Search

Motor vehicle-related injuries to adolescents continue to be of paramount importance to society. Since the original policy statement on the teenaged driver was published in 1996, there have been substantial changes in many state laws and much new research on this topic. There is a need to provide pediatricians with up-to-date information and materials to facilitate appropriate counseling and anticipatory guidance. This statement describes why teenagers are at greater risk of motor vehicle-related injuries, suggests topics suitable for office-based counseling, describes innovative programs, and proposes preventive interventions for pediatricians, parents, legislators, educators, and other child advocates. [ABSTRACT FROM AUTHOR]

We present a neural network framework for learning a survival model to predict a time-to-event outcome while simultaneously learning a topic model that reveals feature relationships. In particular, we model each subject as a distribution over "topics", where a topic could, for instance, correspond to an age group, a disorder, or a disease. The presence of a topic in a subject means that specific clinical features are more likely to appear for the subject. Topics encode information about related features and are learned in a supervised manner to predict a time-to-event outcome. Our framework supports combining many different topic and survival models; training the resulting joint survival-topic model readily scales to large datasets using standard neural net optimizers with minibatch gradient descent. For example, a special case is to combine LDA with a Cox model, in which case a subject's distribution over topics serves as the input feature vector to the Cox model. We explain how to address practical implementation issues that arise when applying these neural survival-supervised topic models to clinical data, including how to visualize results to assist clinical interpretation. We study the effectiveness of our proposed framework on seven clinical datasets on predicting time until death as well as hospital ICU length of stay, where we find that neural survival-supervised topic models achieve competitive accuracy with existing approaches while yielding interpretable clinical topics that explain feature relationships. Our code is available at: https://github.com/georgehc/survival-topics., Competing Interests: Declaration of competing interest All authors have participated in (a) conception and design, or analysis and interpretation of the data; (b) drafting the article or revising it critically for important intellectual content; and (c) approval of the final version. This manuscript has not been submitted to, nor is under review at, another journal or other publishing venue. The authors have no affiliation with any organization with a direct or indirect financial interest in the subject matter discussed in the manuscript, (Copyright © 2024 The Author(s). Published by Elsevier B.V. All rights reserved.)

Textual data often describe events in time but frequently contain little information about their specific timing, whereas complementary structured data streams may have precise timestamps but may omit important contextual information. We investigate the problem in healthcare, where we produce clinician annotations of discharge summaries, with access to either unimodal (text) or multimodal (text and tabular) data, (i) to determine event interval timings and (ii) to train multimodal language models to locate those events in time. We find our annotation procedures, dashboard tools, and annotations result in high-quality timestamps. Specifically, the multimodal approach produces more precise timestamping, with uncertainties of the lower bound, upper bounds, and duration reduced by 42% (95% CI 34-51%), 36% (95% CI 28-44%), and 13% (95% CI 10-17%), respectively. In the classification version of our task, we find that, trained on our annotations, our multimodal BERT model outperforms unimodal BERT model and Llama-2 encoder-decoder models with improvements in F1 scores for upper (10% and 61%, respectively) and lower bounds (8% and 56%, respectively). The code for the annotation tool and the BERT model is available (link).

Reliable extraction of temporal relations from clinical notes is a growing need in many clinical research domains. Our work introduces typed markers to the task of clinical temporal relation extraction. We demonstrate that the addition of medical entity information to clinical text as tags with context sentences then input to a transformer-based architecture can outperform more complex systems requiring feature engineering and temporal reasoning. We propose several strategies of typed marker creation that incorporate entity type information at different granularities, with extensive experiments to test their effectiveness. Our system establishes the best result on I2B2, a clinical benchmark dataset for temporal relation extraction, with a F1 at 83.5% that provides a substantial 3.3% improvement over the previous best system.

With COVID-19 now pervasive, identification of high-risk individuals is crucial. Using data from a major healthcare provider in Southwestern Pennsylvania, we develop survival models predicting severe COVID-19 progression. In this endeavor, we face a tradeoff between more accurate models relying on many features and less accurate models relying on a few features aligned with clinician intuition. Complicating matters, many EHR features tend to be under-coded degrading the accuracy of smaller models. In this study we develop two sets of high-performance risk scores: (i) an unconstrained model built from all available features; and (ii) a pipeline that learns a small set of clinical concepts before training a risk predictor. Learned concepts boost performance over the corresponding features (C-index 0.858 vs. 0.844) and demonstrate improvements over (i) when evaluated out-of-sample (subsequent time periods). Our models outperform previous works (C-index 0.844-0.872 vs. 0.598-0.810)., (©2022 AMIA - All rights reserved.)

We consider the problem of predicting how the likelihood of an outcome of interest for a patient changes over time as we observe more of the patient's data. To solve this problem, we propose a supervised contrastive learning framework that learns an embedding representation for each time step of a patient time series. Our framework learns the embedding space to have the following properties: (1) nearby points in the embedding space have similar predicted class probabilities, (2) adjacent time steps of the same time series map to nearby points in the embedding space, and (3) time steps with very different raw feature vectors map to far apart regions of the embedding space. To achieve property (3), we employ a nearest neighbor pairing mechanism in the raw feature space. This mechanism also serves as an alternative to "data augmentation", a key ingredient of contrastive learning, which lacks a standard procedure that is adequately realistic for clinical tabular data, to our knowledge. We demonstrate that our approach outperforms state-of-the-art baselines in predicting mortality of septic patients (MIMIC-III dataset) and tracking progression of cognitive impairment (ADNI dataset). Our method also consistently recovers the correct synthetic dataset embedding structure across experiments, a feat not achieved by baselines. Our ablation experiments show the pivotal role of our nearest neighbor pairing.

Diagnostic coding, or ICD coding, is the task of assigning diagnosis codes defined by the ICD (International Classification of Diseases) standard to patient visits based on clinical notes. The current process of manual ICD coding is time-consuming and often error-prone, which suggests the need for automatic ICD coding. However, despite the long history of automatic ICD coding, there have been no standardized frameworks for benchmarking ICD coding models. We open-source an easy-to-use tool named AnEMIC , which provides a streamlined pipeline for preprocessing, training, and evaluating for automatic ICD coding. We correct errors in preprocessing by existing works, and provide key models and weights trained on the correctly preprocessed datasets. We also provide an interactive demo performing real-time inference from custom inputs, and visualizations drawn from explainable AI to analyze the models. We hope the framework helps move the research of ICD coding forward and helps professionals explore the potential of ICD coding. The framework and the associated code are available here.

Background and Aims: The time lag encountered when accessing health-care data is one major barrier to implementing opioid overdose prediction measures in practice. Little is known regarding how one's opioid overdose risk changes over time. We aimed to identify longitudinal patterns of individual predicted overdose risks among Medicaid beneficiaries after initiation of opioid prescriptions., Design, Setting and Participants: A retrospective cohort study in Pennsylvania, USA among Pennsylvania Medicaid beneficiaries aged 18-64 years who initiated opioid prescriptions between July 2017 and September 2018 (318 585 eligible beneficiaries (mean age = 39 ± 12 years, female = 65.7%, White = 62.2% and Black = 24.9%)., Measurements: We first applied a previously developed and validated machine-learning algorithm to obtain risk scores for opioid overdose emergency room or hospital visits in 3-month intervals for each beneficiary who initiated opioid therapy, until disenrollment from Medicaid, death or the end of observation (December 2018). We performed group-based trajectory modeling to identify trajectories of these predicted overdose risk scores over time., Findings: Among eligible beneficiaries, 0.61% had one or more occurrences of opioid overdose in a median follow-up of 15 months. We identified five unique opioid overdose risk trajectories: three trajectories (accounting for 92% of the cohort) had consistent overdose risk over time, including consistent low-risk (63%), consistent medium-risk (25%) and consistent high-risk (4%) groups; another two trajectories (accounting for 8%) had overdose risks that substantially changed over time, including a group that transitioned from high- to medium-risk (3%) and another group that increased from medium- to high-risk over time (5%)., Conclusions: More than 90% of Medicaid beneficiaries in Pennsylvania USA with one or more opioid prescriptions had consistent, predicted opioid overdose risks over 15 months. Applying opioid prediction algorithms developed from historical data may not be a major barrier to implementation in practice for the large majority of individuals., (© 2022 Society for the Study of Addiction.)

Background: Little is known about whether machine-learning algorithms developed to predict opioid overdose using earlier years and from a single state will perform as well when applied to other populations. We aimed to develop a machine-learning algorithm to predict 3-month risk of opioid overdose using Pennsylvania Medicaid data and externally validated it in two data sources (ie, later years of Pennsylvania Medicaid data and data from a different state)., Methods: This prognostic modelling study developed and validated a machine-learning algorithm to predict overdose in Medicaid beneficiaries with one or more opioid prescription in Pennsylvania and Arizona, USA. To predict risk of hospital or emergency department visits for overdose in the subsequent 3 months, we measured 284 potential predictors from pharmaceutical and health-care encounter claims data in 3-month periods, starting 3 months before the first opioid prescription and continuing until loss to follow-up or study end. We developed and internally validated a gradient-boosting machine algorithm to predict overdose using 2013-16 Pennsylvania Medicaid data (n=639 693). We externally validated the model using (1) 2017-18 Pennsylvania Medicaid data (n=318 585) and (2) 2015-17 Arizona Medicaid data (n=391 959). We reported several prediction performance metrics (eg, C-statistic, positive predictive value). Beneficiaries were stratified into risk-score subgroups to support clinical use., Findings: A total of 8641 (1·35%) 2013-16 Pennsylvania Medicaid beneficiaries, 2705 (0·85%) 2017-18 Pennsylvania Medicaid beneficiaries, and 2410 (0·61%) 2015-17 Arizona beneficiaries had one or more overdose during the study period. C-statistics for the algorithm predicting 3-month overdoses developed from the 2013-16 Pennsylvania training dataset and validated on the 2013-16 Pennsylvania internal validation dataset, 2017-18 Pennsylvania external validation dataset, and 2015-17 Arizona external validation dataset were 0·841 (95% CI 0·835-0·847), 0·828 (0·822-0·834), and 0·817 (0·807-0·826), respectively. In external validation datasets, 71 361 (22·4%) of 318 585 2017-18 Pennsylvania beneficiaries were in high-risk subgroups (positive predictive value of 0·38-4·08%; capturing 73% of overdoses in the subsequent 3 months) and 40 041 (10%) of 391 959 2015-17 Arizona beneficiaries were in high-risk subgroups (positive predictive value of 0·19-1·97%; capturing 55% of overdoses). Lower risk subgroups in both validation datasets had few individuals (≤0·2%) with an overdose., Interpretation: A machine-learning algorithm predicting opioid overdose derived from Pennsylvania Medicaid data performed well in external validation with more recent Pennsylvania data and with Arizona Medicaid data. The algorithm might be valuable for overdose risk prediction and stratification in Medicaid beneficiaries., Funding: National Institute of Health, National Institute on Drug Abuse, National Institute on Aging., Competing Interests: Declaration of interests W-HL-C and WFG are named as inventors in one preliminary patent (U1195.70174US00) filing from the University of Florida and University of Pittsburgh for use of the machine learning algorithm for opioid risk prediction in Medicare described in this Article. W-HL-C , WFG, DLW, and C-YC are recipients of a grant from the National Institute on Aging (NIA; R21 AG060308). W-HL-C , WFG , JMD, AJG, GC, JLH, CCK, DCM, QY, JW, and HHZ are recipients of a grant from the US National Institute on Drug Abuse (NIDA; R01DA044985). W-HL-C declares grants from the Richard King Mellon Foundation–University of Pittsburgh, University of Florida Clinical and Translational Science Institute, the US National Institute of Mental Health (1R03MH114503-01 and R01MH121907), Pharmaceutical Research and Manufacturers of America Foundation, NIDA (1R01DA050676-01A1 and R01DA044985), Veterans Affairs (VA) Merit 1 (I01HX002191-01A2), and Merck, Sharp & Dohme and Bristol Myers Squibb. CKK declares grants from Lilly, Pfizer, GSK, Cumberland Pharmaceuticals, AbbVie, and EMD Serono; consultant fees from EMD Serono, Express Scripts, and Regeneron; an advisory board role with EMD Serono, Thusane, Regeneron, Taiwan Lipisome Company, Amzell, LG Chem, and Novartis; payment or honoraria from Focus Communications and PRIME Education; and being on the Data and Safety Monitoring Committee for Kolon Tissue Gene and on the board of directors for the International Chinese Osteoarthritis Research Society. WFG declares grant or contract to his institution from Richard King Mellon Foundation. JMD declares salary support from the Pennsylvania Department of Human Services, Richard King Mellon Foundation, and the US National Institute of Health (NIH)–NIDA (R01DA048019). AJG declares grants or contracts from NIH and the VA; royalties from UpToDate; and other financial and non-financial interests from American Society of Addiction Medicine, Association for Multidisciplinary Education and Research in Substance Use and Addiction, and the International Society of Addiction Journal Editors. JCW declares grants or contracts from Allegheny Health Network, Carnegie Mellon University, and University of Pittsburgh Medical Center; payment or honoraria from St Jude's; and receipt of equipment, materials, or other services from Amazon Web Services and Azure. DLW declares grant funding from Merck, Sharp & Dohme, and NIH–NIDA (1R01DA050676-01A1). All other authors declare no competing interests., (Copyright © 2022 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY-NC-ND 4.0 license. Published by Elsevier Ltd.. All rights reserved.)

Spelling correction is a particularly important problem in clinical natural language processing because of the abundant occurrence of misspellings in medical records. However, the scarcity of labeled datasets in a clinical context makes it hard to build a machine learning system for such clinical spelling correction. In this work, we present a probabilistic model of correcting misspellings based on a simple conditional independence assumption, which leads to a modular decomposition into a language model and a corruption model. With a deep character-level language model trained on a large clinical corpus, and a simple edit-based corruption model, we can build a spelling correction model with small or no real data. Experimental results show that our model significantly outperforms baselines on two healthcare spelling correction datasets.

Since the COVID-19 pandemic began, the United States's case fatality rate (CFR) has plummeted. Using national and Florida data, we unpack the drop in CFR between April and December 2020, accounting for such confounders as expanded testing, age distribution shift, and detection-to-death lags. Guided by the insight that treatment improvements in this period should correspond to decreases in hospitalization fatality rate (HFR), and using a block-bootstrapping procedure to quantify uncertainty, we find that although treatment improvements do not follow the same trajectory in Florida and nationally (with Florida undergoing a comparatively severe second peak), by December, significant improvements are observed both in Florida and nationally (at least 17% and 55% respectively). These estimates paint a more realistic picture of improvements than the drop in aggregate CFR (70.8%-91.1%). We publish a website where users can apply our analyses to selected demographics, regions, and dates of interest., (©2021 AMIA - All rights reserved.)

The COVID-19 pandemic presented enormous data challenges in the United States. Policy makers, epidemiological modelers, and health researchers all require up-to-date data on the pandemic and relevant public behavior, ideally at fine spatial and temporal resolution. The COVIDcast API is our attempt to fill this need: Operational since April 2020, it provides open access to both traditional public health surveillance signals (cases, deaths, and hospitalizations) and many auxiliary indicators of COVID-19 activity, such as signals extracted from deidentified medical claims data, massive online surveys, cell phone mobility data, and internet search trends. These are available at a fine geographic resolution (mostly at the county level) and are updated daily. The COVIDcast API also tracks all revisions to historical data, allowing modelers to account for the frequent revisions and backfill that are common for many public health data sources. All of the data are available in a common format through the API and accompanying R and Python software packages. This paper describes the data sources and signals, and provides examples demonstrating that the auxiliary signals in the COVIDcast API present information relevant to tracking COVID activity, augmenting traditional public health reporting and empowering research and decision-making., Competing Interests: The authors declare no competing interest., (Copyright © 2021 the Author(s). Published by PNAS.)

Background: Survivors of opioid overdose have substantially increased mortality risk, although this risk is not evenly distributed across individuals. No study has focused on predicting an individual's risk of death after a nonfatal opioid overdose., Objective: To predict risk of death after a nonfatal opioid overdose., Design and Participants: This retrospective cohort study included 9686 Pennsylvania Medicaid beneficiaries with an emergency department or inpatient claim for nonfatal opioid overdose in 2014-2016. The index date was the first overdose claim during this period., Exposures, Main Outcome, and Measures: Predictor candidates were measured in the 180 days before the index overdose. Primary outcome was 180-day all-cause mortality. Using a gradient boosting machine model, we classified beneficiaries into six subgroups according to their risk of mortality (< 25th percentile of the risk score, 25th to < 50th, 50th to < 75th, 75th to < 90th, 90th to < 98th, ≥ 98th). We then measured receipt of medication for opioid use disorder (OUD), risk mitigation interventions (e.g., prescriptions for naloxone), and prescription opioids filled in the 180 days after the index overdose, by risk subgroup., Key Results: Of eligible beneficiaries, 347 (3.6%) died within 180 days after the index overdose. The C-statistic of the mortality prediction model was 0.71. In the highest risk subgroup, the observed 180-day mortality rate was 20.3%, while in the lowest risk subgroup, it was 1.5%. Medication for OUD and risk mitigation interventions after overdose were more commonly seen in lower risk groups, while opioid prescriptions were more likely to be used in higher risk groups (both p trends < .001)., Conclusions: A risk prediction model performed well for classifying mortality risk after a nonfatal opioid overdose. This prediction score can identify high-risk subgroups to target interventions to improve outcomes among overdose survivors.

Objective: To develop and validate a machine-learning algorithm to improve prediction of incident OUD diagnosis among Medicare beneficiaries with ≥1 opioid prescriptions., Methods: This prognostic study included 361,527 fee-for-service Medicare beneficiaries, without cancer, filling ≥1 opioid prescriptions from 2011-2016. We randomly divided beneficiaries into training, testing, and validation samples. We measured 269 potential predictors including socio-demographics, health status, patterns of opioid use, and provider-level and regional-level factors in 3-month periods, starting from three months before initiating opioids until development of OUD, loss of follow-up or end of 2016. The primary outcome was a recorded OUD diagnosis or initiating methadone or buprenorphine for OUD as proxy of incident OUD. We applied elastic net, random forests, gradient boosting machine, and deep neural network to predict OUD in the subsequent three months. We assessed prediction performance using C-statistics and other metrics (e.g., number needed to evaluate to identify an individual with OUD [NNE]). Beneficiaries were stratified into subgroups by risk-score decile., Results: The training (n = 120,474), testing (n = 120,556), and validation (n = 120,497) samples had similar characteristics (age ≥65 years = 81.1%; female = 61.3%; white = 83.5%; with disability eligibility = 25.5%; 1.5% had incident OUD). In the validation sample, the four approaches had similar prediction performances (C-statistic ranged from 0.874 to 0.882); elastic net required the fewest predictors (n = 48). Using the elastic net algorithm, individuals in the top decile of risk (15.8% [n = 19,047] of validation cohort) had a positive predictive value of 0.96%, negative predictive value of 99.7%, and NNE of 104. Nearly 70% of individuals with incident OUD were in the top two deciles (n = 37,078), having highest incident OUD (36 to 301 per 10,000 beneficiaries). Individuals in the bottom eight deciles (n = 83,419) had minimal incident OUD (3 to 28 per 10,000)., Conclusions: Machine-learning algorithms improve risk prediction and risk stratification of incident OUD in Medicare beneficiaries., Competing Interests: We have read the journal's policy and the authors of this manuscript have the following competing interests: Dr. Kwoh has received honoraria from AbbVie and EMD Serono and has provided consulting services for Astellas, Thusane, and Novartis, EMD Serono and Express Scripts. I confirm that this does not alter our adherence to PLOS ONE policies on sharing data and materials.

We consider the task of producing a useful clustering of healthcare providers from their clinical action signature- their drug, procedure, and billing codes. Because high-dimensional sparse count vectors are challenging to cluster, we develop a novel autoencoder framework to address this task. Our solution creates a low-dimensional embedded representation of the high-dimensional space that preserves angular relationships and assigns examples to clusters while optimizing the quality of this clustering. Our method is able to find a better clustering than under a two-step alternative, e.g., projected K means/medoids, where a representation is learned and then clustering is applied to the representation. We demonstrate our method's characteristics through quantitative and qualitative analysis of real and simulated data, including in several real-world healthcare case studies. Finally, we develop a tool to enhance exploratory analysis of providers based on their clinical behaviors., (©2019 AMIA - All rights reserved.)

Importance: Sepsis is a heterogeneous syndrome. Identification of distinct clinical phenotypes may allow more precise therapy and improve care., Objective: To derive sepsis phenotypes from clinical data, determine their reproducibility and correlation with host-response biomarkers and clinical outcomes, and assess the potential causal relationship with results from randomized clinical trials (RCTs)., Design, Settings, and Participants: Retrospective analysis of data sets using statistical, machine learning, and simulation tools. Phenotypes were derived among 20 189 total patients (16 552 unique patients) who met Sepsis-3 criteria within 6 hours of hospital presentation at 12 Pennsylvania hospitals (2010-2012) using consensus k means clustering applied to 29 variables. Reproducibility and correlation with biological parameters and clinical outcomes were assessed in a second database (2013-2014; n = 43 086 total patients and n = 31 160 unique patients), in a prospective cohort study of sepsis due to pneumonia (n = 583), and in 3 sepsis RCTs (n = 4737)., Exposures: All clinical and laboratory variables in the electronic health record., Main Outcomes and Measures: Derived phenotype (α, β, γ, and δ) frequency, host-response biomarkers, 28-day and 365-day mortality, and RCT simulation outputs., Results: The derivation cohort included 20 189 patients with sepsis (mean age, 64 [SD, 17] years; 10 022 [50%] male; mean maximum 24-hour Sequential Organ Failure Assessment [SOFA] score, 3.9 [SD, 2.4]). The validation cohort included 43 086 patients (mean age, 67 [SD, 17] years; 21 993 [51%] male; mean maximum 24-hour SOFA score, 3.6 [SD, 2.0]). Of the 4 derived phenotypes, the α phenotype was the most common (n = 6625; 33%) and included patients with the lowest administration of a vasopressor; in the β phenotype (n = 5512; 27%), patients were older and had more chronic illness and renal dysfunction; in the γ phenotype (n = 5385; 27%), patients had more inflammation and pulmonary dysfunction; and in the δ phenotype (n = 2667; 13%), patients had more liver dysfunction and septic shock. Phenotype distributions were similar in the validation cohort. There were consistent differences in biomarker patterns by phenotype. In the derivation cohort, cumulative 28-day mortality was 287 deaths of 5691 unique patients (5%) for the α phenotype; 561 of 4420 (13%) for the β phenotype; 1031 of 4318 (24%) for the γ phenotype; and 897 of 2223 (40%) for the δ phenotype. Across all cohorts and trials, 28-day and 365-day mortality were highest among the δ phenotype vs the other 3 phenotypes (P < .001). In simulation models, the proportion of RCTs reporting benefit, harm, or no effect changed considerably (eg, varying the phenotype frequencies within an RCT of early goal-directed therapy changed the results from >33% chance of benefit to >60% chance of harm)., Conclusions and Relevance: In this retrospective analysis of data sets from patients with sepsis, 4 clinical phenotypes were identified that correlated with host-response patterns and clinical outcomes, and simulations suggested these phenotypes may help in understanding heterogeneity of treatment effects. Further research is needed to determine the utility of these phenotypes in clinical care and for informing trial design and interpretation.

Importance: Current approaches to identifying individuals at high risk for opioid overdose target many patients who are not truly at high risk., Objective: To develop and validate a machine-learning algorithm to predict opioid overdose risk among Medicare beneficiaries with at least 1 opioid prescription., Design, Setting, and Participants: A prognostic study was conducted between September 1, 2017, and December 31, 2018. Participants (n = 560 057) included fee-for-service Medicare beneficiaries without cancer who filled 1 or more opioid prescriptions from January 1, 2011, to December 31, 2015. Beneficiaries were randomly and equally divided into training, testing, and validation samples., Exposures: Potential predictors (n = 268), including sociodemographics, health status, patterns of opioid use, and practitioner-level and regional-level factors, were measured in 3-month windows, starting 3 months before initiating opioids until loss of follow-up or the end of observation., Main Outcomes and Measures: Opioid overdose episodes from inpatient and emergency department claims were identified. Multivariate logistic regression (MLR), least absolute shrinkage and selection operator-type regression (LASSO), random forest (RF), gradient boosting machine (GBM), and deep neural network (DNN) were applied to predict overdose risk in the subsequent 3 months after initiation of treatment with prescription opioids. Prediction performance was assessed using the C statistic and other metrics (eg, sensitivity, specificity, and number needed to evaluate [NNE] to identify one overdose). The Youden index was used to identify the optimized threshold of predicted score that balanced sensitivity and specificity., Results: Beneficiaries in the training (n = 186 686), testing (n = 186 685), and validation (n = 186 686) samples had similar characteristics (mean [SD] age of 68.0 [14.5] years, and approximately 63% were female, 82% were white, 35% had disabilities, 41% were dual eligible, and 0.60% had at least 1 overdose episode). In the validation sample, the DNN (C statistic = 0.91; 95% CI, 0.88-0.93) and GBM (C statistic = 0.90; 95% CI, 0.87-0.94) algorithms outperformed the LASSO (C statistic = 0.84; 95% CI, 0.80-0.89), RF (C statistic = 0.80; 95% CI, 0.75-0.84), and MLR (C statistic = 0.75; 95% CI, 0.69-0.80) methods for predicting opioid overdose. At the optimized sensitivity and specificity, DNN had a sensitivity of 92.3%, specificity of 75.7%, NNE of 542, positive predictive value of 0.18%, and negative predictive value of 99.9%. The DNN classified patients into low-risk (76.2% [142 180] of the cohort), medium-risk (18.6% [34 579] of the cohort), and high-risk (5.2% [9747] of the cohort) subgroups, with only 1 in 10 000 in the low-risk subgroup having an overdose episode. More than 90% of overdose episodes occurred in the high-risk and medium-risk subgroups, although positive predictive values were low, given the rare overdose outcome., Conclusions and Relevance: Machine-learning algorithms appear to perform well for risk prediction and stratification of opioid overdose, especially in identifying low-risk subgroups that have minimal risk of overdose.

Background: Motor vehicle crashes (MVCs) are the leading cause of death among teenagers. Little is known about the content of US paediatrician counselling about teen driving., Objective: To examine US paediatrician knowledge, attitudes, and counselling patterns regarding teen driving., Methods: A random sample questionnaire was mailed to American Academy of Pediatrics members in 2009 (n=1606; response=875 (55%)). Analysis was limited to 596 paediatricians who provide adolescent checkups. Questions addressed counselling and attitudes towards roles in promoting safe driving. Logistic regression assessed the relationship between counselling topics and practice characteristics., Results: Most (89%) respondents provide some counselling about driving. Two topics commonly discussed by paediatricians were seatbelts (87%) and alcohol use (82%). Less frequently discussed were: cell phones (47%), speeding (43%), and dangers of transporting teen passengers (41%). Topics rarely discussed were: night driving (21%), graduated driver licensing laws (13%), safe cars (9%), driver education (9%), fatigue (25%), and parental limit setting (23%). Only 10% ever recommend a parent-teen driver agreement. Paediatricians who had a patient injured or killed in an MVC were more likely to discuss night driving (OR=2.86). Physicians caring for a high proportion of adolescents (OR=1.83) or patients with private insurance (OR=1.85) counsel more about the risks of driving with teen passengers., Conclusions: Paediatricians in the USA support counselling on teen driving during routine office visits, but omit many important risk factors. Few recommend parent-teen driver agreements. Methods that help clinicians efficiently and effectively counsel families about teen driving should be developed.

Learning temporal dependencies between variables over continuous time is an important and challenging task. Continuous-time Bayesian networks effectively model such processes but are limited by the number of conditional intensity matrices, which grows exponentially in the number of parents per variable. We develop a partition-based representation using regression trees and forests whose parameter spaces grow linearly in the number of node splits. Using a multiplicative assumption we show how to update the forest likelihood in closed form, producing efficient model updates. Our results show multiplicative forests can be learned from few temporal trajectories with large gains in performance and scalability.

Electronic health records (EHRs) are an emerging relational domain with large potential to improve clinical outcomes. We apply two statistical relational learning (SRL) algorithms to the task of predicting primary myocardial infarction. We show that one SRL algorithm, relational functional gradient boosting, outperforms propositional learners particularly in the medically-relevant high recall region. We observe that both SRL algorithms predict outcomes better than their propositional analogs and suggest how our methods can augment current epidemiological practices.

Systemic administration of chemotherapeutic agents results in indiscriminate drug distribution and severe toxicity. Here we report a technology potentially overcoming these shortcomings through encapsulation and cancer cell-specific targeting of chemotherapeutics in bacterially derived 400 nm minicells. We discovered that minicells can be packaged with therapeutically significant concentrations of chemotherapeutics of differing charge, hydrophobicity, and solubility. Targeting of minicells via bispecific antibodies to receptors on cancer cell membranes results in endocytosis, intracellular degradation, and drug release. This affects highly significant tumor growth inhibition and regression in mouse xenografts and case studies of lymphoma in dogs despite administration of minute amounts of drug and antibody; a factor critical for limiting systemic toxicity that should allow the use of complex regimens of combination chemotherapy.

Background: Geocoding methods vary among spatial epidemiology studies. Errors in the geocoding process and differential match rates may reduce study validity. We compared two geocoding methods using 8,157 Washington State addresses. The multi-stage geocoding method implemented by the state health department used a sequence of local and national reference files. The single-stage method used a single national reference file. For each address geocoded by both methods, we measured the distance between the locations assigned by each method. Area-level characteristics were collected from census data, and modeled as predictors of the discordance between geocoded address coordinates., Results: The multi-stage method had a higher match rate than the single-stage method: 99% versus 95%. Of 7,686 addresses were geocoded by both methods, 96% were geocoded to the same census tract by both methods and 98% were geocoded to locations within 1 km of each other by the two methods. The distance between geocoded coordinates for the same address was higher in sparsely populated and low poverty areas, and counties with local reference files., Conclusion: The multi-stage geocoding method had a higher match rate than the single-stage method. An examination of differences in the location assigned to the same address suggested that study results may be most sensitive to the choice of geocoding method in sparsely populated or low-poverty areas.

Motor vehicle-related injuries to adolescents continue to be of paramount importance to society. Since the original policy statement on the teenaged driver was published in 1996, there have been substantial changes in many state laws and much new research on this topic. There is a need to provide pediatricians with up-to-date information and materials to facilitate appropriate counseling and anticipatory guidance. This statement describes why teenagers are at greater risk of motor vehicle-related injuries, suggests topics suitable for office-based counseling, describes innovative programs, and proposes preventive interventions for pediatricians, parents, legislators, educators, and other child advocates.

Circumstances of injury were abstracted from police reports for 1835 convenience store robberies that occurred during 1992 or 1993 in selected metropolitan areas of seven eastern states. Subset analyses were performed using the data (758 robberies) from four states with relatively complete risk factor information. The purpose of this study was to estimate the risk of injury in a robbery situation for various risk factors. The overall risk of employee robbery-related injury could not be estimated because the probability of robbery is unknown. Of the 1835 robberies, 59% of the total robberies occurred at nighttime (9 p.m. to 3 a.m.), 47% occurred in stores previously robbed in the study period, 63% involved the use of a firearm, and 12% were associated with an injury to at least one employee. In the subset analysis of 758 robberies in four states, the employee probability of injury in a robbery was lower with firearm use compared with no weapon or use of a blunt instrument, and the probability of severe injury (defined as death, or an injury necessitating a trip to a hospital) was lower with a firearm compared with the use of a blunt instrument. However, all five fatalities were firearm-related. Other factors that were associated with a lower probability of employee injury included robbery occurrence in stores that had been robbed multiple times, compared with stores robbed only once; having 1 to 999 dollars stolen, compared with having no money stolen; and the presence of a customer(s) in the store at the time of the robbery. The employee risk of injury was not significantly different between one- (0.106) and multiple-employee (0.111) stores. Similarly, the employee risk of severe injury was not significantly different between one- (0.029) and multiple-employee stores (0.022). We conclude that there are several potential risk factors for employee injury in convenience store robberies, some of which are amenable to interventions. Further research on these factors and their relationship to employee injury is indicated.

The diagnosis of acute suppurative otitis media is not as easy and straightforward as it may seem. Many of the signs and symptoms in children with acute otitis media are also observed in children without it. Furthermore, several of the "classic" findings of acute otitis media, such as fever and earache, are often absent, even in cases confirmed by myringotomy. An otoscope with a fresh bulb and a good power source, as well as a view of the tympanic membrane that is not obstructed by cerumen, are essential to making the diagnosis of acute otitis media. A bulging, cloudy, immobile tympanic membrane is highly associated with otitis media. Erythema of the eardrum alone, however, is often the result of viral infection, crying or attempts to remove cerumen and should not be the sole basis for the diagnosis of acute otitis media. To avoid the common problem of overdiagnosing acute otitis media, the clinician should consider the predictive values of the various symptoms and physical examination findings associated with ear infections.

Rather than using home phototherapy (HP), many pediatricians admit neonates to the hospital for the treatment of unconjugated hyperbilirubinemia. This study investigates the concerns and experiences of pediatricians related to HP use. A questionnaire was sent to 150 pediatricians in the Philadelphia, PA, area and 94 returned questionnaires. Of the responders, 30 of 94 had used HP. Twenty-nine users and 62 of 64 nonusers answered detailed questions regarding HP. Few practitioners follow American Academy of Pediatrics guidelines in selection of HP candidates. Among 29 pediatricians, prematurity (7), Rh incompatibility (13), positive direct Coombs (21), and ABO incompatibility (22) were not considered contraindications, although the American Academy of Pediatrics recommends all as contraindications. Almost all (29/30) HP users identified at least one problem with home treatment. The most commonly reported problems were parental anxiety, blood testing difficulties, and reimbursement issues. Nonusers (62) were asked their reasons for not using HP. The most common replies related to concerns about noncompliance (25), medical-legal issues (23), and administrative difficulties (22), none of which were cited as problems by HP users.

A questionnaire designed to elicit information about the work environment, knowledge, and stresses of pediatric residency program directors was mailed to the 235 member programs of the Association of Pediatric Program Directors (APPD). At the time that the 187 respondents (80% return rate) assumed responsibility for their training programs, many rated their knowledge of various aspects of residency program administration as "poor." The respondents indicated that a lack of time, the pressures of too many other academic responsibilities, and a fear of not "filling" all positions in the National Intern Matching Program created much personal stress. Most program directors felt that educational conferences designed to teach educational methods, and administrative skills, and provide technical information knowledge necessary for residency program supervision would be beneficial.

A pilot study of the levels of stress among residents was conducted in three departments in a university hospital prior to initiating a programme in stress management for residents. The Beck Depression Inventory, which is a brief, standardized self-report measure of depression, was given to residents in anaesthesiology, paediatrics and psychiatry. Six additional questions were asked about the functioning of peers and services residents would like to have available. Of the 113 residents surveyed, 16% were experiencing a mild mood disturbance. The researchers feel this is probably under-reported. Residents felt that about 15% of their colleagues were emotionally impaired; 10% may have a drug and/or alcohol problem; 12% were having marital problems. Eighty per cent of all residents studied said that they would attend support groups if they existed. Approximately 60% thought coping skills/stress management seminars would be useful, and 30% of the paediatric and anaesthesiology residents (60% of the psychiatry residents) said they would use confidential individual psychotherapy if it were available.

Questionnaires were mailed to 116 female physicians who completed their pediatric training within the last 5 years. The questionnaire contained items focusing on (1) maternal complications, (2) health of the infant, and (3) attitudes toward the pregnancy experience. Thirty-five (38%) of the 93 respondents were pregnant during their residency. Seventeen complications occurred in 14 (36%) of the 39 pregnancies. Of the 17 complications, 6 (35%) were serious enough to require hospitalization. Complications resulted in time lost from work in 40% (15/39) of the pregnancies. When asked to describe the experience of being pregnant during residency, only 24% felt that it was "pleasant," while 52% said the experience was "tolerable" and 24% found it "miserable"! Thirty-seven percent (13/35) stated that they would not get pregnant during residency if they had it to do over again.

Physicians who are excessively underconfident or overconfident about their knowledge may have impaired clinical judgement. Confidence weighting of multiple choice examinations asks test-takers to state how confident they are that the answers they selected are correct. This previously described method allows the examinee to receive 'overconfidence' and 'underconfidence' scores. This study was designed to test the hypothesis that these scores would correlate with faculty assessment of pediatric residents' confidence level as observed in the clinical setting. Thirty-three pediatric residents took an examination of general pediatric knowledge using confidence weighting method. Percentage of questions answered correctly ranged from 40 to 81%. There was an association between increasing overconfidence and lower examination scores (r = 0.58; p = 0.001). Increasing overconfidence was also associated with decreasing underconfidence (r = 0.38; p = 0.04). Five faculty members, the program director and the chief resident were asked to rate their perceptions of the residents' confidence on a Likert-type scale. The period of observation ranged from 9 months to 3 years. Linear regression demonstrated an association between underconfidence indices and observed confidence in the clinical setting (r = 0.39; p = 0.03). In addition, three of four residents who left the program had either over- or underconfidence indices greater than one standard deviation from the mean. These results indicate that the multiple choice examination with confidence weighting can predict residents who will be judged as underconfident by clinical preceptors. This finding is important in light of our impression that such house officers often have difficulties later during their residencies.

Of 1,379 children 4 years of age and older who were patients in nine medical centers across the country, 346 (25.1%) were found to be enuretic. Their parents considered heavy sleeping and emotional problems as the main causes of enuresis in children; physical causes rarely were believed to be important. "Waking up the child" from sleep to urinate and "talking about the problem" were the main methods that parents used to treat enuresis. Very few used medication. Parents with a grade school level of education punish bed-wetting children at twice the rate of high school- and college-educated parents. Physicians need to be more aware that enuresis is an important problem for parents and that there are many widely held beliefs about cause and management.

Thirty-five delusional depressed patients were treated for either 28 or 35 days with amitriptyline. The 12 responders could not be differentiated from the nonresponders on a variety of demographic and clinical characteristics. Patients with amitriptyline+nortriptyline plasma levels above 250 ng/ml were significantly more likely to be responders than were patients with levels below that value (p less than .05). A review of the relevant literature revealed that, although some delusional depressives do respond to treatment with tricyclic antidepressants, the presence of delusions is a predictor of poor response to tricyclic antidepressants.

The problem of condyloma acuminata in prepubertal children seems to be more common today than in the past. Transmission of the etiology virus may occur during delivery, from close family contact, or from sexual encounters. Thirty cases of condyloma acuminata in young children have been reported, to which we add four new cases. This article reviews the 34 cases with regard to age and sex, the nature and location of the lesions, and the mode of transmission of the virus. Various modes of therapy are discussed. Because of the possibility that children with condyloma acuminata have been sexually abused, a thorough medical and social evaluation is essential.

If the initial use of amoxicillin fails to cure otitis media, a family will be burdened with medical costs for a second medication, additional office visit fees and time lost from employment. For some families the initial choice of a more expensive but more effective antibiotic may be more cost effective. Using a decision analysis approach we compared the cost effectiveness of amoxicillin and cefaclor. As the amoxicillin efficacy rate decreases or as parental salaries are increased, the cost-effective advantage moves towards cefaclor. With the decision analysis method described, a practitioner can use costs and efficacy rates relevant to his or her practice to determine the most cost-effective initial antibiotic for a child with otitis media.

Questionnaires were used to survey 1,435 parents and 446 physicians in order to determine and compare attitudes and beliefs about enuresis. Although both groups thought that bed-wetting is a maturational problem, the parent group thought emotional causes were important and were less likely to accept small bladder size as an etiology. Parents thought that children should be dry at a much younger age than did the physicians (2.75 vs 5.13 years, respectively). Only 63% of parents thought that medical intervention is a good way to deal with a child's bed-wetting, yet 87% of the physicians suggested medical evaluation. A comparison of the various methods used to stop bed-wetting indicated that parents use waking the child, reassurance and talking with the child, restricting fluids, and punishment significantly more often than physicians. Although many physicians prescribe medication, only 6.6% of the parents thought that medicines are a "very good way" to treat enuresis. When developing a treatment plan for a child with enuresis, the physician should recognize the wide differences between parental and physician attitudes toward this common problem of childhood.

The authors investigated the pharmacological treatment of delusional depression by assigning patients on a random double-blind basis to amitriptyline alone, perphenazine alone, or a combination of the two. Fourteen (78%) of the 18 patients assigned to amitriptyline plus perphenazine were responders, compared with seven (41%) of 17 patients treated with amitriptyline alone and three (19%) of the 16 patients treated with perphenazine alone. The combination of amitriptyline and perphenazine was clearly superior (p less than .01).

This study investigated the pharmacological treatment of delusional depressives by assigning patients on a random double-blind basis to amitriptyline (AT) alone, perphenazine (PER) alone, or AT + PER. As reported in an earlier paper, the combination of AT + PER was the superior treatment with a response rate of 78% compared with 41% for AT alone and 19% for PER alone. This difference is both clinically and statistically significant (p less than 0.01). In this second article, the authors report that the patients on AT + PER had higher plasma levels of AT and its pharmacologically active metabolite nortriptyline (NT) than the patients on AT alone. Although the response rate had a direct relationship to AT + NT plasma levels, an analysis of variance showed that the group of patients treated with AT + PER still had a significantly higher response rate, even after controlling for the plasma levels of AT + NT (p less than 0.05). There was also a correlation between PER plasma levels and clinical response.

The dorsal ventricular ridge is a subcortical structure receiving sensory information from the thalamus in reptiles. In the red-eared turtle, Pseudemys scripta elegans, it contains four cytoarchitectonic areas each characterized by distinct thalamic projections. This is an electron microscopic study of one of these, the dorsal area, which receives its thalamic input from the tectorecipient nucleus rotundus. It contains four concentric zones, internal to the ependymal zone, each of which is distinguished by the distribution of spiny and aspiny neurons. The ependymal zone of dorsal area contains tanycytes whose tails extend into zones 2 and 4. Synapses, usually with asymmetric junctional complexes and round synaptic vesicles, occur on these processes. Zone 1 neurons have fusiform somata and dendrites that parallel the ventricular surface. Their cytoplasm contains rough endoplasmic reticulum located primarily in Nissl bodies, lipofuchsin granules, multivesicular bodies, extensive arrays of Golgi apparatus, and large numbers of mitochondria. Synapses occur mainly on dendritic spines and shafts of zone 1 neurons and less frequently on somata. The majority have round vesicles and asymmetric junctional complexes. In contrast to those in zone 1, neurons in zones 2 and 4 have large amounts of rough endoplasmic reticulum, giving their cytoplasm an electron-dense quality. Synapses occur mainly on spines and shafts of zone 2 and 4 neurons. As in zone 1, the majority have round synaptic vesicles and contain asymmetric junctional complexes. Zones 2 and 4 contain clusters of neurons distributed among isolated neurons. The clusters are larger and less frequent in zone 2. Protoplasmic and fibrous glial processes, axon boutons, dendrites, and axon fascicles surround the neuron clusters. Though less numerous, the same structures also occur inside the clusters. Most synapses inside the clusters have round synaptic vesicles, asymmetric junctional complexes, and occur mainly on spines. Some neurons in clusters have somata whose plasma membranes are in direct apposition. In contrast to dorsal ventricular ridge in snakes, no specialized intercellular contacts were seen between somata in clusters.

This paper describes a new system for resident selection that uses a descriptive rating scale and a microcomputer to create a rank order list for the National Resident Matching Program. The methods utilized to analyze the time efficiency, interrater reliability, and predictive validity of the system are reported. Interrater reliability was found to be high. The rank list was shown to correlate with both clinical performance of Pl-1 and Pl-2 pediatric residents and the American Board of Pediatrics In-training Examination. Correlations for the Pl-2 year were stronger than for the Pl-1 year.