The effect of neighborhood and individual characteristics on pediatric critical illness.

The relationship between neighborhood/individual characteristics and pediatric intensive care unit (PICU) outcomes is largely unexplored. We hypothesized that individual-level racial/ethnic minority status and neighborhood-level low socioeconomic status and minority concentration would adversely affect children's severity of illness on admission to the PICU. Retrospective analyses (1/1/2007-5/23/2011) of clinical, geographic, and demographic data were conducted at an academic, tertiary children's hospital PICU. Clinical data included age, diagnosis, insurance, race/ethnicity, Pediatric Index of Mortality 2 score on presentation to the PICU (PIM2), and mortality. Residential addresses were geocoded and linked with 2010 US Census tract data using geographic information systems geocoding techniques. Repeated measures models to predict PIM2 and mortality were constructed using three successive models with theorized covariates including the patient's race/ethnicity, the predominant neighborhood racial/ethnic group, interactions between patient race/ethnicity and neighborhood race/ethnicity, neighborhood socioeconomic status, and insurance type. Of the 5,390 children, 57.8% were Latino and 70.1% possessed government insurance. Latino children (ß = 0.31; p < 0.01), especially Latino children living in a Latino ethnic enclave (ß = 1.13; p < 0.05), had higher PIM2 scores compared with non-Latinos. Children with government insurance (ß = 0.29; p < 0.01) had higher PIM2 scores compared to children with other payment types and median neighborhood income was inversely associated with PIM2 scores (ß = -0.04 per $10,000/year of income; p < 0.05). Lower median neighborhood income, Latino ethnicity, Latino children living in a predominantly Latino neighborhood, and children possessing government insurance were associated with a higher severity of illness on PICU admission. The reasons why these factors affect critical illness severity require further exploration.

Hemilaminectomy for thoracolumbar Hansen Type I intervertebral disk disease in ambulatory dogs with or without neurologic deficits: 39 cases (2008-2010).

OBJECTIVE: To describe prognostic factors, outcome, and time to recovery among ambulatory dogs having hemilaminectomy for Hansen Type I intervertebral disk disease. STUDY DESIGN: Retrospective case series. ANIMALS: Dogs (n = 38; 39 hemilaminectomies). METHODS: Medical records (January 2008-May 2010) on all dogs that had hemilaminectomy for Hansen Type I intervertebral disk disease were reviewed. Records for dogs that were ambulatory preoperatively were analyzed for signalment, duration and severity of signs, presence of neurologic deficits, and postoperative outcome. Dogs were categorized based on Frankel score and subcategorized by their level of conscious proprioceptive (CP) deficit. Postoperatively, time to ambulation and to regain normal CP responses was recorded. Results for each group were compared using a χ(2) test and considered significant when P < .05. Recovery times were analyzed using a Cox proportional hazards model. RESULTS: Seven dogs were categorized as modified Frankel grade I preoperatively and 32 dogs as grade II with varying levels of deficits (1 of these dogs had previously been operated as grade II and was reoperated again as grade II). Increasing degree of CP deficit preoperatively was significantly correlated with longer time to ambulation (P = .005) as well as longer time to CP normal (P = .01). Duration of signs was not significantly correlated with time to ambulation or neurologic recovery for either grade I or II dogs. CONCLUSIONS: Most dogs recovered well with surgical decompression. Increasing degree of deficits preoperatively is significantly correlated with longer recovery time.

Multitask learning and benchmarking with clinical time series data.

Health care is one of the most exciting frontiers in data mining and machine learning. Successful adoption of electronic health records (EHRs) created an explosion in digital clinical data available for analysis, but progress in machine learning for healthcare research has been difficult to measure because of the absence of publicly available benchmark data sets. To address this problem, we propose four clinical prediction benchmarks using data derived from the publicly available Medical Information Mart for Intensive Care (MIMIC-III) database. These tasks cover a range of clinical problems including modeling risk of mortality, forecasting length of stay, detecting physiologic decline, and phenotype classification. We propose strong linear and neural baselines for all four tasks and evaluate the effect of deep supervision, multitask training and data-specific architectural modifications on the performance of neural models.

Author Correction: Do no harm: a roadmap for responsible machine learning for health care.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Do no harm: a roadmap for responsible machine learning for health care.

Interest in machine-learning applications within medicine has been growing, but few studies have progressed to deployment in patient care. We present a framework, context and ultimately guidelines for accelerating the translation of machine-learning-based interventions in health care. To be successful, translation will require a team of engaged stakeholders and a systematic process from beginning (problem formulation) to end (widespread deployment).

Causal Phenotype Discovery via Deep Networks.

The rapid growth of digital health databases has attracted many researchers interested in using modern computational methods to discover and model patterns of health and illness in a research program known as computational phenotyping. Much of the work in this area has focused on traditional statistical learning paradigms, such as classification, prediction, clustering, pattern mining. In this paper, we propose a related but different paradigm called causal phenotype discovery, which aims to discover latent representations of illness that are causally predictive. We illustrate this idea with a two-stage framework that combines the latent representation learning power of deep neural networks with state-of-the-art tools from causal inference. We apply this framework to two large ICU time series data sets and show that it can learn features that are predictively useful, that capture complex physiologic patterns associated with critical illnesses, and that are potentially more clinically meaningful than manually designed features.

Sim•TwentyFive: an interactive visualization system for data-driven decision support.

Clinicians at the bedside are increasingly overwhelmed by an inundation of information and must rely largely on pattern recognition and professional experience to comprehend complex clinical data and treat their patients in a timely manner. Traditional decision support systems are based on rules and predictive models and often fail to take advantage of increasingly large digital clinical data stores available in real-time. We propose an alternative approach to delivering data-driven decision support based on an interactive system for exploring and visualizing a context of physiologically similar patients from a database. Here we present Sim•TwentyFive, a highly flexible, responsive, intuitive prototype with a comprehensive set of interaction techniques that effectively reduces the cognitive burden of querying, exploring, analyzing and comparing similar past patient episodes. Quantitative performance tests and anonymous summative evaluations from PICU physicians indicated that Sim•TwentyFive is an efficient, intuitive and clinically-useful tool.