Clinical Triage Research Articles

EP036 Rapid clinical photography triage for suspected cancer referrals – A pilot study

EP036 Rapid clinical photography triage for suspected cancer referrals – A pilot study

Detecting functional impairment with the Digital Clock and Recall.

Distinguishing between mild cognitive impairment (MCI) and early dementia requires both neuropsychological and functional assessment that often relies on caregivers' insights. Contacting a patient's caregiver can be time-consuming in a physician's already-filled workday. To assess the utility of a brief, machine learning (ML)-enabled digital cognitive assessment, the Digital Clock and Recall (DCR), for detecting functional dependence. We evaluated whether the DCR can help identify individuals at risk of functional deficits as measured by the informant-rated Functional Activities Questionnaire (FAQ) in older individuals including cognitively unimpaired, MCI, and dementia likely due to Alzheimer's disease. The DCR scaled well with FAQ scores, and ML classifiers trained on multimodal DCR features demonstrated strong performance in predicting functional impairment on a held-out test set. Differences in FAQ scores between DCR-predicted classes were comparable across key demographic groups. The DCR can streamline the clinical decision-making, triage, and intervention planning associated with functional impairment in primary care.

The role of advanced clinical practitioners in gastroenterology: a phenomenological perspective from practice and research

This comprehensive review explores the pivotal role of advanced clinical practitioners (ACPs) in gastroenterology care and the integration of phenomenological approaches in both practice and research domains. ACPs in gastroenterology play multifaceted roles encompassing clinical assessment, triage, diagnosis, treatment and patient management. By adopting a phenomenological approach in practice, ACPs can transcend traditional biomedical models, prioritise patients' lived experiences and provide personalised, holistic care. In research, phenomenological methods offer valuable insights into patients' subjective experiences, informing evidence-based practice and contributing to the advancement of gastroenterology care. Through illustrative examples and critical analysis, this review highlights the significance of integrating phenomenological approaches in enhancing patient outcomes and improving the quality of care in gastroenterology services.

Longitudinal transcriptional changes reveal genes from the natural killer cell-mediated cytotoxicity pathway as critical players underlying COVID-19 progression

Patients present a wide range of clinical severities in response severe acute respiratory syndrome coronavirus 2 infection, but the underlying molecular and cellular reasons why clinical outcomes vary so greatly within the population remains unknown. Here, we report that negative clinical outcomes in severely ill patients were associated with divergent RNA transcriptome profiles in peripheral immune cells compared with mild cases during the first weeks after disease onset. Protein–protein interaction analysis indicated that early-responding cytotoxic natural killer cells were associated with an effective clearance of the virus and a less severe outcome. This innate immune response was associated with the activation of select cytokine–cytokine receptor pathways and robust Th1/Th2 cell differentiation profiles. In contrast, severely ill patients exhibited a dysregulation between innate and adaptive responses affiliated with divergent Th1/Th2 profiles and negative outcomes. This knowledge forms the basis of clinical triage that may be used to preemptively detect high-risk patients before life-threatening outcomes ensue.

Longitudinal transcriptional changes reveal genes from the natural killer cell-mediated cytotoxicity pathway as critical players underlying COVID-19 progression.

Patients present a wide range of clinical severities in response severe acute respiratory syndrome coronavirus 2 infection, but the underlying molecular and cellular reasons why clinical outcomes vary so greatly within the population remains unknown. Here, we report that negative clinical outcomes in severely ill patients were associated with divergent RNA transcriptome profiles in peripheral immune cells compared with mild cases during the first weeks after disease onset. Protein-protein interaction analysis indicated that early-responding cytotoxic natural killer cells were associated with an effective clearance of the virus and a less severe outcome. This innate immune response was associated with the activation of select cytokine-cytokine receptor pathways and robust Th1/Th2 cell differentiation profiles. In contrast, severely ill patients exhibited a dysregulation between innate and adaptive responses affiliated with divergent Th1/Th2 profiles and negative outcomes. This knowledge forms the basis of clinical triage that may be used to preemptively detect high-risk patients before life-threatening outcomes ensue.

Use of artificial intelligence-powered ECG to differentiate between cardiac and pulmonary pathologies in patients with acute dyspnoea in the emergency department

BackgroundAcute dyspnoea is common in acute care settings. However, identifying the origin of dyspnoea in the emergency department (ED) is often challenging. We aimed to investigate whether our artificial intelligence...

SP1.8 - Reducing the burden to deliver remote postoperative surveillance using machine learning for surgical-site infection (SSI)

Abstract Aims Remote postoperative wound (RPW) surveillance has been shown to improve the delivery of care and minimise the harm to patients from surgical-site infection (SSI). The burden for healthcare staff to deliver is a major barrier to implementation, therefore we aimed to develop novel methods for automated assessment of RPW surveillance to reduce this. Methods This was a secondary analysis of two interventional studies on RPW surveillance: “Tracking wound infection with smartphone technology” (TWIST) and “ImplementatioN of Remote Surgical wOund Assessment During the coviD-19 pandEmic” (INROADE). Adult general surgery patients could submit images of their surgical wound(s), and patient-reported outcome measures (PROMs) of SSI for 30-days postoperatively. A multi-input neural network (MNN) was developed to predict a clinical diagnosis of SSI within 48h based on symptoms and wound images. Performance was evaluated using area under the curve (AUC) and externally validated. Results There were 1540 submissions containing PROMs (48h SSI rate = 3.7%, n=57) and 2615 images (48h SSI rate = 3.1%, n=82). The MNN exceeded performance of the component models within the external validation subset (AUC: 0.94, 95% CI: 0.89-0.99) and remained equivalent to clinician triage (AUC: 0.92, 95% CI: 0.90-0.94). Usage to screen out “low-risk” responses prior to clinical triage was estimated to reduce the staff-time required to deliver (6.3h vs 25.5h). Conclusion Automated assessment can be successfully deployed within RPW surveillance pathways to reduce the burden on staff to deliver without compromising care, and allow resources to be appropriately directed to those at greatest risk of SSI.

Contributions of individual qSOFA elements to assessment of severity and for prediction of mortality

Background The quick sequential [sepsis-related] organ failure assessment (qSOFA) acts as a prompt to consider possible sepsis. The contributions of individual qSOFA elements to assessment of severity and for prediction of mortality remain unknown. Methods A total of 3974 patients with community-acquired pneumonia were recruited to an observational prospective cohort study. The area under the receiver operating characteristic curve (AUROC), odds ratio, relative risk and Youden’s index were employed to assess discrimination. Results Respiratory rate ≥22/min demonstrated the most superior diagnostic value, indicated by largest odds ratio, relative risk and AUROC, and maximum Youden’s index for mortality. However, the indices for altered mentation and systolic blood pressure (SBP) ≤100 mm Hg decreased notably in turn. The predictive validities of respiratory rate ≥22/min, altered mentation and SBP ≤100 mm Hg were good, adequate and poor for mortality, indicated by AUROC (0.837, 0.734 and 0.671, respectively). Respiratory rate ≥22/min showed the strongest associations with SOFA scores, pneumonia severity index, hospital length of stay and costs. However, SBP ≤100 mm Hg was most weakly correlated with the indices. Conclusions Respiratory rate ≥22/min made the greatest contribution to parsimonious qSOFA to assess severity and predict mortality. However, the contributions of altered mentation and SBP ≤100 mm Hg decreased strikingly in turn. It is the first known prospective evidence of the contributions of individual qSOFA elements to assessment of severity and for prediction of mortality, which might have implications for more accurate clinical triage decisions.

“YOU'RE NOT BLEEDING SO PHONE 111!”: HOW NHS 111 EXPERIENCES INFORM THE DECISION TO ATTEND THE EMERGENCY DEPARTMENT WITH LOW BACK PAIN

Purpose and backgroundLow back pain (LBP), a leading cause of chronic disability, is associated with high and disproportionate socioeconomic costs. Clinical assessment and triage via the NHS 111 service aims to reduce pressure on the urgent and emergency care system. Studies show that many attend the emergency department (ED) against advice, when they could be better managed elsewhere. The aim of this research was to explore how patients with LBP perceive and interact with NHS 111, and how this may have influenced their decision to attend ED.Methods and ResultsThis was a secondary analysis of a cross-sectional, qualitative dataset of 47 patients (26M:21F, aged 23–79 years) purposively sampled with LBP, who attended 1 of 4 EDs in the UK in 2021, (during the pandemic). The participants took part in online, semi-structured interviews, mean duration 45-minutes (range 23–156 minutes), within six weeks of their ED visit. The interviews were audio-recorded, transcribed verbatim and thematically analysed.Three key themes were identified: purpose, process and performance. The themes showed mixed understanding and low awareness of the purpose of NHS 111, despite its strong triage role. Long waits for call backs however, sometimes resulted in patients attending ED, along with previous negative perspectives and experiences.ConclusionThis research suggests enhanced visibility and knowledge of NHS 111 would increase its use by people with LBP. Further exploration comparing outcomes of callers to NHS 111 with those who seek care elsewhere, could help evaluate how NHS 111 can best help people with LBP and prevent unnecessary ED attendance.No conflicts of interest Sources of fundingFunding for primary data: Health Education England & National Institute of Health and Care Research (ICA-CDRF-2018-04-ST2-040)

A new magnetic resonance imaging-based PUMCH classification system for congenital cervical malformations: devising a standardised diagnosis pathway

ObjectivesTo develop an innovative magnetic resonance imaging (MRI)-based PUMCH (Peking Union Medical College Hospital) classification system aimed at standardising the diagnosis of congenital cervical malformations (CCMs) by identifying their distinctive MRI features.MethodsSeventy-nine consecutive patients with CCM underwent pre-treatment pelvic MRI; three experienced gynaecological radiologists retrospectively analysed these images. Qualitative assessments included Rock et al’s classification; PUMCH classification; haematometra; cervical signal features; ovarian endometriosis; haematosalpinx; and uterine, vaginal, urinary, and musculoskeletal malformations. Quantitative assessments involved the uterine volume, sagittal cervical length, and maximum ovarian cross-sectional area. The surgical treatment types were also recorded. Statistical methods were used to incorporate differences in clinical features and surgical methods into our classification.ResultsMorphologically, CCMs were categorised into three types: type I (53%) was characterised by the presence of a cervix with visible cervical canals; type II (23%) featured an existing cervix with concealed cervical canals; and type III (24%) indicated cervical aplasia, which involves a blind end in the lower part of the uterine corpus. Haematometra was significantly more prevalent in patients with type I CCM than in those with type II (p < 0.001). There were three cervical signal patterns: no signal (27%), no evident layer differentiation (21%), and multi-layer differentiation with haematocele (52%). Most patients (94%) had complete vaginal atresia. Type I CCM patients had a higher likelihood of regaining normal uterovaginal anatomy compared to types II and III.ConclusionsOur proposed PUMCH classification system has a high potential for enhancing the efficiency of clinical diagnosis among patients with CCM.Critical relevance statementThe proposed new PUMCH classification promised to elevate the conventional diagnostic trajectory for congenital cervical malformations, offering a valuable framework to refine the selection and planning of surgical interventions, thereby enhancing overall clinical efficacy.Key PointsEffective classification of congenital cervical malformations is desirable to optimise the diagnostic process.We presented a PUMCH classification of congenital cervical malformations using pelvic MRI.The new classification significantly aids clinical triage for congenital cervical malformations.Graphical

542 Reducing the Burden to Deliver Remote Postoperative Surveillance Using Machine Learning for Surgical-Site Infection (SSI)

Abstract Aim To develop novel methods for automated assessment of patient-generated data from remote postoperative wound (RPW) surveillance to reduce the burden of clinical triage for in-person assessment. Method This was a secondary analysis of an interventional study on RPW surveillance: “ImplementatioN of Remote Surgical wOund Assessment During the coviD-19 pandEmic” (INROADE). INROADE included adult patients undergoing gastrointestinal surgery, who could submit images of their surgical wound(s), and patient-reported symptoms of SSI for 30-days postoperatively. Separate models were developed to predict a clinical diagnosis of SSI within 48h based on patient-reported symptoms (logistic regression) and wound images (convolutional neural networks [CNN]). Model performance was evaluated using area under the curve (AUC), with 95% confidence intervals. The impact of implementation of automated assessment was simulated within the RPW surveillance pathway to “rule-out” responses needing clinician triage. Results Of 200 patients enrolled, there were 1529 responses with 3.3% (n=51) submitted within 48h of an SSI. There was excellent discrimination using logistic regression with patient-reported symptoms (AUC: 0.830, 0.765-0.894). 2,125 images were used to develop the CNN, which showed moderate discrimination to identify 48h SSI (AUC: 0.615, 0.556-0.674). Usage to screen out “low-risk” responses prior to clinical triage was estimated to half the staff-time to deliver (12.2h vs 25.5h), while maintaining diagnostic accuracy (AUC: 0.760 [0.685-0.834] vs 0.789 [0.717-0.861]). Conclusions Automated assessment can be successfully deployed within RPW surveillance pathways to reduce the burden on staff to deliver without compromising care and allow resources to be appropriately directed to those at greatest risk of SSI.

Area of admission and short-term care: is it still possible after the COVID pandemic?

Dear Editor, Since 2011, the Internal Medical Unit of Ospedale dell’Angelo, Mestre, has been equipped with an area of eight beds which served as a short-stay unit called area di accoglimento e degenza breve (area of admission and short-term care, ADB). All patients coming from the emergency room (99%) passed through the ADB area. Here, doctors and nurses carried out an initial clinical care triage to identify patients with features that made the discharge possible within 72 hours. [...]

Machine learning based prediction models for analyzing risk factors in patients with acute abdominal pain: a retrospective study.

Acute abdominal pain (AAP) is a common symptom presented in the emergency department (ED), and it is crucial to have objective and accurate triage. This study aims to develop a machine learning-based prediction model for AAP triage. The goal is to identify triage indicators for critically ill patients and ensure the prompt availability of diagnostic and treatment resources. In this study, we conducted a retrospective analysis of the medical records of patients admitted to the ED of Wuhan Puren Hospital with acute abdominal pain in 2019. To identify high-risk factors, univariate and multivariate logistic regression analyses were used with thirty-one predictor variables. Evaluation of eight machine learning triage prediction models was conducted using both test and validation cohorts to optimize the AAP triage prediction model. Eleven clinical indicators with statistical significance (p < 0.05) were identified, and they were found to be associated with the severity of acute abdominal pain. Among the eight machine learning models constructed from the training and test cohorts, the model based on the artificial neural network (ANN) demonstrated the best performance, achieving an accuracy of 0.9792 and an area under the curve (AUC) of 0.9972. Further optimization results indicate that the AUC value of the ANN model could reach 0.9832 by incorporating only seven variables: history of diabetes, history of stroke, pulse, blood pressure, pale appearance, bowel sounds, and location of the pain. The ANN model is the most effective in predicting the triage of AAP. Furthermore, when only seven variables are considered, including history of diabetes, etc., the model still shows good predictive performance. This is helpful for the rapid clinical triage of AAP patients and the allocation of medical resources.

Liquid biopsy multiomics enable clinical triage for early colorectal cancer diagnosis.

e15599 Background: Colorectal cancer (CRC) is the second most common cause of cancer death in the United State. Despite advancements in screening, diagnosis, and treatment, the disease remains a formidable public health issue. In 2023, approximately 153020 individuals were diagnosed with CRC and 52550 will die from the disease. In order to meet early diagnostic, we developed this multiomics assay to enable clinical triage for liquid biopsy screening. Methods: Xenonucleic acid (XNAs) are artificial genetic polymers that retain the Watson-Crick base-pairing capability and exhibit high chemical and biological stability. They can be employed as molecular clamps in RT-qPCR due to the stronger hybridizing/binding capability seen in XNA/DNA duplexes than DNA/DNA duplexes. We are developing an XNA-based multiplex RT-qPCR assay for APC, BRAF, CTNNB1, KRAS, NRAS, PIK3CA, SMAD4 and TP53 gene mutation panel and CRC 9 genes methylation panel combined with multiplex protein panel, called ColoscapeTM Plus. We enrolled in FIT positive subjects of both genders, aged 50-74, attending the CRC screening program, we tested 152 plasma samples, 75 health, 50 AA and 27 early CRC with this multiomics assay. Results: The assay results showed that its sensitivity was about 44% (95% CI: 0.302- 0.587) and specificity about 80% (95%CI: 0.689- 0.880) for advanced adenoma (AA). For CRC stage I and II, its sensitivity is about 63% and 78% with 80% specificity separately. Conclusions: The primary data has shown that this low cost and rapid turnaround time multiomics assay has a comparable sensitivity for early colorectal cancer diagnostic.

Keeping health care professionals healthy: Systematic strategies for mitigation, wellness, and burnout prevention.

9013 Background: Health care professionals are deeply committed to the compassionate care of their patients. However, the same level of caring and compassion is not typically extended to colleagues, much less toward themselves. Escalating issues include: burnout, demoralization, personal conflicts, early retirement, and high turnover. Literature and training offer few suggestions to proactively address these complexities. There is a lack of attention to creating truly healthy work environments, with strategic infrastructure to enhance leadership, communication, and relationship among team members that directly impact wellness, engagement, and productivity. We will present the Staff Leadership Model (SLM) that integrates courageous and compassionate communication to foster healthier work environments, including how to implement and sustain the necessary infrastructure. Methods: In 4 large academic cancer settings, evolving iterations of the SLM were developed since 1995, focusing on maximizing inherent human capacities. In 2007, the Department of Supportive Care Medicine (DSCM), an interdisciplinary department at City of Hope (COH), systematically integrated the SLM. Specific processes, such as onboarding and skills training, regularly scheduled, open and honest communication, ways of working agreements, and performance evaluations that actively support value-driven behaviors, were implemented. Tailored interpersonal strategies were implemented, such as small group democratic meeting structure that equalizes all members, and problem-solving skills to deal with conflict. Clinicians received annual Press Ganey-administered surveys that measured clinician engagementas well as clinician resilience, measuring the ability to disconnect from work and ability to find meaning in their work, providing early warning indicators of burnout. Results: Sustained success has been demonstrated: 1) 5 years top tier clinician/staff engagement scores, with increase from 4.24 in 2022 to 4.32 in 2023 (+0.14 vs. overall COH) ; 2) 3.96 resilience index score vs. 3.80 overall COH in 2023 (3.23 decompression score vs. COH 2.94 and 4.69 activation score vs. COH 4.65); 3) Growth of the department from 25 professionals in 2007 to 170+ in 2024; 4) A clearly-defined clinical triage created through the expertise of the professionals; and 5) Staff creativity and leadership resulting in the creation and consistent growth of innovative, market-differentiating programs. Conclusions: Results support the relationship between the SLM and healthier work environments that increase engagement and decrease burnout, as reflected by the consistently high scores. The resilience index is a new survey measure with one year data. Cause-effect relation will continue to be evaluated. Overall, SLM creates a culture of health, wellness, support, and optimized individuals and teams.

Malignant risk of pediatric Bethesda category III thyroid nodules subcategorized by nuclear atypia and other: A single institution experience.

The 2023 Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) divides AUS diagnoses into two major subcategories: atypia of undetermined significance (AUS) nuclear atypia (AUS-N) and other (AUS-O). This study aims to compare the histological outcome and malignant rate of pediatric AUS thyroid nodules classified into AUS-N and AUS-O subcategories. A search of our institutional electronic pathology database for the period from January 2012 to July 2023 was conducted to identify pediatric (<21 years old) thyroid nodules that were interpreted as AUS and subsequently had surgery. Cases were further divided into AUS-N and AUS-O subcategories. Results of follow-up surgical resections were collected. The malignant rate was calculated and compared between AUS-N and AUS-O groups. The study identified 62 thyroid nodules from 58 pediatric patients. Among these nodules, 29 and 33 were subcategorized as AUS-N and AUS-O, respectively. Both groups exhibited a female predominance and displayed a similar nodule size distribution. Histological analysis revealed 15 carcinomas in AUS-N nodules, including 11 cases of classic papillary thyroid carcinoma (PTC) and four cases of follicular type of PTC. In contrast, in the AUS-O group, a total of five carcinomas were documented, including two PTCs and three oncocytic thyroid carcinomas. Notably, the malignant rate of AUS-N nodules (52%) is significantly higher than that of AUS-O nodules (15%) (p=.002). In pediatric AUS thyroid nodules, the malignant risk in AUS-N is significantly higher than that in AUS-O. These findings may guide more appropriate clinical triage and/or improve management of pediatric patients with AUS thyroid nodules.

The Diagnostic Performance of Lung Sonography to Assess Pregnant Patients Infected With SARS-CoV-2 Based on Varied Clinical Severity

Objective: The aim of this study was to assess the diagnostic accuracy of lung ultrasonography (LUS) and a local clinical triage protocol complicated by varied clinical severities of COVID-19, in pregnant patients who had computed tomography (CT) and reverse transcription polymerase chain reaction (RT-PCR) results. Materials and Methods: This retrospective cohort study was conducted in a tertiary maternity unit, during the pandemic, between March 2020 and November 2021. The pregnant patients included in the study were diagnostically categorized as asymptomatic, mild, moderate, severe, and critical, upon hospital admission and their presenting disease severity. All of the patients, as well as those with mild and asymptomatic severity, LUS were evaluated accuracy based on their RT-PCR results. Results: In total, 1531 pregnant patients underwent a primary emergency department clinical evaluation, in the pandemic clinic. The sensitivity, specificity, positive predictive value, and negative predictive values for LUS were 70.5%, 72.45%, 88.1%, and 45.8%, respectively. Conclusion: In this large cohort of patients, LUS was shown to be quick, safe, effective, and reproducible diagnostic tool for the screening of COVID-19 pneumonia, in this obstetric practice. This clinical evidence also provided a new set of triage strategies, during the pandemic, for pregnant patients.

Assessing the efficacy of clinical triage in the prostate cancer best practice pathway

Objective: The Prostate Cancer UK timed prostate cancer diagnostic pathway is a new UK pathway to deliver a diagnosis within 28 days for patients with suspected prostate cancer. We aim to review our experience of implementing this pathway, in particular the success of clinical triage. Subjects/patients (or materials) and methods: We reviewed consecutive referrals to the prostate cancer best timed pathway from September 2019 to September 2020. Patients were triaged either straight to test multiparametric magnetic resonance imaging (mpMRI) or straight to outpatient clinic. Inappropriate triage was defined as either: patients referred straight to test mpMRI who were diagnosed with prostate cancer and managed with watchful waiting or androgen deprivation therapy (ADT) or patients who were referred straight to clinic who subsequently had an mpMRI and underwent radical treatment. Results: A total of 154/155 (99.6%) of patients were appropriately triaged in the straight to mpMRI group. A total of 124/132 (94.0%) of patients were appropriately triaged in the straight to clinic group (patients who required review to ascertain if mpMRI was appropriate and then received radical treatment). Conclusion: The implementation of a consultant supervised triage process has been successful in avoiding unnecessary mpMRIs or unnecessary clinic appointments before a patient received a diagnosis or reassuring test. We advise that outcomes are audited regularly to ensure the effective use of resources. Level of evidence: 2C

What Should Healthcare Systems Consider When Modernizing Call Centers? Early Considerations From the Veterans Health Administration.

Growing numbers of hospitals and payers are using call centers to answer patients' clinical and administrative questions, schedule appointments, address billing issues, and offer supplementary care during public health emergencies and national disasters. In 2020, the Veterans Health Administration (VA) implemented VA Health Connect, an enterprise-wide initiative to modernize call centers. VA Health Connect is designed to improve the care experience with the convenience, flexibility, and simplicity of a single toll-free number connected to a range of 24/7 virtual services. The services are organized into four areas: administrative guidance for scheduling and general inquiries; pharmacy support for medication matters; clinical triage for evaluation of symptoms and recommended care; and virtual visits with providers for urgent and episodic care. Through a qualitative evaluation of VA Health Connect, we sought to identify the factors that affected the development of this program and to compile considerations to support the implementation of other enterprise-wide initiatives. The evaluation team interviewed 29 clinical and administrative leads from across the VA. These leads were responsible for the modernization of their local service networks. PhD-level qualitative methodologists conducted the interviews, asking participants to reflect on barriers and facilitators to modernization and implementation. The team employed a rapid qualitative analytic approach commonly used in healthcare research to distill robust results. A review of the early implementation of VA Health Connect found: (1) deadlines proved challenging but provided momentum for the initiative; (2) a balance between standardized processes and local adaptations facilitated implementation; (3) attention to staffing, hiring, and training of call center staff before implementation expedited workflows; (4) establishing national and local leadership commitment to the innovation from the onset increased team cohesion and efficacy; and (5) anticipating information technology infrastructure needs prevented delays to modernization and implementation. Our findings suggest that healthcare systems would benefit from anticipating likely obstacles (e.g., delays in software implementations and negotiations with unions), thus providing ample time to secure leadership buy-in and identify local champions, communicating early and often, and supporting flexible implementation to meet local needs. VA leadership can use this evaluation to refine implementation, and it could also have important implications for regulators, federal health exchanges, insurers, and other healthcare systems when determining resource levels for call centers.

Interlaboratory comparison of high-throughput protein biomarker assay quantifications for radiation exposure classification.

In the event of a widespread radiological incident, thousands of individuals will require rapid assessment of exposure using validated biodosimetry assays to inform clinical triage. In this scenario, multiple biodosimetry laboratories may be necessary for large-volume sample processing. To meet this need, we have developed a high-throughput assay for the rapid measurement of intracellular protein biomarkers in human peripheral blood samples using an Imaging Flow Cytometry (IFC) platform. The objective of this work was to harmonize and validate the reproducibility of our blood biomarker assay for radiation exposure across three IFC instruments, two located at Columbia University (CU) and the third at Health Canada. The Center for Radiological Research (CRR) at CU served as the central laboratory and reference instrument, where samples were prepared in triplicate, labeled with two radiation responsive leukocyte biomarkers (BAX and phosphor-p53 (Ser37)), and distributed for simultaneous interrogation by each IFC. Initial tests showed that significantly different baseline biomarker measurements were generated on each instrument when using the same acquisition settings, suggesting that harmonization of signal intensities is necessary. Subsequent tests harmonized biomarker measurements after irradiation by modulating laser intensity using two reference materials: unstained samples and standardized rainbow beads. Both methods generated measurements on each instrument without significant differences between the new and references instruments, allowing for the use of one master template to quantify biomarker expression across multiple instruments. Deming regression analyses of 0-5 Gy dose-response curves showed overall good correlation of BAX and p53 values across new and reference instruments. While Bland-Altman analyses indicated low to moderate instrument biases, ROC Curve analyses ultimately show successful discrimination between exposed and unexposed samples on each instrument (AUC values > 0.85).