A Hybrid Physical-Digital Simulation Laboratory to Expedite Context-Aware Design and Usability Testing in Digital Health

Fit within existing physical and digitalised workflows is a critical aspect of digital health software usability. Early, iterative exploration of contextual usability issues is complicated by barriers of access to healthcare settings. The Validitron SimLab is a new facility for digital health prototyping that augments immersive, realistic physical environments with a digital sandbox allowing new and existing software to be easily set up and tested in the physical space.

Clinical Experiences of Working with Patient-Generated Health Data in Primary Care

This research aims to provide insight into the GP experience with patient-generated health data (PGHD) in a virtual care visit. Despite the prevalence of wearables, including smartwatches, the acceptability of generated data in primary care is understudied. The result of this study from mixed-method analysis showed the basic capabilities of PGHD to enhance clinical decision-making and positive impact on collaboration with the patient. The impact of PGHD on clinician satisfaction was not determined, highlighting the importance of rigorous methodology in future research.

Adapting an Environmental Scan for 'Insights Reporting': Learnings from an Online Brain Cancer Peer Support Platform

When developing a digital health solution, product owners, healthcare professionals, researchers, IT teams, and consumers require timely, accurate contextual information to inform solution development. Insights Reporting can rapidly draw together information from literature, end users and existing technology to inform the development process. This was the case when creating an online brain cancer peer support platform where solution development was conducted in parallel with contextual information synthesis. This paper discusses the novel adaptation of an environmental scan methodology using codesign and multiple layers of qualitative rigor, to create Insights Reporting. This seven-step process can be completed in two months and results in salient points of knowledge that can rapidly inform the design of a solution, creating a shared understanding of a digital health phenomenon. Project members noted that Insights Reporting surfaces previously inaccessible knowledge, catalyzes decision-making and allows all stakeholders to influence the report agenda, affirming principles of digital health equity.

Virtual emergency care in Victoria: Stakeholder perspectives of strengths, weaknesses, and barriers and facilitators of service scale-up

BACKGROUND

Virtual emergency services have been proposed as an alternative service model to conventional in-person emergency department attendance.

METHODS

Twenty participants were interviewed: 10 emergency medicine physicians, 4 health care consumers, and 6 other health care professionals. Conventional content analysis was performed on the interview transcriptions to identify perceived strengths and weaknesses of the VED, and barriers and facilitators to scaling-up the VED.

RESULTS

VEDs are perceived as a convenient approach to provide and receive emergency care while ensuring safety and quality of care, however some patients may still need to attend the ED in person for physical assessments. There is currently a lack of evidence, guidelines, and resources to support their implementation. Most of the potential and existing barriers and facilitators for scaling-up the VED were related to their effectiveness, reach and adoption. Broader public health contextual factors were viewed as barriers, while potential actions to address resources and costs could be facilitators.

CONCLUSIONS

VEDs were viewed as a convenient service model to provide care, can not replace all in-person visits. Current policies and guidelines are insufficient for wider implementation. Most of the barriers and facilitators for its scaling-up were related to VED effectiveness and delivery.

Identifying Public Healthcare Priorities in Virtual Care for Older Adults: A Participatory Research Study

There has been increasing adoption and implementation of virtual healthcare in recent years, especially with COVID-19 impacting the world. As a result, virtual care initiatives may not undergo stringent quality control processes to ensure that they are appropriate to their context and meet sector needs. The two objectives of this study were to identify virtual care initiatives for older adults currently in use in Victoria and virtual care challenges that could be prioritised for further investigation and scale-up and to understand why certain virtual care initiatives and challenges are prioritised over others for investigation and scale-up.

METHODS

This project used an Emerging Design approach. A survey of public health services in the state of Victoria in Australia was first carried out, followed by the co-production of research and healthcare priorities with key stakeholders in the areas of primary care, hospital care, consumer representation, research, and government. The survey was used to gather existing virtual care initiatives for older adults and any associated challenges. Co-production processes consisted of individual ratings of initiatives and group-based discussions to identify priority virtual care initiatives and challenges to be addressed for future scale-up. Stakeholders nominated their top three virtual initiatives following discussions.

RESULTS

Telehealth was nominated as the highest priority initiative type for scaling up, with virtual emergency department models of care nominated as the highest priority within this category. Remote monitoring was voted as a top priority for further investigations. The top virtual care challenge was data sharing across services and settings, and the user-friendliness of virtual care platforms was nominated as the top priority for further investigation.

CONCLUSIONS

Stakeholders prioritised public health virtual care initiatives that are easy to adopt and address needs that are perceived to be more immediate (acute more so than chronic care). Virtual care initiatives that incorporate more technology and integrated elements are valued, but more information is needed to inform their potential scale-up.

Virtual Care Initiatives for Older Adults in Australia: Scoping Review

BACKGROUND

There has been a rapid shift toward the adoption of virtual health care services in Australia. It is unknown how widely virtual care has been implemented or evaluated for the care of older adults in Australia.

OBJECTIVE

We aimed to review the literature evaluating virtual care initiatives for older adults across a wide range of health conditions and modalities and identify key challenges and opportunities for wider adoption at both patient and system levels in Australia.

METHODS

A scoping review of the literature was conducted. We searched MEDLINE, Embase, PsycINFO, CINAHL, AgeLine, and gray literature (January 1, 2011, to March 8, 2021) to identify virtual care initiatives for older Australians (aged ≥65 years). The results were reported according to the World Health Organization's digital health evaluation framework.

RESULTS

Among the 6296 documents in the search results, we identified 94 that reported 80 unique virtual care initiatives. Most (69/80, 89%) were at the pilot stage and targeted community-dwelling older adults (64/79, 81%) with chronic diseases (52/80, 65%). The modes of delivery included videoconference, telephone, apps, device or monitoring systems, and web-based technologies. Most initiatives showed either similar or better health and behavioral outcomes compared with in-person care. The key barriers for wider adoption were physical, cognitive, or sensory impairment in older adults and staffing issues, legislative issues, and a lack of motivation among providers.

CONCLUSIONS

Virtual care is a viable model of care to address a wide range of health conditions among older adults in Australia. More embedded and integrative evaluations are needed to ensure that virtually enabled care can be used more widely by older Australians and health care providers.

Fostering the use of Learning Health Systems through a fellowship program for interprofessional clinicians

Introduction

Methods

Results

Conclusion

The Rapid Development of Virtual Care Tools in Response to COVID-19: Case Studies in Three Australian Health Services

BACKGROUND

News of the impact of COVID-19 around the world delivered a brief opportunity for Australian health services to plan new ways of delivering care to large numbers of people while maintaining staff safety through greater physical separation. The rapid pivot to telemedicine and virtual care provided immediate and longer term benefits; however, such rapid-cycle development also created risks.

OBJECTIVE

The aim of this study was to understand the sociotechnical aspects of the rapid-cycle development of seven different COVID-19 virtual care tools, and to identify enablers, barriers, and risks at three health services in Victoria, Australia.

METHODS

A qualitative, embedded, multiple case study design was adopted. Researchers from three health services collaborated with university researchers who were independent from those health services to gather and analyze structured interview data from key people involved in either clinical or technical aspects of designing and deploying seven different virtual care tools.

RESULTS

The overall objectives of each health service reflected the international requirements for managing large numbers of patients safely but remotely and for protecting staff. However, the governance, digital maturity, and specific use cases at each institution shaped the methodology and specific outcomes required. Dependence on key individuals and their domain knowledge within an existing governance framework generally enabled rapid deployment, but sometimes posed barriers. Existing relationships with technical service developers enabled strong solutions, which in some cases were highly scalable. Conventional project methodologies such as steering committees, scope, budget control, tight functional specification, consumer engagement and codesign, universal accessibility, and postimplementation evaluation were ignored almost universally in this environment.

CONCLUSIONS

These three health services took a variety of approaches to the rapid-cycle development of virtual care tools to meet their urgent needs for triaging and remote monitoring during the first year of the COVID-19 pandemic. Their experiences provided insights into many social and technical barriers and enablers to the development of virtual care tools. If these are addressed proactively, they will improve clinical governance and technical management of future virtual care. Some changes can be made within individual health services, while others entail health system policy reforms. Enhancing the environment for virtual care tool design and implementation now will yield returns not only during future health emergencies but also in many more routine care settings.

Resources for Improving Pediatric Readiness and Quality of Care in Rural Communities and Emergency Departments

OBJECTIVES

To share the process and products of an 8-year, federally funded grant from the Health Resources and Services Administration Emergency Medical Services for Children program to increase pediatric emergency readiness and quality of care provided in rural communities located within 2 underserved local emergency medical services agencies (LEMSAs) in Northern California.

METHODS

In 2 multicounty LEMSAs with 24 receiving hospital emergency departments, we conducted focus groups and interviews with patients and parents, first responders, receiving hospital personnel, and other community stakeholders. From this, we (a regional, urban children's hospital) provided a variety of resources for improving the regionalization and quality of pediatric emergency care provided by prehospital providers and healthcare staff at receiving hospitals in these rural LEMSAs.

RESULTS

From this project, we provided resources that included regularly scheduled pediatric-specific training and education programs, pediatric-specific quality improvement initiatives, expansion of telemedicine services, and cultural competency training. We also enhanced community engagement and investment in pediatric readiness.

CONCLUSIONS

The resources we provided from our regional, urban children's hospital to 2 rural LEMSAs facilitated improvements in a regionalized system of care for critically ill and injured children. Our shared resources framework can be adapted by other regional children's hospitals to increase readiness and quality of pediatric emergency care in rural and underserved communities and LEMSAs.

Implementability of healthcare interventions: an overview of reviews and development of a conceptual framework

Background

Implementation research may play an important role in reducing research waste by identifying strategies that support translation of evidence into practice. Implementation of healthcare interventions is influenced by multiple factors including the organisational context, implementation strategies and features of the intervention as perceived by people delivering and receiving the intervention. Recently, concepts relating to perceived features of interventions have been gaining traction in published literature, namely, acceptability, fidelity, feasibility, scalability and sustainability. These concepts may influence uptake of healthcare interventions, yet there seems to be little consensus about their nature and impact. The aim of this paper is to develop a testable conceptual framework of implementability of healthcare interventions that includes these five concepts.

Methods

A multifaceted approach was used to develop and refine a conceptual framework of implementability of healthcare interventions. An overview of reviews identified reviews published between January 2000 and March 2021 that focused on at least one of the five concepts in relation to a healthcare intervention. These findings informed the development of a preliminary framework of implementability of healthcare interventions which was presented to a panel of experts. A nominal group process was used to critique, refine and agree on a final framework.

Results

A total of 252 publications were included in the overview of reviews. Of these, 32% were found to be feasible, 4% reported sustainable changes in practice and 9% were scaled up to other populations and/or settings. The expert panel proposed that scalability and sustainability of a healthcare intervention are dependent on its acceptability, fidelity and feasibility. Furthermore, acceptability, fidelity and feasibility require re-evaluation over time and as the intervention is developed and then implemented in different settings or with different populations. The final agreed framework of implementability provides the basis for a chronological, iterative approach to planning for wide-scale, long-term implementation of healthcare interventions.

Conclusions

We recommend that researchers consider the factors acceptability, fidelity and feasibility (proposed to influence sustainability and scalability) during the preliminary phases of intervention development, evaluation and implementation, and iteratively check these factors in different settings and over time.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13012-021-01171-7.

Design and evaluation of a Women in American Medical Informatics Association (AMIA) leadership program

The objective is to report on the design and evaluation of the inaugural Women in AMIA Leadership Program. A year-long leadership curriculum was developed. Survey responses were summarized with descriptive statistics and quotes selected. Twenty-four scholars participated in the program. There was a significant increase in perceived achievement of learning objectives after the program (P < .0001). The largest improvement was in leadership confidence and presence in work interactions (modal answer Neutral in presurvey from 21 responses rose to Agree in postsurvey from 24 responses). Most (92% of 13) scholars clarified leadership vision and goals and (83% of 18) would be Very Likely to recommend the program to others. The goals of the program-developing women's leader identity, increasing networks, and accumulating experience for future programs-were achieved. The second leadership program is on its way in the United States and Australia. This study may benefit organizations seeking to develop leadership programs for women in informatics and digital health.

Examining the association of social risk with heart failure readmission in the Veterans Health Administration

BACKGROUND

Previous research has found that social risk factors are associated with an increased risk of 30-day readmission. We aimed to assess the association of 5 social risk factors (living alone, lack of social support, marginal housing, substance abuse, and low income) with 30-day Heart Failure (HF) hospital readmissions within the Veterans Health Affairs (VA) and the impact of their inclusion on hospital readmission model performance.

METHODS

We performed a retrospective cohort study using chart review and VA and Centers for Medicare and Medicaid Services (CMS) administrative data from a random sample of 1,500 elderly (≥ 65 years) Veterans hospitalized for HF in 2012. Using logistic regression, we examined whether any of the social risk factors were associated with 30-day readmission after adjusting for age alone and clinical variables used by CMS in its 30-day risk stratified readmission model. The impact of these five social risk factors on readmission model performance was assessed by comparing c-statistics, likelihood ratio tests, and the Hosmer-Lemeshow goodness-of-fit statistic.

RESULTS

The prevalence varied among the 5 risk factors; low income (47 % vs. 47 %), lives alone (18 % vs. 19 %), substance abuse (14 % vs. 16 %), lacks social support (2 % vs. <1 %), and marginal housing (< 1 % vs. 3 %) among readmitted and non-readmitted patients, respectively. Controlling for clinical factors contained in CMS readmission models, a lack of social support was found to be associated with an increased risk of 30-day readmission (OR 4.8, 95 %CI 1.35-17.88), while marginal housing was noted to decrease readmission risk (OR 0.21, 95 %CI 0.03-0.87). Living alone (OR: 0.9, 95 %CI 0.64-1.26), substance abuse (OR 0.91, 95 %CI 0.67-1.22), and having low income (OR 1.01, 95 %CI 0.77-1.31) had no association with HF readmissions. Adding the five social risk factors to a CMS-based model (age and comorbid conditions; c-statistic 0.62) did not improve model performance (c-statistic: 0.62).

CONCLUSIONS

While a lack of social support was associated with 30-day readmission in the VA, its prevalence was low. Moreover, the inclusion of some social risk factors did not improve readmission model performance. In an integrated healthcare system like the VA, social risk factors may have a limited effect on 30-day readmission outcomes.

Adaptation of an NLP system to a new healthcare environment to identify social determinants of health

Social determinants of health (SDoH) are increasingly important factors for population health, healthcare outcomes, and care delivery. However, many of these factors are not reliably captured within structured electronic health record (EHR) data. In this work, we evaluated and adapted a previously published NLP tool to include additional social risk factors for deployment at Vanderbilt University Medical Center in an Acute Myocardial Infarction cohort. We developed a transformation of the SDoH outputs of the tool into the OMOP common data model (CDM) for re-use across many potential use cases, yielding performance measures across 8 SDoH classes of precision 0.83 recall 0.74 and F-measure of 0.78.

Examining the Interfacility Variation of Social Determinants of Health in the Veterans Health Administration

Introduction

Recently, numerous studies have linked social determinants of health (SDoH) with clinical outcomes. While this association is well known, the interfacility variability of these risk favors within the Veterans Health Administration (VHA) is not known. Such information could be useful to the VHA for resource and funding allocation. The aim of this study is to explore the interfacility variability of 5 SDoH within the VHA.

Methods

In a cohort of patients (aged ≥ 65 years) hospitalized at VHA acute care facilities with either acute myocardial infarction (AMI), heart failure (HF), or pneumonia in 2012, we assessed (1) the proportion of patients with any of the following five documented SDoH: lives alone, marginal housing, alcohol use disorder, substance use disorder, and use of substance use services, using administrative diagnosis codes and clinic stop codes; and (2) the documented facility-level variability of these SDoH. To examine whether variability was due to regional coding differences, we assessed the variation of living alone using a validated natural language processing (NLP) algorithm.

Results

The proportion of veterans admitted for AMI, HF, and pneumonia with SDoH was low. Across all 3 conditions, lives alone was the most common SDoH (2.2% [interquartile range (IQR), 0.7-4.7]), followed by substance use disorder (1.3% [IQR, 0.5-2.1]), and use of substance use services (1.2% [IQR, 0.6-1.8]). Using NLP, the proportion of hospitalized veterans with lives alone was higher for HF (14.4% vs 2.0%, P < .01), pneumonia (11% vs 1.9%, P < .01), and AMI (10.2% vs 1.4%, P < .01) compared with International Classification of Diseases, Ninth Edition codes. Interfacility variability was noted with both administrative and NLP extraction methods.

Conclusions

The presence of SDoH in administrative data among patients hospitalized for common medical issues is low and variable across VHA facilities. Significant facility-level variation of 5 SDoH was present regardless of extraction method.

Abstract 155: Comparative Effectiveness of Carotid Endarterectomy Compared to Medical Therapy Among Patients With Asymptomatic Carotid Stenosis

Background: Carotid endarterectomy (CEA) reduces stroke risk compared to medical therapy alone among patients with asymptomatic carotid stenosis. CEA involves a tradeoff between higher perioperative short-term risks in exchange for a lower long-term risk of stroke. However, overall declines in stroke rates raise concerns that CEA may no longer be a preferred treatment. We examined the effectiveness of CEA compared to medical therapy (MT) among asymptomatic patients in preventing stroke and stroke-death within 5 years of follow-up.

Methods: We identified Veterans ≥65 years old with carotid stenosis (n=2712 CEA and n=2509 MT patients) who did not have a history of stroke or transient ischemic attack. We propensity score-matched MT patients to CEA patients to control for baseline confounding and used methods to mimic analyses from the Asymptomatic Carotid Stenosis Trial, the last published trial to compare CEA to MT. We accounted for “immortal time” bias by randomizing patients to CEA and MT groups and censoring patients if their actual treatment became inconsistent with the arm in which they were randomized (e.g., patient received CEA, but was randomized to MT). We accounted for the informative censoring by estimating time-dependent inverse probability of censoring weights using measured covariates (demographics and 72 time-varying comorbidities). We computed weighted Kaplan-Meier (KM) curves and estimated the risk of stroke/stroke-death in each group over 5 years of follow-up.

Results: The observed stroke or death rate (perioperative complications) within 30 days in the CEA arm was 3%. The 5-year risk were similar among patients randomized to CEA 5.5% (95% CI, 4.3%-6.7%) versus MT 7.6% (95% CI,5.9%-9.2%) (risk difference, -2.1%, 95% CI -4%- 0%) with little difference in the KM curves (logrank p=0.2).

Conclusion: CEA was not superior to MT in a community sample of Veterans after 5 years of follow-up, suggesting that CEA may no longer be the preferred treatment strategy.

Management of acute upper gastrointestinal bleeding: an update for the general physician

Acute upper gastrointestinal bleed (AUGIB) is one of the most common medical emergencies in the UK, with roughly one presentation every 6 min. Despite advances in therapeutics and endoscopy provision, mortality following AUGIB over the last two decades has remained high, with over 9,000 deaths annually in the UK; consequently, several national bodies have published UK-relevant guidelines. Despite this, the 2015 UK National Confidential Enquiry into Patient Outcome and Death in AUGIB highlighted variations in practice, raised concerns regarding suboptimal patient care and released a series of recommendations. This review paper incorporates the latest available evidence and UK-relevant guidelines to summarise the optimal pre-endoscopic, endoscopic, and post-endoscopic approach to and management of non-variceal and variceal AUGIB that will be of practical value to both general physicians and gastroenterologists.

Abstract WP317: Door-In-Door-Out (DIDO) - Identifying Causes of Interfacility Transfer Delays

Background: Delays in endovascular therapy can lead to worse outcomes in stroke patients. Interfacility delays when transferring patients to thrombectomy-capable hospitals are common. We sought to identify causes of such delays and create a standardized transfer process to improve transfer times.

Methods: A 15-hospital system formed a multidisciplinary team to assess the current state of DIDO processes from data on previous transfers, evaluate current transfer algorithms, and analyze baseline performance. Individual hospital transfer algorithms were compared.

Results: Data was analyzed for 47 patients transferred from 12 facilities to thrombectomy-capable centers from June 2017 to July 2018 who received ischemic stroke treatment. 71% were from a metro area within 50 miles of a thrombectomy-capable hospital and 29% were from outside the metro area, up to 400 miles away. 80% were transported by air by 9 different agencies. DIDO times were: 27% less than 90 min; 27% 90 to 120 min; 44% greater than 120 min. There was minimal opportunity for improvement based on: day of week, NIHSS, and treatment with alteplase. Areas for improvement include: timeliness of door to CT (x-=13 vs 35 min), time to CT read (x-=8 vs 16 min), time to alteplase (x-=48 vs 115 min), and time to transfer (x-=125 vs 148 min), time from arrival to consult with thrombectomy capable centers (x-= 53 vs 86 min), time from consult to dispatch (x-=19 vs 26 min), and time of transfer agency on scene (x-=22 vs 37 min) for metro vs rural hospitals.

Conclusions: Identified causes of delays include decision to transfer and initiation of dispatch to coordinate transport. A difference in timeliness initial diagnosis and treatment of patients outside metro areas was identified. Last known well (LKW) to arrival for transferred patients revealed a need to educate sending facilities on the expanded treatment windows- only 3 patients had a LKW to arrival time greater than 4.5 hours.

Using clinical Natural Language Processing for health outcomes research: Overview and actionable suggestions for future advances

The importance of incorporating Natural Language Processing (NLP) methods in clinical informatics research has been increasingly recognized over the past years, and has led to transformative advances. Typically, clinical NLP systems are developed and evaluated on word, sentence, or document level annotations that model specific attributes and features, such as document content (e.g., patient status, or report type), document section types (e.g., current medications, past medical history, or discharge summary), named entities and concepts (e.g., diagnoses, symptoms, or treatments) or semantic attributes (e.g., negation, severity, or temporality). From a clinical perspective, on the other hand, research studies are typically modelled and evaluated on a patient- or population-level, such as predicting how a patient group might respond to specific treatments or patient monitoring over time. While some NLP tasks consider predictions at the individual or group user level, these tasks still constitute a minority. Owing to the discrepancy between scientific objectives of each field, and because of differences in methodological evaluation priorities, there is no clear alignment between these evaluation approaches. Here we provide a broad summary and outline of the challenging issues involved in defining appropriate intrinsic and extrinsic evaluation methods for NLP research that is to be used for clinical outcomes research, and vice versa. A particular focus is placed on mental health research, an area still relatively understudied by the clinical NLP research community, but where NLP methods are of notable relevance. Recent advances in clinical NLP method development have been significant, but we propose more emphasis needs to be placed on rigorous evaluation for the field to advance further. To enable this, we provide actionable suggestions, including a minimal protocol that could be used when reporting clinical NLP method development and its evaluation.

Abstract 192: Trends in Carotid Imaging in the VA Health System Following Choosing Wisely: An Interrupted Time Series Analysis

Background: The majority of carotid imaging is performed on asymptomatic patients, those without stroke or transient ischemic attack, for whom the evidence for carotid revascularization is limited. In 2013, three Choosing Wisely campaigns identified carotid imaging for syncope (American Academy of Neurology), screening (American Academy of Family Physicians) and preoperatively before cardiac surgery (Society of Thoracic Surgeons) as low-value. The impact of these recommendations is unknown.

Methods: We compared annual rates of carotid imaging in the VA Health System between 2007 and 2016. As administrative carotid imaging codes are non-specific, we used natural language processing to identify carotid imaging, including ultrasound, computed tomography angiography (CTA) and magnetic resonance angiography (MRA). We reviewed the study reason listed in radiology records to develop a lexicon of test indications for identifying imaging for syncope, preoperative testing and carotid bruits. We conducted an interrupted time series analysis to determine whether trends in overall and indication-specific asymptomatic carotid imaging changed, comparing imaging rates prior to the recommendations (2007-2012) and following their release (2014-2016).

Results: We identified 773,044 carotid images for asymptomatic indications (mean age, 69.9 years; 96.5% men) including 720,961 ultrasounds, 28,979 CTAs, and 24,539 MRAs. From 2007 to 2016 the annual rate of asymptomatic imaging decreased from 10.79 to 10.35 images per 1,000 veterans (rate ratio 0.96; P<0.001). Annual rates of imaging for bruits also decreased while imaging for syncope and preoperative evaluation increased (rate ratios: 0.62, 1.08 and 1.59, all P<.0001) (Figure). Rates of asymptomatic imaging declined by 1.01% annually (95% CI -1.18% to -0.84%) prior to Choosing Wisely, with a change to a flat rate in the post-period (0.12%; 95% CI -0.37% to 0.62%). Trends in annual rates of imaging for syncope and preoperative evaluation increased following Choosing Wisely, while the rate of decrease of imaging for bruits diminished.

Conclusions: The release of 3 Choosing Wisely guidelines targeting carotid imaging did not reduce overall or indication specific rates of asymptomatic carotid imaging in the VA Health System.

Abstract WMP92: Towards High-Precision Stroke Classification Using Natural Language Processing

Introduction: Case identification for genetic epidemiologic stroke studies relies upon diagnostic codes or laborious hand-adjudication. Natural language processing (NLP) can extract discerning characteristics of a patient phenotype from clinical texts to support high-precision phenotyping. In this pilot study, we aimed to support large-scale analysis of stroke pedigrees by developing an NLP algorithm to accurately identify stroke patients from the electronic health record.

Methods: We randomly sampled 500 stroke cases (400 ischemic, IS, and 100 hemorrhagic, HS) and 100 non-stroke neurologic controls from a neuroICU database; 2 content experts validated diagnoses. We divided the dataset into train (n=400) and test (n=200) sets and applied 5-fold cross validation analyses to the train set. We leveraged 2 text-processing approaches: n-grams (creating word features of 1-4 word window lengths from clinical notes, e.g., 1-gram = “tpa”) and NLP that adds ConText to n-grams (e.g., negation ). In both approaches, we applied feature selection to identify the most informative n-grams (top 1% ranked by X 2 ) associated with the diagnoses. We trained each classifier using varying note types: 1) radiology, 2) radiology + discharge, 3) radiology + discharge + others and report mean positive predictive value (PPV) for IS vs HS/Other and HS vs IS/Other, using a) n-grams and b) NLP.

Results: Using n-grams alone, there was consistently high PPV for discerning all Stroke vs Other for all text sources (Figure). N-gram PPV improved dramatically for IS vs HS/Other by adding more note types. NLP consistently out-performed n-grams alone, with high PPV for all Stroke vs Other and for IS vs HS/Other. Both approaches performed poorly for HS vs IS/Other.

Conclusions: NLP is a promising method to accurately phenotype stroke cases. Determining the optimal number of notes required will allow for more streamlined data processing. Future work will focus on improving discrimination of HS vs IS/Other.

Prevention of allograft HCV recurrence with peri-transplant human monoclonal antibody MBL-HCV1 combined with a single oral direct-acting antiviral: A proof-of-concept study

Patients with active hepatitis C virus (HCV) infection at transplantation experience rapid allograft infection, increased risk of graft failure and accelerated fibrosis. MBL-HCV1, a neutralizing human monoclonal antibody (mAb) targeting the HCV envelope, was combined with a licensed oral direct-acting antiviral (DAA) to prevent HCV recurrence post-transplant in an open-label exploratory efficacy trial. Eight subjects received MBL-HCV1 beginning on the day of transplant with telaprevir initiated between days 3 and 7 post-transplantation. Following FDA approval of sofosbuvir, two subjects received MBL-HCV1 starting on the day of transplant with sofosbuvir initiated on day 3. Combination treatment was administered for 8-12 weeks or until the stopping rule for viral rebound was met. The primary endpoint was undetectable HCV RNA at day 56 with exploratory endpoints of sustained virologic response (SVR) at 12 and 24 weeks post-treatment. Both subjects receiving mAb and sofosbuvir achieved SVR24. Four of eight subjects in the mAb and telaprevir group met the primary endpoint; one subject achieved SVR24 and three subjects relapsed 2-12 weeks post-treatment. The other four subjects experienced viral breakthrough. There were no serious adverse events related to study treatment. This proof-of-concept study demonstrates that peri-transplant immunoprophylaxis combined with a single oral direct-acting antiviral in the immediate post-transplant period can prevent HCV recurrence.

Developing a web-based SKOS editor

Background

The Simple Knowledge Organization System (SKOS) was introduced to the wider research community by a 2005 World Wide Web Consortium (W3C) working draft, and further developed and refined in a 2009 W3C recommendation. Since then, SKOS has become the de facto standard for representing and sharing thesauri, lexicons, vocabularies, taxonomies, and classification schemes. In this paper, we describe the development of a web-based, free, open-source SKOS editor built for the development, curation, and management of small to medium-sized lexicons for health-related Natural Language Processing (NLP).

Results

The web-based SKOS editor allows users to create, curate, version, manage, and visualise SKOS resources. We tested the system against five widely-used, publicly-available SKOS vocabularies of various sizes and found that the editor is suitable for the development and management of small to medium-size lexicons. Qualitative testing has focussed on using the editor to develop lexical resources to drive NLP applications in two domains. First, developing a lexicon to support an Electronic Health Record-based NLP system for the automatic identification of pneumonia symptoms. Second, creating a taxonomy of lexical cues associated with Diagnostic and Statistical Manual of Mental Disorders (DSM-5) diagnoses with the goal of facilitating the automatic identification of symptoms associated with depression from short, informal texts.

Conclusions

The SKOS editor we have developed is — to the best of our knowledge — the first free, open-source, web-based, SKOS editor capable of creating, curating, versioning, managing, and visualising SKOS lexicons.

Electronic supplementary material

The online version of this article (doi:10.1186/s13326-015-0043-z) contains supplementary material, which is available to authorized users.

A genomic case study of mixed fibrolamellar hepatocellular carcinoma

We report the first comprehensive genomic analysis of a case of mixed conventional and fibrolamellar HCC (mFL-HCC). This study confirms the expression of DNAJB1:PRKACA, a fusion previously associated with pure FL-HCC but not conventional HCC, in mFL-HCC. These results indicate the DNAJB1:PRKACA fusion has diagnostic utility for both pure and mixed FL-HCC.

Background

Mixed fibrolamellar hepatocellular carcinoma (mFL-HCC) is a rare liver tumor defined by the presence of both pure FL-HCC and conventional HCC components, represents up to 25% of cases of FL-HCC, and has been associated with worse prognosis. Recent genomic characterization of pure FL-HCC identified a highly recurrent transcript fusion (DNAJB1:PRKACA) not found in conventional HCC.

Patients and Methods

We performed exome and transcriptome sequencing of a case of mFL-HCC. A novel BAC-capture approach was developed to identify a 400 kb deletion as the underlying genomic mechanism for a DNAJB1:PRKACA fusion in this case. A sensitive Nanostring Elements assay was used to screen for this transcript fusion in a second case of mFL-HCC, 112 additional HCC samples and 44 adjacent non-tumor liver samples.

Results

We report the first comprehensive genomic analysis of a case of mFL-HCC. No common HCC-associated mutations were identified. The very low mutation rate of this case, large number of mostly single-copy, long-range copy number variants, and high expression of ERBB2 were more consistent with previous reports of pure FL-HCC than conventional HCC. In particular, the DNAJB1:PRKACA fusion transcript specifically associated with pure FL-HCC was detected at very high expression levels. Subsequent analysis revealed the presence of this fusion in all primary and metastatic samples, including those with mixed or conventional HCC pathology. A second case of mFL-HCC confirmed our finding that the fusion was detectable in conventional components. An expanded screen identified a third case of fusion-positive HCC, which upon review, also had both conventional and fibrolamellar features. This screen confirmed the absence of the fusion in all conventional HCC and adjacent non-tumor liver samples.

Conclusion

These results indicate that mFL-HCC is similar to pure FL-HCC at the genomic level and the DNAJB1:PRKACA fusion can be used as a diagnostic tool for both pure and mFL-HCC.

Should ALPPS be Used for Liver Resection in Intermediate-Stage HCC?

BACKGROUND

Extended liver resections in patients with hepatocellular carcinoma (HCC) are problematic due to hepatitis, fibrosis, and cirrhosis. Associating liver partition with portal vein ligation for staged hepatectomy (ALPPS) has been promoted as a novel method to induce hypertrophy for patients with extensive colorectal liver metastases, but outcomes in HCC have not been well investigated.

METHODS

All patients registered in the international ALPPS Registry ( www.alpps.org ) from 2010 to 2015 were studied. Hypertrophy of the future liver remnant, perioperative morbidity and mortality, age, overall survival, and other parameters were compared between patients with HCC and patients with colorectal liver metastases (CRLM).

RESULTS

The study compared 35 patients with HCC and 225 patients with CRLM. The majority of patients undergoing ALPPS for HCC fall into the intermediate-stage category of the Barcelona clinic algorithm. In this study, hypertrophy was rapid and extensive for the HCC patients, albeit lower than for the CRLM patients (47 vs. 76 %; p < 0.002). Hypertrophy showed a linear negative correlation with the degrees of fibrosis. The 90-day mortality for ALPPS used to treat HCC was almost fivefold higher than for CRLM (31 vs. 7 %; p < 0.001). Multivariate analysis showed that patients older than 61 years had a significantly reduced overall survival (p < 0.004).

CONCLUSION

The ALPPS approach induces a considerable hypertrophic response in HCC patients and allows resection of intermediate-stage HCC, albeit at the cost of a 31 % perioperative mortality rate. The use of ALPPS for HCC remains prohibitive for most patients and should be performed only for a highly selected patient population younger than 60 years with low-grade fibrosis.

Atrial fibrillation in patients undergoing liver transplantation-a single-center experience

BACKGROUND

As the prevalence of atrial fibrillation rises with age and older patients increasingly receive transplants, the perioperative management of this common arrhythmia and its impact on outcomes in liver transplantation is of relevance.

METHODS

Retrospective review of 757 recipients of liver transplantation from January 2002 through December 2011.

RESULTS

Nineteen recipients (2.5%) had documented pre-transplantation atrial fibrillation. Sixteen patients underwent liver and 3 a combined liver-kidney transplantation. Three patients died within 30 days (84.2% 1-month survival) and another 3 within 1 year of transplantation (68.4% 1-year survival). Compared with patients without atrial fibrillation, the relative risk of death in the atrial fibrillation group was 5.29 at 1 month (P = .0034; 95% confidence interval [CI], 1.73-16.18) and 3.28 at 1 year (P = .0008; 95% CI, 1.63-6.59). Time to extubation and intensive care unit (ICU) and hospital readmissions were not different from the control cohort. Rapid ventricular response requiring treatment occurred in 4 patients during surgery and 7 after surgery, resulting in 3 ICU and 3 hospital readmissions.

CONCLUSIONS

The results suggest that patients with atrial fibrillation may be at increased risk of mortality after liver transplantation. Optimization of medical therapy may decrease ICU and hospital readmission due to rapid ventricular response.

Using twitter to examine smoking behavior and perceptions of emerging tobacco products

BACKGROUND

Social media platforms such as Twitter are rapidly becoming key resources for public health surveillance applications, yet little is known about Twitter users' levels of informedness and sentiment toward tobacco, especially with regard to the emerging tobacco control challenges posed by hookah and electronic cigarettes.

OBJECTIVE

To develop a content and sentiment analysis of tobacco-related Twitter posts and build machine learning classifiers to detect tobacco-relevant posts and sentiment towards tobacco, with a particular focus on new and emerging products like hookah and electronic cigarettes.

METHODS

We collected 7362 tobacco-related Twitter posts at 15-day intervals from December 2011 to July 2012. Each tweet was manually classified using a triaxial scheme, capturing genre, theme, and sentiment. Using the collected data, machine-learning classifiers were trained to detect tobacco-related vs irrelevant tweets as well as positive vs negative sentiment, using Naïve Bayes, k-nearest neighbors, and Support Vector Machine (SVM) algorithms. Finally, phi contingency coefficients were computed between each of the categories to discover emergent patterns.

RESULTS

The most prevalent genres were first- and second-hand experience and opinion, and the most frequent themes were hookah, cessation, and pleasure. Sentiment toward tobacco was overall more positive (1939/4215, 46% of tweets) than negative (1349/4215, 32%) or neutral among tweets mentioning it, even excluding the 9% of tweets categorized as marketing. Three separate metrics converged to support an emergent distinction between, on one hand, hookah and electronic cigarettes corresponding to positive sentiment, and on the other hand, traditional tobacco products and more general references corresponding to negative sentiment. These metrics included correlations between categories in the annotation scheme (phihookah-positive=0.39; phi(e-cigs)-positive=0.19); correlations between search keywords and sentiment (χ²₄=414.50, P<.001, Cramer's V=0.36), and the most discriminating unigram features for positive and negative sentiment ranked by log odds ratio in the machine learning component of the study. In the automated classification tasks, SVMs using a relatively small number of unigram features (500) achieved best performance in discriminating tobacco-related from unrelated tweets (F score=0.85).

CONCLUSIONS

Novel insights available through Twitter for tobacco surveillance are attested through the high prevalence of positive sentiment. This positive sentiment is correlated in complex ways with social image, personal experience, and recently popular products such as hookah and electronic cigarettes. Several apparent perceptual disconnects between these products and their health effects suggest opportunities for tobacco control education. Finally, machine classification of tobacco-related posts shows a promising edge over strictly keyword-based approaches, yielding an improved signal-to-noise ratio in Twitter data and paving the way for automated tobacco surveillance applications.

An ISDS-Based Initiative for Conventions for Biosurveillance Data Analysis Methods

Objective

The panel will present the problem of standardizing analytic methods for public health disease surveillance, enumerate goals and constraints of various stakeholders, and present a straw-man framework for a conventions group.

Introduction

Twelve years into the 21st century, after publication of hundreds of articles and establishment of numerous biosurveillance systems worldwide, there is no agreement among the disease surveillance community on most effective technical methods for public health data monitoring. Potential utility of such methods includes timely anomaly detection, threat corroboration and characterization, follow-up analysis such as case linkage and contact tracing, and alternative uses such as providing supplementary information to clinicians and policy makers.

Several factors have impeded establishment of analytical conventions. As immediate owners of the surveillance problem, public health practitioners are overwhelmed and understaffed. Goals and resources differ widely among monitoring institutions, and they do not speak with a single voice. Limited funding opportunities have not been sufficient for cross-disciplinary collaboration driven by these practitioners. Most academics with the expertise and luxury of method development cannot access surveillance data. Lack of data access is a formidable obstacle to developers and has caused talented statisticians, data miners, and other analysts to abandon the field. The result is that older research is neglected and repeated, literature is flooded with papers of varying utility, and the decision-maker seeking realistic solutions without detailed technical knowledge faces a difficult task.

Regarding conventions, the disease surveillance community can learn from older, more established disciplines, but it also poses some unique challenges. The general problem is that disease surveillance lies on the fringe of disparate fields (biostatistics, statistical process control, data mining, and others), and poses problems that do not adequately fit conventional approaches in these disciplines.

In its eighth year, the International Society of Disease Surveillance is well positioned to address the standardization problem because its membership represents the involved stakeholders including progressive programs worldwide as well as resource-limited settings, and also because best practices in disease surveillance is fundamental to its mission. The proposed panel is intended to discuss how an effective, sustainable technical conventions group might be maintained and how it could support stakeholder institutions.

Methods

Members of a Technical Conventions Group would have experience and dedication to advancing the science of disease surveillance. Primary functions would include:

Specify and communicate technical problems facing professionals involved in routine monitoring of population health. Alternative use applications would also be considered, such as the use of epidemiological findings to inform clinical diagnoses.

Independently evaluate the utility of proposed analytical solutions to well-defined problems in public health surveillance and confer approval or certification, perhaps on several levels, such as whether results can be replicated with shareable data. Approved solutions might be restricted to commonly available software such as the R language or Microsoft EXCEL.

Facilitate sharing of tools and methodologies to evaluate methods and to visualize their results

The framework to be discussed in the proposed panel would be a means of keeping open lines of collaboration and idea-sharing. Overcoming obstacles toward this goal is worthy of a conference panel discussion whether or not it concludes that a conventions group is a viable approach.

Results

Three 15-minute panelist talks are proposed:

Background: in-depth description of the dimensions of the problem above

Constraints facing public health practitioners and requirements for practical analytic tools

Strawman conventions group: role, logistics, inclusiveness, methods of communicating with stakeholders and related organizations and producing/disseminating output.

For the 45 minutes of discussion, the first 15–20 will invite reactions to the first two talks to sharpen the scope of the effort. The remainder of the session will cover the advisability, feasibility, and logistics of an ISDS-based conventions group, and modify the straw-man group concept.

Atrial laceration caused by removal of a transjugular intrahepatic portosystemic shunt necessitates emergent cardiopulmonary bypass during liver transplant: a case report

In situ transjugular intrahepatic portosystemic shunting (TIPS) can complicate liver transplantation. We present a case where an intraoperative attempt to remove a malpositioned TIPS resulted in atrial laceration. Massive transfusion and emergent institution of cardiopulmonary bypass allowed patient resuscitation and completion of surgery. We describe our surgical and anesthesiologic management, and discuss the absence of criteria to predict when TIPS may become adherent to the inferior vena cava or the right atrium and difficult to remove.

A Phase I study of an intravesically administered immunotoxin targeting EpCAM for the treatment of nonmuscle-invasive bladder cancer in BCGrefractory and BCG-intolerant patients

PURPOSE

A Phase I dose-escalation study was performed to determine the maximum tolerated dose (MTD) of the immunotoxin VB4-845 in patients with nonmuscle-invasive bladder cancer (NMIBC) refractory to or intolerant of bacillus Calmette-Guerin (BCG). Secondary objectives included evaluation of the safety, tolerability, pharmacokinetics, immunogenicity, and efficacy of VB4-845.

PATIENTS AND METHODS

Sixty-four patients with Grade 2 or 3, stage Ta or T1 transitional cell carcinoma or in situ carcinoma, either refractory to or intolerant of BCG therapy, were enrolled. Treatment was administered in ascending dose cohorts ranging from 0.1 to 30.16 mg. After receiving weekly instillations of VB4-845 to the bladder via catheter for 6 consecutive weeks, patients were followed for 4-6 weeks post-therapy and assessed at week 12.

RESULTS

An MTD was not determined, as a dose-limiting toxicity was not identified over the dose range tested. VB4-845 therapy was safe and well tolerated with most adverse events reported as mild; as a result, no patients were removed from the study in response to toxicity. By the end of the study, the majority of patients had developed antibodies to the exotoxin portion of VB4-845. A complete response was achieved in 39% of patients at the 12-week time point.

CONCLUSIONS

VB4-845 dosed on a weekly basis for 6 weeks was very well tolerated at all dose levels. Although an MTD was not determined at the doses administered, VB4-845 showed evidence of an antitumor effect that warrants further clinical investigation for the treatment of NMIBC in this patient population.

A comparison of two approaches to text processing: facilitating chart reviews of radiology reports in electronic medical records

Chart review is central to health services research. Text processing, which analyzes free-text fields through automated methods, can facilitate this process. We compared precision and accuracy of NegEx and SQLServer 2008 Free-Text Search in identifying acute fractures in radiology reports.The term "fracture" was included in 23,595 radiology reports from the Veterans Aging Cohort Study. Four hundred reports were randomly selected and manually reviewed for acute fractures to establish a gold standard. Reports were then processed by SQLServer and NegEx. Results were compared to the gold standard to determine accuracy, precision, recall, and F-statistic.NegEx and the gold standard identified acute fractures in 13 reports. SQLServer identified 2 in a report-based analysis (precision: 1.00; accuracy: 0.97; recall: 0.15; F-statistic: 0.26), and 12 in a sentence-by-sentence analysis (precision: 1.00; recall: 0.92; accuracy: 0.92; F-statistic: 0.96).Text-processing tools utilizing basic database or programming skills are comparable, precise, and accurate in identifying reports for review.

Oleanolic Acid, a plant triterpenoid, significantly improves survival and function of islet allograft

BACKGROUND.: Oleanolic acid (OA) is a ubiquitous triterpenoid, with potent antioxidant and anti-inflammatory properties. Here, we tested whether these combined properties of OA can prevent nonimmunologic primary nonfunctioning and immunologic phenomena ascribed to graft rejection hence prolong islet allograft survival. METHODS.: Islet transplants were performed under kidney capsule of streptozotocin-induced diabetic C57BL/6 mice with BALB/c islets. Recipients were treated with 0.5 mg/day of OA intraperitoneally, and serum samples were collected once in 2 days and used for luminex, ELISA, and donor-specific antibody screening. Transplanted mice were killed at different time intervals to obtain splenocytes and kidney samples for ELISPOT, mixed leukocyte reaction, and immunohistochemical studies. RESULTS.: After transplantation, the decrement of blood glucose was significantly faster in mice receiving OA less than 2+/-1 days compared with untreated (4+/-2 days). OA prolonged survival of transplanted islets up to 23+/-3 days and reversed diabetes even with 250 islets. Treatment group showed increased serum interleukin (IL)-10 (twofold) and decreased inducible protein-10 and IL-4 (threefold) in luminex. Significantly reduced frequency of interferon-gamma (4.5-fold), IL-4 (3.5-fold), IL-2 (2.3-fold), and IL-17 (fourfold) producing T-cell populations were found in ELISPOT. OA-treated grafts had significant reduced and delayed infiltration of CD4+ and CD8+ T cells. OA also delayed donor-specific antibody generation up to 19 days after transplantation. Combined treatment with cyclosporine A, OA further prolonged the islet allograft survival to 34+/-3 days. CONCLUSIONS.: In conclusion, OA is an attractive, dietary nontoxic plant triterpenoid, which suppresses the production of proinflammatory cytokines and delays graft-specific immune responses to prolong islet allograft survival.

Analysis of a failed clinical decision support system for management of congestive heart failure

In order to increase compliance with The Joint Commission's Congestive Heart Failure Core Measures, a rule based clinical decision support system (CDSS) was developed and deployed at a community hospital in our health system. We evaluated the performance of the CDSS in identifying patients with primary congestive heart failure (CHF)and identified problems encountered with its introduction. Performance of the CDSS was compared against a manual review of records of patients with diagnosis of primary CHF. The CDSS had a sensitivity of 0.79 and PPV of 0.11. The CDSS issued multiple alerts for majority of the patients(74%). The number of alerts issued for patients without primary CHF was large, and for a majority of patients (63%) physicians did not respond to alerts the first time. The CDSS performed poorly and was eventually withdrawn but provided insight into a subsequently successful method for managing CHF.

Activated effector and memory T cells contribute to circulating sCD30: potential marker for islet allograft rejection

T-cell activation up-regulates CD30 resulting in an increase in serum soluble CD30 (sCD30). CD4+ T cells, a major source for sCD30, play a significant role in the pathogenesis of rejection. In this study, sCD30 was measured pre- and posttransplant in mouse islet allograft models and human islet allograft recipients. sCD30 was measured by ELISA in diabetic C57BL/6, CD4Knockout (KO) and CD8KO islet allograft recipients. sCD30 increased significantly prior to rejection (1.8 +/- 1 days) in 80% of allograft recipients. Sensitization with donor splenocytes, or a second graft, further increased sCD30 (282.5 +/- 53.5 for the rejecting first graft vs. 374.6 +/- 129 for the rejecting second graft) prior to rejection suggesting memory CD4+ T cells contribute to sCD30. CD4KO failed to reject islet allograft and did not demonstrate sCD30 increase. CD8KO showed elevated (227 +/- 107) sCD30 (1 day) prior to rejection. High pretransplant sCD30 (>20 U/ml) correlated with poor outcome in human islet allograft recipients. Further, increase in sCD30 posttransplant preceded (3-4 months) loss of islet function. We conclude that sCD30 is released from activated CD4 T cells prior to islet allograft rejection and monitoring sCD30 can be a valuable adjunct in the follow-up of islet transplant recipients.

Toward vocabulary control for chief complaint

The chief complaint (CC) is the data element that documents the patient's reason for visiting the emergency department (ED). The need for a CC vocabulary has been acknowledged at national meetings and in multiple publications, but to our knowledge no groups have specifically focused on the requirements and development plans for a CC vocabulary. The national consensus meeting "Towards Vocabulary Control for Chief Complaint" was convened to identify the potential uses for ED CC and to develop the framework for CC vocabulary control. The 10-point consensus recommendations for action were 1) begin to develop a controlled vocabulary for CC, 2) obtain funding, 3) establish an infrastructure, 4) work with standards organizations, 5) address CC vocabulary characteristics for all user communities, 6) create a collection of CC for research, 7) identify the best candidate vocabulary for ED CCs, 8) conduct vocabulary validation studies, 9) establish beta test sites, and 10) plan publicity and marketing for the vocabulary.

Heuristic sample selection to minimize reference standard training set for a part-of-speech tagger

Part-of-speech tagging represents an important first step for most medical natural language processing (NLP) systems. The majority of current statistically-based POS taggers are trained using a general English corpus. Consequently, these systems perform poorly on medical text. Annotated medical corpora are difficult to develop because of the time and labor required. We investigated a heuristic-based sample selection method to minimize annotated corpus size for retraining a Maximum Entropy (ME) POS tagger. We developed a manually annotated domain specific corpus (DSC) of surgical pathology reports and a domain specific lexicon (DL). We sampled the DSC using two heuristics to produce smaller training sets and compared the retrained performance against (1) the original ME modeled tagger trained on general English, (2) the ME tagger retrained on the DL, and (3) the MedPost tagger trained on MEDLINE abstracts. RESULTS showed that the ME tagger retrained with a DSC was superior to the tagger retrained with the DL, and also superior to MedPost. Heuristic methods for sample selection produced performance equivalent to use of the entire training set, but with many fewer sentences. Learning curve analysis showed that sample selection would enable an 84% decrease in the size of the training set without a decrement in performance. We conclude that heuristic sample selection can be used to markedly reduce human annotation requirements for training of medical NLP systems.

Molecular correlates associated with a phase II study of temsirolimus (CCI-779) in patients with metastatic or recurrent endometrial cancer—NCIC IND 160

3003

Background: PTEN is a tumor suppressor gene and mutations in PTEN causing loss of protein expression/function may play a significant role in the pathogenesis of EC. Loss of PTEN protein expression has been reported in 26–80% of EC and leads to deregulated PI3K/Akt/mTOR signalling which may give neoplastic cells a survival advantage by enhancing angiogenesis, protein translation and cell cycle progression. Inhibition of mTOR, a protein kinase downstream of the PI3K/Akt pathway and target of rapamycins, inhibits proliferation of EC cell lines and formation of EC in PTEN−/+ heterozygous mice. We have evaluated temsirolimus (T) an ester derivative of rapamycin that inhibits mTOR given the frequent loss of PTEN in human EC. Methods: A two stage, phase II study has been conducted to evaluate single agent activity of T in women with recurrent or metastatic EC (chemotherapy naïve, upto 1 prior line of hormonal therapy). Treatment was 25mg i.v. weekly. One cycle is defined as 4 weeks of therapy. Thirty one patients (pts) have been registered; 23 are evaluable for toxicity and 19 for response. Results: Preliminary results have demonstrated encouraging activity and the trial fulfilled predefined criteria for activity. Of 19 pts evaluable for response, 5 have had a confirmed partial response (PR) (26%) and 12 have stable disease (SD) as best response (63%). Two pts had progressive disease (PD) (11%). PTEN, phosphorylated (p) mTOR and p-S6 protein (immunohistochemistry) results are available on 9 pts to date. The preliminary results indicate that responses and stable disease are seen in pts irrespective of PTEN status. Loss of p-mTOR was evident in tumor cells from all pts (range 75–99% loss) and did not correlate with response. Phosphorylated S6 was low in tumor cells in the one pt who had PD (5%), and levels were higher in pts with PR and SD (mean and median 30%). Conclusions: We conclude that Temsirolimus has encouraging single agent activity for in recurrent and metastatic EC, and the findings indicate this is irrespective of PTEN status. We are currently evaluating activity of T in pts previously treated with chemotherapy and molecular correlates in additional specimens.

No significant financial relationships to disclose.

Continuous treatment of Barrett's oesophagus patients with proton pump inhibitors up to 13 years: observations on regression and cancer incidence

BACKGROUND

There is little evidence that treatment of patients with Barrett's oesophagus with proton pump inhibitors over periods up to 6 years results in major regression of Barrett's oesophagus.

AIM

To determine if longer periods of treatment with proton pump inhibitors lead to significant regression of Barrett's oesophagus, and to determine the incidence of oesophageal adenocarcinoma in the proton pump inhibitor-treated patients.

METHODS

We analysed prospectively-collected data on Barrett's oesophagus patients treated with proton pump inhibitors for 1-13 years.

RESULTS

188 patients with Barrett's oesophagus and intestinal metaplasia, were treated for 1-13 years with a proton pump inhibitor (966 years of treatment; mean 5.1 years). No change in length was seen during treatment but 48% of patients developed squamous islands (25% after 1-3 years; 100% at 12-13 years). Squamous islands correlated with treatment duration and male sex but not with proton pump inhibitor dose or patient age. Six patients developed dysplasia and three males developed adenocarcinoma during treatment (cancer incidence 0.31%).

CONCLUSIONS

Proton-pump inhibitor treatment over 1-13 years does not shorten the Barrett's oesophagus segment but squamous islands appear in many patients. The incidence of oesophageal adenocarcinoma was low in these proton pump inhibitor-treated patients compared with published series.

Uterine papillary serous carcinoma: evaluation of multimodality treatment with abdominopelvic radiotherapy and chemotherapy

The aim of this study was to compare overall survival (OS), progression-free survival (PFS), and relapse patterns between different modalities of treatment for uterine papillary serous carcinoma (UPSC). A retrospective review of 124 patients with pathologically confirmed UPSC was performed, of whom, 97 patients were eligible for study. Postoperative treatment groups included adjuvant radiotherapy consisting of whole abdomen and a pelvic boost (abdominopelvic radiotherapy [APRT]) (55 patients), paclitaxel and carboplatin chemotherapy (CT) for six cycles followed by APRT (18 patients), CT only (5 patients), and 19 patients were observed without postoperative adjuvant therapy. Three-year OS was 81% and 63% for the CT followed by APRT and APRT alone, respectively (P= 0.11). After controlling for stage, the group treated with APRT alone had significantly decreased OS and PFS compared to the CT/APRT group (HR 3.6; 1.3-9.8; P= 0.01) and (HR 2.9; 95% CI 1.1-7.3; P= 0.03), respectively. Within the limitations of a retrospective study, the results of this study indicate that multimodality postoperative treatment with paclitaxel and a platinum-based CT followed by APRT may increase the survival of patients with UPSC. However, further prospective studies using these combined modalities are needed to confirm these findings.