eScholarship@UMassChan
eScholarship@UMassChan is a digital archive for UMass Chan Medical School's research and scholarship, including journal articles, theses, datasets and more. We welcome submissions from our faculty, staff, and students. eScholarship@UMassChan is a service of the Lamar Soutter Library, Worcester, MA, USA. See also our open access journal publishing services.
Questions? See the Help menu or contact escholarship@umassmed.edu.
Communities in eScholarship@UMassChan
Select a community to browse its collections.
Recent Publications
Publication Unknown Hospice Enrollment Among Medicare Beneficiaries Hospitalized for Chronic Obstructive Pulmonary Disease(UMass Chan Medical School, 2026-05-07)Rationale: Chronic obstructive pulmonary disease (COPD) is associated with substantial morbidity and mortality, reduced quality of life, and high rates of acute care utilization, particularly near the end of life. Although hospice services improve quality of life for patients and caregivers and reduce acute care use, hospice remains underutilized among patients with COPD. Further research is needed to identify factors associated with hospice enrollment in this population. Objectives: To describe the use of hospice following a COPD hospitalization and to identify patient- and hospital-level factors associated with enrollment Methods: We performed a retrospective cohort study using claims data from Medicare fee-for-service beneficiaries hospitalized for a COPD exacerbation in 2014 at 2,650 U.S. acute care hospitals. Our primary outcome was hospice enrollment within 30-days of hospital discharge. We used a hierarchical logistic regression model with hospital-level random effects, adjusting for both patient- and hospital-level variables, to identify factors independently associated with hospice enrollment. Model estimates were used to calculate hospital-specific risk-standardized hospice enrollment rates. The contribution of hospital-level effects to the likelihood of hospice enrollment was quantified using the median odds ratio (MOR). Results: 234,540 patients were included in the analysis, of whom 12,695 (5.4%) enrolled in hospice within 30 days of hospital discharge. Receipt of invasive (aOR 2.66, 95% confidence interval [CI] 2.49-2.84) and non-invasive ventilation (aOR 2.64, 95% CI 2.51-2.78) during the index admission, presence of metastatic cancer (aOR 2.49, 95% CI 2.33-2.65) and use of supplemental oxygen prior to admission (aOR 1.47, 95% CI 1.41-1.52) were the factors most strongly associated with hospice enrollment. Risk-standardized rates of hospice enrollment ranged from 2.9% to 13.3%. Larger hospitals, those in urban areas, and hospitals in the South had higher risk-standardized rates of hospice enrollment than their counterparts. The hospital median odds ratio was 1.45 (95% CI 1.41-1.49) comparable to several patient-level factors associated with hospice enrollment. Conclusions: Among Medicare beneficiaries hospitalized for COPD, approximately 1 in 20 patients enroll in hospice in the month following a COPD hospitalization. Despite adjustment for individual-level differences among patients, substantial variation in hospice enrollment rates remains, suggesting that unmeasured hospital factors contribute to hospice utilization. Additional research is needed to identify strategies to increase hospice use among this patient population.Publication Unknown Jointly-hic: joint decomposition of contact frequency maps captures salient features of genome architecture across tissues and development(2026-05-02)Chromosome conformation capture methods, such as Hi-C, have been used to profile chromosome organization from a wide variety of biosamples and conditions; however, existing methods for analyzing such datasets have disadvantages for large-scale integrative studies of long-range interactions. To address this shortcoming, we introduce an analytical framework, jointly-hic, that computes harmonized projections across arbitrarily many contact frequency matrices, suitable for integrative studies of compartmentalization and long-range interactions. Our approach produces robust and directly comparable first and higher-order principal component scores that collectively capture biologically meaningful information beyond traditional A/B compartment scores.Publication Unknown Evaluating model generalizability for suicide attempt risk prediction: traditional machine vs deep learning(2026-04-30)Suicide remains a leading cause of death and a significant public health concern in the United States. A majority (83%) of suicide decedents had a healthcare visit within the prior 365 days, presenting unique opportunities to utilize healthcare data for AI-based interventions. While previous works applied machine learning (ML) to analyze healthcare records for suicide attempt risk prediction (SARP), they lack external validation. Additionally, advantages of deep learning (DL) over ML for tabular SARP remains understudied. We performed external validation of a state-of-the-art SARP model from the Mental Health Research Network using over 750,000 UMass Memorial Health patient encounters. We further compared ML vs DL, assessing cross-setting healthcare generalizability. We found existing models did not generalize well, ML significantly outperformed DL on most metrics, and DL achieved higher sensitivity. These findings underscore the need for developing robust, generalizable SARP models for diverse healthcare contexts, improving identification of individuals at risk.Publication Unknown The potential role for GLP-1 receptor agonists in cancer survivorship: a narrative review(2026-04-30)INTRODUCTION: Cancer survivors face a high burden of comorbidities which negatively influence long-term outcomes. Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) may augment these outcomes, but their role in this population remains understudied. METHODS: We conducted a literature review of randomized controlled trials, observational studies, retrospective cohort studies, pre-clinical trials, and reviews related to GLP-1 RAs and cancer survivors. RESULTS: We first delineate the GLP-1 RA mechanisms of action in glycemic control, inflammation, cardiovascular protection, neuroprotection, nephroprotection, and cancer proliferation. Then, we highlight the potential clinical benefits that these drugs could have in cancer survivors based on available studies, including metabolism, cardiovascular risk, lymphedema and mortality. Finally, we address ongoing safety concerns for GLP-1 RAs, particularly focusing on their potential use in cancer survivors. DISCUSSION: Given the limited evidence to date, further research is needed to evaluate the long-term benefits and safety of GLP-1 RA use in cancer survivors.Publication Unknown The Relationship Between Body Mass Index and Skull Soft Tissues: Implications for Bone Conduction Coupling(2026-04-29)OBJECTIVE: Quantify mastoid soft-tissue thickness on temporal bone computed tomography (CT) and evaluate associations with body mass index (BMI), age, and sex. STUDY DESIGN: Retrospective cross-sectional study. SETTING: Tertiary referral center. PATIENTS: Adults with temporal bone CT and a recorded BMI within 12 months. MAIN OUTCOME MEASURES: Mean mastoid soft-tissue thickness measured at 6 sites (3 per ear) on multiplanar reconstructions. RESULTS: One hundred forty-seven CT scans (mean age: 57.4±22.0 y; 46.3% male; mean BMI: 30.2±9.9 kg/m 2 ). The mean mastoid thickness was 9.7±4.1 mm (range: 3.4 to 27.0 mm). BMI was associated with greater mastoid thickness (adjusted β =0.30 mm per kg/m 2 , 95% CI: 0.25-0.34; P <0.001). Male sex was associated with thicker tissue (adjusted difference: 1.37 mm, 95% CI: 0.46-2.27; P =0.003), whereas age was not independently associated. Interaural means were highly correlated ( r =0.89), with a negligible paired difference. Interrater ICC for mean mastoid thickness was 0.995 (n=59). CONCLUSIONS: Mastoid soft-tissue thickness varies widely and increases predictably with BMI. BMI may serve as an accessible clinical proxy for the soft-tissue envelope at bone conduction coupling sites, with potential implications for the interpretation of bone conduction audiometry and the planning of skin-drive and magnet-retained bone conduction hearing devices.