TCCC Decision Support With Machine Learning Prediction of Hemorrhage Risk, Shock Probability.

INTRODUCTION: Expected future delays in evacuation during near-peer conflicts in remote locales are expected to require extended care including prolonged field care over hours to days. Such delays can increase potential complications, such as insufficient blood flow (shock), bloodstream infection (sepsis), internal bleeding (hemorrhage), and require more complex treatment beyond stabilization. The Trauma Triage Treatment and Training Decision Support (4TDS) system is a real-time decision support system to monitor casualty health and identify such complications. The 4TDS software prototype operates on an Android smart phone or tablet configured for use in the DoD Nett Warrior program. It includes machine learning models to evaluate trends in six vital signs streamed from a sensor placed on a casualty to identify shock probability, internal hemorrhage risk, and need for a massive transfusion. MATERIALS AND METHODS: The project team used a mixed methods approach to create and evaluate the system including literature review, rapid prototyping, design requirements review, agile development, an algorithm "silent test," and usability assessments with novice to expert medics from all three services. RESULTS: Both models, shock (showing an accuracy of 0.83) and hemorrhage/massive transfusion protocol, were successfully validated using externally collected data. All usability assessment participants completed refresher training scenarios and were able to accurately assess a simulated casualty's condition using the phone prototype. Mean responses to statements on evaluation criteria [e.g., fit with Tactical Combat Casualty Care (TCCC), ease of use, and decision confidence] fell at five or above on a 7-point scale, indicating strong support. CONCLUSIONS: Participatory design ensured 4TDS and machine learning models reflect medic and clinician mental models and work processes and built support among potential users should the system transition to operational use. Validation results can support 4TDS readiness for FDA 510k clearance as a Class II medical device.

Validation of a Machine Learning Model for Early Shock Detection.

OBJECTIVES: The objectives of this study were to test in real time a Trauma Triage, Treatment, and Training Decision Support (4TDS) machine learning (ML) model of shock detection in a prospective silent trial, and to evaluate specificity, sensitivity, and other estimates of diagnostic performance compared to the gold standard of electronic medical records (EMRs) review. DESIGN: We performed a single-center diagnostic performance study. PATIENTS AND SETTING: A prospective cohort consisted of consecutive patients aged 18 years and older who were admitted from May 1 through September 30, 2020 to six Mayo Clinic intensive care units (ICUs) and five progressive care units. MEASUREMENTS AND MAIN RESULTS: During the study time, 5,384 out of 6,630 hospital admissions were eligible. During the same period, the 4TDS shock model sent 825 alerts and 632 were eligible. Among 632 hospital admissions with alerts, 287 were screened positive and 345 were negative. Among 4,752 hospital admissions without alerts, 78 were screened positive and 4,674 were negative. The area under the receiver operating characteristics curve for the 4TDS shock model was 0.86 (95% CI 0.85-0.87%). The 4TDS shock model demonstrated a sensitivity of 78.6% (95% CI 74.1-82.7%) and a specificity of 93.1% (95% CI 92.4-93.8%). The model showed a positive predictive value of 45.4% (95% CI 42.6-48.3%) and a negative predictive value of 98.4% (95% CI 98-98.6%). CONCLUSIONS: We successfully validated an ML model to detect circulatory shock in a prospective observational study. The model used only vital signs and showed moderate performance compared to the gold standard of clinician EMR review when applied to an ICU patient cohort.

Genome-wide analysis of 944 133 individuals provides insights into the etiology of haemorrhoidal disease.

OBJECTIVE: Haemorrhoidal disease (HEM) affects a large and silently suffering fraction of the population but its aetiology, including suspected genetic predisposition, is poorly understood. We report the first genome-wide association study (GWAS) meta-analysis to identify genetic risk factors for HEM to date. DESIGN: We conducted a GWAS meta-analysis of 218 920 patients with HEM and 725 213 controls of European ancestry. Using GWAS summary statistics, we performed multiple genetic correlation analyses between HEM and other traits as well as calculated HEM polygenic risk scores (PRS) and evaluated their translational potential in independent datasets. Using functional annotation of GWAS results, we identified HEM candidate genes, which differential expression and coexpression in HEM tissues were evaluated employing RNA-seq analyses. The localisation of expressed proteins at selected loci was investigated by immunohistochemistry. RESULTS: We demonstrate modest heritability and genetic correlation of HEM with several other diseases from the GI, neuroaffective and cardiovascular domains. HEM PRS validated in 180 435 individuals from independent datasets allowed the identification of those at risk and correlated with younger age of onset and recurrent surgery. We identified 102 independent HEM risk loci harbouring genes whose expression is enriched in blood vessels and GI tissues, and in pathways associated with smooth muscles, epithelial and endothelial development and morphogenesis. Network transcriptomic analyses highlighted HEM gene coexpression modules that are relevant to the development and integrity of the musculoskeletal and epidermal systems, and the organisation of the extracellular matrix. CONCLUSION: HEM has a genetic component that predisposes to smooth muscle, epithelial and connective tissue dysfunction.

Decision Support for Tactical Combat Casualty Care Using Machine Learning to Detect Shock.

INTRODUCTION: The emergence of more complex Prolonged Field Care in austere settings and the need to assist inexperienced providers' ability to treat patients create an urgent need for effective tools to support care. We report on a project to develop a phone-/tablet-based decision support system for prehospital tactical combat casualty care that collects physiologic and other clinical data and uses machine learning to detect and differentiate shock manifestation. MATERIALS AND METHODS: Software interface development methods included literature review, rapid prototyping, and subject matter expert design requirements reviews. Machine learning algorithm methods included development of a model trained on publicly available Medical Information Mart for Intensive Care data, then on de-identified data from Mayo Clinic Intensive Care Unit. RESULTS: The project team interviewed 17 Army, Air Force, and Navy medical subject matter experts during design requirements review sessions. They had an average of 17 years of service in military medicine and an average of 4 deployments apiece and all had performed tactical combat casualty care on live patients during deployment. Comments provided requirements for shock identification and management in prehospital settings, including support for indication of shock probability and shock differentiation. The machine learning algorithm based on logistic regression performed best among other algorithms we tested and was able to predict shock onset 90 minutes before it occurred with better than 75% accuracy in the test dataset. CONCLUSIONS: We expect the Trauma Triage, Treatment, and Training Decision Support system will augment a medic's ability to make informed decisions based on salient patient data and to diagnose multiple types of shock through remotely trained, field deployed ML models.

Preparing for an Influenza Pandemic: Hospital Acceptance Study of Filtering Facepiece Respirator Decontamination Using Ultraviolet Germicidal Irradiation.

OBJECTIVES: Predictions estimate supplies of filtering facepiece respirators (FFRs) would be limited in the event of a severe influenza pandemic. Ultraviolet decontamination and reuse (UVDR) is a potential approach to mitigate an FFR shortage. A field study sought to understand healthcare workers' perspectives and potential logistics issues related to implementation of UVDR methods for FFRs in hospitals. METHODS: Data were collected at three hospitals using a structured guide to conduct 19 individual interviews, 103 focus group interviews, and 285 individual surveys. Data were then evaluated using thematic analysis to reveal key themes. RESULTS: Data revealed noteworthy variation in FFR use across the sample, along with preferences and requirements for the use of UVDR, unit design, and FFR reuse. Based on a scale of 1 (low) to 10 (high), the mean perception of safety in a high mortality pandemic wearing no FFR was 1.25 of 10, wearing an FFR for an extended period without decontamination was 4.20 of 10, and using UVDR was 7.72 of 10. CONCLUSIONS: In addition to technical design and development, preparation and training will be essential to successful implementation of a UVDR program. Ultraviolet decontamination and reuse program design and implementation must account for actual clinical practice, compliance with regulations, and practical financial considerations to be successfully adopted so that it can mitigate potential FFR shortages in a pandemic.

Improving Clinician Decisions and Communication in Critical Care Using Novel Information Technology.

INTRODUCTION: The electronic medical record (EMR) is presumed to support clinician decisions by documenting and retrieving patient information. Research shows that the EMR variably affects patient care and clinical decision making. The way information is presented likely has a significant impact on this variability. Well-designed representations of salient information can make a task easier by integrating information in useful patterns that clinicians use to make improved clinical judgments and decisions. Using Cognitive Systems Engineering methods, our research team developed a novel health information technology (NHIT) that interfaces with the EMR to display salient clinical information and enabled communication with a dedicated text-messaging feature. The software allows clinicians to customize displays according to their role and information needs. Here we present results of usability and validation assessments of the NHIT. MATERIALS AND METHODS: Our subjects were physicians, nurses, respiratory therapists, and physician trainees. Two arms of this study were conducted, a usability assessment and then a validation assessment. The usability assessment was a computer-based simulation using deceased patient data. After a brief five-minute orientation, the usability assessment measured individual clinician performance of typical tasks in two clinical scenarios using the NHIT. The clinical scenarios included patient admission to the unit and patient readiness for surgery. We evaluated clinician perspective about the NHIT after completing tasks using 7-point Likert scale surveys. In the usability assessment, the primary outcome was participant perceptions about the system's ease of use compared to the legacy system.A subsequent cross-over, validation assessment compared performance of two clinical teams during simulated care scenarios: one using only the legacy IT system and one using the NHIT in addition to the legacy IT system. We oriented both teams to the NHIT during a 1-hour session on the night before the first scenario. Scenarios were conducted using high-fidelity simulation in a real burn intensive care unit room. We used observations, task completion times, semi-structured interviews, and surveys to compare user decisions and perceptions about their performance. The primary outcome for the validation assessment was time to reach accurate (correct) decision points. RESULTS: During the usability assessment, clinicians were able to complete all tasks requested. Clinicians reported the NHIT was easier to use and the novel information display allowed for easier data interpretation compared to subject recollection of the legacy EMR.In the validation assessment, a more junior team of clinicians using the NHIT arrived at accurate diagnoses and decision points at similar times as a more experienced team. Both teams noted improved communication between team members when using the NHIT and overall rated the NHIT as easier to use than the legacy EMR, especially with respect to finding information. CONCLUSIONS: The primary findings of these assessments are that clinicians found the NHIT easy to use despite minimal training and experience and that it did not degrade clinician efficiency or decision-making accuracy. These findings are in contrast to common user experiences when introduced to new EMRs in clinical practice.

Recurrent chromosomal and epigenetic alterations in oral squamous cell carcinoma and its putative premalignant condition oral lichen planus.

Head and neck squamous cell carcinoma (HNSCC) affects about 700.000 individuals per year worldwide with oral squamous cell carcinoma (OSCC) as a major subcategory. Despite a comprehensive treatment concept including surgery, radiation, and chemotherapy the 5-year survival rate is still only about 50 percent. Chronic inflammation is one of the hallmarks of carcinogenesis. Until now, little is known about the premalignant status of oral lichen planus (OLP) and molecular alterations in OLP are still poorly characterized. Our study aims to delineate differential DNA methylation patterns in OLP, OSCC, and normal oral mucosa. By applying a bead chip approach, we identified altered chromosomal patterns characteristic for OSCC while finding no recurrent alterations in OLP. In contrast, we identified numerous alterations in the DNA methylation pattern in OLP, as compared to normal controls, that were also present in OSCC. Our data support the hypothesis that OLP is a precursor lesion of OSCC sharing multiple epigenetic alterations with OSCC.

Aberrant DNA methylation of ADAMTS16 in colorectal and other epithelial cancers.

BACKGROUND: ADAMs (a disintegrin and metalloproteinase) have long been associated with tumor progression. Recent findings indicate that members of the closely related ADAMTS (ADAMs with thrombospondin motifs) family are also critically involved in carcinogenesis. Gene silencing through DNA methylation at CpG loci around e.g. transcription start or enhancer sites is a major mechanism in cancer development. Here, we aimed at identifying genes of the ADAM and ADAMTS family showing altered DNA methylation in the development or colorectal cancer (CRC) and other epithelial tumors. METHODS: We investigated potential changes of DNA methylation affecting ADAM and ADAMTS genes in 117 CRC, 40 lung cancer (LC) and 15 oral squamous-cell carcinoma (SCC) samples. Tumor tissue was analyzed in comparison to adjacent non-malignant tissue of the same patients. The methylation status of 1145 CpGs in 51 ADAM and ADAMTS genes was measured with the HumanMethylation450 BeadChip Array. ADAMTS16 protein expression was analyzed in CRC samples by immunohistochemistry. RESULTS: In CRC, we identified 72 CpGs in 18 genes which were significantly affected by hyper- or hypomethylation in the tumor tissue compared to the adjacent non-malignant tissue. While notable/frequent alterations in methylation patterns within ADAM genes were not observed, conspicuous changes were found in ADAMTS16 and ADAMTS2. To figure out whether these differences would be CRC specific, additional LC and SCC tissue samples were analyzed. Overall, 78 differentially methylated CpGs were found in LC and 29 in SCC. Strikingly, 8 CpGs located in the ADAMTS16 gene were commonly differentially methylated in all three cancer entities. Six CpGs in the promoter region were hypermethylated, whereas 2 CpGs in the gene body were hypomethylated indicative of gene silencing. In line with these findings, ADAMTS16 protein was strongly expressed in globlet cells and colonocytes in control tissue but not in CRC samples. Functional in vitro studies using the colorectal carcinoma cell line HT29 revealed that ADAMTS16 expression restrained tumor cell proliferation. CONCLUSIONS: We identified ADAMTS16 as novel gene with cancer-specific promoter hypermethylation in CRC, LC and SCC patients implicating ADAMTS16 as potential biomarker for these tumors. Moreover, our results provide evidence that ADAMTS16 may have tumor suppressor properties.

Comparing the Workload Perceptions of Identifying Patient Condition and Priorities of Care Among Burn Providers in Three Burn ICUs.

Multidisciplinary rounds (MDRs) in the burn intensive care unit serve as an efficient means for clinicians to assess patient status and establish patient care priorities. Both tasks require significant cognitive work, the magnitude of which is relevant because increased cognitive work of task completion has been associated with increased error rates. We sought to quantify this workload during MDR using the National Aeronautics and Space Administration Task Load Index (NASA-TLX). Research staff at three academic regional referral burn centers administered the NASA-TLX to clinicians during MDR. Clinicians assessed their workload associated with 1) "Identify(ing) if the patient is better, same, or worse than yesterday" and 2) "Identify(ing) the most important objectives of care for the patient today." Data were collected on clinician type, years of experience, and hours of direct patient care. Surveys were administered to 116 total clinicians, 41 physicians, 25 nurses, 13 medical students, and 37 clinicians in other roles. Clinicians with less experience reported more cognitive work when completing both tasks (P < .005). Clinicians in the "others" group (respiratory therapists, dieticians, pharmacists, etc.) reported less cognitive work than all other groups for both tasks (P < .05). The NASA-TLX was an effective tool for collecting perceptions of cognitive workload associated with MDR. Perceived cognitive work varied by clinician type and experience level when completing two key tasks. Less experience was associated with increased perceived work, potentially increasing mental error rates, and increasing risk to patients. Creating tools or work processes to reduce cognitive work may improve clinician performance.

Developing a Cognitive and Communications Tool for Burn Intensive Care Unit Clinicians.

BACKGROUND: Burn Intensive Care Unit (BICU) work is necessarily complex and depends on clinician actions, resources, and variable patient responses to interventions. Clinicians use large volumes of data that are condensed in time, but separated across resources, to care for patients. Correctly designed health information technology (IT) systems may help clinicians to treat these patients more efficiently, accurately, and reliably. We report on a 3-year project to design and develop an ecologically valid IT system for use in a military BICU. METHODS: We use a mixed methods Cognitive Systems Engineering approach for research and development. Observations, interviews, artifact analysis, survey, and thematic analysis methods were used to reveal underlying factors that mold the work environment and affect clinician decisions that may affect patient outcomes. Participatory design and prototyping methods have been used to develop solutions. RESULTS: We developed 39 requirements for the IT system and used them to create three use cases to help developers better understand how the system might support clinician work to develop interface prototypes. We also incorporated data mining functions that offer the potential to aid clinicians by recognizing patterns recognition of clinically significant events, such as incipient sepsis. The gaps between information sources and accurate, reliable, and efficient clinical decision that we have identified will enable us to create scenarios to evaluate prototype systems with BICU clinicians, to develop increasingly improved designs, and to measure outcomes. CONCLUSION: The link from data to analyses, requirements, prototypes, and their evaluation ensures that the solution will reflect and support work in the BICU as it actually occurs, improving staff efficiency and patient care quality.

The mixed blessings of smart infusion devices and health care IT.

From July to October 2009, a team of human factors researchers evaluated the use of a commercially available infusion device among nurses at a tertiary care hospital in the Midwest. The study's purpose was to determine the factors that may influence the adoption and "best practice" use of smart infusion devices by identifying the human, technological, environmental, and/or organizational factors and to describe how they support or impede safe practices. The study's aim was to show how technology and individual and team behavior influence each other, as well as care performance and outcomes. Research team members shadowed nursing personnel as they performed routine care activities, and conducted cognitive task analysis interviews with nurses, an engineer, and a pharmacist. They identified key themes, and then made several systematic passes through the data to identify all instances of each theme and to collect examples and illustrative quotes. Although staff members were positive in their comments about the smart pump, observations and interviews revealed discrepancies between prescriptions and infusions, and "workarounds" to cope with the mismatch between interface design and actual care requirements. Despite "smart pump" capabilities, situations continue such as the need for clinicians to perform calculations in order to deliver medications. These workarounds, which make them and patients vulnerable to adverse outcomes, confirm prior published research by Cook, Nemeth, Nunnally, Hollnagel, and Woods. The team provided recommendations based on findings for training and interface design.

Between choice and chance: the role of human factors in acute care equipment decisions.

OBJECTIVES: We report on a human factors evaluation project at a major urban teaching hospital that was intended to use human factors methods to assist the selection of a new infusion device among 4 commercially available models. METHODS: The project provided an expert evaluation of the pumps, collected data on programming each pump by a sample of practitioners, tabulated recent adverse event reports in the US Food and Drug Administration Manufacturer and User Device Experience database, and observed actual use in intensive care and hematology/oncology units. RESULTS: Programming by clinicians showed no correlation between clinical experience and ability to program any of the pumps under consideration. Field observations reflected diverse use patterns across services that required ease of use pumps did not offer. Upon review of a final candidate pump, purchasing preferences superceded clinical considerations. CONCLUSIONS: Equipment and systems that are intended for use by clinicians must necessarily reflect an understanding of actual clinical practice to be well suited for use at the sharp (operator) end. However, purchase decisions for medical equipment including infusion devices are typically made by hospital staff members who are experienced in administrative and clinical matters but have no expertise in the evaluation of complex equipment. This project demonstrates how collaboration by human factors and clinical professionals can inform equipment decisions and assist clinician performance to improve patient safety. It also reveals how technical decisions that directly influence anesthesia staff performance and patient safety are subject to organizational factors such as social and political pressure.

Creating resilient IT: how the sign-out sheet shows clinicians make healthcare work.

Information technology (IT) systems have been described as brittle and prone to automation surprises. Recent reports of information system failure, particularly computerized physician order entry (CPOE) systems, shows the result of such IT failure in actual practice. Such mismatches with healthcare work requirements require improvement to IT research and development. Resilience is the feature of some systems that makes it possible for them to respond to sudden, unanticipated demands for performance and return to normal operation quickly, with minimum decrement in performance. Workers create resilience at healthcare's sharp end by daily confronting constraints and obstacles that need to be surmounted in order to accomplish results. The resident sign-out sheet is described as an example of resilience. Efforts to develop successful IT systems for healthcare's sharp end must incorporate properties that reflect work domain traits, as the sign-out sheet shows.

Mapping cognitive work: the way out of healthcare IT system failures.

The failure of automation to improve clinical performance is likely rooted in the design concepts on which IT systems are based. Current systems provide clinicians with specific direction about how to care for individual patients. This is much like the specific, detailed, complicated, and narrow trip route driving directions that can be obtained from various web sites. Daily healthcare work rarely has the certainty that makes such directions useful. Rather than directions, useful healthcare automation is likely to have characteristics of a map. Clinicians could use its depictions of available routes, obstacles, and distances between the current and goal locations in order to choose routes and to track progress toward goals. Such representations are likely to be quite different than those currently incorporated in healthcare automation. We demonstrate the concept of creating maps and using constraints as the basis for the design of healthcare automation.