of the information sources available to patients when researching their graft options.

Development and Validation of a Machine Learning COVID-19 Veteran (COVet) Deterioration Risk Score.

BACKGROUND AND OBJECTIVE: To develop the COVid Veteran (COVet) score for clinical deterioration in Veterans hospitalized with COVID-19 and further validate this model in both Veteran and non-Veteran samples. No such score has been derived and validated while incorporating a Veteran sample. DERIVATION COHORT: Adults (age ≥ 18 yr) hospitalized outside the ICU with a diagnosis of COVID-19 for model development to the Veterans Health Administration (VHA) (n = 80 hospitals). VALIDATION COHORT: External validation occurred in a VHA cohort of 34 hospitals, as well as six non-Veteran health systems for further external validation (n = 21 hospitals) between 2020 and 2023. PREDICTION MODEL: eXtreme Gradient Boosting machine learning methods were used, and performance was assessed using the area under the receiver operating characteristic curve and compared with the National Early Warning Score (NEWS). The primary outcome was transfer to the ICU or death within 24 hours of each new variable observation. Model predictor variables included demographics, vital signs, structured flowsheet data, and laboratory values. RESULTS: A total of 96,908 admissions occurred during the study period, of which 59,897 were in the Veteran sample and 37,011 were in the non-Veteran sample. During external validation in the Veteran sample, the model demonstrated excellent discrimination, with an area under the receiver operating characteristic curve of 0.88. This was significantly higher than NEWS (0.79; p < 0.01). In the non-Veteran sample, the model also demonstrated excellent discrimination (0.86 vs. 0.79 for NEWS; p < 0.01). The top three variables of importance were eosinophil percentage, mean oxygen saturation in the prior 24-hour period, and worst mental status in the prior 24-hour period. CONCLUSIONS: We used machine learning methods to develop and validate a highly accurate early warning score in both Veterans and non-Veterans hospitalized with COVID-19. The model could lead to earlier identification and therapy, which may improve outcomes.

Relationship between Jo-1 B Cell Epitope Profile and Clinical Features of Anti-Synthetase Syndrome.

OBJECTIVE: Anti-histidyl-transfer RNA synthetase (Jo-1) antibodies are associated with myositis as well as different extramuscular organ complications comprising the anti-synthetase syndrome. This study aimed to clarify the relationship between anti-Jo-1 epitope recognition patterns and specific clinical features of this syndrome. METHODS: B cell epitope mapping was performed via enzyme-linked immunosorbent assay in 180 patients who were anti-Jo-1 antibody-positive using overlapping peptides/protein fragments spanning the amino-terminal 151 amino acids of Jo-1 as substrate antigens. Statistical associations with clinical features were assessed through rank-sum, correlation, and cluster analyses. RESULTS: The level of reactivity against subfragments spanning amino acids 1-151 of Jo-1 paralleled that of full-length Jo-1, confirming the immunodominance of this amino-terminal region. The corresponding frequencies of reactivity to peptides 1 (amino acids [aa] 1-21), 3 (aa 27-47), 4 (aa 40-60), 10 (aa 118-138), and 11 (aa 131-151) were 6.1%, 42.5%, 6.8%, 6.7%, and 20.3%. While anti-full-length Jo-1 antibodies were significantly associated with Raynaud phenomenon, anti-fragment A2 (aa 1-60) and A3 (aa 1-90) antibodies were associated with proximal muscle weakness, Raynaud phenomenon, arthritis, and sicca syndrome. Anti-fragment A4 (aa 1-120) and A5 (aa 1-151) antibodies were also associated with sicca syndrome. Peptide 1 (aa 1-21) antibodies were associated with Raynaud phenomenon and dysphagia. Whereas anti-peptide 3 (aa 27-47) antibodies were also linked to Raynaud phenomenon, anti-peptide 9 (aa 105-125) antibodies were associated with mechanic's hands. CONCLUSION: Autoantibodies targeting different amino-terminal subfragments and/or peptides of Jo-1 were associated with specific clinical features of the anti-synthetase syndrome, demonstrating the biomarker potential of B cell epitope profiling in this disorder.

Clinical features associated with the presence of anti-Ro52 and anti-Ro60 antibodies in Jo-1 antibody-positive anti-synthetase syndrome.

Introduction: Anti-SSA antibodies target two unrelated proteins, Ro52 (E3 ligase) and Ro60 (RNA binding protein). Previous studies indicate that anti-Ro52 antibodies are frequently associated with various myositis-specific autoantibodies (MSAs)-including anti-tRNA synthetase antibodies-and that the coexistence of MSAs and anti-Ro52 antibodies may portend worse clinical outcomes. Although not well-described in the setting of myositis, work from our animal model of HRS (histidyl-tRNA synthetase)-induced myositis suggests that anti-Ro60 antibodies may also be linked to specific MSAs such as anti-HRS/Jo-1. We therefore aimed to demonstrate the prevalence and clinical characteristics of Ro52 and Ro60 antibody positivity in patients possessing Jo-1 antibodies. Methods: To establish the immunological link between anti-synthetase, anti-Ro52, and anti-Ro60 antibodies, we evaluated the relative titers of these antibodies in blood and bronchoalveolar lavage fluid (BALF) of mice following immunization with HRS/Jo-1. In parallel, we used ELISA-based approaches to assess sera from 177 anti-Jo1 antibody-positive patients for the presence of anti-Ro52 and/or anti-Ro60 antibodies. We then determined statistical associations between co-existing anti-Jo-1, anti-Ro52, and/or anti-Ro60 antibodies and clinical manifestations associated with the anti-synthetase syndrome. Results: Mice immunized with HRS had higher levels of anti-Ro52 and anti-Ro60 antibodies in serum and BALF than PBS-immunized mice. In 177 anti-Jo-1 antibody-positive patients, the prevalence of anti-Ro52 and anti-Ro60 antibodies was 36% and 15%, respectively. The frequency of dry eye/dry mouth, interstitial pneumonia, and pulmonary events over time differed between patients with various combinations of anti-Ro52 and anti-Ro60 antibodies. While anti-Ro52 antibodies generally correlated with statistically significant increases in each of these clinical manifestations, the presence of Ro60 antibodies alone was associated with decreased frequency of ILD. Discussion: Anti-Ro52 and/or anti-Ro60 antibodies are often co-expressed with anti-Jo1 antibodies, defining clinical subsets with different disease course/outcomes.

Antibodies to Malondialdehyde-Acetaldehyde Adduct Are Associated With Prevalent and Incident Rheumatoid Arthritis-Associated Interstitial Lung Disease in US Veterans.

OBJECTIVE: The objective of this study is to determine the associations of protein-specific anti-malondialdehyde-acetaldehyde (MAA) antibodies with prevalent and incident rheumatoid arthritis-interstitial lung disease (RA-ILD). METHODS: Within a multicenter, prospective cohort of US veterans with RA, RA-ILD was validated by medical record review of clinical diagnoses, chest imaging, and pathology. Serum antibodies to MAA-albumin, MAA-collagen, MAA-fibrinogen, and MAA-vimentin (IgA, IgM, and IgG) were measured by a standardized enzyme-linked immunosorbent assay. Associations of anti-MAA antibodies with prevalent and incident RA-ILD were assessed using multivariable regression models adjusting for established RA-ILD risk factors. RESULTS: Among 2,739 participants with RA (88% male, mean age of 64 years), there were 114 with prevalent and 136 with incident RA-ILD (average time to diagnosis: 6.6 years). Higher IgM anti-MAA-collagen (per 1 SD: adjusted odds ratio [aOR] 1.28, 95% confidence interval [CI] 1.02-1.61), IgA anti-MAA-fibrinogen (aOR 1.48, 95% CI 1.14-1.92), and IgA (aOR 1.78, 95% CI 1.34-2.37) and IgG (aOR 1.48, 95% CI 1.14-1.92) anti-MAA-vimentin antibodies were associated with prevalent RA-ILD. In incident analyses, higher IgA (per one SD: adjusted hazards ratio [aHR] 1.40, 95% CI 1.11-1.76) and IgM (aHR 1.29, 95% CI 1.04-1.60) anti-MAA-albumin antibody concentrations were associated with increased ILD risk. Participants with IgA (aHR 2.13, 95% CI 1.16-3.90) or IgM (aHR 1.98, 95% CI 1.08-3.64) anti-MAA-albumin antibody concentrations in the highest quartile had an approximately two-fold increased risk of incident RA-ILD. Across all isotypes, anti-MAA-fibrinogen, anti-MAA-collagen, and anti-MAA-vimentin antibodies were not significantly associated with incident RA-ILD. CONCLUSION: Protein-specific anti-MAA antibodies to collagen, fibrinogen, and vimentin were associated with prevalent RA-ILD. IgA and IgM anti-MAA-albumin antibodies were associated with a higher risk of incident RA-ILD. These findings suggest that MAA modifications and resultant immune responses may contribute to RA-ILD pathogenesis.

Perioperative non-opioid analgesia strategies after high tibial osteotomy: a systematic review of prospective studies.

PURPOSE: Little is known about the optimal analgesia regimen after HTO. Thus, this study systematically reviewed the literature on clinical and patient-reported outcomes of pain management strategies for patients after HTO. METHODS: A comprehensive search of the PubMed, Cochrane CENTRAL, and CINAHL databases was conducted from inception through September 2023. Studies were included if they evaluated pain reduction with analgesia strategies after HTO and were excluded if they did not report pain control outcomes. RESULTS: Five studies with 217 patients were included. Patients with a multimodal intraoperative injection cocktail to the knee, femoral nerve block (FNB), or adductor canal block (ACB) for HTO had significant improvement in visual analog scale (VAS) and numerical rating scale (NRS) scores in the first 12 h postoperatively compared to controls. Patients on duloxetine had significantly lower NRS scores at 1, 7, and 14 days postoperatively and significantly lower nonsteroidal anti-inflammatory drug (NSAID) usage throughout the two-week postoperative period than the control group. Patients receiving an ACB had significantly lower opioid consumption than controls at 12 h postoperative. In patients with an FNB or ACB, no significant difference in quadriceps strength or time to straight leg raise postoperatively was observed compared to controls. CONCLUSION: A multimodal periarticular injection cocktail, FNB, or an ACB effectively reduces pain on the first day after HTO, with an ACB able to reduce opioid consumption on the first postoperative day. Duloxetine combined with an ACB effectively decreases pain for two weeks postoperatively while reducing NSAID consumption in patients after HTO. LEVEL OF EVIDENCE: IV.

Pain medication beliefs in individuals with headache.

OBJECTIVE: To evaluate pain medication beliefs in a community sample of individuals with headache. BACKGROUND: Previous studies of medication adherence for individuals with headache have identified a high rate of prescription nonfulfillment, frequent medication discontinuation, and widely varying levels of medication-related satisfaction. Still, there is a limited understanding of how these individuals view their medications and their relationships with health-care providers. Insight into these perceptions could prove useful in explaining medication adherence behaviors. METHODS: In this secondary analysis of a cross-sectional study, data from N = 215 adults with headache were analyzed. Participants completed the Pain Medication Attitudes Questionnaire (PMAQ), Center for Epidemiologic Studies Depression Scale (CES-D), State-Trait Anxiety Inventory Form Y-2, Weekly Stress Inventory Short Form, and Migraine Disability Scale. These participants also provided a list of their current pain medications. RESULTS: Using the PMAQ, participants could be characterized as having medication beliefs that were "trusting and unconcerned" (n = 83/215 [38.6%]), "skeptical and somewhat worried" (n = 99/215 [46.0%]), or "skeptical and concerned" (n = 33/215 [15.3%]). Individuals with skeptical and concerned beliefs expressed elevated concerns (z > 1.15) about side effects, scrutiny, perceived need, tolerance, withdrawal, and addiction. Individuals who were trusting and unconcerned expressed low levels (z < -0.40) of these beliefs. Increasing levels of mistrust and medication concerns were correlated with higher depression scores on the CES-D, with values ranging from r = 0.23 to r = 0.38. CONCLUSIONS: Subgroups of pain medication beliefs were identified, including two groups of patients with at least some concerns about their medical providers. Beliefs ranged from a lack of concern about using pain medications to worries about scrutiny and harm. It is unclear if poor experiences with pain medications cause these beliefs or if they prevent individuals from effectively utilizing medications. Additionally, more negative beliefs about pain medications were associated with more depressive symptoms.

Multicenter Development and Prospective Validation of eCARTv5: A Gradient Boosted Machine Learning Early Warning Score.

OBJECTIVE: Early detection of clinical deterioration using machine learning early warning scores may improve outcomes. However, most implemented scores were developed using logistic regression, only underwent retrospective validation, and were not tested in important subgroups. Our objective was to develop and prospectively validate a gradient boosted machine model (eCARTv5) for identifying clinical deterioration on the wards. DESIGN: Multicenter retrospective and prospective observational study. SETTING: Inpatient admissions to the medical-surgical wards at seven hospitals in three health systems for model development (2006-2022) and at 21 hospitals from three health systems for retrospective (2009-2023) and prospective (2023-2024) external validation. PATIENTS: All adult patients hospitalized at each participating health system during the study years. INTERVENTIONS: None MEASUREMENTS AND MAIN RESULTS: Predictor variables (demographics, vital signs, documentation, and laboratory values) were used in a gradient boosted trees algorithm to predict intensive care unit transfer or death in the next 24 hours. The developed model (eCART) was compared to the Modified Early Warning Score (MEWS) and the National Early Warning Score (NEWS) using the area under the receiver operating characteristic curve (AUROC). The development cohort included 901,491 admissions, the retrospective validation cohort included 1,769,461 admissions, and the prospective validation cohort included 205,946 admissions. In retrospective validation, eCART had the highest AUROC (0.835; 95%CI 0.834, 0.835), followed by NEWS (0.766 (95%CI 0.766, 0.767)), and MEWS (0.704 (95%CI 0.703, 0.704)). eCART's performance remained high (AUROC ≥0.80) across a range of patient demographics, clinical conditions, and during prospective validation. CONCLUSIONS: We developed eCART, which performed better than the NEWS and MEWS retrospectively, prospectively, and across a range of subgroups. These results served as the foundation for Food and Drug Administration clearance for its use in identifying deterioration in hospitalized ward patients.

Development and external validation of deep learning clinical prediction models using variable-length time series data.

OBJECTIVES: To compare and externally validate popular deep learning model architectures and data transformation methods for variable-length time series data in 3 clinical tasks (clinical deterioration, severe acute kidney injury [AKI], and suspected infection). MATERIALS AND METHODS: This multicenter retrospective study included admissions at 2 medical centers that spanned 2007-2022. Distinct datasets were created for each clinical task, with 1 site used for training and the other for testing. Three feature engineering methods (normalization, standardization, and piece-wise linear encoding with decision trees [PLE-DTs]) and 3 architectures (long short-term memory/gated recurrent unit [LSTM/GRU], temporal convolutional network, and time-distributed wrapper with convolutional neural network [TDW-CNN]) were compared in each clinical task. Model discrimination was evaluated using the area under the precision-recall curve (AUPRC) and the area under the receiver operating characteristic curve (AUROC). RESULTS: The study comprised 373 825 admissions for training and 256 128 admissions for testing. LSTM/GRU models tied with TDW-CNN models with both obtaining the highest mean AUPRC in 2 tasks, and LSTM/GRU had the highest mean AUROC across all tasks (deterioration: 0.81, AKI: 0.92, infection: 0.87). PLE-DT with LSTM/GRU achieved the highest AUPRC in all tasks. DISCUSSION: When externally validated in 3 clinical tasks, the LSTM/GRU model architecture with PLE-DT transformed data demonstrated the highest AUPRC in all tasks. Multiple models achieved similar performance when evaluated using AUROC. CONCLUSION: The LSTM architecture performs as well or better than some newer architectures, and PLE-DT may enhance the AUPRC in variable-length time series data for predicting clinical outcomes during external validation.

Reported Unmet Hearing Aid Need in Older People With Dementia: The US National Core Indicators Survey.

OBJECTIVES: Hearing aids have important health benefits for older adults with Alzheimer disease and related dementias (ADRD); however, hearing aid adoption in this group is low. This study aimed to determine where to target hearing aid interventions for American long-term care recipients with ADRD by examining the association of ADRD and residence type with respondent-reported unmet hearing aid need. DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used data from the United States National Core Indicators-Aging and Disabilities survey (2015-2019) for long-term care recipients aged ≥65 years. METHODS: We used multivariable logistic regression to model the likelihood of reporting unmet hearing aid need conditional on ADRD status and residence type (own/family house or apartment, residential care, or nursing facility/home), adjusting for sociodemographic factors and response type (self vs proxy). RESULTS: Of the 25,492 respondents [median (IQR) age, 77 (71, 84) years; 7074 (27.8%) male], 5442 (21.4%) had ADRD and 3659 (14.4%) owned hearing aids. Residence types were 17,004 (66.8%) own/family house or apartment, 4966 (19.5%) residential care, and 3522 (13.8%) nursing home. Among non-hearing aid owners, ADRD [adjusted odds ratio (AOR) 0.90, 95% CI 0.80-1.0] and residence type were associated with respondent-reported unmet hearing aid need. Compared to the nursing home reference group, respondents in their own/family home (AOR 1.85, 95% CI 1.61-2.13) and residential care (AOR 1.30, 95% CI 1.10-1.53) were more likely to report unmet hearing aid need. This pattern was significantly more pronounced in people with ADRD than in those without, stemming from an interaction between ADRD and residence type. CONCLUSIONS AND IMPLICATIONS: American long-term care recipients with ADRD living in their own/family home are more likely to report unmet hearing aid need than those with ADRD in institutional and congregate settings. This information can inform the design and delivery of hearing interventions for older adults with ADRD.

The Effect of Lower Limb Alignment on Tibiofemoral Joint Contact Biomechanics after Medial Meniscus Posterior Root Repair: A Finite-Element Analysis.

INTRODUCTION: The purpose of this study was to determine how variations in lower limb alignment affect tibiofemoral joint contact biomechanics in the setting of medial meniscus posterior root tear (MMPRT) and associated root repair. METHODS: A finite-element model of an intact knee joint was developed. Limb alignments ranging from 4° valgus to 8° varus were simulated under a 1,000 N compression load applied to the femoral head. For the intact, MMPRT, and root repair conditions, the peak contact pressure (PCP), total contact area, mean and maximum local contact pressure (LCP) elevation, and total area of LCP elevation of the medial tibiofemoral compartment were quantified. RESULTS: The PCP and total contact area of the medial compartment in the intact knee increased from 2.43 MPa and 361 mm 2 at 4° valgus to 9.09 MPa and 508 mm 2 at 8° of varus. Compared with the intact state, in the MMPRT condition, medial compartment PCP was greater and the total contact area smaller for all alignment conditions. Root repair roughly restored PCPs in the medial compartment; however, this ability was compromised in knees with increasing varus alignment. Specifically, elevations in PCP relative to the intact state increased with increasing varus, as did the total contact area with LCP elevation. After root repair, medial compartment PCP remained elevated above the intact state at all degrees tested, ranging from 0.05 MPa at 4° valgus to 0.27 MPa at 8° of varus, with overall PCP values increasing from 2.48 to 9.09 MPa. For varus alignment greater than 4°, root repair failed to reduce the total contact area with LCP elevation relative to the MMPRT state. DISCUSSION: Greater PCPs and areas of LCP elevation in varus knees may reduce the clinical effectiveness of root repair in delaying or preventing the development of tibiofemoral osteoarthritis.

A novel structured debriefing program for consensus determinations of in-hospital cardiac arrest predictability and preventability.

AIM: Hospital rapid response systems aim to stop preventable cardiac arrests, but defining preventability is a challenge. We developed a multidisciplinary consensus-based process to determine in-hospital cardiac arrest (IHCA) preventability based on objective measures. METHODS: We developed an interdisciplinary ward IHCA debriefing program at an urban quaternary-care academic hospital. This group systematically reviewed all IHCAs weekly, reaching consensus determinations of the IHCA's cause and preventability across three mutually exclusive categories: 1) unpredictable (no evidence of physiologic instability < 1 h prior to and within 24 h of the arrest), 2) predictable but unpreventable (meeting physiologic instability criteria in the setting of either a poor baseline prognosis or a documented goals of care conversation) or 3) potentially preventable (remaining cases). RESULTS: Of 544 arrests between 09/2015 and 11/2023, 339 (61%) were deemed predictable by consensus, with 235 (42% of all IHCAs) considered potentially preventable. Potentially preventable arrests disproportionately occurred on nights and weekends (70% vs 55%, p = 0.002) and were more frequently respiratory than cardiac in etiology (33% vs 15%, p < 0.001). Despite similar rates of ROSC across groups (67-70%), survival to discharge was highest in arrests deemed unpredictable (31%), followed by potentially preventable (21%), and then those deemed predictable but unpreventable which had the lowest survival rate (16%, p = 0.007). CONCLUSIONS: Our IHCA debriefing procedures are a feasible and sustainable means of determining the predictability and potential preventability of ward cardiac arrests. This approach may be useful for improving quality benchmarks and care processes around pre-arrest clinical activities.