Dartmouth-Hitchcock Medical Center Research Articles

Asynchronous Consultations in Geriatric Psychiatry: Experiences in a Rural Academic Health System

More Americans are living longer and entering old age than ever before. The number of adults age 65 years and older in the United States (US) is projected to grow to more than 83.7 million by 2050. Today's 65-year-olds can expect to live, on average, another 19.4 years. However, the outcomes of extending longevity have not kept pace with the US healthcare system's ability to improve mental and cognitive well-being. The 2012 Institute of Medicine (IOM) report estimates that 14%–20% of the elderly population is afflicted by a mental health or substance use condition, including dementia-related behavioral and psychiatric symptoms. Despite this clear growing demographic need, the number of physicians formally trained in geriatric psychiatry and geriatric medicine is shrinking, limiting the availability of specialty care to this vulnerable demographic. This disparity is magnified in rural settings and/or historically disadvantaged areas/populations where access to geriatric psychiatrists and geriatric medicine specialists may be severely limited, if available at all.This session will cover the development, implementation, and preliminary outcome measures of two asynchronous geriatric psychiatry consultation programs, one EMR based e-consultation program and one on-demand telephone consultation program, that have been developed by geriatric psychiatrists at Dartmouth-Hitchcock Medical Center. Our aim is to demonstrate how similar asynchronous geriatric psychiatry consultation programs could be feasibly implemented at essentially any site or within any health system to begin to address the limited availability of specialty geriatric psychiatry expertise, especially in rural and/or historically disadvantaged areas/populations.

Correlates of level of satisfaction among primary health care workers in Albania

Aim: The aim of this study was to explore the level of satisfaction of primary health care staff in Albania and the factors associated with it. Methods: A cross-sectional study was conducted in Tirana city, the Albanian capital, from 11 November 2020 until 25 November 2020. Among all health centers (HCs) and health centers of specialties (HCSs) of Tirana Municipality, there were selected randomly a HC in rural areas, a HC in urban areas and one HCS. All the staff (doctors and nurses) being present at the time of data collection was interviewed, using an international standardized tool (the Dartmouth-Hitchcock Medical Center instrument) assessing the satisfaction with various elements of the work in PHC, validated in Albanian. A total of 102 PHC staff were included in the study. Binary logistic regression was used to assess the association of staff satisfaction with independent factors. Results: The aspects of work most appreciated by PHC staff (% satisfied or very satisfied) were: respectful treatment by colleagues (78.2%), staff morale and their positive attitude towards work (73.2%). The most disliked aspects of work by PHC staff (% dissatisfied or very dissatisfied) were: current salary (60.8%), stress at work (38.3%), physical and medical infrastructure in the institution (27%). Staff in rural HCs, older staff, females and nurses and family doctors are more likely to be satisfied compared to their respective colleagues. Conclusion: Our findings suggest various factors associated with the satisfaction of PHC staff in Albania. These findings could be guiding future efforts aiming to improve the work conditions of the professionals working in primary health care in Albania.

Medical Students and the Informal Radiology Curriculum: Adopting the Emergency Radiology Triage Assistant Program (ER-TAP) Amid COVID-19

The Emergency Radiology Triage Assistant Program (ER-TAP) at the XXX was established in 2010 and has offered medical students paid employment to work as triage assistants in the emergency radiology reading room. ER-TAP is a radiology call assistant program which was modeled after the program first developed by the Yale Department of Radiology (1). Similar programs have been implemented at Dartmouth-Hitchcock Medical Center and the Hospital of the University of Pennsylvania (2,3). ER-TAP at XX offers medical students an opportunity to work weekdays and holidays from 6:00 pm to midnight.

ALIGNING GERIATRIC EMERGENCY CARE WITH THE 4MS OF THE AGE-FRIENDLY HEALTH SYSTEM FRAMEWORK

Abstract In 2019 Dartmouth-Hitchcock Medical Center (DHMC) was recognized by the Institute for Healthcare Improvement as a level 2 Age Friendly Health System (AFHS). Over the past two years DHMC leveraged the strengths of the AFHS to become a level 1 American College of Emergency Physicians (ACEP) accredited geriatric emergency department. Leveraging the paradigm of 4M care including: What Matter’s Most, Mobility, Medication and Mentation, DHMC achieved alignment between the geriatric emergency department accreditation and the AFHS framework. Pathways focusing on the 4Ms have been identified as a model to improve transitions of care post ED. This talk will focus on lessons learned and outcome data specific to the creation of the 4M pathways and transitions of care,

THE IMPLEMENTATION OF GERIATRIC EMERGENCY DEPARTMENT ACCREDITATION (GEDA) IN RURAL CRITICAL ACCESS HOSPITALS

Abstract The largely rural setting of Northern New England offers unique challenges to implementing improved acute care for the growing geriatric population. Northern New England is one of the United States’ most rapidly aging regions, with Vermont and New Hampshire being the second and third oldest US states respectively by median age (U.S. Census 2017). There is a need to expand innovations in geriatric emergency medicine to reach older adults in rural areas such as Northern New England. Dartmouth-Hitchcock Medical Center and the West Health are collaborating on a project leveraging telehealth to extend the reach of a GED to rural hospitals, as well as investigate the opportunities for scaling and sustaining this concept to other rural facilities across Northern New England and throughout the country. This symposium will focus on our experience implementing a hub and spoke model to achieve our goal of improving the care of older adults in rural emergency departments.

Abstract 11336: Optimized Machine Learning Models by Identification of Candidate Predictors for 30-Day Readmission Risk Prediction Among Patients Hospitalized for Acute Myocardial Infarction

Introduction: Acute myocardial infarction (AMI) is a leading cause of death in the US. More than 800,000 adults experience an AMI each year, and up to 20% of the patients are rehospitalized within 30 days. Predicting 30-day readmission accurately could help clinicians identify high-risk patients and tailor treatment accordingly. Hypothesis: The neural network will generate the best performing model for predicting 30-day readmissions following an AMI. Methods: The cohort included adults hospitalized with incident AMI (6,506) at Dartmouth Hitchcock Medical Center between 2011-2016. Patients who died before discharge (236) were excluded, leaving 6,270 patients. Our outcome was 30-day hospital readmissions. Candidate features were selected using an Extra Tree Classifier. Selected features included demographics and clinical characteristics. Multiple imputation using K-Nearest Neighbors was performed. Five machine learning models were implemented: logistic regression, LASSO, neural network, random forest, and XGBoost. Models were evaluated on a hold-out test-set, using AUROC. Results: Among the cohort, about 36% were female, the majority were white and non-Hispanic, and the 30-day readmission rate was 6.2%. Neural network was the best performing model with an AUROC of 0.80 (95% CI: 0.75, 0.84). However, the XGBoost and random forest models performed similarly with AUROCs of 0.80 (95% CI: 0.76, 0.85) and 0.79 (95% CI: 0.75, 0.84), respectively. The logistic regression and LASSO models performed poorly, with AUROCs below 0.5, respectively. Important predictor variables were major depression, race, and antidepressant at discharge. Conclusions: Neural network, random forest, and XGBoost models predicted 30-day readmission following an AMI with good performance. Across these models, depression was an important feature for predicting 30-day readmissions. Results reinforce the importance of mental health among patients hospitalized for AMI.

S462 Impact of Esophageal Biopsies During Index Endoscopy for Food Bolus Impaction on Rates of Patient Outpatient Follow-Up and Recurrence of Food Bolus Impaction: A Retrospective Study

Introduction: Food bolus impaction (FBI) is a gastroenterological emergency requiring urgent upper endoscopy, of which up to half of cases may be due to eosinophilic esophagitis (EoE). EoE is defined by eosinophilic infiltration in the esophageal mucosa without secondary causes, which can lead to inflammatory and structural complications. Patients with EoE should have outpatient follow up as it can be effectively managed with proton pump inhibitors, swallowed topical steroids, and food elimination diets. This study aims to determine baseline rate of esophageal biopsy during index endoscopy for FBI, as recommended by American College of Gastroenterology clinical guidelines, and evaluate the impact of biopsies on rate of outpatient follow-up and repeat FBI presentations. Methods: This retrospective cohort chart review study evaluated patients over age 18 who presented with acute FBI at Dartmouth-Hitchcock Medical Center from 2015 to 2020. Patients with incidentally found food boluses and with non-food foreign body impactions were excluded. Simple count and chi-squared analyses were performed, with a p-value < 0.05 indicating significance. Results: 122 patients met inclusion criteria, with a mean age of 64 years old and 65% being male. Approximately half had a prior FBI though only 20% had a previous outpatient gastroenterology appointment. Baseline rate of esophageal biopsy during index endoscopy was 16%, of which 10% revealed a new diagnosis of EoE. After index endoscopy, 55% presented for outpatient follow-up and 51% underwent repeat endoscopy. 11% presented again to the emergency department for a recurrent FBI, with 31% of those patients having previously declined an outpatient visit. A greater percentage (65%) of patients with index biopsy completed an outpatient follow-up compared to those without a biopsy (53%) (p=0.32). There was minimal difference between those who had a biopsy and presented again with emergent FBI (15%) compared to those without a biopsy who re-presented (10%) (p=0.49). Conclusion: The baseline rate of esophageal biopsy during initial endoscopy was low, suggesting missed opportunities to diagnose EoE early. A greater number of index biopsy cases followed up in outpatient clinic, though was not statistically significant. Findings could be affected by low volume of index biopsies. Further research is needed to identify barriers to outpatient follow-up and treatment adherence for patients diagnosed with EoE on index biopsy.

Vivianne Tawfik, M.D., Ph.D., a Recipient of the 2022 James E. Cottrell, M.D., Presidential Scholar Award

Vivianne Tawfik, M.D., Ph.D., a Recipient of the 2022 James E. Cottrell, M.D., Presidential Scholar Award

Predicting oncogene mutations of lung cancer using deep learning and histopathologic features on whole-slide images

Lung cancer is a leading cause of death in both men and women globally. The recent development of tumor molecular profiling has opened opportunities for targeted therapies for lung adenocarcinoma (LUAD) patients. However, the lack of access to molecular profiling or cost and turnaround time associated with it could hinder oncologists' willingness to order frequent molecular tests, limiting potential benefits from precision medicine. In this study, we developed a weakly supervised deep learning model for predicting somatic mutations of LUAD patients based on formalin-fixed paraffin-embedded (FFPE) whole-slide images (WSIs) using LUAD subtypes-related histological features and recent advances in computer vision. Our study was performed on a total of 747 hematoxylin and eosin (H&E) stained FFPE LUAD WSIs and the genetic mutation data of 232 patients who were treated at Dartmouth-Hitchcock Medical Center (DHMC). We developed our convolutional neural network-based models to analyze whole slides and predict five major genetic mutations, i.e., BRAF, EGFR, KRAS, STK11, and TP53. We additionally used 111 cases from the LUAD dataset of the CPTAC-3 study for external validation. Our model achieved an AUROC of 0.799 (95% CI: 0.686–0.904) and 0.686 (95% CI: 0.620–0.752) for predicting EGFR genetic mutations on the DHMC and CPTAC-3 test sets, respectively. Predicting TP53 genetic mutations also showed promising outcomes. Our results demonstrated that H&E stained FFPE LUAD whole slides could be utilized to predict oncogene mutations, such as EGFR, indicating that somatic mutations could present subtle morphological characteristics in histology slides, where deep learning-based feature extractors can learn such latent information.

Idiopathic Spinal Subarachnoid and Subdural Hemorrhage in a Patient With Eosinophilic Granulomatosis and Polyangiitis

BACKGROUND AND IMPORTANCE: Development of concomitant spontaneous subdural hematoma (SSDH) and spontaneous subarachnoid hemorrhage (SSAH) in the spine without inciting trauma, iatrogenic event, or underlying cerebrovascular malformation is rare. It is rarer still for these to present in a patient with eosinophilic granulomatosus with polyangiitis. CLINICAL PRESENTATION: Herein, we present the case of a 73-year-old woman with a complex medical history, including possible eosinophilic granulomatosus with polyangiitis, who was admitted to the medical intensive care unit at Dartmouth-Hitchcock Medical Center and was found to have concominant SSAH and SSDH. CONCLUSION: This case serves as a review of the common and uncommon symptoms of IS-SAH with SSDH to improve recognition of this acute neurosurgical emergency in the medical intensive care setting.

Machine Learning Approaches for Hospital Acquired Pressure Injuries: A Retrospective Study of Electronic Medical Records.

BackgroundMany machine learning heuristics integrate well with Electronic Medical Record (EMR) systems yet often fail to surpass traditional statistical models for biomedical applications.ObjectiveWe sought to compare predictive performances of 12 machine learning and traditional statistical techniques to predict the occurrence of Hospital Acquired Pressure Injuries (HAPI).MethodsEMR information was collected from 57,227 hospitalizations acquired from Dartmouth Hitchcock Medical Center (April 2011 to December 2016). Twelve classification algorithms, chosen based upon classic regression and recent machine learning techniques, were trained to predict HAPI incidence and performance was assessed using the Area Under the Receiver Operating Characteristic Curve (AUC).ResultsLogistic regression achieved a performance (AUC = 0.91 ± 0.034) comparable to the other machine learning approaches. We report discordance between machine learning derived predictors compared to the traditional statistical model. We visually assessed important patient-specific factors through Shapley Additive Explanations.ConclusionsMachine learning models will continue to inform clinical decision-making processes but should be compared to traditional modeling approaches to ensure proper utilization. Disagreements between important predictors found by traditional and machine learning modeling approaches can potentially confuse clinicians and need to be reconciled. These developments represent important steps forward in developing real-time predictive models that can be integrated into EMR systems to reduce unnecessary harm.

Learning Outcomes from an Academic Internal Medicine Morbidity and Mortality Conference.

Morbidity and Mortality (M&M) conferences are widely utilized clinical teaching forums across the USA. Recent literature demonstrates heterogeneity in the educational objectives of M&M, with prior authors suggesting a variety of overarching purposes of the conference, including teaching quality improvement methods; promotion of patient safety; enhancement of clinical knowledge and skills; and reflection on humanistic aspects of medical practice. There is less information in the published literature regarding learning outcomes of M&M participants. The goal of this study was to describe learning outcomes from the Internal Medicine Morbidity, Mortality, and Improvement conference at Dartmouth-Hitchcock Medical Center. Directed content analysis of learning outcomes statements. Internal medicine physicians who requested Maintenance of Certification credit for one or more conference sessions during the 2017-2018 academic year. Two independent reviewers analyzed 347 learning outcomes statements submitted by 49 conference participants. The reviewers used an iterative process to code content themes (the category of medical or healthcare delivery knowledge learned) and learning implementation domains (the context or manner in which learning would be applied), and to identify comments regarding the shared experience of practicing medicine. Seventy-eight percent of comments described learning related to clinical knowledge and skills, and 28%, 34%, and 9% of comments described learning related to clinical reasoning skills, systems knowledge, and the need for systems change, respectively. Most conferences generated learning within a variety of themes and across multiple domains. Sixty-four percent of conferences included at least one reflection on the shared experience of practicing medicine. Participants derived several types of learning from this Internal Medicine M&M conference. Although clinical knowledge and skills represent the most common type of learning, the conference also produces rich learning in other domains as well as reflections on the humility, challenges, and meaning of being a physician.

Electron flash-RT program in clinical setting for human translation.

e13596 Background: A FLASH-RT program was established at Dartmouth-Hitchcock Medical Center in minimally-modified clinical setting by joint efforts of biomedical engineering, radiation oncology, radiation biology and medical physics teams. Various projects on dosimetry, chemical sensing, molecular profiling, software/hardware development, and translational studies have been conducted. The aim is to share logistical considerations and experience on running a FLASH-RT program to support institution-wide academic activities with an ultimate goal of treating human patients with FLASH-RT in 2022. Methods: A linac was converted in the clinical setting by qualified engineers to deliver an ultra-high dose rate (UHDR) electron beam. Routine safety and dosimetry checks were done by physicists for every reversible conversion. Long-term record-keeping and retrospective surveys were carried out to demonstrate the feasibility, safety, stability and accuracy of this dual-purpose (FLASH and conventional RT) approach. Comprehensive failure mode and effects analysis (FMEA) has been completed to systematically evaluate safety related considerations. A treatment planning system (TPS) has been developed in Varian Eclipse to facilitated comparative studies. The FLASH-capable linac has been utilized as shared resource to support institution-wide academic activities as well as normal clinical treatments. Results: With its safety (no accident or FLASH-related malfunction), flexibility (&gt; 100 conversions in 2 years), reliability (̃6000 hours in flash mode and ̃5x105 Gy accumulative dose delivered at isocenter) and accuracy (̃5% conversion-to-conversion variations) demonstrated by commissioning, long-term user experience and comprehensive FMEA analysis, the FLASH-RT platform has been actively utilized for researches in six major categories 1) FLASH beam dosimetry; 2) real-time beam delivery monitoring and control; 3) oximetry and chemical sensing; 4) preclinical/translational small/large animal treatment with tumor control and normal tissue complication endpoints, 5) treatment plan and delivery optimization; 6) the design of phase I/II trials. Key findings in each category will be reported. Conclusions: A FLASH-RT program in clinical setting is established at Dartmouth with joint efforts, promoting collaborative projects to advance FLASH-RT to clinical treatment. The system has been reliably utilized for over two years for mechanistic as well as translational studies and support a phase I/II trial treating cutaneous lymphoma with eFLASH-RT.

Bullous Pemphigoid: Readability, Quality, and Treatments of Online Health Resources

Bullous Pemphigoid: Readability, Quality, and Treatments of Online Health Resources

A pilot study assessing the feasibility and acceptability of Project UPLIFT adapted for patients with psychogenic nonepileptic seizures

BackgroundPeople with psychogenic nonepileptic seizures (PNES) are at elevated risk of multiple psychiatric comorbidities. Current treatment gaps highlight the need for time-limited, distance-delivered interventions that can be cost-effectively delivered to patients with PNES. Project UPLIFT is a self-management program addressing mood problems in epilepsy that has not previously been adapted for PNES. The purpose of this study was to assess the feasibility and acceptability of UPLIFT for patients with PNES. MethodsProject UPLIFT was minimally adapted by a multidisciplinary research group at Dartmouth-Hitchcock Medical Center. Participants were recruited into a nonrandomized pilot study and were assessed at baseline, after completing the intervention, and at 1-month follow-up. ResultsThe intervention was adapted and delivered without additional modification. A total of eight participants enrolled in the trial, and six participated in the first session. All six participants who started UPLIFT completed the intervention, with a 98% attendance record across the eight sessions. At follow-up, 100% reported that UPLIFT provided useful tools for daily life. All participants continued using UPLIFT after completing the program, and 83% felt it helped with seizure management. Preliminary results provide proof-of-concept for future efficacy trials. ConclusionThis study provides initial support for the feasibility and acceptability of Project UPLIFT, minimally adapted for patients with PNES. As a time-limited program that can be distance-delivered to groups of patients, UPLIFT may be well-suited for the healthcare environment brought on by the COVID-19 pandemic.

Understanding secondary overtriage for neurosurgical patients in a rural tertiary care setting

ObjectiveExcessive interfacility transfer to a tertiary care facility of minimally injured patients for subspecialty evaluation leads to overuse of resources and is referred to as secondary overtriage (SO). Little is known regarding the epidemiology of SO in rural settings, particularly for patients with a mild head injury who may be safely managed without admission to level I trauma centers. MethodsIn order to determine the rate of SO for neurosurgical patients with Glasgow Coma Scale (GCS) of 13–15 referred to a rural level 1 trauma center, we conducted a retrospective chart review of 224 patients evaluated for potential transfer to Dartmouth-Hitchcock Medical Center from January 1, 2014 through December 31, 2014. SO was defined as any admission where a patient was transferred from an outside facility, had a length of stay shorter than 48 h, did not require neurosurgical intervention, and was alive at the time of discharge. ResultsOf the 224 patients evaluated, 163 patients were transferred. Of the 163 patients included in this study, 43 (26.4%) met criteria for SO, 59 (36.2%) patients met criteria for intervention, and 61 (37.2%) patients met criteria for observation. ConclusionsApproximately a quarter of the total patients who are transferred to a rural level I trauma center for neurosurgical evaluation are minimally injured, do not require neurosurgical intervention, and are discharged within 48 h of presentation. Management at their referring facility with remote neurosurgical consultation is likely safe in this population. Understanding the rate of SO in neurosurgical patients and risk factors present in this group can better guide future transfer policies at rural medical centers.

Two weeks of twice-daily prism adaptation treatment does not improve posture or gait in Parkinson\u2019s disease: a double-blind randomized controlled trial

BackgroundGait difficulties in Parkinson’s disease have been related to problems shifting the center of gravity forward. We previously showed reduced forward stepping latencies for people with Parkinson’s disease after one session of adaptation to upward visual shifts, which produces downward motor after-effects and potentially shifts the center of gravity forward. Here we tested if repeated prism adaptation improved gait and postural control in Parkinson’s disease through a parallel, double-blind, randomized, sham-controlled trial.MethodsWe recruited participants with idiopathic Parkinson’s disease aged 40–85 and meeting any one of three clinical criteria: (1) Hoehn and Yahr Stage II.5–IV; (2) scoring > 0 on the gait, freezing of gait, and/or postural stability items of the Movement Disorder Society Unified Parkinson’s Disease Rating Scale; or (3) Timed Up and Go > 12 s. Sealed envelope style randomization allocated participants to two weeks of twice-daily prism adaptation or sham treatment. Participants, care givers, and those assessing the outcomes were blinded to group assignment. Primary outcomes were changes in postural control measured using the Berg Balance Scale and the Limits of Stability, Sensory Organization, and Motor Control tests from the Smart EquiTest system. Secondary outcomes included other physiotherapy and questionnaire measures. Outcomes were assessed at the Dartmouth Hitchcock Medical Center immediately before and after the treatment period, with further long-term postal follow-up over 3 months. Outcomes were analyzed using analyses of variance with follow-up t tests.ResultsEighteen participants were allocated to undergo prism adaptation, of which sixteen were analyzed. Thirteen participants were allocated to undergo sham treatment, and all were analyzed. The prism adaptation group showed increased forward stepping velocity on the Limits of Stability test (pre: M=2.33, SEM=0.24; post: M=2.88, SEM=0.26; t(15)=3.2, p=.005, d=.819). The sham group showed no such change (pre: M=2.13, SEM=0.22; 1d post: M=2.24, SEM=0.22; t(13)=.636, p=.537, d=.176). However, there were no group differences for any other outcome measures and no indications that prism adaptation produced functional improvements in posture, gait, or activities of daily living.ConclusionsPrism adaptation does not improve gait or postural control in Parkinson’s disease.Trial registrationClinicalTrials.govNCT02380859. Registered prospectively on 5 March 2015.

Abstract 13049: Peripheral Artery Disease In Patients Undergoing Transcatheter Aortic Valve Replacement In Northern New England

Background: Transcatheter aortic valve replacement (TAVR) is now more common than surgical aortic valve replacement (SAVR) for severe aortic valve stenosis. In the general population, the prevalence of peripheral artery disease (PAD) in patients undergoing transfemoral TAVR is 25%, with higher risk of adverse outcomes including bleeding, readmission, myocardial infarction, and all-cause mortality. The objective of this study was to examine the prevalence of peripheral artery disease in patients undergoing TAVR at Dartmouth-Hitchcock Medical Center (DHMC). Methods: We performed a retrospective observational study of 569 patients undergoing TAVR between 1/1/2012-4/8/2020. PAD was defined as presence of aorto-iliac, femoral-popliteal, below-knee tibial, reno-vascular, mesenteric, cerebrovascular, aortic disease, or upper extremity arterial disease noted in the patient’s chart prior to TAVR. Results: Of 569 TAVR patients, 19% (n=109) had concomitant PAD. Of these, 70 patients (64%) were established with a vascular specialist and 54 (50%) had a prior intervention relating to their disease. Figure 1 characterizes the distribution and location of PAD among patients. Conclusion: In a Northern New England population, 1 in 5 patients undergoing TAVR had concomitant PAD. Cerebrovascular and peripheral vascular disease were most common. Understanding PAD epidemiology is of paramount importance with respect to vascular access and longitudinal management in a vulnerable aortic stenosis population.

Thrombin Generation As a Predictive Biomarker for Venous Thromboembolism in Patients with Pancreatic and Lung Cancer Undergoing Systemic Therapy

Introduction:Venous thromboembolism (VTE) is common in patients with active malignancy. Whilethromboprophylaxis can mitigate this risk, current guidelines do not support routine use as the benefit ismodest and maybe negated by an increase in bleeding complications. However, there is significantvariation in VTE risk within the cancer population, thus prophylaxis can be considered in high-riskpatients. The Khorana score (KS) is validated to predict chemotherapy-associated VTE; however, it hasseveral limitations including variable performance based on tumor type as well as a static riskcategorization, based on pre-chemotherapy laboratory data, rather than a continuous assessment.Thrombin generation (TG) is an emerging biomarker that assesses global coagulation activation andpredicted increased VTE risk in cancer patients in the Vienna Cancer and Thrombosis Study (PMID:21464402). This study enrolled a heterogeneous cancer population, however, peak TG was not reportedby tumor type nor were the TG levels monitored over time in response to systemic therapy. The primaryaim of this study was to assess the relationship between TG (both peak TG and endogenous thrombinpotential, ETP) and KS. Secondary outcomes were to evaluate the impact of systemic therapy on peakTG and ETP levels over time and to assess the relationship between TG and clinical outcomes.Methods:This was a prospective study that enrolled adults with newly diagnosed, locally advanced or metastaticadenocarcinoma of the lung or pancreas. All patients received their care at Dartmouth HitchcockMedical Center. Those with a history of active VTE or use of full dose anticoagulant within 30 days priorto enrollment were excluded. After informed consent, KS was calculated and blood was collected insodium citrate tubes at 3 different time points (at initiation of therapy, and at the beginning of the 2 ndand 3 rd cycles of systemic therapy). Platelet-poor plasma was prepared by centrifugation and stored at -80°C until analysis by calibrated automated thrombogram (CAT; Thrombinoscope BV, Maastricht,Netherlands) using 1 pM tissue factor and 4 uM phospholipids to trigger coagulation reactions.Measurements were performed in triplicate for each specimen, and raw data were converted to peakTG and ETP. Information about VTE events, response to treatment and survival was obtained by chartreview. Mean and standard deviation were calculated for continuous variables and unpaired T test wasused for statistical analysis.Results:We report the results from 32 participants (17 lung, 15 pancreas). The majority of patients hadmetastatic disease (94%) and all received systemic therapy. The median age was 67 and 56% ofparticipants were male. The KS breakdown for the cohort was: KS1, 25%; KS2, 53%; KS3, 19%; KS4, 3%.Mean peak TG was 279, 352, 487 and 325 nmole and mean ETP was 1320, 1653, 2252, 1726 nmole/minfor KS 1, 2, 3 and 4, respectively. Initial peak TG and ETP levels were significantly higher in the KS ≥2group compared to those with KS=1 (peak TG: 384 ±135 vs 279 ±82, p 0.047; Initial ETP: 1806 ±632 vs1320 ±286, p 0.045). There were 8 VTE events (25%) with all but one occurring in the pancreatic cancercohort; all events occurred in KS 2 and 3 groups (75% and 25% respectively). No statistically significant difference was observed for initial peak TG or ETP in those with VTE vs those without (Peak TG: 342±171 vs 363 ±119, p 0.701; ETP: 1736 ±720 vs 1667 ±571, p 0.783). Both peak TG and ETP decreased inresponse to systemic therapy (initial peak TG vs final: 376 ±115 vs 225 ±125, p 0.0001; initial ETP vs final:1799 ±498 vs 1254 ±423, p 0.0003). No significant difference in survival was noted based on initial ETPlevel of <1500 nmole/min vs ≥1500 nmole/min (574 ±620 vs 287 ±227, p 0.069).Conclusions:There appears to be an association between KS and peak TG and ETP. In addition, both peak TG and ETPdeclined in response to systemic therapy, suggesting that the degree of coagulation activation is relatedto tumor burden. Our findings in this small study support further investigation of TG for VTE riskassessment in cancer patients. Future strategies incorporating TG into risk stratification models mayallow oncologists to better identify the population that may benefit most from pharmacologicthromboprophylaxis.Funding: Northern New England Clinical Oncology DisclosuresNo relevant conflicts of interest to declare.

27971 Single institution experience with a dermatology hospitalist service

The dermatology hospitalist model was first defined by the Society of Dermatology Hospitalists in 2009; however, scant evidence exists to demonstrate its institutional benefit. Dartmouth-Hitchcock Medical Center established a formal dermatology consult service in early 2019, for which we named a director, ensured faculty continuity, and enacted earlier rounding times to communicate with primary teams. Here, we examine the value of an inpatient dermatology consult service at our institution. Both utilization and value of the service were determined for the year 2019 through analysis of the number of consults, consulting team, provisional diagnosis of primary team, final diagnosis of the dermatology service, performance of inpatient biopsy, and scheduled follow-up for hospitalized patients.