Effectiveness of an Artificial Intelligence-Enabled Intervention for Detecting Clinical Deterioration.

Importance: Inpatient clinical deterioration is associated with substantial morbidity and mortality but may be easily missed by clinicians. Early warning scores have been developed to alert clinicians to patients at high risk of clinical deterioration, but there is limited evidence for their effectiveness. Objective: To evaluate the effectiveness of an artificial intelligence deterioration model-enabled intervention to reduce the risk of escalations in care among hospitalized patients using a study design that facilitates stronger causal inference. Design, Setting, and Participants: This cohort study used a regression discontinuity design that controlled for confounding and was based on Epic Deterioration Index (EDI; Epic Systems Corporation) prediction model scores. Compared with other observational research, the regression discontinuity design facilitates causal analysis. Hospitalized adults were included from 4 general internal medicine units in 1 academic hospital from January 17, 2021, through November 16, 2022. Exposure: An artificial intelligence deterioration model-enabled intervention, consisting of alerts based on an EDI score threshold with an associated collaborative workflow among nurses and physicians. Main Outcomes and Measures: The primary outcome was escalations in care, including rapid response team activation, transfer to the intensive care unit, or cardiopulmonary arrest during hospitalization. Results: During the study, 9938 patients were admitted to 1 of the 4 units, with 963 patients (median [IQR] age, 76.1 [64.2-86.2] years; 498 males [52.3%]) included within the primary regression discontinuity analysis. The median (IQR) Elixhauser Comorbidity Index score in the primary analysis cohort was 10 (0-24). The intervention was associated with a -10.4-percentage point (95% CI, -20.1 to -0.8 percentage points; P = .03) absolute risk reduction in the primary outcome for patients at the EDI score threshold. There was no evidence of a discontinuity in measured confounders at the EDI score threshold. Conclusions and Relevance: Using a regression discontinuity design, this cohort study found that the implementation of an artificial intelligence deterioration model-enabled intervention was associated with a significantly decreased risk of escalations in care among inpatients. These results provide evidence for the effectiveness of this intervention and support its further expansion and testing in other care settings.

A wing-assisted incline running exercise regime during rearing increases initial flight velocity during descent in adult white- and brown-feathered laying hens.

Domestic laying hens rely primarily on their hindlimbs for terrestrial locomotion. Although they perform flapping flight, they appear to use maximal power during descent and thus may lack control for maneuvering and avoiding injuries on landing. This in turn may result in injury in open rearing systems. Wing-assisted incline running (WAIR) requires a bird to use its wings to assist the hindlimbs during climbing of an incline, and training in WAIR may therefore provide a useful method to increase a hen's power reserve and control for flight. We subjected hens to an exercise regimen involving inclines to induce WAIR for 16 wk during rearing. We then measured wing and body kinematics during aerial descent from a 155 cm platform. We hypothesized that birds reared with exercise would be better able to modulate their wing and body kinematics for making slower, more-controlled descent and landing. Brown-feathered birds exhibited greater wing beat frequencies than white-feathered birds, which is consistent with the higher wing loading of brown-feathered birds and WAIR-trained birds exhibited greater initial flight velocities compared to control birds. This may indicate that WAIR training provided an improved capacity to modulate flight velocity and strengthen the leg muscles. Providing incline exercises during rearing may therefore improve welfare for adult laying hens as greater initial flight velocity should reduce the power required for supporting body weight in the air and allow a hen to direct her excess power toward maneuvering.

Reduction of wing area affects estimated stress in the primary flight muscles of chickens.

In flying birds, the pectoralis (PECT) and supracoracoideus (SUPRA) generate most of the power required for flight, while the wing feathers create the aerodynamic forces. However, in domestic laying hens, little is known about the architectural properties of these muscles and the forces the wings produce. As housing space increases for commercial laying hens, understanding these properties is important for assuring safe locomotion. We tested the effects of wing area loss on mass, physiological cross-sectional area (PCSA), and estimated muscle stress (EMS) of the PECT and SUPRA in white-feathered laying hens. Treatments included Unclipped (N = 18), Half-Clipped with primaries removed (N = 18) and Fully-Clipped with the primaries and secondaries removed (N = 18). The mass and PCSA of the PECT and SUPRA did not vary significantly with treatment. Thus, laying hen muscle anatomy may be relatively resistant to changes in external wing morphology. We observed significant differences in EMS among treatments, as Unclipped birds exhibited the greatest EMS. This suggests that intact wings provide the greatest stimulus of external force for the primary flight muscles.

Trends in Pediatric Primary Care Visits During the COVID-19 Pandemic: Opportunity to Address Adolescent Behavioral Health Through Telemedicine.

BACKGROUND AND OBJECTIVES: The COVID-19 pandemic impacted the volume and nature of pediatric primary care visits nationwide. This study aimed to identify trends in pediatric visits at our institution during the pandemic to reveal opportunities to improve care of children and adolescents. METHODS: We performed a retrospective chart review of all pediatric visits conducted at a single family medicine clinic within a large academic medical center in Northern California from January 1, 2019, through September 30, 2021. Data collected for each visit included age, sex, type of visit (preventive or problem-focused), reason for visit (if problem-focused), and mode of visit (in-person or telehealth). We analyzed data using descriptive statistics and χ2 tests. RESULTS: A total of 4,844 pediatric visits occurred during the study period. Visit volume dropped 9% from 2019 to 2020 and recovered to prepandemic levels in 2021. During the study period from 2019 to 2021, the percentage of problem-focused visits increased from 30% to 37% (P=.008) among adolescents, driven largely by an increase in the percentage of behavioral health visits from 14% to 29% (P<.001). We found no significant changes in the age or sex of patients seen. Telemedicine visit volume decreased from 2020 to 2021 in all age categories except for adolescents, which remained stable at 43% of all visits. CONCLUSIONS: A sharp increase in behavioral health concerns among adolescents stands out as the most notable impact of COVID-19 on pediatric care at our institution. Our findings raise questions about how behavioral health care can be optimized for adolescents in the postpandemic era.

Voice Assistants' Responses to Questions About the COVID-19 Vaccine: National Cross-sectional Study.

BACKGROUND: Artificial intelligence-powered voice assistants (VAs), such as Apple Siri, Google Assistant, and Amazon Alexa, interact with users in natural language and are capable of responding to simple commands, searching the internet, and answering questions. Despite being an increasingly popular way for the public to access health information, VAs could be a source of ambiguous or potentially biased information. OBJECTIVE: In response to the ongoing prevalence of vaccine misinformation and disinformation, this study aims to evaluate how smartphone VAs respond to information- and recommendation-seeking inquiries regarding the COVID-19 vaccine. METHODS: A national cross-sectional survey of English-speaking adults who owned a smartphone with a VA installed was conducted online from April 22 to 28, 2021. The primary outcomes were the VAs' responses to 2 questions: "Should I get the COVID vaccine?" and "Is the COVID vaccine safe?" Directed content analysis was used to assign a negative, neutral, or positive connotation to each response and website title provided by the VAs. Statistical significance was assessed using the t test (parametric) or Mann-Whitney U (nonparametric) test for continuous variables and the chi-square or Fisher exact test for categorical variables. RESULTS: Of the 466 survey respondents included in the final analysis, 404 (86.7%) used Apple Siri, 53 (11.4%) used Google Assistant, and 9 (1.9%) used Amazon Alexa. In response to the question "Is the COVID vaccine safe?" 419 (89.9%) users received a direct response, of which 408 (97.3%) had a positive connotation encouraging users to get vaccinated. Of the websites presented, only 5.3% (11/207) had a positive connotation and 94.7% (196/207) had a neutral connotation. In response to the question "Should I get the COVID vaccine?" 93.1% (434/466) of users received a list of websites, of which 91.5% (1155/1262) had a neutral connotation. For both COVID-19 vaccine-related questions, there was no association between the connotation of a response and the age, gender, zip code, race or ethnicity, and education level of the respondent. CONCLUSIONS: Our study found that VAs were much more likely to respond directly with positive connotations to the question "Is the COVID vaccine safe?" but not respond directly and provide a list of websites with neutral connotations to the question "Should I get the COVID vaccine?" To our knowledge, this is the first study to evaluate how VAs respond to both information- and recommendation-seeking inquiries regarding the COVID-19 vaccine. These findings add to our growing understanding of both the opportunities and pitfalls of VAs in supporting public health information dissemination.

The AI Will See You Now: Feasibility and Acceptability of a Conversational AI Medical Interviewing System.

BACKGROUND: Primary care physicians (PCPs) are often limited in their ability to collect detailed medical histories from patients, which can lead to errors or delays in diagnosis. Recent advances in artificial intelligence (AI) show promise in augmenting current human-driven methods of collecting personal and family histories; however, such tools are largely unproven. OBJECTIVE: The main aim of this pilot study was to evaluate the feasibility and acceptability of a conversational AI medical interviewing system among patients. METHODS: The study was conducted among adult patients empaneled at a family medicine clinic within a large academic medical center in Northern California. Participants were asked to test an AI medical interviewing system, which uses a conversational avatar and chatbot to capture medical histories and identify patients with risk factors. After completing an interview with the AI system, participants completed a web-based survey inquiring about the performance of the system, the ease of using the system, and attitudes toward the system. Responses on a 7-point Likert scale were collected and evaluated using descriptive statistics. RESULTS: A total of 20 patients with a mean age of 50 years completed an interview with the AI system, including 12 females (60%) and 8 males (40%); 11 were White (55%), 8 were Asian (40%), and 1 was Black (5%), and 19 had at least a bachelor's degree (95%). Most participants agreed that using the system to collect histories could help their PCPs have a better understanding of their health (16/20, 80%) and help them stay healthy through identification of their health risks (14/20, 70%). Those who reported that the system was clear and understandable, and that they were able to learn it quickly, tended to be younger; those who reported that the tool could motivate them to share more comprehensive histories with their PCPs tended to be older. CONCLUSIONS: In this feasibility and acceptability pilot of a conversational AI medical interviewing system, the majority of patients believed that it could help clinicians better understand their health and identify health risks; however, patients were split on the effort required to use the system, and whether AI should be used for medical interviewing. Our findings suggest areas for further research, such as understanding the user interface factors that influence ease of use and adoption, and the reasons behind patients' attitudes toward AI-assisted history-taking.

Voice Assistants and Cancer Screening: A Comparison of Alexa, Siri, Google Assistant, and Cortana.

Despite increasing interest in how voice assistants like Siri or Alexa might improve health care delivery and information dissemination, there is limited research assessing the quality of health information provided by these technologies. Voice assistants present both opportunities and risks when facilitating searches for or answering health-related questions, especially now as fewer patients are seeing their physicians for preventive care due to the ongoing pandemic. In our study, we compared the 4 most widely used voice assistants (Amazon Alexa, Apple Siri, Google Assistant, and Microsoft Cortana) and their ability to understand and respond accurately to questions about cancer screening. We show that there are clear differences among the 4 voice assistants and that there is room for improvement across all assistants, particularly in their ability to provide accurate information verbally. In order to ensure that voice assistants provide accurate information about cancer screening and support, rather than undermine efforts to improve preventive care delivery and population health, we suggest that technology providers prioritize partnership with health professionals and organizations.

From Code to Bedside: Implementing Artificial Intelligence Using Quality Improvement Methods.

Despite increasing interest in how artificial intelligence (AI) can augment and improve healthcare delivery, the development of new AI models continues to outpace adoption in existing healthcare processes. Integration is difficult because current approaches separate the development of AI models from the complex healthcare environments in which they are intended to function, resulting in models developed without a clear and compelling use case and not tested or scalable in a clinical setting. We propose that current approaches and traditional research methods do not support successful AI implementation in healthcare and outline a repeatable mixed-methods approach, along with several examples, that facilitates uptake of AI technologies into human-driven healthcare processes. Unlike traditional research, these methods do not seek to control for variation, but rather understand it to learn how a technology will function in practice coupled with user-centered design techniques. This approach, leveraging design thinking and quality improvement methods, aims to increase the adoption of AI in healthcare and prompt further study to understand which methods are most successful for AI implementations.

Salience-Based Edge Selection in Flicker and Binocular Color Vision.

A test cross that flickers between light yellow and dark blue at 5 to 8Hz looks apparently yellow on a dark gray surround and apparently blue on a light gray surround (flicker augmented contrast). The achromatic surround cannot be inducing the perceived colors. Instead, the visual system selects the more salient apparent color with the higher Michelson contrast. The same is true for dichoptic vision. When one eye views a steady, light yellow cross and the other eye views a congruent steady dark blue cross, the binocular combination of colors looks apparently yellow on a dark gray surround and apparently blue on a light gray surround. Thus, when competing stimuli are distributed over time (flicker) or space (dichoptic vision), the visual system overweights the stimulus with the higher contrast. To see objects clearly, we accept the best view of any object and downplay inferior alternatives.

Preliminary Evaluation of Washington State's Early Intervention Program for First-Episode Psychosis.

OBJECTIVE: Early intervention programs are designed to address the needs of youths experiencing first-episode psychosis (FEP). Washington State developed New Journeys, a network of coordinated specialty care programs for FEP. In this study, the authors have outlined components of the New Journeys model and preliminary findings since its initial implementation. METHODS: Youths and young adults diagnosed as having psychosis (N=112) completed measures at and after intake on a range of mental health assessments and functional outcomes for the first 12 months of treatment. Administrative data including state-funded emergency department and psychiatric hospitalizations were assessed 24 months before and after intake. Generalized estimating equations were used to assess change over time on multiple measures of mental health status. RESULTS: Compared with their condition at intake, clients had significant decreases in symptoms of anxiety (ß=-2.48, p<0.001), psychotic experiences (ß=-3.37, p<0.05), and clinician-rated psychotic symptoms (ß=-1.47, p<0.05) during treatment. Additionally, quality of life (ß=-5.95, p<0.001) and school attendance (odds ratio=1.42, p<0.05) significantly improved during treatment. Administrative data indicated that postintake, clients were less likely to visit the emergency department for psychiatric reasons (ß=0.22, p<0.05), utilize community psychiatric inpatient services (ß=0.31, p<0.001), and utilize public assistance (ß=0.71, p<0.05) compared with 24 months before intake. CONCLUSIONS: New Journeys clients experienced improved clinical and functional outcomes during their first year of treatment, and rates of state-funded service utilization decreased during their treatment.

High levels of sarcospan are well tolerated and act as a sarcolemmal stabilizer to address skeletal muscle and pulmonary dysfunction in DMD.

Duchenne muscular dystrophy (DMD) is a genetic disorder that causes progressive muscle weakness, ultimately leading to early mortality in affected teenagers and young adults. Previous work from our lab has shown that a small transmembrane protein called sarcospan (SSPN) can enhance the recruitment of adhesion complex proteins to the cell surface. When human SSPN is expressed at three-fold levels in mdx mice, this increase in adhesion complex abundance improves muscle membrane stability, preventing many of the histopathological changes associated with DMD. However, expressing higher levels of human SSPN (ten-fold transgenic expression) causes a severe degenerative muscle phenotype in wild-type mice. Since SSPN-mediated stabilization of the sarcolemma represents a promising therapeutic strategy in DMD, it is important to determine whether SSPN can be introduced at high levels without toxicity. Here, we show that mouse SSPN (mSSPN) can be overexpressed at 30-fold levels in wild-type mice with no deleterious effects. In mdx mice, mSSPN overexpression improves dystrophic pathology and sarcolemmal stability. We show that these mice exhibit increased resistance to eccentric contraction-induced damage and reduced fatigue following exercise. mSSPN overexpression improved pulmonary function and reduced dystrophic histopathology in the diaphragm. Together, these results demonstrate that SSPN overexpression is well tolerated in mdx mice and improves sarcolemma defects that underlie skeletal muscle and pulmonary dysfunction in DMD.

The importance of lake-specific characteristics for water quality across the continental United States.

Lake water quality is affected by local and regional drivers, including lake physical characteristics, hydrology, landscape position, land cover, land use, geology, and climate. Here, we demonstrate the utility of hypothesis testing within the landscape limnology framework using a random forest algorithm on a national-scale, spatially explicit data set, the United States Environmental Protection Agency's 2007 National Lakes Assessment. For 1026 lakes, we tested the relative importance of water quality drivers across spatial scales, the importance of hydrologic connectivity in mediating water quality drivers, and how the importance of both spatial scale and connectivity differ across response variables for five important in-lake water quality metrics (total phosphorus, total nitrogen, dissolved organic carbon, turbidity, and conductivity). By modeling the effect of water quality predictors at different spatial scales, we found that lake-specific characteristics (e.g., depth, sediment area-to-volume ratio) were important for explaining water quality (54-60% variance explained), and that regionalization schemes were much less effective than lake specific metrics (28-39% variance explained). Basin-scale land use and land cover explained between 45-62% of variance, and forest cover and agricultural land uses were among the most important basin-scale predictors. Water quality drivers did not operate independently; in some cases, hydrologic connectivity (the presence of upstream surface water features) mediated the effect of regional-scale drivers. For example, for water quality in lakes with upstream lakes, regional classification schemes were much less effective predictors than lake-specific variables, in contrast to lakes with no upstream lakes or with no surface inflows. At the scale of the continental United States, conductivity was explained by drivers operating at larger spatial scales than for other water quality responses. The current regulatory practice of using regionalization schemes to guide water quality criteria could be improved by consideration of lake-specific characteristics, which were the most important predictors of water quality at the scale of the continental United States. The spatial extent and high quality of contextual data available for this analysis makes this work an unprecedented application of landscape limnology theory to water quality data. Further, the demonstrated importance of lake morphology over other controls on water quality is relevant to both aquatic scientists and managers.

Rituximab therapy for myopathy associated with anti-signal recognition particle antibodies: a case series.

OBJECTIVE: The myopathy associated with anti-signal recognition particle (anti-SRP) is a severe necrotizing immune-mediated disease characterized by rapidly progressive proximal muscle weakness, markedly elevated serum creatine kinase (CK) levels, and poor responsiveness to traditional immunosuppressive therapies. Reports on the efficacy of B cell depletion therapy for anti-SRP-associated myopathy are mixed. We describe 8 patients with anti-SRP-associated myopathy and their response to treatment with the anti-CD20 monoclonal antibody rituximab. METHODS: We identified 8 patients with myopathy who tested positive for anti-SRP antibodies by immunoprecipitation and were treated with rituximab as part of clinical care. We reviewed their medical records to assess clinical, serologic, and histologic characteristics and response to therapy. In 5 patients, serum was collected before and after rituximab therapy. Autoantibodies were detected by immunoprecipitation and quantitated by densitometry, and the percent decreases in anti-SRP autoantibody levels were calculated. RESULTS: Six of 8 patients who had been refractory to standard immunosuppressive therapy demonstrated improved manual muscle strength and/or decline in CK levels as early as 2 months after rituximab treatment. Three patients sustained the response for 12-18 months after initial dosing. All of the patients were continued on adjunctive corticosteroids, but doses were substantially reduced after rituximab. Quantitative levels of serum anti-SRP antibodies also decreased after rituximab treatment. CONCLUSION: B cell depletion therapy with rituximab is effective for patients with myopathy associated with anti-SRP. The substantial decrease in anti-SRP antibody levels after rituximab treatment also suggests that B cells and anti-SRP antibodies may play a role in the pathogenesis of this myopathy.

A novel autoantibody recognizing 200-kd and 100-kd proteins is associated with an immune-mediated necrotizing myopathy.

OBJECTIVE: Myofiber necrosis without prominent inflammation is a nonspecific finding in patients with dystrophies and toxic or immune-mediated myopathies. However, the etiology of a necrotizing myopathy is often obscure, and the question of which patients would benefit from immunosuppression remains unanswered. The aim of this study was to identify novel autoantibodies in patients with necrotizing myopathy. METHODS: Muscle biopsy specimens and serum samples were available for 225 patients with myopathy. Antibody specificities were determined by performing immunoprecipitations from (35)S-methionine-labeled HeLa cell lysates. Selected biopsy specimens were stained for membrane attack complex, class I major histocompatibility complex (MHC), and endothelial cell marker CD31. RESULTS: Muscle biopsy specimens from 38 of 225 patients showed predominantly myofiber necrosis. Twelve of these patients had a known autoantibody association with or other etiology for their myopathy. Sixteen of the remaining 26 sera immunoprecipitated 200-kd and 100-kd proteins; this specificity was observed in only 1 of 187 patients without necrotizing myopathy. Patients with the anti-200/100 autoantibody specificity had proximal weakness (100%), high creatine kinase levels (mean maximum 10,333 IU/liter), and an irritable myopathy on electromyography (88%). Sixty-three percent of these patients had been exposed to statins prior to the onset of weakness. All patients responded to immunosuppressive therapy, and many experienced a relapse of weakness when the medication was tapered. Immunohistochemical studies showed membrane attack complex on small blood vessels in 6 of 8 patients and on the surface of non-necrotic myofibers in 4 of 8 patients. Five of 8 patients had abnormal capillary morphology, and 4 of 8 patients expressed class I MHC on the surface of non-necrotic myofibers. CONCLUSION: An anti-200/100-kd specificity defines a subgroup of patients with necrotizing myopathy who previously were considered to be autoantibody negative. We propose that these patients have an immune-mediated myopathy that is frequently associated with prior statin use and should be treated with immunosuppressive therapy.

Clinical profile of anti-PL-12 autoantibody. Cohort study and review of the literature.

BACKGROUND: The antisynthetase syndrome consists of interstitial lung disease (ILD), arthritis, myositis, fever, mechanic's hands, and Raynaud phenomenon in the presence of an antisynthetase autoantibody, most commonly anti-Jo-1. It is believed that all the antisynthetases are associated with a similar clinical profile, but definitive data in this diverse group are lacking. The purpose of this study was to examine the clinical profile of anti-PL-12, an antisynthetase autoantibody directed against alanyl-transfer RNA synthetase. METHODS: Thirty-one subjects with anti-PL-12 autoantibody were identified from the databases at the Medical University of South Carolina, the University of Pittsburgh Medical Center, Johns Hopkins Medical Center, and Brigham and Women's Hospital. The medical charts were reviewed and the following data were recorded: demographic information; pulmonary and rheumatologic symptoms; connective tissue disease (CTD) diagnoses; serologic autoantibody findings; CT scan results; BAL findings; pulmonary function test results; lung histopathology; and treatment interventions. RESULTS: The median age at symptom onset was 51 years; 81% were women and 52% were African American. Ninety percent of anti-PL-12-positive patients had ILD, 65% of whom presented initially to a pulmonologist. Ninety percent of anti-PL-12-positive patients had an underlying CTD. Polymyositis and dermatomyositis were the most common underlying diagnoses. Raynaud phenomenon occurred in 65% of patients, fever in 45% of patients, and mechanic's hands in 16% of patients. Test results for the presence of antinuclear antibody were positive in 48% of cases. CONCLUSIONS: Anti-PL-12 is strongly associated with the presence of ILD, but less so with myositis and arthritis. Idiopathic ILD diagnosed as idiopathic pulmonary fibrosis may, in fact, be associated with anti-PL-12 and be a "forme fruste" of an underlying autoimmune disorder.