Dartmouth-Hitchcock Medical Center Research Articles

Barriers to Engagement with Gastrointestinal Health Management

BackgroundAlthough behavioral interventions have been effective in gastrointestinal (GI) conditions, barriers exist in implementing these interventions into clinical practice. The majority of previously published studies have focused on workforce limitations and have not considered individual and social determinants of health (SDoH) factors that can impact engagement in GI behavioral healthcare.AimsTo characterize barriers to engagement in appointment attendance and health management, explore individual and SDoH factors impacting GI behavioral healthcare engagement, and identify barriers that occur more often for patients with SDoH-related vulnerability (low health literacy and/or financial insecurity).MethodsA survey was distributed to adult patients who had been seen in Gastroenterology at the Dartmouth-Hitchcock Medical Center from June 2022 to December 2022.ResultsOne hundred participants [mean age = 58 years, 57.1% women] completed the survey. SDoH vulnerability was present in 32.3% of the population. For the entire sample, 73% reported at least one barrier to accessing care and 75% reported at least one factor which impacted health management. Those with SDoH vulnerability reported significantly more barriers to attending appointments and to managing health. In addition, they were significantly more likely to endorse physical health problems, difficulty affording medical bills, pain, mobility issues, trauma experiences, significant stress, and difficulty with concentration.ConclusionsPatients in a GI clinic reported multiple barriers to accessing care and participating in health management. Innovative, multi-level strategies are needed to address barriers to ensure that all patients are able to obtain quality GI behavioral health services.

Implementation of a Multidisciplinary Team to Improve the Quality of Care for Older Surgical Patients at a Rural Academic Medical Center.

For the first time in history, people age older than 65 years make up >20% of the non-metro population, compared with 16% of the metro population. From 2010 to 2020 the nonmetro population age older than 65 years grew by 22%, while the working-age population declined by 4.9%, and the population aged under 18 years declined by 5.7%.1,2 Multidisciplinary geriatric surgical programs are an increasingly recognized approach to the care of older surgical patients and preliminary data suggest they can reduce length of stay. Although rural areas have the greatest proportion of patients age older than 65 years, implementation of such programs faces special challenges in rural settings with limited resources. Dartmouth-Hitchcock Medical Center is one of the most rural academic centers in the United States. Challenges include a shortage of geriatric-trained providers, long distances to access primary care and subspecialists, and extremely limited postacute care options and skilled nursing facility beds. To address the unique needs of our provider and patient population we began with a development period where we conducted stakeholder interviews. Using these data, we mapped out a workflow and developed pilot projects to address different portions of the workflow, such as preoperative screening for frailty and cognitive impairment, interdisciplinary weekly case conferences, proactive case management, delirium and geriatric surgery postoperative pathway order sets, and a variety of tools for reorientation and delirium management. Herein we describe the process of development and pragmatic clinical implementation of geriatric-focused care for older surgical patients in our rural tertiary center, including some of the main challenges we faced and the strategies we undertook to overcome them, and some of our early patient centered and clinical outcomes. This information may assist other institutions as they design geriatric-focused surgical programs to address the growing population of older adults and the need for compliance with state legislation. The clinical program described is not a research study, and the outcome data we report is for the purpose of description, and should not be interpreted as a rigorous research investigation of the effect of our intervention.

Abstract PO4-01-14: Effects of neoadjuvant endocrine therapy on early-stage breast cancer: a retrospective cohort study of patients during the COVID19 pandemic

Abstract Background: At the height of the COVID19 pandemic, many surgical procedures across the country including at Dartmouth-Hitchcock Medical Center (DHMC) were postponed to redirect resources. The American Society of Breast Surgeons shared guidelines to manage patients experiencing surgical delays. For estrogen receptor positive (ER+), HER2- early-stage tumors, recommendations were to use neoadjuvant endocrine therapy (NET). NET is usually under-utilized for locally advanced hormone receptor-positive breast cancer (BC) despite literature concluding that chemotherapy and NET have comparable clinical response rates but with lower toxicity for the latter. The COVID19 pandemic thus presented a unique occasion to extend the use of NET for localized BC. Herein, we examined the rate of adoption of NET and associated patient/tumor characteristics at DHMC during the lockdown. Methods: A retrospective analysis of patients diagnosed with early-stage ER+, HER2- BC between December 2019 and June 2020. Data extracted from chart review included age, menopausal status, tumor stage/grade, body mass index (BMI), and adherence to adjuvant endocrine therapy (ET). A “delay in surgery” was defined as days between surgical consult and surgery over 14 days. Descriptive statistics were applied to data collected on patient/tumor characteristics, and the number of patients accepting or declining NET. Results: 109 cases were identified within the study period, with 42 found with surgical delay. The median age of the delayed group was 62 and the majority of patients were post-menopausal. 36 patients received NET with most started on an aromatase inhibitor. Median BMI was 28.5. Median duration of treatment was 39.5 days. Three cases were noted to have some decrease in cellularity for pathologic partial response and four had no definite response. The majority of patients on NET had no change in pathological grade. Out of the 36 patients who had NET, 30 (83.3%) remain on adjuvant ET. In contrast, only 59% (43 out of 73 patients) of those who did not receive NET were on adjuvant ET at the median follow up time of 2.75 years post-therapy initiation. Of note, 16 (out of 73) patients who were not on NET were either lost to follow up, had their care transferred to another health network or were not offered ET. Results are reported in Tables 1 and 2. Conclusions: Delays in surgery during the COVID19 crisis resulted in increased use of NET in early-stage BC. Impact of NET use includes possible increase in long-term adherence to adjuvant ET with a trend towards statistical significance (odds ratio of 3.25, p=0.08). A larger sample size is needed to evaluate this finding and a prospective trial is in progress (NCT04568616). NET is a viable treatment option during surgical delays and may help increase long-term adherence to adjuvant ET. Table. Patient and tumor characteristics Table. Number of patients on adjuvant ET Citation Format: Mary Chamberlin, Todd Miller, Marie Anne Christine Buteau, Steven Tau. Effects of neoadjuvant endocrine therapy on early-stage breast cancer: a retrospective cohort study of patients during the COVID19 pandemic [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO4-01-14.

How a Scoville aneurysm clip fought in the Cold War and helped to establish neurosurgery in Iceland.

It can be said that the specialty of neurosurgery in Iceland had its beginnings on November 30, 1971, with the arrival of a huge American C-130 Hercules aircraft. It was carrying a small package containing Scoville aneurysm clips. They were sent to the late Bjarni Hannesson (1938-2013), who had received his neurosurgical training in 1967-1971 at the Dartmouth-Hitchcock Medical Center (then known as Mary Hitchcock Memorial Hospital and located in Hanover, New Hampshire). He used one to clip the right posterior communicating artery aneurysm of a 34-year-old fisherman, who recovered well. The apparent reason for the use of such a huge aircraft for such a small payload is to be found in the sociocultural politics of the Cold War. It involved the continued presence of the American base at Keflavík, where the C-130 landed. The base was under pressure to be closed by Iceland's left-leaning, nominally communist government. The C-130's arrival generated welcome publicity for the continued operation of the American base, which is still there.

Evaluating the Impact of Telehealth on Carbon Footprint During Three Phases of the Pandemic at a Rural Academic Medical Center.

Background: Climate change is primarily driven by greenhouse gases, such as carbon dioxide (CO2). Telehealth visits have been found to mitigate carbon emissions by reducing patient and physician transport. Dartmouth Hitchcock Medical Center (DHMC) is the most rural academic medical center in the country, serving a population where the majority of patients reach the hospital by car. No large study or systematic review has evaluated the impact of telehealth visits on CO2 emissions (CO2e) across multiple specialties in a purely rural setting. Further, no sizable rurally focused study has compared CO2e avoided during the various stages of the pandemic. Methods: We extracted data for all outpatient telehealth visits at DHMC from three periods: prepandemic, early pandemic, and late pandemic. The extracted data included the pandemic stage of the virtual visit, the type of visit (video or telephone), the specialty, and the distance from the patient's home to DHMC. Results: The total CO2e avoided among all three pandemic stages analyzed in this study was 23,658,898 kg (n = 251,832). During period 1, the mean driving distance = 159.0 miles; CO2e avoided per encounter = 128.3 kg; period 2, mean distance = 84.85 miles; average CO2e avoided per encounter = 68.47 CO2e kg; and period 3, mean distance = 112.9 miles; average CO2e avoided per encounter = 91.08 kg. Conclusions: This data supported long distances to the medical center and large savings in CO2e avoided across multiple specialties that spanned all pandemic periods. Further, this level of averted emissions could translate to over $3M in saved fuel costs and the avoidance of six excess deaths. While discussions of the future of telehealth commonly focus on access, use cases, technology, costs, and satisfaction, the impact on carbon footprint is an additional important metric, particularly in largely rural regions.

People with IBD evidence more microarousals during sleep architecture assessments

ObjectivePoor sleep is common in inflammatory bowel disease (IBD) and may be associated with overall worse disease outcomes. While the sleep/IBD literature is growing, the data are often self-reported. Further,...

Abstract 16941: Breast Artery Calcification for Primary Prevention of Major Adverse Cardiovascular Outcomes: A New Risk Stratification Tool for Women

Introduction: Reductions in cardiovascular mortality among women lag behind men. There is a critical need for improved primary prevention of major adverse cardiovascular events (MACE) in women. We examined if breast artery calcification (BAC), detected on screening mammograms, prognosticates long-term MACE. Methods: In this retrospective study of consecutive women age 40-75 years without coronary disease from 2011-12, followed for 10 years at Dartmouth-Hitchcock Medical Center, we recorded baseline mammograms, clinical parameters and 10-year MACE risk was estimated via the pooled cohort equation (PCE). The association of BAC and MACE was examined by a multivariable, interactive logistic regression model with MACE as dependent variable, adjusting for confounders and an interaction of BAC with 3 strata of PCE predicted risk: low (<5%), moderate (5-7.5%) and high (>7.5%). The primary endpoint was MACE over 10 years. Results: Among 1216 women, 19% had BAC. Women with BAC were older, had more hypertension and diabetes, and a higher baseline PCE score. Over 10 years of follow-up, 21% of women with BAC developed MACE compared to 11% women without BAC (p<0.001). This association was independent of risk factors and statin use. Among women with low (< 5%) risk of MACE, those with BAC had an 18% incidence of MACE, compared to 7% in women without BAC (OR 2.8, 95% CI 1.5 - 5.0) (Fig 1A). The incremental increase in MACE in patients with BAC decreased with higher PCE-predicted risk. Peak interaction occurred at a PCE-predicted risk of 10% (Fig 1B). Among women with a PCE-predicted risk <10%, BAC was a strong predictor of MACE with an attributable fraction of 20% (Fig 1B). Conclusions: BAC is a strong, independent predictor of MACE, particularly in women at low risk by PCE. Although PCE is used widely to risk stratify women for MACE, our study suggests it severely underestimates risk in women with a low PCE score. Further prospective trials are needed to validate these findings.

Association of neurocognitive disorders with morbidity and mortality in older adults undergoing major surgery in the USA: a retrospective, population-based, cohort study

Neurocognitive disorders become increasingly common as patients age, and increasing numbers of surgical interventions are done on older patients. The aim of this study was to understand the clinical characteristics and outcomes of surgical patients with neurocognitive disorders in the USA in order to guide future targeted interventions for better care. This retrospective cohort study used claims data for US Medicare beneficiaries aged 65 years and older with a record of inpatient admission for a major diagnostic or therapeutic surgical procedure between Jan 1, 2017, and Dec 31, 2018. Data were retrieved through a data use agreement between Dartmouth Hitchcock Medical Center and US Centers for Medicare and Medicaid Services via the Research Data Assistance Center. The exposure of interest was the presence of a pre-existing neurocognitive disorder as defined by diagnostic code within 3 years of index hospital admission. The primary outcome was mortality at 30 days, 90 days, and 365 days from date of surgery among all patients with available data. Among 5 263 264 Medicare patients who underwent a major surgical procedure, 767 830 (14·59%) had a pre-existing neurocognitive disorder and 4 495 434 (85·41%) had no pre-existing neurocognitive disorder. Adjusting for demographic factors and comorbidities, patients with a neurocognitive disorder had higher 30-day (hazard ratio 1·24 [95% CI 1·23-1·25]; p<0·0001), 90-day (1·25 [1·24-1·26]; p<0·0001), and 365-day mortality (1·25 [1·25-1·26]; p<0·0001) compared with patients without a neurocognitive disorder. Our findings suggest that the presence of a neurocognitive disorder is independently associated with an increased risk of mortality. Identification of a neurocognitive disorder before surgery can help clinicians to better disclose risks and plan for patient care after hospital discharge. Department of Anesthesiology and Perioperative Medicine at Dartmouth Hitchcock Medical Center.

A-054 Comparison of a Reformulated Vitamin D Immunoassay, Its Predecessor, and Two Contemporary Formulations

Abstract Background We took advantage of our verification of a new biotin-resistant 25-hydroxyvitamin D (vitamin D) immunoassay formulation to compare analytical performance of the outgoing formulation and assays from Roche, Diasorin and Beckman. Liquid chromatography with tandem mass spectrometry (LCMS) was an independent arbiter of vitamin D concentration and provided quantification of vitamin D2 and vitamin D3. Methods Twenty-five specimens from patients on high-dose vitamin D2 supplementation (50 000 IU/week) and fifty additional specimens spanning our measurable range were identified by the Dartmouth Hitchcock Analytics Institute. Aliquots were tested by the Roche cobas Vitamin D Total II and Vitamin D Total III assays, deidentified, split, and shipped to other study sites for measurement by LCMS, the Diasorin Liaison 25 OH Vitamin D Total assay, and the Beckman Coulter Access 25(OH) Vitamin D Total assay. Estimated vitamin D2 was calculated by subtracting the LCMS-measured vitamin D3 result from the immunoassay total vitamin D result. Bias calculations assume LCMS measurements as the true concentration. Statistical analyses were performed in R. This study was approved by the Dartmouth Hitchcock Medical Center Institutional Review Board. Results No significant differences were observed between platforms in absolute or relative bias in total vitamin D or estimated vitamin D2. Clinical classification was unanimous in 62/75 specimens. For thirteen specimens in which at least one disagreement was observed, immunoassay results for discrepant classification were no more than 10.2 ng/mL from the classification cutoff (median 2 ng/mL). Two specimens with 68 and 70 ng/mL vitamin D were classified as toxic (&amp;gt;100 ng/mL) by one immunoassay each. Conclusion Current vitamin D immunoassays continue to improve over their predecessors and offer good agreement across platforms. Vitamin D2 recovery no longer represents a clinically-concerning obstacle. These immunoassays display positive bias for total vitamin D relative to LCMS but generally result in correct clinical classification.

Phylogenetic lineages and antimicrobial resistance determinants of clinical Klebsiella oxytoca spanning local to global scales.

Klebsiella oxytoca is an opportunistic pathogen causing serious nosocomial infections. Knowledge about the population structure and diversity of healthcare-associated K. oxytoca from a genomic standpoint remains limited. Here, we characterized the phylogenetic relationships and genomic characteristics of 20 K. oxytoca sensu stricto isolates recovered from bloodstream infections at the Dartmouth-Hitchcock Medical Center, New Hampshire, USA from 2017 to 2021. Results revealed a diverse population consisting of 15 sequence types (STs) that together harbored 10 variants of the intrinsic beta-lactamase gene bla OXY-2, conferring resistance to penicillins. Similar sets of antimicrobial resistance (AMR) determinants reside in multiple distinct lineages, with no one lineage dominating the local population. To place the New Hampshire K. oxytoca in a broader context, we compared them to 304 publicly available genomes of clinical isolates from 18 countries. This global clinical K. oxytoca sensu stricto population is represented by over 65 STs that together harbored resistance genes against 14 antimicrobial classes, including eight bla OXY-2 variants. Three dominant STs in the global population (ST2, ST176, ST199) circulate across multiple countries and were also present in the New Hampshire population. The global K. oxytoca population is genetically diverse, but there is evidence for broad dissemination of a few lineages carrying distinct set of AMR determinants. Our findings reveal the clinical diversity of K. oxytoca sensu stricto and its importance in surveillance efforts aimed at monitoring the evolution of this drug-resistant nosocomial pathogen. IMPORTANCE The opportunistic pathogen Klebsiella oxytoca has been increasingly implicated in patient morbidity and mortality worldwide, including several outbreaks in healthcare settings. The emergence and spread of antimicrobial resistant strains exacerbate the disease burden caused by this species. Our study showed that clinical K. oxytoca sensu stricto is phylogenetically diverse, harboring various antimicrobial resistance determinants and bla OXY-2 variants. Understanding the genomic and population structure of K. oxytoca is important for international initiatives and local epidemiological efforts for surveillance and control of drug-resistant K. oxytoca.

Deficiency in the Midst of Abundance: Xerophthalmia Case Series and Vitamin A Levels in a New England Hospital

Purpose: To describe 3 adult cases of keratitis secondary to vitamin A deficiency (VAD) in relation to vitamin A levels Dartmouth-Hitchcock Medical Center (DHMC) population and published literature. Methods: Records of 3 patients with xerophthalmia were reviewed. All serum vitamin A levels obtained at DHMC during the same time period of our 3 cases (2019–2020) were analyzed. Outcomes were examined by age and range of deficiency. Results: Three patients, with short gut syndrome, chronic esophagitis, and alcohol use disorder, presented with xerophthalmia over 1 year. Serum vitamin A levels were 6.4 μg/dL, 16.1 μg/dL, and undetectable (&lt;5.0 μg/dL), respectively (normal: 32.5–78.0 μg/dL). Findings ranged from conjunctival keratinization to corneal perforation. Corneal cultures in patient 3 grew methicillin-sensitive Staphylococcus aureus. The ocular surface improved significantly in 2 patients following vitamin A supplementation. Two patients died during treatment. Analysis of 1596 vitamin A levels at DHMC revealed 431 patients with VAD, including 158 levels at or below those of our presented symptomatic cases. Conclusions: Vitamin A deficiency can occur in adults in high resource settings and lead to severe ocular morbidities, and is commonly associated with comorbidities such as alcohol use disorder and gastrointestinal disease. Vitamin A supplementation improved ocular findings in 2 of our patients. VAD was identified in 431 patients at DHMC over one year, indicating a surprisingly large population of patients at risk for xerophthalmia.

Machine Learning-Based Prediction of Distant Recurrence in Invasive Breast Carcinoma Using Clinicopathological Data: A Cross-Institutional Study.

Breast cancer is the most common type of cancer worldwide. Alarmingly, approximately 30% of breast cancer cases result in disease recurrence at distant organs after treatment. Distant recurrence is more common in some subtypes such as invasive breast carcinoma (IBC). While clinicians have utilized several clinicopathological measurements to predict distant recurrences in IBC, no studies have predicted distant recurrences by combining clinicopathological evaluations of IBC tumors pre- and post-therapy with machine learning (ML) models. The goal of our study was to determine whether classification-based ML techniques could predict distant recurrences in IBC patients using key clinicopathological measurements, including pathological staging of the tumor and surrounding lymph nodes assessed both pre- and post-neoadjuvant therapy, response to therapy via standard-of-care imaging, and binary status of adjuvant therapy administered to patients. We trained and tested four clinicopathological ML models using a dataset (144 and 17 patients for training and testing, respectively) from Duke University and validated the best-performing model using an external dataset (8 patients) from Dartmouth Hitchcock Medical Center. The random forest model performed better than the C-support vector classifier, multilayer perceptron, and logistic regression models, yielding AUC values of 1.0 in the testing set and 0.75 in the validation set (p < 0.002) across both institutions, thereby demonstrating the cross-institutional portability and validity of ML models in the field of clinical research in cancer. The top-ranking clinicopathological measurement impacting the prediction of distant recurrences in IBC were identified to be tumor response to neoadjuvant therapy as evaluated via SOC imaging and pathology, which included tumor as well as node staging.

Impact of area deprivation index on clinical outcomes of adults with acute myeloid leukemia at a rural academic medical center.

e18615 Background: Living in a disadvantaged neighborhood may lead to survival disparities amongst patients with acute myeloid leukemia (AML). The Area Deprivation Index (ADI) is a measure of socioeconomic disadvantage by decile used to find discrepancies in healthcare outcomes. Increasing ADI negatively impacts ovarian cancer survival, with each decile increase in state or national ADI corresponding with a 9-10% greater risk of death (Hufnagel D et al., Cancer Epidemiology 2021). Higher ADI was correlated with decreased rates of adjuvant therapy for localized pancreatic cancer (Mora et al., American Journal of Surgery 2021). There are no American studies on ADI and AML. This study aims to identify the impact of socioeconomic disadvantage on survival outcomes for patients with AML in a rural population. Methods: A retrospective review of 180 adults with AML treated at Dartmouth Hitchcock Medical Center (DHMC) between 1/1/2010 - 12/31/2020. Patients with APML, primary treatment outside of DHMC, and those who opted for comfort measures only at diagnosis were excluded. Demographic data, including cytogenetics, treatment type, response to treatment, and survival at one year were recorded. Addresses from the medical record were used to determine state-level ADI using the University of Wisconsin Neighborhood Atlas. Results: Eighty-nine patients were included (64% male; 93% White). Median age was 65 years. Fifty-one patients (57%) had high risk disease, 31 intermediate risk (35%), 5 favorable risk (6%), and unknown risk for 2 (2%). Sixty-two patients (70%) received intensive chemotherapy [7+3] while 27 (30%) received less intensive therapy [hypomethylating therapy]. Twenty-six patients (29%) underwent transplantation. ADI decile was not associated with survival (p = 0.62). Furthermore, when ADI was divided into three groups (deciles 1-3, 4-6, 7-10) there was no difference in survival (p = 0.345). There was no effect of travel distance to DHMC [median 81 minutes] on OS (p = 0.619). There was a significant difference in survival outcomes between the two treatment types (p = 0.001), with median 25.7 vs. 7.1 months in the intensive vs. less intensive groups. This survival advantage was seen on multivariate analysis (p = 0.03). Females had better OS compared to males (p = 0.02) and older age was a negative predictor of outcome (p = 0.007). In those who achieved remission with induction, transplant led to improved survival (p = 0.04). Conclusions: There was no significant difference in treatment outcomes based on ADI, potentially due to equitable quality of care delivered regardless of socioeconomic disparities. This may be due to strong nurse navigation, social work support, specialists from the primary site who provide care in the community, and access to transplantation. Future studies to delineate the impact of various interventions for those with high ADI may be beneficial in treatment planning.

Neoadjuvant endocrine therapy for early-stage breast cancer: An institutional experience during the COVID19 pandemic.

e12625 Background: Neoadjuvant Endocrine therapy (NET) is under-utilized for locally advanced hormone receptor + breast cancer (BC) despite literature to date concluding that chemotherapy and NET have comparable response rates but with lower toxicity for the latter. The COVID19 pandemic presented a unique occasion to extend the use of NET for localized BC. Many surgical procedures across the country, including at Dartmouth-Hitchcock Medical Center (DHMC) were postponed in order to redirect resources. The American Society of Breast Surgeons released guidelines to manage patients in which surgery was delayed. For Estrogen Receptor + (ER+), HER2- early-stage tumors, the guidelines recommended NET as first-line treatment. Herein we review the rate of adoption of NET and patient/tumor characteristics associated with NET adoption at DHMC during the lockdown. Methods: A Retrospective analysis of patients diagnosed with early-stage ER+, HER2- BC between December 2019 and June 2020. Inclusion criteria included patients with non-metastatic hormone receptor + BC. Data extracted from chart review included age, menopausal status, tumor stage/grade, body mass index (BMI). A delay in surgery was defined as days between surgical consult and surgery over 14 days. Descriptive statistics will be applied to data collected on patient/tumor characteristics, and the number of patients accepting or declining NET. Results: 109 cases were identified within the study period with 42 cases designated as experiencing delays in surgery. The median age of the group (n = 42) was 62 and the majority was post-menopausal. 36 patients received NET with most started on an aromatase inhibitor. Median BMI was 28.5. Median duration of treatment was 39.5 days. Tumor response to NET was not clearly described in the pathology reports. Three cases were noted to have some decrease in cellularity for pathologic partial response and four had no definite response. Imaging modalities were compared to pathology size and MRI was found to be a fair predictor of size obtained from pathology specimens with a correlation metric of 0.78 indicating that it may serve as the best proxy for pre-treatment size. Pre-surgical MRIs were obtained in 51.3% of cases. The majority of patients on NET had no change in the grade of their pathology. During the course of the pandemic long-term adherence of adjuvant endocrine therapy decreased from 78% for those diagnosed in early 2020 to 56% for those diagnosed in late 2020. Conclusions: The COVID19 crisis was an unprecedented challenge to healthcare systems. Delays in surgery resulted in increased use of NET in early-stage BC. We plan to integrate other data including impact of NET on long-term adherence to endocrine therapy. Change in clinical vs pathologic stage will be reported for those with or without NET. This experience may help guide care during healthcare crises or in resources limited settings.

Clinician education on climate change and health: virtual learning community models

Clinician education on climate change and health: virtual learning community models

Outpatient intravenous diuresis in a rural setting: safety, efficacy, and outcomes.

To evaluate the safety, efficacy, and outcomes of outpatient intravenous diuresis in a rural setting and compare it to urban outcomes. A single-center study was conducted on 60 patients (131 visits) at the Dartmouth-Hitchcock Medical Center (DHMC) from 1/2021-12/2022. Demographics, visit data, and outcomes were collected and compared to urban outpatient IV centers, and inpatient HF hospitalizations from DHMC FY21 and national means. Descriptive statistics, T-tests and chi-squares were used. The mean age was 70 ± 13 years, 58% were male, and 83% were NYHA III-IV. Post-diuresis, 5% had mild-moderate hypokalemia, 16% had mild worsening of renal function, and 3% had severe worsening of renal function. No hospitalizations occurred due to adverse events. The mean infusion-visit urine output was 761 ± 521 ml, and post-visit weight loss was -3.9 ± 5.0 kg. No significant differences were observed between HFpEF and HFrEF groups. 30-day readmissions were similar to urban outpatient IV centers, DHMC FY21, and the national mean (23.3% vs. 23.5% vs. 22.2% vs. 22.6%, respectively; p = 0.949). 30-day mortality was similar to urban outpatient IV centers but lower than DHMC FY21 and the national means (1.7% vs. 2.5% vs. 12.3% vs. 10.7%, respectively; p < 0.001). At 60 days, 42% of patients had ≥1 clinic revisit, 41% had ≥1 infusion revisit, 33% were readmitted to the hospital, and two deaths occurred. The clinic avoided 21 hospitalizations, resulting in estimated cost savings of $426,111. OP IV diuresis appears safe and effective for rural HF patients, potentially decreasing mortality rates and healthcare expenses while mitigating rural-urban disparities.

(549) Case Report of an IPP Reservoir Salvage Following Infection

Abstract Introduction Inflatable penile prosthesis (IPP) infections are rare and devastating. IPP infections are treated with either explant or salvage, the latter with either a malleable or inflatable device. Isolated component infections are very rare in the literature. This is the first description of successful treatment of an isolated IPP reservoir infection, with preservation and eventual reuse of the patient’s uninfected pump and cylinders. Objective To report a case of successful IPP salvage after an isolated reservoir infection. Methods This is a case study of IPP reservoir explant and eventual reimplant in a 51-year-old male, with preservation of the implant pump and cylinders. Past medical history is notable for HIV, hypertension, and hyperlipidemia. The patient had a Coloplast Titan IPP placed penoscrotally in November 2020 for treatment of post-prostatectomy erectile dysfunction. The reservoir was placed ectopically in the patient’s right high submuscular space. Results In January 2021 the patient underwent emergent exploratory laparoscopic washout and drain placement at a different hospital for perforated appendicitis. The IPP reservoir was not encountered during this procedure. In March 2021 he presented to that same hospital with recurrent fevers and right lower quadrant pain. CT showed increased peritoneal stranding and thickening surrounding the IPP reservoir as well as multiple small, complex loculated fluid collections suspicious for abscesses external to the reservoir. There was no evidence of infection of the pump or cylinders on imaging or physical exam. He was transferred to Dartmouth-Hitchcock Medical Center for further care. Conservative treatment with intravenous vancomycin, Zosyn, and fluconazole yielded no change after 48 hours. Three days after initial presentation, the reservoir was explanted during a collaborative procedure with general surgery, during which the appendix remnant was also removed. Intraoperatively the penile prosthesis reservoir was encased in a dense rind that was incised and drained. Copious purulent material was sent for culture, which showed multi-organism aerobic and anerobic bacteria. The reservoir was removed and the remnant tubing to the pump was clamped. The abdomen was irrigated repeatedly with saline, dilute betadine, vancomycin, gentamicin, Zosyn, and amphotericin. The prothesis tubing was cut as low as possible, capped, and wrapped with a paste of vancomycin mixed with Zosyn and amphotericin. There were no intraoperative complications. The patient was discharged two days later. At his one-week postoperative visit, he was readmitted for interventional radiology drain placement due to abdominal abscess recurrence. Once the recurrent abscess had resolved, he underwent successful radiation treatment for recurrent prostate cancer. In March 2022, the reservoir was successfully replaced into the left high submuscular space via counter incision and the dormant cylinders and pump were reconnected. The patient continues to do well, with no signs of recurrent infection and with excellent device function. He is satisfied with the result of surgery. Conclusions To our knowledge this is the only case of isolated IPP reservoir infection successfully treated with reservoir explant and eventual replacement. The original IPP pump and cylinders were preserved without issue and eventually reused. Disclosure Any of the authors act as a consultant, employee or shareholder of an industry for: Coloplast.

Development and Validation of Machine Learning and Electronic Medical Records-Based Characterization of Stiff Person Syndrome (P5-8.001)

<h3>Objective:</h3> To develop and validate an objective framework in identifying medical history items that best characterize the diagnosis of Stiff Person Syndrome (SPS). <h3>Background:</h3> SPS is a rare neurologic disorder with characteristic muscle spasms and painful rigidity. While electronic medical records may contain items that can help refine its diagnostic criteria, the rarity of SPS leads to statistical underpowering in conventional tests. A machine learning approach that leverages underutilized electronic medical records may reveal clinical features that are strongly associated with the diagnosis of SPS. <h3>Design/Methods:</h3> This retrospective cohort study assessed 23 patients carrying SPS diagnoses and 25 controls who were all anti-GAD positive and were treated at Dartmouth Hitchcock Medical Center. After binarizing the 319 clinical features, we employed an iterative machine-learning-based feature selection approach (Contribution Selection Algorithm) to identify variables that best discriminate between patients with SPS vs. anti-GAD positive controls. Each iteration generated SHapley Additive exPlanation (SHAP) values for each binarized feature. Classifier accuracy, AUC, and Matthews Correlation Coefficients (MCC) for the Support Vector Machine models were calculated using repeated stratified 4-fold cross-validation. <h3>Results:</h3> The feature selection algorithm identified depression, pain in joint, hypothyroidism, and GERD as the top 4 predictors of SPS, as the SVM model that employed these features achieved the best predictive performance with a binary classification accuracy of 0.775 (95%CI, 0.759–0.790), an AUC of 0.808 (95%CI, 0.791–0.825), and a MCC of 0.565 (95%CI, 0.534–0.597). <h3>Conclusions:</h3> Our machine learning-based architecture identified features that are strongly associated with SPS. By circumventing the limitations of statistical underpowering and focusing on identifying associations through induction, this approach may act as a powerful tool for generating new hypotheses for rare disorders. This framework complements hypothesis-driven investigations, together forming a closed loop approach to aid diagnostic challenges in neurology. <b>Disclosure:</b> Mr. Park has nothing to disclose. Dr. Song has nothing to disclose.

Abstract LB070: Predicting distant recurrences in invasive breast carcinoma patients using their clinicopathological profiles

Abstract Purpose: To determine whether the classification-based machine learning (ML) or artificially intelligent (AI) techniques can predict distant recurrence flags (yes or no) in invasive breast cancer patients using the data comprising several clinicopathological measurements such as pathological staging of tumor and surrounding nodes deemed both pre and post neoadjuvant (i.e. either chemo, radiation or hormone based) therapy including imaging based therapy-response as well and the status of adjuvant therapy (Chemo or Anti-Her2/Neu antibody) therapy were administered to patients post resection of their tumor to minimize the possibility of diseases-recurrence in them. Method: The clinicopathological data from a retrospective study of 900 breast cancer patients at Duke university (posted at The Cancer Imaging Archive, i.e. TCIA) was tapped for our study. These patients received neoadjuvant therapies such as Chemo-, Radiation or Endocrine Hormone based therapy, had their responses evaluated through imaging and pathological staging (both tumor and nodal), and received Adjuvant therapies (such as Chemo or Anti-Her2/Neu antibody therapy) post tumor resection. The patient entries pertaining to ungraded tumor responses from therapies or those labelled as ‘Not applicable (NA)’ for any of the aforementioned clinicopathological parameters were filtered out retaining 161 patient-entries which were further split into train and test sets in 90:10 ratio where 90% were used for training (ntrain=144) and 10% for testing (ntest=17) of several ML models. Classification based machine learning models such as Random Forest (RF), C-Support Vector Classification (SVC) and Supervised Neural Network (aka Multi-Layer Perceptron, i.e. MLP) were employed to train and test aforementioned clinicopathological parameters to predict distant recurrence flags or labels (i.e. Yes or No). The training of the model was conducted using ImaGene software. Out of the three aforementioned models, RF seemed to have performed best with grid search activated over tree-depth hyperparameter to predict distant recurrence flags in test patients (ntest=17) at AUC=1.0 (p&amp;lt;0.002). Further, the validation of the model was conducted using external clinicopathological dataset of 20 patients from Dartmouth-Hitchcock (DH) Medical center which yielded AUC&amp;gt;0.75. Conclusion: Random Forest model trained using grid search through tree-depth hyperparameter can predict distant recurrence flags (Yes or No) in invasive breast cancer patients at AUC in the range of 0.75 1.0 across multiple institutions, thereby aiding portability of models and advancing the idea of multi-instutional ML/AI model validations in Invasive Breast Carcinoma subtype of Breast Cancer. Citation Format: Shrey S. Sukhadia, Kristen Muller, Adrienne Workman, Shivashankar Nagaraj, Olivier Gevaert. Predicting distant recurrences in invasive breast carcinoma patients using their clinicopathological profiles [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 2 (Clinical Trials and Late-Breaking Research); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(8_Suppl):Abstract nr LB070.

How Do Postgraduate Trainees Prepare for New Clinical Learning Environments.

Transitioning to a new clinical learning environment (CLE) requires learning new skills, roles, teams, workflows, and culture. We previously identified activities and questions to guide orientation in the categories of who, what, where, when, why, and how. There is limited literature about how learners plan for this transition. Describe how postgraduate trainees prepare for a clinical rotation, based on qualitative analysis of narrative responses within a simulated orientation experience. In June 2018, incoming residents and fellows in multiple specialties at Dartmouth Hitchcock Medical Center completed an online simulated orientation that asked how they intended to prepare for their first rotation. We used directed content analysis to code their anonymously collected responses using the orientation activities and question categories from our prior study. We used open coding to describe additional themes. Narrative responses were available for 97% (116 of 120) of learners. While 46% (53 of 116) of learners listed preparations related to what happens in the CLE, responses fitting into other question categories were less frequent: who (9%, 11 of 116), where (7%, 8 of 116), when (4%, 5 of 116), why (<1%, 1 of 116), and how (0%, 0 of 116). Learners also infrequently described activities to aid the transition: reading orientation materials (11%, 13 of 116), speaking with a colleague (11%, 13 of 116), and arriving early (3%, 3 of 116). They more frequently commented on content reading (40%, 46 of 116), asking for advice (28%, 33 of 116), and self-care (12%, 14 of 116). When describing preparation for a new CLE, residents focused on tasks in the what category more than understanding the system and learning goals in other categories.