Dr. Ross is a vascular surgeon and research scientist. She graduated from Stanford University School of Medicine in 2011 and completed her vascular surgery 0+5 residency at Stanford University School of Medicine in 2018. During her residency, she completed a two-year post-doctoral fellowship in biomedical informatics. Her current research focuses on using machine learning and electronic health records for early disease identification, precision medicine, and evaluating opportunities to engage in patient education beyond the clinic.
- Vascular Surgery
- Preventative health
- Peripheral vascular disease
- Carotid disease
- Venous disease
- Vascular and endovascular treatment of abdominal and thoracic aortic aneurysms
- Vascular trauma
Honors & Awards
US-UK Fulbright Scholar, US-UK Fulbright Commission (2008-09)
Soros Fellow, Paul & Daisy Soros Fellowship for New Americans (2008-2010)
Association for Academic Surgery Young Investigators Award, Association for Academic Surgery (AAS) (2018-19)
Society of University Surgeons Junior Faculty Award, Society of University Surgeons (SUS) (2018-2019)
Residency: Stanford University Vascular Surgery Fellowship (2018) CA
Post Doc, Stanford University School of Medicine, Biomedical informatics (2015)
Medical Education: Stanford University School of Medicine Registrar (2011) CA
MSc, London School of Hygiene and Tropical Medicine, London School of Economics, Health Policy, Planning and Financing (2009)
BA, Stanford University, Human Biology (2004)
Evaluation of an Electronic Health Record-based Screening Tool for Peripheral Artery Disease
This protocol represents a pilot randomized-controlled trial evaluating the effect of an electronic health record (EHR)-based peripheral artery disease (PAD) screening tool on rates of new non-invasive testing, diagnosis and treatment of PAD over a 6-month period. An EHR-based PAD screening tool will be applied to the Stanford EHR, which will generate a group of patients of varying risks of having undiagnosed PAD. Patients with the highest risk of having undiagnosed PAD will then be evaluated for inclusion in this study. 1:1 randomization will be performed on a consecutive basis until study enrollment is completed (25 patients per arm). Physicians of patients randomized to the intervention arm will be sent notification via an EHR message detailing the patient's risk of undiagnosed PAD and suggestions for referral to vascular medicine for risk assessment and/or non-invasive ankle brachial index (ABI) testing. The primary outcome is number of patients receiving ABI testing for PAD at 6 months, with secondary outcomes including number of new PAD diagnoses, number of new referrals to cardiovascular specialists (vascular medicine, vascular surgery, and/or cardiology) and number of patients receiving initiation of new cardiovascular medications (anti-platelet agents, statins, and/or antihypertensive agents).
Stanford is currently not accepting patients for this trial. For more information, please contact Elsie Ross, MD, MSc, 650-723-5477.
Evaluation of regional variations in length of stay after elective, uncomplicated carotid endarterectomy in North America.
Journal of vascular surgery
OBJECTIVE: The objective of this study was to evaluate factors affecting regional variation in length of stay (LOS) after elective, uncomplicated carotid endarterectomy (CEA).METHODS: Data were obtained from the Vascular Quality Initiative database and included patients with complete data who received elective CEA without complications between 2012 and 2017 across 18 regions in North America and 294 centers. The main outcome measure was LOS >1day after surgery (LOS >1 postoperative day [POD]). Using least absolute shrinkage and selection operator regression, multivariable modeling, and mixed-effects general linear modeling, we evaluated whether regional variations in LOS were independent of demographic, clinical, or center-related factors and to what extent these factors accounted for postoperative variation in LOS.RESULTS: A total of 36,004 patients were included. Mean postprocedure LOS was 1.6± 6.6days. Overall, 24% of patients had an LOS >1 POD. After adjustment for important demographic, clinical, and center-related factors, the region in which a patient was treated independently and significantly affected LOS after elective, uncomplicated CEA. Region and center of treatment accounted for 18% of LOS variation. Demographic, clinical, and surgical factors accounted for another 32% of variation in LOS. Of these factors, postoperative discharge to a facility other than home (odds ratio [OR], 6.3; confidence interval [CI], 5.2-7.6), use of intravenous (IV) vasoactive agents (OR, 3.2; CI, 3-3.4), intraoperative drain placement (OR, 1.4; CI, 1.3-1.55), and female sex (OR, 1.4; CI, 1.3-1.5) were associated with longer LOS. Factors associated with LOS ≤1 POD included preoperative aspirin (OR, 0.88; CI, 0.8-0.96) and statin use (OR, 0.9; CI, 0.83-0.98), high surgeon volume (highest quartile: OR, 0.68; CI, 0.5-0.87), and completion evaluation after CEA (eg, Doppler, ultrasound; OR, 0.87; CI, 0.8-0.95). We also found that use of IV vasoactive medications varied significantly across regions, independent of demographic and clinical factors.CONCLUSIONS: Significant regional variation in LOS exists after elective, uncomplicated CEA even after controlling for a wide range of important factors, indicating that there remain unmeasured causes of longer LOS in some regions. Even so, modification of certain clinical practices may reduce overall LOS. Regional differences in use of IV vasoactive medications not driven by clinical factors warrant further analysis, given the strong association with longer LOS.
View details for DOI 10.1016/j.jvs.2019.02.071
View details for PubMedID 31280981
Diagnosis and management of external iliac endofibrosis: A case report
JOURNAL OF VASCULAR NURSING
2019; 37 (2): 86–90
External iliac artery endofibrosis is an uncommon, nonatherosclerotic disease seen in endurance cyclists. It is poorly identified by providers. These otherwise healthy patients usually present with symptoms of arterial insufficiency, such as thigh or buttock pain, loss of power, or weakness occurring during strenuous exercises. These symptoms subside rapidly with rest. As these patients lack traditional risk factors of peripheral artery disease, their symptoms are often overlooked or are attributed to other etiologies, resulting in mismanagement and delayed treatment. In this case study, we report our experience with the successful management of a 48-year-old male who is a longstanding, avid cyclist. He self-referred to our institution after extensive research of providers familiar with his problem and at the recommendation of other cyclists with similar experiences. The patient underwent a successful left external iliac to common femoral artery endarterectomy and patch angioplasty. Three months after operation, he returned to cycling and, for the most part, has remained without symptoms.
View details for DOI 10.1016/j.jvn.2018.11.008
View details for Web of Science ID 000469492800003
View details for PubMedID 31155167
- A Comprehensive Evaluation of Lifestyle and Social Factors Related to Peripheral Artery Disease Events in a Large Longitudinal Study MOSBY-ELSEVIER. 2019: E54–E55
Predicting Future Cardiovascular Events in Patients With Peripheral Artery Disease Using Electronic Health Record Data.
Circulation. Cardiovascular quality and outcomes
2019; 12 (3): e004741
BACKGROUND: Patients with peripheral artery disease (PAD) are at risk of major adverse cardiac and cerebrovascular events. There are no readily available risk scores that can accurately identify which patients are most likely to sustain an event, making it difficult to identify those who might benefit from more aggressive intervention. Thus, we aimed to develop a novel predictive model-using machine learning methods on electronic health record data-to identify which PAD patients are most likely to develop major adverse cardiac and cerebrovascular events.METHODS AND RESULTS: Data were derived from patients diagnosed with PAD at 2 tertiary care institutions. Predictive models were built using a common data model that allowed for utilization of both structured (coded) and unstructured (text) data. Only data from time of entry into the health system up to PAD diagnosis were used for modeling. Models were developed and tested using nested cross-validation. A total of 7686 patients were included in learning our predictive models. Utilizing almost 1000 variables, our best predictive model accurately determined which PAD patients would go on to develop major adverse cardiac and cerebrovascular events with an area under the curve of 0.81 (95% CI, 0.80-0.83).CONCLUSIONS: Machine learning algorithms applied to data in the electronic health record can learn models that accurately identify PAD patients at risk of future major adverse cardiac and cerebrovascular events, highlighting the great potential of electronic health records to provide automated risk stratification for cardiovascular diseases. Common data models that can enable cross-institution research and technology development could potentially be an important aspect of widespread adoption of newer risk-stratification models.
View details for PubMedID 30857412
Evaluation of Cell Therapy on Exercise Performance and Limb Perfusion in Peripheral Artery Disease: The CCTRN Patients with Intermittent Claudication Injected with ALDH Bright Cells (PACE) Trial.
Atherosclerotic peripheral artery disease affects 8% to 12% of Americans >65 years of age and is associated with a major decline in functional status, increased myocardial infarction and stroke rates, and increased risk of ischemic amputation. Current treatment strategies for claudication have limitations. PACE (Patients With Intermittent Claudication Injected With ALDH Bright Cells) is a National Heart, Lung, and Blood Institute-sponsored, randomized, double-blind, placebo-controlled, phase 2 exploratory clinical trial designed to assess the safety and efficacy of autologous bone marrow-derived aldehyde dehydrogenase bright (ALDHbr) cells in patients with peripheral artery disease and to explore associated claudication physiological mechanisms.All participants, randomized 1:1 to receive ALDHbr cells or placebo, underwent bone marrow aspiration and isolation of ALDHbr cells, followed by 10 injections into the thigh and calf of the index leg. The coprimary end points were change from baseline to 6 months in peak walking time (PWT), collateral count, peak hyperemic popliteal flow, and capillary perfusion measured by magnetic resonance imaging, as well as safety.A total of 82 patients with claudication and infrainguinal peripheral artery disease were randomized at 9 sites, of whom 78 had analyzable data (57 male, 21 female patients; mean age, 66±9 years). The mean±SEM differences in the change over 6 months between study groups for PWT (0.9±0.8 minutes; 95% confidence interval [CI] -0.6 to 2.5; P=0.238), collateral count (0.9±0.6 arteries; 95% CI, -0.2 to 2.1; P=0.116), peak hyperemic popliteal flow (0.0±0.4 mL/s; 95% CI, -0.8 to 0.8; P=0.978), and capillary perfusion (-0.2±0.6%; 95% CI, -1.3 to 0.9; P=0.752) were not significant. In addition, there were no significant differences for the secondary end points, including quality-of-life measures. There were no adverse safety outcomes. Correlative relationships between magnetic resonance imaging measures and PWT were not significant. A post hoc exploratory analysis suggested that ALDHbr cell administration might be associated with an increase in the number of collateral arteries (1.5±0.7; 95% CI, 0.1-2.9; P=0.047) in participants with completely occluded femoral arteries.ALDHbr cell administration did not improve PWT or magnetic resonance outcomes, and the changes in PWT were not associated with the anatomic or physiological magnetic resonance imaging end points. Future peripheral artery disease cell therapy investigational trial design may be informed by new anatomic and perfusion insights.URL: http://www.clinicaltrials.gov. Unique identifier: NCT01774097.
View details for DOI 10.1161/CIRCULATIONAHA.116.025707
View details for PubMedID 28209728
Enhanced Quality Measurement Event Detection: An Application to Physician Reporting.
EGEMS (Washington, DC)
2017; 5 (1): 5
The wide-scale adoption of electronic health records (EHR)s has increased the availability of routinely collected clinical data in electronic form that can be used to improve the reporting of quality of care. However, the bulk of information in the EHR is in unstructured form (e.g., free-text clinical notes) and not amenable to automated reporting. Traditional methods are based on structured diagnostic and billing data that provide efficient, but inaccurate or incomplete summaries of actual or relevant care processes and patient outcomes. To assess the feasibility and benefit of implementing enhanced EHR- based physician quality measurement and reporting, which includes the analysis of unstructured free- text clinical notes, we conducted a retrospective study to compare traditional and enhanced approaches for reporting ten physician quality measures from multiple National Quality Strategy domains. We found that our enhanced approach enabled the calculation of five Physician Quality and Performance System measures not measureable in billing or diagnostic codes and resulted in over a five-fold increase in event at an average precision of 88 percent (95 percent CI: 83-93 percent). Our work suggests that enhanced EHR-based quality measurement can increase event detection for establishing value-based payment arrangements and can expedite quality reporting for physician practices, which are increasingly burdened by the process of manual chart review for quality reporting.
View details for PubMedID 29881731
The use of machine learning for the identification of peripheral artery disease and future mortality risk.
Journal of vascular surgery
2016; 64 (5): 1515-1522 e3
A key aspect of the precision medicine effort is the development of informatics tools that can analyze and interpret "big data" sets in an automated and adaptive fashion while providing accurate and actionable clinical information. The aims of this study were to develop machine learning algorithms for the identification of disease and the prognostication of mortality risk and to determine whether such models perform better than classical statistical analyses.Focusing on peripheral artery disease (PAD), patient data were derived from a prospective, observational study of 1755 patients who presented for elective coronary angiography. We employed multiple supervised machine learning algorithms and used diverse clinical, demographic, imaging, and genomic information in a hypothesis-free manner to build models that could identify patients with PAD and predict future mortality. Comparison was made to standard stepwise linear regression models.Our machine-learned models outperformed stepwise logistic regression models both for the identification of patients with PAD (area under the curve, 0.87 vs 0.76, respectively; P = .03) and for the prediction of future mortality (area under the curve, 0.76 vs 0.65, respectively; P = .10). Both machine-learned models were markedly better calibrated than the stepwise logistic regression models, thus providing more accurate disease and mortality risk estimates.Machine learning approaches can produce more accurate disease classification and prediction models. These tools may prove clinically useful for the automated identification of patients with highly morbid diseases for which aggressive risk factor management can improve outcomes.
View details for DOI 10.1016/j.jvs.2016.04.026
View details for PubMedID 27266594
View details for PubMedCentralID PMC5079774
The Promise and Challenge of Induced Pluripotent Stem Cells for Cardiovascular Applications.
JACC. Basic to translational science
2016; 1 (6): 510-523
The recent discovery of human-induced pluripotent stem cells (iPSCs) has revolutionized the field of stem cells. iPSCs have demonstrated that biological development is not an irreversible process and that mature adult somatic cells can be induced to become pluripotent. This breakthrough is projected to advance our current understanding of many disease processes and revolutionize the approach to effective therapeutics. Despite the great promise of iPSCs, many translational challenges still remain. In this article, we review the basic concept of induction of pluripotency as a novel approach to understand cardiac regeneration, cardiovascular disease modeling and drug discovery. We critically reflect on the current results of preclinical and clinical studies using iPSCs for these applications with appropriate emphasis on the challenges facing clinical translation.
View details for DOI 10.1016/j.jacbts.2016.06.010
View details for PubMedID 28580434
- National Comparison of Hybrid and Open Repair for Aortoiliac-Femoral Occlusive Disease MOSBY-ELSEVIER. 2016: 551
- Use of Predictive Analytics for the Identification of Latent Vascular Disease and Future Adverse Cardiac Events MOSBY-ELSEVIER. 2016: 28S–29S
Use of Machine Learning to Accurately Predict Adverse Events in Patients with Peripheral Artery Disease Using Electronic Health Record Data
SAGE PUBLICATIONS LTD. 2016: 290
View details for Web of Science ID 000377101000015
Statin Intensity or Achieved LDL? Practice-based Evidence for the Evaluation of New Cholesterol Treatment Guidelines
2016; 11 (5)
The recently updated American College of Cardiology/American Heart Association cholesterol treatment guidelines outline a paradigm shift in the approach to cardiovascular risk reduction. One major change included a recommendation that practitioners prescribe fixed dose statin regimens rather than focus on specific LDL targets. The goal of this study was to determine whether achieved LDL or statin intensity was more strongly associated with major adverse cardiac events (MACE) using practice-based data from electronic health records (EHR).We analyzed the EHR data of more than 40,000 adult patients on statin therapy between 1995 and 2013. Demographic and clinical variables were extracted from coded data and unstructured clinical text. To account for treatment selection bias we performed propensity score stratification as well as 1:1 propensity score matched analyses. Conditional Cox proportional hazards modeling was used to identify variables associated with MACE.We identified 7,373 adults with complete data whose cholesterol appeared to be actively managed. In a stratified propensity score analysis of the entire cohort over 3.3 years of follow-up, achieved LDL was a significant predictor of MACE outcome (Hazard Ratio 1.1; 95% confidence interval, 1.05-1.2; P < 0.0004), while statin intensity was not. In a 1:1 propensity score matched analysis performed to more aggressively control for covariate balance between treatment groups, achieved LDL remained significantly associated with MACE (HR 1.3; 95% CI, 1.03-1.7; P = 0.03) while treatment intensity again was not a significant predictor.Using EHR data we found that on-treatment achieved LDL level was a significant predictor of MACE. Statin intensity alone was not associated with outcomes. These findings imply that despite recent guidelines, achieved LDL levels are clinically important and LDL titration strategies warrant further investigation in clinical trials.
View details for DOI 10.1371/journal.pone.0154952
View details for Web of Science ID 000376882500009
View details for PubMedID 27227451
View details for PubMedCentralID PMC4881915
Learning statistical models of phenotypes using noisy labeled training data.
Journal of the American Medical Informatics Association
Traditionally, patient groups with a phenotype are selected through rule-based definitions whose creation and validation are time-consuming. Machine learning approaches to electronic phenotyping are limited by the paucity of labeled training datasets. We demonstrate the feasibility of utilizing semi-automatically labeled training sets to create phenotype models via machine learning, using a comprehensive representation of the patient medical record.We use a list of keywords specific to the phenotype of interest to generate noisy labeled training data. We train L1 penalized logistic regression models for a chronic and an acute disease and evaluate the performance of the models against a gold standard.Our models for Type 2 diabetes mellitus and myocardial infarction achieve precision and accuracy of 0.90, 0.89, and 0.86, 0.89, respectively. Local implementations of the previously validated rule-based definitions for Type 2 diabetes mellitus and myocardial infarction achieve precision and accuracy of 0.96, 0.92 and 0.84, 0.87, respectively.We have demonstrated feasibility of learning phenotype models using imperfectly labeled data for a chronic and acute phenotype. Further research in feature engineering and in specification of the keyword list can improve the performance of the models and the scalability of the approach.Our method provides an alternative to manual labeling for creating training sets for statistical models of phenotypes. Such an approach can accelerate research with large observational healthcare datasets and may also be used to create local phenotype models.
View details for DOI 10.1093/jamia/ocw028
View details for PubMedID 27174893
View details for PubMedCentralID PMC5070523
Factors impacting follow-up care after placement of temporary inferior vena cava filters
27th Annual Meeting of the Western-Vascular-Society
MOSBY-ELSEVIER. 2013: 440–45
Rates of inferior vena cava (IVC) filter retrieval have remained suboptimal, in part because of poor follow-up. The goal of our study was to determine demographic and clinical factors predictive of IVC filter follow-up care in a university hospital setting.We reviewed 250 consecutive patients who received an IVC filter placement with the intention of subsequent retrieval between March 2009 and October 2010. Patient demographics, clinical factors, and physician specialty were evaluated. Multivariate logistic regression analysis was performed to identify variables predicting follow-up care.In our cohort, 60.7% of patients received follow-up care; of those, 93% had IVC filter retrieval. Major indications for IVC filter placement were prophylaxis for high risk surgery (53%) and venous thromboembolic event with contraindication and/or failure of anticoagulation (39%). Follow-up care was less likely for patients discharged to acute rehabilitation or skilled nursing facilities (P < .0001), those with central nervous system pathology (eg, cerebral hemorrhage or spinal fracture; P < .0001), and for those who did not receive an IVC filter placement by a vascular surgeon (P < .0001). In a multivariate analysis, discharge home (odds ratio [OR], 4.0; 95% confidence interval [CI], 1.99-8.2; P < .0001), central nervous system pathology (OR, 0.46; 95% CI, 0.22-0.95; P = .04), and IVC filter placement by the vascular surgery service (OR, 4.7; 95% CI, 2.3-9.6; P < .0001) remained independent predictors of follow-up care. Trauma status and distance of residence did not significantly impact likelihood of patient follow-up.Service-dependent practice paradigms play a critical role in patient follow-up and IVC filter retrieval rates. Nevertheless, specific patient populations are more prone to having poorer rates of follow-up. Such trends should be factored into institutional quality control goals and patient-centered care.
View details for DOI 10.1016/j.jvs.2012.12.085
View details for PubMedID 23588109
Effect of chronic red cell transfusion therapy on vasculopathies and silent infarcts in patients with sickle cell disease
AMERICAN JOURNAL OF HEMATOLOGY
2011; 86 (1): 104-106
Regular, chronic red cell transfusions (CTX) have been shown to be effective prophylaxis against stroke in sickle cell disease (SCD) in those at risk. Because serial brain imaging is not routinely performed, little is known about the impact of CTX on silent infarcts (SI) and cerebral vascular pathology. Thus, we retrospectively evaluated the magnetic resonance imaging reports of a cohort of SCD patients who were prescribed CTX for either primary or secondary stroke prophylaxis. Seventeen patients with Hb SS were included (mean age 15 years, mean follow-up 4.3 years). Eight patients were on CTX for primary prophylaxis. New SI occurred in 17.6% of patients corresponding to an SI rate of 5.42 per 100 patient-years. Vasculopathy of the cerebral arteries was present in 65% of patients and progressed in 63% of these patients. Those who developed progressive vasculopathy were on CTX for an average of 8 years before lesions progressed. Patients on CTX for secondary prophylaxis had more SIs and evidence of progressive vascular disease than patients on CTX for primary prophylaxis. We conclude that adherence to CTX does not necessarily prevent SI or halt cerebral vasculopathy progression, especially in those with a history of overt stroke.
View details for DOI 10.1002/ajh.21901
View details for PubMedID 21117059