High Satisfaction With a Virtual Assistant for Plastic Surgery Frequently Asked Questions.

BACKGROUND: Most of a surgeon's office time is dedicated to patient education, preventing an appropriate patient-physician relationship. Telephone-accessed artificial intelligent virtual assistants (AIVAs) that simulate a human conversation and answer preoperative frequently asked questions (FAQs) can be effective solutions to this matter. An AIVA capable of answering preoperative plastic surgery-related FAQs has previously been described by the authors. OBJECTIVES: The aim of this paper was to determine patients' perception and satisfaction with an AIVA. METHODS: Twenty-six adult patients from a plastic surgery service answered a 3-part survey consisting of: (1) an evaluation of the answers' correctness, (2) their agreement with the feasibility, usefulness, and future uses of the AIVA, and (3) a section on comments. The first part made it possible to measure the system's accuracy, and the second to evaluate perception and satisfaction. The data were analyzed with Microsoft Excel 2010 (Microsoft Corporation, Redmond, WA). RESULTS: The AIVA correctly answered the patients' questions 98.5% of the time, and the topic with the lowest accuracy was "nausea." Additionally, 88% of patients agreed with the statements of the second part of the survey. Thus, the patients' perception was positive and overall satisfaction with the AIVA was high. Patients agreed the least with using the AIVA to select their surgical procedure. The comments provided improvement areas for subsequent stages of the project. CONCLUSIONS: The results show that patients were satisfied and expressed a positive experience with using the AIVA to answer plastic surgery FAQs before surgery. The system is also highly accurate.

Improving Patient Selection and Prioritization for Hospital at Home Through Predictive Modeling.

Hospital at home is designed to offer patients hospital level care in the comfort of their own home. The process by which clinicians select eligible patients that are clinically and socially appropriate for this model of care requires labor-intensive manual chart reviews. We addressed this problem by providing a predictive model, web application, and data pipeline that produces an eligibility score based on a set of clinical and social factors that influence patients' success in the program. Providers used this predictive model to prioritize the order in which they perform chart reviews and patient screenings. Training performance area under the curve (AUC) was 0.77. Testing 'in production' had an AUC of 0.75. Admission criteria in training rapidly changed over the course of the study due to the novelty of the clinical model. The current algorithm successfully identified many inconsistencies in enrollment and has streamlined the process of patient identification.

Improving the delivery of palliative care through predictive modeling and healthcare informatics.

OBJECTIVE: Access to palliative care (PC) is important for many patients with uncontrolled symptom burden from serious or complex illness. However, many patients who could benefit from PC do not receive it early enough or at all. We sought to address this problem by building a predictive model into a comprehensive clinical framework with the aims to (i) identify in-hospital patients likely to benefit from a PC consult, and (ii) intervene on such patients by contacting their care team. MATERIALS AND METHODS: Electronic health record data for 68 349 inpatient encounters in 2017 at a large hospital were used to train a model to predict the need for PC consult. This model was published as a web service, connected to institutional data pipelines, and consumed by a downstream display application monitored by the PC team. For those patients that the PC team deems appropriate, a team member then contacts the patient's corresponding care team. RESULTS: Training performance AUC based on a 20% holdout validation set was 0.90. The most influential variables were previous palliative care, hospital unit, Albumin, Troponin, and metastatic cancer. The model has been successfully integrated into the clinical workflow making real-time predictions on hundreds of patients per day. The model had an "in-production" AUC of 0.91. A clinical trial is currently underway to assess the effect on clinical outcomes. CONCLUSIONS: A machine learning model can effectively predict the need for an inpatient PC consult and has been successfully integrated into practice to refer new patients to PC.

Deploying Predictive Models In A Healthcare Environment - An Open Source Approach.

Despite dramatic progress in the application of predictive modeling and data mining techniques to problems in modern medicine, a major challenge facing technical practitioners is that of delivering models to clinicians. We have developed an easily implementable framework for publishing predictive models written in R or Python in a way that allows them to be consumed by practically any downstream clinical application, as well as allowing them to be reused in a wide variety of environments without modification. The approach makes models available as web services embedded in containers and uses only open source technology. We provide a template, practical explanation and discussion of involved technologies for a model production framework. We currently use this framework to deliver a model for predicting readmission to hospital following discharge to skilled nursing facilities. The flexibility and simplicity of this methodology will allow it to be readily adopted at a wide variety of institutions. We also provide source code for an example model.

Comparison of Whole-Genome Sequencing Methods for Analysis of Three Methicillin-Resistant Staphylococcus aureus Outbreaks.

Whole-genome sequencing (WGS) can provide excellent resolution in global and local epidemiological investigations of Staphylococcus aureus outbreaks. A variety of sequencing approaches and analytical tools have been used; it is not clear which is ideal. We compared two WGS strategies and two analytical approaches to the standard method of SmaI restriction digestion pulsed-field gel electrophoresis (PFGE) for typing S. aureus Forty-two S. aureus isolates from three outbreaks and 12 reference isolates were studied. Near-complete genomes, assembled de novo with paired-end and long-mate-pair (8 kb) libraries were first assembled and analyzed utilizing an in-house assembly and analytical informatics pipeline. In addition, paired-end data were assembled and analyzed using a commercial software package. Single nucleotide variant (SNP) analysis was performed using the in-house pipeline. Two assembly strategies were used to generate core genome multilocus sequence typing (cgMLST) data. First, the near-complete genome data generated with the in-house pipeline were imported into the commercial software and used to perform cgMLST analysis. Second, the commercial software was used to assemble paired-end data, and resolved assemblies were used to perform cgMLST. Similar isolate clustering was observed using SNP calling and cgMLST, regardless of data assembly strategy. All methods provided more discrimination between outbreaks than did PFGE. Overall, all of the evaluated WGS strategies yielded statistically similar results for S. aureus typing.

VCF-Miner: GUI-based application for mining variants and annotations stored in VCF files.

Next-generation sequencing platforms are widely used to discover variants associated with disease. The processing of sequencing data involves read alignment, variant calling, variant annotation and variant filtering. The standard file format to hold variant calls is the variant call format (VCF) file. According to the format specifications, any arbitrary annotation can be added to the VCF file for downstream processing. However, most downstream analysis programs disregard annotations already present in the VCF and re-annotate variants using the annotation provided by that particular program. This precludes investigators who have collected information on variants from literature or other sources from including these annotations in the filtering and mining of variants. We have developed VCF-Miner, a graphical user interface-based stand-alone tool, to mine variants and annotation stored in the VCF. Powered by a MongoDB database engine, VCF-Miner enables the stepwise trimming of non-relevant variants. The grouping feature implemented in VCF-Miner can be used to identify somatic variants by contrasting variants in tumor and in normal samples or to identify recessive/dominant variants in family studies. It is not limited to human data, but can also be extended to include non-diploid organisms. It also supports copy number or any other variant type supported by the VCF specification. VCF-Miner can be used on a personal computer or large institutional servers and is freely available for download from http://bioinformaticstools.mayo.edu/research/vcf-miner/.

Pertussis outbreak, southeastern Minnesota, 2012.

OBJECTIVE: To describe clinical and laboratory findings from the 2012 southeastern Minnesota pertussis outbreak. PATIENTS AND METHODS: Patients were selected for 2 parts of the study. In the first part, nasopharyngeal swabs from a convenience sample of 265 unique patients were used for both the clinician-requested polymerase chain reaction (PCR) test and culture. B pertussis isolates were tested for macrolide susceptibility and typed using whole genome sequencing and pulsed-field gel electrophoresis. Pertactin gene sequences were analyzed to identify pertactin-deficient B pertussis. In the second part, all patients seen at Mayo Clinic in Rochester, Minnesota, who had PCR results positive for Bordetella pertussis or Bordetella parapertussis between January 1, 2012, and December 31, 2012, were analyzed for patient demographic features and vaccination records. RESULTS: One hundred sixty patients had results positive for B pertussis, and 21 patients had results positive for B parapertussis. Among the 265 swabs cultured, B pertussis was detected by both culture and PCR in 11. One swab was positive for B pertussis by culture alone, and 13 were positive by PCR alone. Polymerase chain reaction detected B pertussis more frequently than did culture (P=.001). No macrolide resistance was detected. All 12 isolates tested had an altered pertactin gene, including 9 with a signal sequence deletion, 2 with insertion sequence disruptions, and 1 with a premature stop codon. Nine and 3 isolates were pertactin types prn1 and prn2, respectively. Whole genome sequencing and pulsed-field gel electrophoresis detected the presence of multiple B pertussis strains. The mean age of patients with pertussis was younger than that of those without pertussis (15.6 and 25.5 years, respectively; P=.002). Compared with those whose test results were negative for B pertussis, fewer patients with positive results had received whole-cell pertussis vaccine (P=.02). In the subgroup who had received acellular vaccine exclusively, the time since the most recent pertussis vaccination in those with results positive for B pertussis was longer than that in those with negative results (1363 vs 1010 days; P=.004). CONCLUSION: The 2012 pertussis outbreak in southeastern Minnesota included multiple strains of B pertussis, all putatively lacking pertactin. Our findings may indicate decreased efficacy of (and waning immunity from) acellular vaccines as contributors to the outbreak.

The Biological Reference Repository (BioR): a rapid and flexible system for genomics annotation.

MOTIVATION: The Biological Reference Repository (BioR) is a toolkit for annotating variants. BioR stores public and user-specific annotation sources in indexed JSON-encoded flat files (catalogs). The BioR toolkit provides the functionality to combine and retrieve annotation from these catalogs via the command-line interface. Several catalogs from commonly used annotation sources and instructions for creating user-specific catalogs are provided. Commands from the toolkit can be combined with other UNIX commands for advanced annotation processing. We also provide instructions for the development of custom annotation pipelines. AVAILABILITY AND IMPLEMENTATION: The package is implemented in Java and makes use of external tools written in Java and Perl. The toolkit can be executed on Mac OS X 10.5 and above or any Linux distribution. The BioR application, quickstart, and user guide documents and many biological examples are available at http://bioinformaticstools.mayo.edu.

The fast changing landscape of sequencing technologies and their impact on microbial genome assemblies and annotation.

BACKGROUND: The emergence of next generation sequencing (NGS) has provided the means for rapid and high throughput sequencing and data generation at low cost, while concomitantly creating a new set of challenges. The number of available assembled microbial genomes continues to grow rapidly and their quality reflects the quality of the sequencing technology used, but also of the analysis software employed for assembly and annotation. METHODOLOGY/PRINCIPAL FINDINGS: In this work, we have explored the quality of the microbial draft genomes across various sequencing technologies. We have compared the draft and finished assemblies of 133 microbial genomes sequenced at the Department of Energy-Joint Genome Institute and finished at the Los Alamos National Laboratory using a variety of combinations of sequencing technologies, reflecting the transition of the institute from Sanger-based sequencing platforms to NGS platforms. The quality of the public assemblies and of the associated gene annotations was evaluated using various metrics. Results obtained with the different sequencing technologies, as well as their effects on downstream processes, were analyzed. Our results demonstrate that the Illumina HiSeq 2000 sequencing system, the primary sequencing technology currently used for de novo genome sequencing and assembly at JGI, has various advantages in terms of total sequence throughput and cost, but it also introduces challenges for the downstream analyses. In all cases assembly results although on average are of high quality, need to be viewed critically and consider sources of errors in them prior to analysis. CONCLUSION: These data follow the evolution of microbial sequencing and downstream processing at the JGI from draft genome sequences with large gaps corresponding to missing genes of significant biological role to assemblies with multiple small gaps (Illumina) and finally to assemblies that generate almost complete genomes (Illumina+PacBio).

Scenario driven data modelling: a method for integrating diverse sources of data and data streams.

BACKGROUND: Biology is rapidly becoming a data intensive, data-driven science. It is essential that data is represented and connected in ways that best represent its full conceptual content and allows both automated integration and data driven decision-making. Recent advancements in distributed multi-relational directed graphs, implemented in the form of the Semantic Web make it possible to deal with complicated heterogeneous data in new and interesting ways. RESULTS: This paper presents a new approach, scenario driven data modelling (SDDM), that integrates multi-relational directed graphs with data streams. SDDM can be applied to virtually any data integration challenge with widely divergent types of data and data streams. In this work, we explored integrating genetics data with reports from traditional media. SDDM was applied to the New Delhi metallo-beta-lactamase gene (NDM-1), an emerging global health threat. The SDDM process constructed a scenario, created a RDF multi-relational directed graph that linked diverse types of data to the Semantic Web, implemented RDF conversion tools (RDFizers) to bring content into the Sematic Web, identified data streams and analytical routines to analyse those streams, and identified user requirements and graph traversals to meet end-user requirements. CONCLUSIONS: We provided an example where SDDM was applied to a complex data integration challenge. The process created a model of the emerging NDM-1 health threat, identified and filled gaps in that model, and constructed reliable software that monitored data streams based on the scenario derived multi-relational directed graph. The SDDM process significantly reduced the software requirements phase by letting the scenario and resulting multi-relational directed graph define what is possible and then set the scope of the user requirements. Approaches like SDDM will be critical to the future of data intensive, data-driven science because they automate the process of converting massive data streams into usable knowledge.

Next generation models for storage and representation of microbial biological annotation.

BACKGROUND: Traditional genome annotation systems were developed in a very different computing era, one where the World Wide Web was just emerging. Consequently, these systems are built as centralized black boxes focused on generating high quality annotation submissions to GenBank/EMBL supported by expert manual curation. The exponential growth of sequence data drives a growing need for increasingly higher quality and automatically generated annotation. Typical annotation pipelines utilize traditional database technologies, clustered computing resources, Perl, C, and UNIX file systems to process raw sequence data, identify genes, and predict and categorize gene function. These technologies tightly couple the annotation software system to hardware and third party software (e.g. relational database systems and schemas). This makes annotation systems hard to reproduce, inflexible to modification over time, difficult to assess, difficult to partition across multiple geographic sites, and difficult to understand for those who are not domain experts. These systems are not readily open to scrutiny and therefore not scientifically tractable. The advent of Semantic Web standards such as Resource Description Framework (RDF) and OWL Web Ontology Language (OWL) enables us to construct systems that address these challenges in a new comprehensive way. RESULTS: Here, we develop a framework for linking traditional data to OWL-based ontologies in genome annotation. We show how data standards can decouple hardware and third party software tools from annotation pipelines, thereby making annotation pipelines easier to reproduce and assess. An illustrative example shows how TURTLE (Terse RDF Triple Language) can be used as a human readable, but also semantically-aware, equivalent to GenBank/EMBL files. CONCLUSIONS: The power of this approach lies in its ability to assemble annotation data from multiple databases across multiple locations into a representation that is understandable to researchers. In this way, all researchers, experimental and computational, will more easily understand the informatics processes constructing genome annotation and ultimately be able to help improve the systems that produce them.