A Data-Driven Approach to Predicting Septic Shock in the Intensive Care Unit

Early diagnosis of sepsis and septic shock has been unambiguously linked to lower mortality and better patient outcomes. Despite this, there is a strong unmet need for a reliable clinical tool that can be used for large-scale automated screening to identify high-risk patients. We addressed the following questions: Can a novel algorithm to identify patients at high risk of septic shock 24 hours before diagnosis be discovered using available clinical data? What are performance characteristics of this predictive algorithm? Can current metrics for evaluation of sepsis be improved using novel algorithm? Publicly available data from the intensive care unit setting was used to build septic shock and control patient cohorts. Using Bayesian networks, causal relationships between diagnosis groups, procedure groups, laboratory results, and demographic data were inferred. Predictive model for septic shock 24 hours prior to digital diagnosis was built based on inferred causal networks. Sepsis risk scores were augmented by de novo inferred model and performance was evaluated. A novel predictive model to identify high-risk patients 24 hours ahead of time, with area under curve of 0.81, negative predictive value of 0.87, and a positive predictive value as high as 0.65 was built. The specificity of quick sequential organ failure assessment, systemic inflammatory response syndrome, and modified early warning score was improved when augmented with the novel model, whereas no improvements were made to the sequential organ failure assessment score. We used a data-driven, expert knowledge agnostic method to build a screening algorithm for early detection of septic shock. The model demonstrates strong performance in the data set used and provides a basis for expanding this work toward building an algorithm that is used to screen patients based on electronic medical record data in real time.

A Genome Model to Explain Major Features of Neurodevelopmental Disorders in Newborns

The purpose of this study was to test the hypothesis that infections are linked to chromosomal anomalies that cause neurodevelopmental disorders. In children with disorders in the development of their nervous systems, chromosome anomalies known to cause these disorders were compared with foreign DNAs, including known teratogens. Genes essential for neurons, lymphatic drainage, immunity, circulation, angiogenesis, cell barriers, structure, epigenetic and chromatin modifications were all found close together in polyfunctional clusters that were deleted or rearranged in neurodevelopmental disorders. In some patients, epigenetic driver mutations also changed access to large chromosome segments. These changes account for immune, circulatory, and structural deficits that accompany neurologic deficits. Specific and repetitive human DNA encompassing large deletions matched infections and passed rigorous artifact tests. Deletions of up to millions of bases accompanied infection-matching sequences and caused massive changes in human homologies to foreign DNAs. In data from 3 independent studies of private, familial, and recurrent chromosomal rearrangements, massive changes in homologous microbiomes were found and may drive rearrangements and encourage pathogens. At least 1 chromosomal anomaly was found to consist of human DNA fragments with a gap that corresponded to a piece of integrated foreign DNA. Microbial DNAs that match repetitive or specific human DNA segments are thus proposed to interfere with the epigenome and highly active recombination during meiosis, driven by massive changes in human DNA-foreign DNA homologies. Abnormal recombination in gametes produces zygotes containing rare chromosome anomalies that cause neurologic disorders and nonneurologic signs. Neurodevelopmental disorders may be examples of assault on the human genome by foreign DNAs at a critical stage. Some infections may be more likely tolerated because they resemble human DNA segments. Even rare developmental disorders can be screened for homology to infections within altered epigenomes and chromatin structures. Considering effects of foreign DNAs can assist prenatal and genetic counseling, diagnosis, prevention, and early intervention.

LEP: A Statistical Method Integrating Individual-Level and Summary-Level Data of the Same Trait From Different Populations

Statistical approaches for integrating multiple data sets in genome-wide association studies (GWASs) are increasingly important. Proper utilization of more relevant information is expected to improve statistical efficiency in the analysis. Among these approaches, LEP was proposed for joint analysis of individual-level data and summary-level data in the same population by leveraging pleiotropy. The key idea of LEP is to explore correlation of the association status among different data sets while accounting for the heterogeneity. In this commentary, we show that LEP is applicable to integrate individual-level data and summary-level data of the same trait from different populations, providing new insights into the genetic architecture of different populations.

Proceedings from the Digital Innovation in Mental Health Conference, London, 2017

Aims and Scope: The conference aims were two-fold: (1) to explore how digital technology is implemented into personalized and/or group mental health interventions and (2) to promote digital equality through developing culturally sensitive ways of bringing technological innovation to disadvantaged groups. A broad scope of perspectives were welcomed and encouraged, from lived experience, academic, clinical, media, the arts, policy-making, tech innovation, and other perspectives.