The algorithm journey map: a tangible approach to implementing AI solutions in healthcare

When integrating AI tools in healthcare settings, complex interactions between technologies and primary users are not always fully understood or visible. This deficient and ambiguous understanding hampers attempts by healthcare organizations to adopt AI/ML, and it also creates new challenges for researchers to identify opportunities for simplifying adoption and developing best practices for the use of AI-based solutions. Our study fills this gap by documenting the process of designing, building, and maintaining an AI solution called SepsisWatch at Duke University Health System. We conducted 20 interviews with the team of engineers and scientists that led the multi-year effort to build the tool, integrate it into practice, and maintain the solution. This "Algorithm Journey Map" enumerates all social and technical activities throughout the AI solution's procurement, development, integration, and full lifecycle management. In addition to mapping the "who?" and "what?" of the adoption of the AI tool, we also show several 'lessons learned' throughout the algorithm journey maps including modeling assumptions, stakeholder inclusion, and organizational structure. In doing so, we identify generalizable insights about how to recognize and navigate barriers to AI/ML adoption in healthcare settings. We expect that this effort will further the development of best practices for operationalizing and sustaining ethical principles-in algorithmic systems.

Implementation of a Prospective Index-Cluster Sampling Strategy for the Detection of Presymptomatic Viral Respiratory Infection in Undergraduate Students

Background

Index-cluster studies may help characterize the spread of communicable infections in the presymptomatic state. We describe a prospective index-cluster sampling strategy (ICSS) to detect presymptomatic respiratory viral illness and its implementation in a college population.

Methods

We enrolled an annual cohort of first-year undergraduates who completed daily electronic symptom diaries to identify index cases (ICs) with respiratory illness. Investigators then selected 5-10 potentially exposed, asymptomatic close contacts (CCs) who were geographically co-located to follow for infections. Symptoms and nasopharyngeal samples were collected for 5 days. Logistic regression model-based predictions for proportions of self-reported illness were compared graphically for the whole cohort sampling group and the CC group.

Results

We enrolled 1379 participants between 2009 and 2015, including 288 ICs and 882 CCs. The median number of CCs per IC was 6 (interquartile range, 3-8). Among the 882 CCs, 111 (13%) developed acute respiratory illnesses. Viral etiology testing in 246 ICs (85%) and 719 CCs (82%) identified a pathogen in 57% of ICs and 15% of CCs. Among those with detectable virus, rhinovirus was the most common (IC: 18%; CC: 6%) followed by coxsackievirus/echovirus (IC: 11%; CC: 4%). Among 106 CCs with a detected virus, only 18% had the same virus as their associated IC. Graphically, CCs did not have a higher frequency of self-reported illness relative to the whole cohort sampling group.

Conclusions

Establishing clusters by geographic proximity did not enrich for cases of viral transmission, suggesting that ICSS may be a less effective strategy to detect spread of respiratory infection.

Development and Temporal Validation of a Machine Learning Model to Predict Clinical Deterioration

OBJECTIVES

Early warning scores detecting clinical deterioration in pediatric inpatients have wide-ranging performance and use a limited number of clinical features. This study developed a machine learning model leveraging multiple static and dynamic clinical features from the electronic health record to predict the composite outcome of unplanned transfer to the ICU within 24 hours and inpatient mortality within 48 hours in hospitalized children.

METHODS

Using a retrospective development cohort of 17 630 encounters across 10 388 patients, 2 machine learning models (light gradient boosting machine [LGBM] and random forest) were trained on 542 features and compared with our institutional Pediatric Early Warning Score (I-PEWS).

RESULTS

The LGBM model significantly outperformed I-PEWS based on receiver operating characteristic curve (AUROC) for the composite outcome of ICU transfer or mortality for both internal validation and temporal validation cohorts (AUROC 0.785 95% confidence interval [0.780-0.791] vs 0.708 [0.701-0.715] for temporal validation) as well as lead-time before deterioration events (median 11 hours vs 3 hours; P = .004). However, LGBM performance as evaluated by precision recall curve was lesser in the temporal validation cohort with associated decreased positive predictive value (6% vs 29%) and increased number needed to evaluate (17 vs 3) compared with I-PEWS.

CONCLUSIONS

Our electronic health record based machine learning model demonstrated improved AUROC and lead-time in predicting clinical deterioration in pediatric inpatients 24 to 48 hours in advance compared with I-PEWS. Further work is needed to optimize model positive predictive value to allow for integration into clinical practice.

Pediatric reference intervals for endocrine markers in healthy children and adolescents on the Liaison XL (DiaSorin) immunoassay system

OBJECTIVES

Prominent physiological changes occurring throughout childhood and adolescence necessitate the consideration of age and sex in biomarker interpretation. Critical gaps exist in pediatric reference intervals (RIs) for specialized endocrine markers, despite expected influence of growth and development. The current study aimed to establish and/or verify RIs for six specialized endocrine markers on a specialized immunoassay system.

METHODS

Samples were collected from healthy children and adolescents (5 to <19 years) and apparently healthy outpatients (0 to <5 years) as part of the Canadian Laboratory Initiative on Pediatric Reference Intervals (CALIPER). Serum samples were analysed for aldosterone, renin (plasma), thyroglobulin, anti-thyroglobulin, growth hormone, and insulin-like growth factor-1 (IGF-1) on the Liaison XL (DiaSorin) immunoassay platform. RIs (2.5th and 97.5th percentiles) were established for aldosterone, renin, thyroglobulin, anti-thyroglobulin, and growth hormone. Manufacturer-recommended pediatric RIs for IGF-1 were verified.

RESULTS

Age-specific RIs were established for aldosterone, renin, and thyroglobulin, while no age-specific differences were observed for anti-thyroglobulin or growth hormone. IGF-1 was the only endocrine marker studied that demonstrated significant sex-specific differences. Manufacturer-recommended IGF-1 RIs were verified for children aged 6 to <19 years, while those for children aged 0 to <6 years did not verify.

CONCLUSIONS

This study marks the first time that pediatric RIs for aldosterone and renin were established in the CALIPER cohort and highlights the dynamic changes that occur in water and sodium homeostasis during the first years of life. Overall, these data will assist pediatric clinical laboratories in test result interpretation and improve clinical decision-making for patients tested using Liaison immunoassays.

Defining blood gas analysis stability limits across five sample types

Accurate reporting of blood gas samples is dependent upon following proper preanalytical sample handling requirements though there is variation for sample acceptability criteria across institutions. We examined five common sample types (arterial, venous, umbilical arterial, umbilical venous and capillary) stored at either room temperature or on crushed ice in a time series (0, 15, 30, 45, 60, 90, 180, 240 min) and applied local regulatory and/or institutional allowable performance limits to determine the need for cold preservation and/or maximum stability time for pH, pO₂, pCO₂, glucose, lactate, sodium, potassium, chloride, and ionized calcium where applicable in each sample type. Although changes in sample pO₂ and/or lactate values were responsible, in part or in whole, for surpassing the allowable limits in nearly all sample types analyzed, this was not uniformly observed across sample types within the typical time limits that are referenced in literature. Furthermore, we demonstrated that cold preservation may not ubiquitously provide longer stability for blood gas specimens and this is dependent on the sample type and analyte in question. Nevertheless, these results demonstrate the known instability of pO₂ and lactate and suggest that it may be possible to simplify the monitoring of preanalytical conditions by first evaluating pO₂ and lactate in patient blood gas samples if applicable.

Preliminary results of a clinical research and innovation scholarship to prepare medical students to lead innovations in health care

There is consensus amongst national organizations to integrate health innovation and augmented intelligence (AI) into medical education. However, there is scant evidence to guide policymakers and medical educators working to revise curricula. This study presents academic, operational, and domain understanding outcomes for the first three cohorts of participants in a clinical research and innovation scholarship program.

Changes in ambient air quality and atmospheric composition and reactivity in the South East of the UK as a result of the COVID-19 lockdown

The COVID-19 pandemic forced governments around the world to impose restrictions on daily life to prevent the spread of the virus. This resulted in unprecedented reductions in anthropogenic activity, and reduced emissions of certain air pollutants, namely oxides of nitrogen. The UK 'lockdown' was enforced on 23/03/2020, which led to restrictions on movement, social interaction, and 'non-essential' businesses and services. This study employed an ensemble of measurement and modelling techniques to investigate changes in air quality, atmospheric composition and boundary layer reactivity in the South East of the UK post-lockdown. The techniques employed included in-situ gas- and particle-phase monitoring within central and local authority air quality monitoring networks, remote sensing by long path Differential Optical Absorption Spectroscopy and Sentinel-5P's TROPOMI, and detailed 0-D chemical box modelling. Findings showed that de-trended NO₂ concentrations decreased by an average of 14-38% when compared to the mean of the same period over the preceding 5-years. We found that de-trended particulate matter concentrations had been influenced by interregional pollution episodes, and de-trended ozone concentrations had increased across most sites, by up to 15%, such that total O_x levels were roughly preserved. 0-D chemical box model simulations showed the observed increases in ozone concentrations during lockdown under the hydrocarbon-limited ozone production regime, where total NO_x decreased proportionally greater than total non-methane hydrocarbons, which led to an increase in total hydroxyl, peroxy and organic peroxy radicals. These findings suggest a more complex scenario in terms of changes in air quality owing to the COVID-19 lockdown than originally reported and provide a window into the future to illustrate potential outcomes of policy interventions seeking large-scale NO_x emissions reductions without due consideration of other reactive trace species.

Predicting Hospital Utilization and Inpatient Mortality of Patients Tested for COVID-19

Using structured elements from Electronic Health Records (EHR), we seek to: i) build predictive models to stratify patients tested for COVID-19 by their likelihood for hospitalization, ICU admission, mechanical ventilation and inpatient mortality, and ii) identify the most important EHR-based features driving the predictions. We leveraged EHR data from the Duke University Health System tested for COVID-19 or hospitalized between March 11, 2020 and August 24, 2020, to build models to predict hospital admissions within 4 weeks. Models were also created for ICU admissions, need for mechanical ventilation and mortality following admission. Models were developed on a cohort of 86,355 patients with 112,392 outpatient COVID-19 tests or any-cause hospital admissions between March 11, 2020 and June 4, 2020. The four models considered resulted in AUROC=0.838 (CI: 0.832-0.844) and AP=0.272 (CI: 0.260-0.287) for hospital admissions, AUROC=0.847 (CI: 0.839-855) and AP=0.585 (CI: 0.565-0.603) for ICU admissions, AUROC=0.858 (CI: 0.846-0.871) and AP=0.434 (CI: 0.403-0.467) for mechanical ventilation, and AUROC=0.0.856 (CI: 0.842-0.872) and AP=0.243 (CI: 0.205-0.282) for inpatient mortality. Patient history abstracted from the EHR has the potential for being used to stratify patients tested for COVID-19 in terms of utilization and mortality. The dominant EHR features for hospital admissions and inpatient outcomes are different. For the former, age, social indicators and previous utilization are the most important predictive features. For the latter, age and physiological summaries (pulse and blood pressure) are the main drivers.

Detecting national human enteric disease outbreaks linked to animal contact in the United States of America

Enteric pathogens, such as non-typhoidal Salmonella, Campylobacter and Escherichia coli, can reside in the intestinal tract of many animals, including livestock, companion animals, small mammals and reptiles. Often, these animals can appear healthy; nonetheless, humans can become infected after direct or indirect contact, resulting in a substantial illness burden. An estimated 14% of the 3.2 million illnesses that occur in the United States of America (USA) each year from such enteric pathogens are attributable to animal contact. Surveillance for enteric pathogens in the USA includes the compilation and interpretation of both laboratory and epidemiologic data. However, the authors feel that a collaborative, multisectoral and transdisciplinary - or One Health - approach is needed for data collection and analysis, at every level. In addition, they suggest that the future of enteric illness surveillance lies in the development of improved technologies for pathogen detection and characterisation, such as genomic sequencing and metagenomics. In particular, using whole-genome sequencing to compare genetic sequences of enteric pathogens from humans, food, animals and the environment, can help to predict antimicrobial resistance among these pathogens, determine their genetic relatedness and identify outbreaks linked to a common source. In this paper, the authors describe three recent, multi-state human enteric illness outbreaks linked to animal contact in the USA and discuss how integrated disease surveillance was essential to outbreak detection and response. Additional datasharing between public health and animal health laboratories and epidemiologists at the local, national, regional and international level may help to improve surveillance for emerging animal and human health threats and lead to new opportunities for prevention.

A blood-based host gene expression assay for early detection of respiratory viral infection: an index-cluster prospective cohort study

Background

Early and accurate identification of individuals with viral infections is crucial for clinical management and public health interventions. We aimed to assess the ability of transcriptomic biomarkers to identify naturally acquired respiratory viral infection before typical symptoms are present.

Methods

In this index-cluster study, we prospectively recruited a cohort of undergraduate students (aged 18–25 years) at Duke University (Durham, NC, USA) over a period of 5 academic years. To identify index cases, we monitored students for the entire academic year, for the presence and severity of eight symptoms of respiratory tract infection using a daily web-based survey, with symptoms rated on a scale of 0–4. Index cases were defined as individuals who reported a 6-point increase in cumulative daily symptom score. Suspected index cases were visited by study staff to confirm the presence of reported symptoms of illness and to collect biospecimen samples. We then identified clusters of close contacts of index cases (ie, individuals who lived in close proximity to index cases, close friends, and partners) who were presumed to be at increased risk of developing symptomatic respiratory tract infection while under observation. We monitored each close contact for 5 days for symptoms and viral shedding and measured transcriptomic responses at each timepoint each day using a blood-based 36-gene RT-PCR assay.

Findings

Between Sept 1, 2009, and April 10, 2015, we enrolled 1465 participants. Of 264 index cases with respiratory tract infection symptoms, 150 (57%) had a viral cause confirmed by RT-PCR. Of their 555 close contacts, 106 (19%) developed symptomatic respiratory tract infection with a proven viral cause during the observation window, of whom 60 (57%) had the same virus as their associated index case. Nine viruses were detected in total. The transcriptomic assay accurately predicted viral infection at the time of maximum symptom severity (mean area under the receiver operating characteristic curve [AUROC] 0·94 [95% CI 0·92–0·96]), as well as at 1 day (0·87 [95% CI 0·84–0·90]), 2 days (0·85 [0·82–0·88]), and 3 days (0·74 [0·71–0·77]) before peak illness, when symptoms were minimal or absent and 22 (62%) of 35 individuals, 25 (69%) of 36 individuals, and 24 (82%) of 29 individuals, respectively, had no detectable viral shedding.

Interpretation

Transcriptional biomarkers accurately predict and diagnose infection across diverse viral causes and stages of disease and thus might prove useful for guiding the administration of early effective therapy, quarantine decisions, and other clinical and public health interventions in the setting of endemic and pandemic infectious diseases.

Funding

US Defense Advanced Research Projects Agency.

Real-World Integration of a Sepsis Deep Learning Technology Into Routine Clinical Care: Implementation Study

Background

Successful integrations of machine learning into routine clinical care are exceedingly rare, and barriers to its adoption are poorly characterized in the literature.

Objective

This study aims to report a quality improvement effort to integrate a deep learning sepsis detection and management platform, Sepsis Watch, into routine clinical care.

Methods

In 2016, a multidisciplinary team consisting of statisticians, data scientists, data engineers, and clinicians was assembled by the leadership of an academic health system to radically improve the detection and treatment of sepsis. This report of the quality improvement effort follows the learning health system framework to describe the problem assessment, design, development, implementation, and evaluation plan of Sepsis Watch.

Results

Sepsis Watch was successfully integrated into routine clinical care and reshaped how local machine learning projects are executed. Frontline clinical staff were highly engaged in the design and development of the workflow, machine learning model, and application. Novel machine learning methods were developed to detect sepsis early, and implementation of the model required robust infrastructure. Significant investment was required to align stakeholders, develop trusting relationships, define roles and responsibilities, and to train frontline staff, leading to the establishment of 3 partnerships with internal and external research groups to evaluate Sepsis Watch.

Conclusions

Machine learning models are commonly developed to enhance clinical decision making, but successful integrations of machine learning into routine clinical care are rare. Although there is no playbook for integrating deep learning into clinical care, learnings from the Sepsis Watch integration can inform efforts to develop machine learning technologies at other health care delivery systems.

Machine learning for early detection of sepsis: an internal and temporal validation study

OBJECTIVE

Determine if deep learning detects sepsis earlier and more accurately than other models. To evaluate model performance using implementation-oriented metrics that simulate clinical practice.

MATERIALS AND METHODS

We trained internally and temporally validated a deep learning model (multi-output Gaussian process and recurrent neural network [MGP-RNN]) to detect sepsis using encounters from adult hospitalized patients at a large tertiary academic center. Sepsis was defined as the presence of 2 or more systemic inflammatory response syndrome (SIRS) criteria, a blood culture order, and at least one element of end-organ failure. The training dataset included demographics, comorbidities, vital signs, medication administrations, and labs from October 1, 2014 to December 1, 2015, while the temporal validation dataset was from March 1, 2018 to August 31, 2018. Comparisons were made to 3 machine learning methods, random forest (RF), Cox regression (CR), and penalized logistic regression (PLR), and 3 clinical scores used to detect sepsis, SIRS, quick Sequential Organ Failure Assessment (qSOFA), and National Early Warning Score (NEWS). Traditional discrimination statistics such as the C-statistic as well as metrics aligned with operational implementation were assessed.

RESULTS

The training set and internal validation included 42 979 encounters, while the temporal validation set included 39 786 encounters. The C-statistic for predicting sepsis within 4 h of onset was 0.88 for the MGP-RNN compared to 0.836 for RF, 0.849 for CR, 0.822 for PLR, 0.756 for SIRS, 0.619 for NEWS, and 0.481 for qSOFA. MGP-RNN detected sepsis a median of 5 h in advance. Temporal validation assessment continued to show the MGP-RNN outperform all 7 clinical risk score and machine learning comparisons.

CONCLUSIONS

We developed and validated a novel deep learning model to detect sepsis. Using our data elements and feature set, our modeling approach outperformed other machine learning methods and clinical scores.

Prospective and External Evaluation of a Machine Learning Model to Predict In-Hospital Mortality of Adults at Time of Admission

IMPORTANCE

The ability to accurately predict in-hospital mortality for patients at the time of admission could improve clinical and operational decision-making and outcomes. Few of the machine learning models that have been developed to predict in-hospital death are both broadly applicable to all adult patients across a health system and readily implementable. Similarly, few have been implemented, and none have been evaluated prospectively and externally validated.

OBJECTIVES

To prospectively and externally validate a machine learning model that predicts in-hospital mortality for all adult patients at the time of hospital admission and to design the model using commonly available electronic health record data and accessible computational methods.

DESIGN, SETTING, AND PARTICIPANTS

In this prognostic study, electronic health record data from a total of 43 180 hospitalizations representing 31 003 unique adult patients admitted to a quaternary academic hospital (hospital A) from October 1, 2014, to December 31, 2015, formed a training and validation cohort. The model was further validated in additional cohorts spanning from March 1, 2018, to August 31, 2018, using 16 122 hospitalizations representing 13 094 unique adult patients admitted to hospital A, 6586 hospitalizations representing 5613 unique adult patients admitted to hospital B, and 4086 hospitalizations representing 3428 unique adult patients admitted to hospital C. The model was integrated into the production electronic health record system and prospectively validated on a cohort of 5273 hospitalizations representing 4525 unique adult patients admitted to hospital A between February 14, 2019, and April 15, 2019.

MAIN OUTCOMES AND MEASURES

The main outcome was in-hospital mortality. Model performance was quantified using the area under the receiver operating characteristic curve and area under the precision recall curve.

RESULTS

A total of 75 247 hospital admissions (median [interquartile range] patient age, 59.5 [29.0] years; 45.9% involving male patients) were included in the study. The in-hospital mortality rates for the training validation; retrospective validations at hospitals A, B, and C; and prospective validation cohorts were 3.0%, 2.7%, 1.8%, 2.1%, and 1.6%, respectively. The area under the receiver operating characteristic curves were 0.87 (95% CI, 0.83-0.89), 0.85 (95% CI, 0.83-0.87), 0.89 (95% CI, 0.86-0.92), 0.84 (95% CI, 0.80-0.89), and 0.86 (95% CI, 0.83-0.90), respectively. The area under the precision recall curves were 0.29 (95% CI, 0.25-0.37), 0.17 (95% CI, 0.13-0.22), 0.22 (95% CI, 0.14-0.31), 0.13 (95% CI, 0.08-0.21), and 0.14 (95% CI, 0.09-0.21), respectively.

CONCLUSIONS AND RELEVANCE

Prospective and multisite retrospective evaluations of a machine learning model demonstrated good discrimination of in-hospital mortality for adult patients at the time of admission. The data elements, methods, and patient selection make the model implementable at a system level.

A miRNA Host Response Signature Accurately Discriminates Acute Respiratory Infection Etiologies

Background: Acute respiratory infections (ARIs) are the leading indication for antibacterial prescriptions despite a viral etiology in the majority of cases. The lack of available diagnostics to discriminate viral and bacterial etiologies contributes to this discordance. Recent efforts have focused on the host response as a source for novel diagnostic targets although none have explored the ability of host-derived microRNAs (miRNA) to discriminate between these etiologies. Methods: In this study, we compared host-derived miRNAs and mRNAs from human H3N2 influenza challenge subjects to those from patients with Streptococcus pneumoniae pneumonia. Sparse logistic regression models were used to generate miRNA signatures diagnostic of ARI etiologies. Generalized linear modeling of mRNAs to identify differentially expressed (DE) genes allowed analysis of potential miRNA:mRNA relationships. High likelihood miRNA:mRNA interactions were examined using binding target prediction and negative correlation to further explore potential changes in pathway regulation in response to infection. Results: The resultant miRNA signatures were highly accurate in discriminating ARI etiologies. Mean accuracy was 100% [88.8-100; 95% Confidence Interval (CI)] in discriminating the healthy state from S. pneumoniae pneumonia and 91.3% (72.0-98.9; 95% CI) in discriminating S. pneumoniae pneumonia from influenza infection. Subsequent differential mRNA gene expression analysis revealed alterations in regulatory networks consistent with known biology including immune cell activation and host response to viral infection. Negative correlation network analysis of miRNA:mRNA interactions revealed connections to pathways with known immunobiology such as interferon regulation and MAP kinase signaling. Conclusion: We have developed novel human host-response miRNA signatures for bacterial and viral ARI etiologies. miRNA host response signatures reveal accurate discrimination between S. pneumoniae pneumonia and influenza etiologies for ARI and integrated analyses of the host-pathogen interface are consistent with expected biology. These results highlight the differential miRNA host response to bacterial and viral etiologies of ARI, offering new opportunities to distinguish these entities.

Flea market finds and global exports: Four multistate outbreaks of human Salmonella infections linked to small turtles, United States-2015

Zoonotic transmission of Salmonella infections causes an estimated 11% of salmonellosis annually in the United States. This report describes the epidemiologic, traceback and laboratory investigations conducted in the United States as part of four multistate outbreaks of Salmonella infections linked to small turtles. Salmonella isolates indistinguishable from the outbreak strains were isolated from a total of 143 ill people in the United States, pet turtles, and pond water samples collected from turtle farm A, as well as ill people from Chile and Luxembourg. Almost half (45%) of infections occurred in children aged <5 years, underscoring the importance of the Centers for Disease Control and Prevention recommendation to keep pet turtles and other reptiles out of homes and childcare settings with young children. Although only 43% of the ill people who reported turtle exposure provided purchase information, most small turtles were purchased from flea markets or street vendors, which made it difficult to locate the vendor, trace the turtles to a farm of origin, provide education and enforce the United States federal ban on the sale and distribution of small turtles. These outbreaks highlight the importance of improving public awareness and education about the risk of Salmonella from small turtles not only in the United States but also worldwide.

An outbreak of Escherichia coli O157:H7 infections following a dairy education school field trip in Washington state, 2015

On 27 April 2015, Washington health authorities identified Escherichia coli O157:H7 infections associated with dairy education school field trips held in a barn 20-24 April. Investigation objectives were to determine the magnitude of the outbreak, identify the source of infection, prevent secondary illness transmission and develop recommendations to prevent future outbreaks. Case-finding, hypothesis generating interviews, environmental site visits and a case-control study were conducted. Parents and children were interviewed regarding event activities. Odds ratios (OR) and 95% confidence intervals (CI) were computed. Environmental testing was conducted in the barn; isolates were compared to patient isolates using pulsed-field gel electrophoresis (PFGE). Sixty people were ill, 11 (18%) were hospitalised and six (10%) developed haemolytic uremic syndrome. Ill people ranged in age from <1 year to 47 years (median: 7), and 20 (33%) were female. Twenty-seven case-patients and 88 controls were enrolled in the case-control study. Among first-grade students, handwashing (i.e. soap and water, or hand sanitiser) before lunch was protective (adjusted OR 0.13; 95% CI 0.02-0.88, P = 0.04). Barn samples yielded E. coli O157:H7 with PFGE patterns indistinguishable from patient isolates. This investigation provided epidemiological, laboratory and environmental evidence for a large outbreak of E. coli O157:H7 infections from exposure to a contaminated barn. The investigation highlights the often overlooked risk of infection through exposure to animal environments as well as the importance of handwashing for disease prevention. Increased education and encouragement of infection prevention measures, such as handwashing, can prevent illness.

A community approach to mortality prediction in sepsis via gene expression analysis

Improved risk stratification and prognosis prediction in sepsis is a critical unmet need. Clinical severity scores and available assays such as blood lactate reflect global illness severity with suboptimal performance, and do not specifically reveal the underlying dysregulation of sepsis. Here, we present prognostic models for 30-day mortality generated independently by three scientific groups by using 12 discovery cohorts containing transcriptomic data collected from primarily community-onset sepsis patients. Predictive performance is validated in five cohorts of community-onset sepsis patients in which the models show summary AUROCs ranging from 0.765-0.89. Similar performance is observed in four cohorts of hospital-acquired sepsis. Combining the new gene-expression-based prognostic models with prior clinical severity scores leads to significant improvement in prediction of 30-day mortality as measured via AUROC and net reclassification improvement index These models provide an opportunity to develop molecular bedside tests that may improve risk stratification and mortality prediction in patients with sepsis.

Developing a common framework for evaluating the implementation of genomic medicine interventions in clinical care: the IGNITE Network's Common Measures Working Group

Purpose

Implementation research provides a structure for evaluating the clinical integration of genomic medicine interventions. This paper describes the Implementing GeNomics In PracTicE (IGNITE) Network’s efforts to promote: 1) a broader understanding of genomic medicine implementation research; and 2) the sharing of knowledge generated in the network.

Methods

To facilitate this goal the IGNITE Network Common Measures Working Group (CMG) members adopted the Consolidated Framework for Implementation Research (CFIR) to guide their approach to: identifying constructs and measures relevant to evaluating genomic medicine as a whole, standardizing data collection across projects, and combining data in a centralized resource for cross network analyses.

Results

CMG identified ten high-priority CFIR constructs as important for genomic medicine. Of those, eight didn’t have standardized measurement instruments. Therefore, we developed four survey tools to address this gap. In addition, we identified seven high-priority constructs related to patients, families, and communities that did not map to CFIR constructs. Both sets of constructs were combined to create a draft genomic medicine implementation model.

Conclusion

We developed processes to identify constructs deemed valuable for genomic medicine implementation and codified them in a model. These resources are freely available to facilitate knowledge generation and sharing across the field.

Nasopharyngeal Protein Biomarkers of Acute Respiratory Virus Infection

Infection of respiratory mucosa with viral pathogens triggers complex immunologic events in the affected host. We sought to characterize this response through proteomic analysis of nasopharyngeal lavage in human subjects experimentally challenged with influenza A/H3N2 or human rhinovirus, and to develop targeted assays measuring peptides involved in this host response allowing classification of acute respiratory virus infection. Unbiased proteomic discovery analysis identified 3285 peptides corresponding to 438 unique proteins, and revealed that infection with H3N2 induces significant alterations in protein expression. These include proteins involved in acute inflammatory response, innate immune response, and the complement cascade. These data provide insights into the nature of the biological response to viral infection of the upper respiratory tract, and the proteins that are dysregulated by viral infection form the basis of signature that accurately classifies the infected state. Verification of this signature using targeted mass spectrometry in independent cohorts of subjects challenged with influenza or rhinovirus demonstrates that it performs with high accuracy (0.8623 AUROC, 75% TPR, 97.46% TNR). With further development as a clinical diagnostic, this signature may have utility in rapid screening for emerging infections, avoidance of inappropriate antibacterial therapy, and more rapid implementation of appropriate therapeutic and public health strategies.

Mitochondrial permeability transition pore induction is linked to formation of the complex of ATPase C-subunit, polyhydroxybutyrate and inorganic polyphosphate

Mitochondrial permeability transition pore (mPTP) opening allows free movement of ions and small molecules leading to mitochondrial membrane depolarization and ATP depletion that triggers cell death. A multi-protein complex of the mitochondrial ATP synthase has an essential role in mPTP. However, the molecular identity of the central 'pore' part of mPTP complex is not known. A highly purified fraction of mammalian mitochondria containing C-subunit of ATPase (C-subunit), calcium, inorganic polyphosphate (polyP) and polyhydroxybutyrate (PHB) forms ion channels with properties that resemble the native mPTP. We demonstrate here that amount of this channel-forming complex dramatically increases in intact mitochondria during mPTP activation. This increase is inhibited by both Cyclosporine A, an inhibitor of mPTP and Ruthenium Red, an inhibitor of the Mitochondrial Calcium Uniporter. Similar increases in the amount of complex formation occurs in areas of mouse brain damaged by ischemia-reperfusion injury. These findings suggest that calcium-induced mPTP is associated with de novo assembly of a channel comprising C-subunit, polyP and PHB.

Host gene expression classifiers diagnose acute respiratory illness etiology

Acute respiratory infections caused by bacterial or viral pathogens are among the most common reasons for seeking medical care. Despite improvements in pathogen-based diagnostics, most patients receive inappropriate antibiotics. Host response biomarkers offer an alternative diagnostic approach to direct antimicrobial use. This observational cohort study determined whether host gene expression patterns discriminate noninfectious from infectious illness and bacterial from viral causes of acute respiratory infection in the acute care setting. Peripheral whole blood gene expression from 273 subjects with community-onset acute respiratory infection (ARI) or noninfectious illness, as well as 44 healthy controls, was measured using microarrays. Sparse logistic regression was used to develop classifiers for bacterial ARI (71 probes), viral ARI (33 probes), or a noninfectious cause of illness (26 probes). Overall accuracy was 87% (238 of 273 concordant with clinical adjudication), which was more accurate than procalcitonin (78%, P < 0.03) and three published classifiers of bacterial versus viral infection (78 to 83%). The classifiers developed here externally validated in five publicly available data sets (AUC, 0.90 to 0.99). A sixth publicly available data set included 25 patients with co-identification of bacterial and viral pathogens. Applying the ARI classifiers defined four distinct groups: a host response to bacterial ARI, viral ARI, coinfection, and neither a bacterial nor a viral response. These findings create an opportunity to develop and use host gene expression classifiers as diagnostic platforms to combat inappropriate antibiotic use and emerging antibiotic resistance.

A Genomic Signature of Influenza Infection Shows Potential for Presymptomatic Detection, Guiding Early Therapy, and Monitoring Clinical Responses

Early, presymptomatic intervention with oseltamivir (corresponding to the onset of a published host-based genomic signature of influenza infection) resulted in decreased overall influenza symptoms (aggregate symptom scores of 23.5 vs 46.3), more rapid resolution of clinical disease (20 hours earlier), reduced viral shedding (total median tissue culture infectious dose [TCID₅₀] 7.4 vs 9.7), and significantly reduced expression of several inflammatory cytokines (interferon-γ, tumor necrosis factor-α, interleukin-6, and others). The host genomic response to influenza infection is robust and may provide the means for early detection, more timely therapeutic interventions, a meaningful reduction in clinical disease, and an effective molecular means to track response to therapy.

Synergistic neuroprotection by epicatechin and quercetin: Activation of convergent mitochondrial signaling pathways

In view of evidence that increased consumption of epicatechin (E) and quercetin (Q) may reduce the risk of stroke, we have measured the effects of combining E and Q on mitochondrial function and neuronal survival following oxygen-glucose deprivation (OGD). Relative to mouse cortical neuron cultures pretreated (24h) with either E or Q (0.1-10μM), E+Q synergistically attenuated OGD-induced neuronal cell death. E, Q and E+Q (0.3μM) increased spare respiratory capacity but only E+Q (0.3μM) preserved this crucial parameter of neuronal mitochondrial function after OGD. These improvements were accompanied by corresponding increases in cyclic AMP response element binding protein (CREB) phosphorylation and the expression of CREB-target genes that promote neuronal survival (Bcl-2) and mitochondrial biogenesis (PGC-1α). Consistent with these findings, E+Q (0.1 and 1.0μM) elevated mitochondrial gene expression (MT-ND2 and MT-ATP6) to a greater extent than E or Q after OGD. Q (0.3-3.0μM), but not E (3.0μM), elevated cytosolic calcium (Ca(2+)) spikes and the mitochondrial membrane potential. Conversely, E and E+Q (0.1 and 0.3μM), but not Q (0.1 and 0.3μM), activated protein kinase B (Akt). Nitric oxide synthase (NOS) inhibition with L-N(G)-nitroarginine methyl ester (1.0μM) blocked neuroprotection by E (0.3μM) or Q (1.0μM). Oral administration of E+Q (75mg/kg; once daily for 5days) reduced hypoxic-ischemic brain injury. These findings suggest E and Q activate Akt- and Ca(2+)-mediated signaling pathways that converge on NOS and CREB resulting in synergistic improvements in neuronal mitochondrial performance which confer profound protection against ischemic injury.

Comparing reference-based RNA-Seq mapping methods for non-human primate data

Background

The application of next-generation sequencing technology to gene expression quantification analysis, namely, RNA-Sequencing, has transformed the way in which gene expression studies are conducted and analyzed. These advances are of particular interest to researchers studying organisms with missing or incomplete genomes, as the need for knowledge of sequence information is overcome. De novo assembly methods have gained widespread acceptance in the RNA-Seq community for organisms with no true reference genome or transcriptome. While such methods have tremendous utility, computational cost is still a significant challenge for organisms with large and complex genomes.

Results

In this manuscript, we present a comparison of four reference-based mapping methods for non-human primate data. We utilize TopHat2 and GSNAP for mapping to the human genome, and Bowtie2 and Stampy for mapping to the human genome and transcriptome for a total of six mapping approaches. For each of these methods, we explore mapping rates and locations, number of detected genes, correlations between computed expression values, and the utility of the resulting data for differential expression analysis.

Conclusions

We show that reference-based mapping methods indeed have utility in RNA-Seq analysis of mammalian data with no true reference, and the details of mapping methods should be carefully considered when doing so. Critical algorithm features include short seed sequences, the allowance of mismatches, and the allowance of gapped alignments in addition to splice junction gaps. Such features facilitate sensitive alignment of non-human primate RNA-Seq data to a human reference.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-570) contains supplementary material, which is available to authorized users.

Associations between obesogenic risk and depressive symptomatology in Australian adolescents: a cross-sectional study

BACKGROUND

Depression and obesity are significant health concerns currently facing adolescents worldwide. This paper investigates the associations between obesity and related risk behaviours and depressive symptomatology in an Australian adolescent population.

METHODS

Data from the Australian Capital Territory It's Your Move project, an Australian community-based intervention project were used. In 2012, 800 students (440 females, 360 males) aged 11-14 years (M=13.11 years, SD=0.62 years), from 6 secondary schools were weighed and measured and completed a questionnaire which included physical activity, sedentary behaviour and dietary intake. Weight status was defined by WHO criteria. A cut-off score ≥10 on the Short Mood and Feelings Questionnaire indicated symptomatic depression. Logistic regression was used to test associations.

RESULTS

After controlling for potential confounders, results showed significantly higher odds of depressive symptomatology in males (OR=1.22, p<0.05) and females (OR=1.12, p<0.05) who exceeded guidelines for daily screen-time leisure sedentary activities. Higher odds of depressive symptoms were seen in females who consumed greater amounts of sweet drink (OR=1.18, p<0.05), compared to lower female consumers of sweet drinks, and males who were overweight/obese also had greater odds of depressive symptoms (OR=1.83, p<0.05) compared to male normal weight adolescents.

CONCLUSIONS

This study demonstrates the associations between obesogenic risks and depression in adolescents. Further research should explore the direction of these associations and identify common determinants of obesity and depression. Mental health outcomes need to be included in the rationale and evaluation for diet and activity interventions.

Inconsistent associations between sweet drink intake and 2-year change in BMI among Victorian children and adolescents

OBJECTIVE

The aim of this study was to examine whether baseline (T1) or 2-year change in sweet drink intake in children and adolescents was associated with age- and gender-standardized body mass index (BMIz) at time two (T2), 2 years later.

METHODS

Data on 1465 children and adolescents from the comparison groups of two quasi-experimental intervention studies from Victoria, Australia were analysed. At two time points between 2003 and 2008 (mean interval: 2.2 years) height and weight were measured and sweet drink consumption (soft drink and fruit juice/cordial) was assessed.

RESULTS

No association was observed between T1 sweet drink intake and BMIz at T2 among children or adolescents. Children from higher socioeconomic status families who reported an increased intake of sweet drinks at T2 compared with T1 had higher mean BMIz at T2 (β: 0.13, P = 0.05). There was no evidence of a dose-response relationship between sweet drink intake and BMIz. In supplementary analyses, we observed that more frequent usual consumption of fruit juice/cordial was associated with a higher BMIz at T2 among children.

CONCLUSION

This study showed limited evidence of an association between sweet drink intake and BMIz. However, the association is complex and may be confounded by both dietary and activity behaviours.

The 15° face-changing acetabular component for treatment of osteoarthritis secondary to developmental dysplasia of the hip

We report the use of a 15° face-changing cementless acetabular component in patients undergoing total hip replacement for osteoarthritis secondary to developmental dysplasia of the hip. The rationale behind its design and the surgical technique used for its implantation are described. It is distinctly different from a standard cementless hemispherical component as it is designed to position the bearing surface at the optimal angle of inclination, that is, < 45°, while maximising the cover of the component by host bone.

Obesity prevention in the family day care setting: impact of the Romp & Chomp intervention on opportunities for children's physical activity and healthy eating

BACKGROUND

The Romp & Chomp intervention reduced the prevalence of overweight/obesity in pre-school children in Geelong, Victoria, Australia through an intervention promoting healthy eating and active play in early childhood settings. This study aims to determine if the intervention successfully created more health promoting family day care (FDC) environments.

METHODS

The evaluation had a cross-sectional, quasi-experimental design with the intervention FDC service in Geelong and a comparison sample from 17 FDC services across Victoria. A 45-item questionnaire capturing nutrition- and physical activity-related aspects of the policy, socio-cultural and physical environments of the FDC service was completed by FDC care providers (in 2008) in the intervention (n= 28) and comparison (n= 223) samples.

RESULTS

Select results showed intervention children spent less time in screen-based activities (P= 0.03), organized active play (P < 0.001) and free inside play (P= 0.03) than comparison children. There were more rules related to healthy eating (P < 0.001), more care provider practices that supported children's positive meal experiences (P < 0.001), fewer unhealthy food items allowed (P= 0.05), higher odds of staff being trained in nutrition (P= 0.04) and physical activity (P < 0.001), lower odds of having set minimum times for outside (P < 0.001) and organized (P= 0.01) active play, and of rewarding children with food (P < 0.001).

CONCLUSIONS

Romp & Chomp improved the FDC service to one that discourages sedentary behaviours and promotes opportunities for children to eat nutritious foods. Ongoing investment to increase children's physical activity within the setting and improving the capacity and health literacy of care providers is required to extend and sustain the improvements.

Spatial accessibility and availability measures and statistical properties in the food environment

Spatial accessibility is of increasing interest in the health sciences. This paper addresses the statistical use of spatial accessibility and availability indices. These measures are evaluated via an extensive simulation based on cluster models for local food outlet density. We derived Monte Carlo critical values for several statistical tests based on the indices. In particular we are interested in the ability to make inferential comparisons between different study areas where indices of accessibility and availability are to be calculated. We derive tests of mean difference as well as tests for differences in Moran's I for spatial correlation for each of the accessibility and availability indices. We also apply these new statistical tests to a data example based on two counties in South Carolina for various accessibility and availability measures calculated for food outlets, stores, and restaurants.