Alan Ellis
Bio
Durham native Alan Ellis came to NC State in July 2014 after 17 years at UNC-Chapel Hill, which he spent in various research capacities, including 11 years as a research associate at the Cecil G. Sheps Center for Health Services Research. In addition to research, Ellis enjoys writing and teaching.
Website
Education
Ph.D. Social Work University of North Carolina at Chapel Hill 2012
MSW Social Work University of North Carolina at Chapel Hill 1996
B.A. Computer Science and Spanish Dartmouth College 1990
Area(s) of Expertise
Research: mental health services, mental health promotion, child well-being, social skills, methods for comparative effectiveness research, teaching and learning. Teaching: evaluation of social work practice, social justice. Other expertise: conflict resolution.
Publications
- Perioperative Albumin Among Adults Undergoing Thoracic Surgery in the United States: Utilization, Associations With Clinical Outcomes, and Contribution to Hospital Costs , JOURNAL OF CARDIOTHORACIC AND VASCULAR ANESTHESIA (2024)
- Stigmatizing and affirming provider language in medical records on hospitalized patients with opioid use disorder , Journal of Hospital Medicine (2024)
- Feasibility of a best–worst scaling exercise to set priorities for autism research , Health Expectations (2022)
- Rapid and Deferred Help Seeking Among African American Parents of Children With Emotional and Behavioral Difficulties , Psychiatric Services (2022)
- Duration of medical home participation and quality of care for patients with chronic conditions , Health Services Research (2021)
- Evaluation of trends in continuous peripheral nerve block utilization for total knee arthroplasty within and outside the Veterans Affairs Healthcare System , Regional Anesthesia & Pain Medicine (2021)
- Improving Methods for Discrete Choice Experiments to Measure Patient Preferences , (2021)
- Perioperative colloid choice and bleeding in patients undergoing musculoskeletal surgery: An observational administrative database study , International Journal of Critical Illness and Injury Science (2021)
- Propensity Score Weighting and Trimming Strategies for Reducing Variance and Bias of Treatment Effect Estimates: A Simulation Study , American Journal of Epidemiology (2021)
- Propensity score trimming mitigates bias due to covariate measurement error in inverse probability of treatment weighted analyses: A plasmode simulation , Statistics in Medicine (2021)
Grants
Prescription drugs play a crucial role in the prevention and treatment of chronic disease in older adults. The evidence base for the benefit and harm of these treatments comes from experimental (randomized controlled trials, RCTs) and nonexperimental (observational, epidemiologic, or, more recently: real world evidence) studies. For older adults, and especially those with multiple comorbidities, and, consequently, taking multiple drug classes, critically important evidence comes from nonexperimental studies, because these individuals ������������������ particularly those who use the most drugs ������������������ are often excluded from RCTs. Unfortunately, nonexperimental studies often suffer from confounding, including by frailty, as recently evinced by RCTs on statins that failed to confirm nonexperimental findings of protective effects in sepsis-associated acute respiratory distress syndrome, chronic obstructive pulmonary disease, and lung cancer . Confounding by frailty can lead to suboptimal or even harmful treatment decisions (if nonexperimental studies are the only evidence available) or the conduct of costly RCTs that fail to replicate the findings of nonexperimental research. For a timely assessment of drug benefit and harm in older adults and real-world settings, it is therefore vital to develop and apply improved methods to reduce confounding in nonexperimental studies. Funded by R01/R56 AG023178 since 2005, we have achieved substantial advances in knowledge about methods to improve the validity of nonexperimental research. Using both empirical data and extensive simulations, we have developed novel analytic techniques to reduce confounding, including propensity score calibration and the exclusion of patients treated contrary to prediction. We have disseminated our results by means of oral presentations, posters, and workshops/���������symposia and in a series of 112 publications, including 23 in the top epidemiologic journals (AJE and Epidemiology), 23 in the top pharmacoepidemiologic journal (PDS), 10 in Medical Care, and several in top medical journals (JAMA, JNCI, Diabetes Care, and JAGS). We propose to continue building effective tools to address the most significant problem hindering nonexperimental study designs from generating valid answers about beneficial and harmful effects of treatments, i.e., the problem of uncontrolled confounding. While the proposed study will build on our work over the last 15 years through increasing the validity of nonexperimental methods to assess the preventive effects of treatments in older adults, we propose significant and innovative aims. We will address these aims using empirical motivating examples and extensive simulations to improve the validity of nonexperimental comparative effectiveness research (CER). Improved validity of nonexperimental CER will result in more robust evidence about drug benefit and harm in older adults. This work will directly inform clinically relevant treatment decisions, provide timely evidence unavailable from RCTs, and ultimately improve individual and overall health of older adults.
North Carolina State University (NC State) School of Social Work is applying for continuation funding and requesting $1,919,598. NC State University������������������s experienced grant staff have successfully recruited and trained 197 BHWET trainees, including 30 that will graduate this year. Grant staff will establish inter-professional relationships with community agencies, provide evidence-based trauma-informed training to agency staff, faculty and trainees, recruit and train a diverse workforce, and utilize an evidence-based method of program evaluation. Our MSW BHWET graduates enter the social work field well prepared to address behavioral health needs with cultural humility, practice experience, and knowledge of evidence-based treatment modalities. Since 2014, as a HRSA funded site, NC State has risen as a leader in behavioral health engagement and training both statewide and locally. Funds for this proposal will provide high-quality inter-professional training to agency staff, faculty, and MSW students to increase knowledge and access to high quality behavioral services in the state. Our grant staff will administer a training program that focuses on health center collaboration, community engagement, and violence prevention to address the behavioral health needs of the target population, children, adolescents, and transition age youth in NC. Over four years, we will enhance and establish interdisciplinary relationships with community-based partners, recruit 60 new training sites, provide training and support to 116 MSW students, agency staff, and faculty. NC is a state with rural and urban population center where 94 out of 100 counties are flagged for mental health care shortages.
North Carolina (NC) a state with rural and urban areas has been severely harmed by the opioid epidemic. Four areas of the state are in the top 25 of overdose deaths in the country (CDC, 2018). We are partnering with 3 sites in these hard hit counties.The purpose of this proposal is to request $1,337,219 over three years to increase the number of clinical social workers equipped to provide prevention, treatment, and recovery services in high need/high demand areas. Master of social work (MSW) students will train in community-based experiential sites, alongside paraprofessionals, psychologists, pharmacists, nurses, physicians, psychiatrists and others, in the treatment of opioid use disorders (OUD) and substance use disorders (SUD). The proposed OWEP staff will work collaboratively with community-based training sites to place students in high-quality integrated or interprofessional team-based care settings. Measurable objectives include: 1) Recruit 84 students over three years to train in integrated interdisciplinary OUD/SUD sites with at least two other disciplines. 2) Increase the number of OUD/SUD field placements in high demand areas. 3) Develop, increase, and enhance behavioral health training that focuses on field supervisors, faculty, community-based agency staff, family support members, and faculty from collaborating programs by engaging experts in the addiction field. 4) Collect post-graduate employment data, including the National Provider Identifier (NPI), for at least 3 years after program completion. The PI/Program Director has 5 years of experience successfully administering similar HRSA-funded projects and will provide overall grant management, including staff, budgets, reports, and compliance. The Program Coordinator will manage training activities, program tasks, and community relationships and recruit community-based training sites. The Program Evaluator, who also has 5 years of experience on related HRSA grants, will monitor the evaluation plan, collect and analyze data, and disseminate results. The Program Recruiter will recruit applicants from underserved communities to fill the pipeline of students committed to an OUD/SUD career in high-need areas and will also recruit non-traditional partners. OWEP staff will work with key stakeholders, including the NC Harm Reduction Coalition, UNC Center for Health Equity and the Center for Families and Community Engagement to build OUD/SUD prevention, education, and treatment capacity around the state. NC State has submitted for Funding Priority for demonstrated ability to place social work professionals in integrated care settings and Funding Preference for our high rate of placing graduates in medically underserved practice settings. In addition NC State is partnering with site that have HPSA score 16+ and OD rate above 21.7.
Prescription drugs play a central role in preventing and treating chronic disease in older adults. Observational studies can provide important information on the comparative effectiveness of alternative treatments, but such studies suffer from bias due to unmeasured confounding. The parent study-- a collaboration led by UNC with investigators at RTI Health Solutions; Harvard; University of Manchester, UK; and NC State-- focuses on developing methods to reduce unmeasured confounding. Dr. Ellis will participate in regular telephone conferences and correspondence to assist with analysis and interpretation for all analytic aims of the parent R01 project. He will conduct analyses pertaining to the parent project������������������s Aim 1 (to develop guidance for defining study populations with equipoise), Aim 3 (to explore negative control outcomes), Aim 4 (to assess the effect of prioritizing covariates on covariate-balancing propensity score estimation), and Aim 6 (to develop methods for variable prioritization for coarsened exact matching in pharmacoepidemiology). Dr. Ellis will also participate in dissemination of findings from the parent project.
Community-engaged research requires capturing stakeholder preferences, and a well-known method of capturing patient preferences is the stated choice experiment (SCE). SCEs typically present multiple sets of alternatives and allow stakeholders to indicate which alternatives they prefer. For example, in scenarios involving different cancer treatments, the alternatives might have attributes such as effectiveness, risk of adverse effects, out-of-pocket cost, treatment setting (inpatient vs. outpatient), and procedure type (medical vs. surgical), among others. When a group of people responds to several different scenarios that vary with respect to these attributes, the resulting data make it possible to describe not only people������������������s preferences within attributes (e.g., preference for medical procedures over surgical procedures), but also how much weight they assign to different attributes (for example, the degree to which cost is more important than treatment setting or vice versa). However, SCE studies suffer from two related problems: (1) the lack of standard methods for determining sample size requirements and (2) the ability to yield ����������������statistically significant��������������� results that fail to provide meaningful information about the preferences of a larger population. This study aims to remedy these problems through simulations based on actual stakeholder preference data. We will implement two different methods for sample size determination and describe their sensitivity to assumption violations and design choices. Further, we will develop more accurate methods of measuring the level of uncertainty about the estimates being generated. These approaches will result in guidelines for sample size in SCE pilot studies, as well as additional reporting guidelines for SCEs. These outcomes will improve our ability to capture stakeholder preferences and use that information in healthcare decision-making.
Dr. Ellis will perform data management and analysis tasks and assist with dissemination of findings. Dr. Ellis will be using existing access to a multihospital database (Premier, Inc.) to help answer the following questions. These questions align with one of the two main aims of the DVAMC PSCI - which is the reduction of Perioperative Opioid Related Adverse Drug Events. The questions are: (1) What were the recent patterns of perioperative opioid use, non-opioid analgesia, and anesthesia - among patients undergoing Joint Replacement Surgery - at ~500 non-VA hospitals between Q4 2008 and Q3 2014? (2) How did institutional rates of Perioperative Opioid Related Adverse Drug Events (PORADEs) vary after adjusting for inter-facility differences in patient-mix ? (3) What patterns of perioperative opioid use, non-opioid analgesia, and anesthesia did hospitals that were "positive deviants" have? i.e., Among hospitals where observed rates of PORADEs were significantly lower than what would be expected based on patient-mix, what were the prevailing patterns of practice as discernible using Premier's administrative-financial data? Dr. Ellis's work will set the stage for determining whether the patterns of practice at VA hospitals compare favorably with non-VA hospitals for the most common major operation performed in both systems - Joint Replacement Surgery.
Dr. Ellis will perform data management and analysis tasks and assist with dissemination of findings. Dr. Ellis will be using existing access to a multihospital database (Premier, Inc.) to help answer the following questions. These questions align with one of the two main aims of the DVAMC PSCI - which is the reduction of Perioperative Opioid Related Adverse Drug Events. The questions are: (1) What were the recent patterns of perioperative opioid use, non-opioid analgesia, and anesthesia - among patients undergoing Joint Replacement Surgery - at ~500 non-VA hospitals between Q4 2008 and Q3 2014? (2) How did institutional rates of Perioperative Opioid Related Adverse Drug Events (PORADEs) vary after adjusting for inter-facility differences in patient-mix ? (3) What patterns of perioperative opioid use, non-opioid analgesia, and anesthesia did hospitals that were "positive deviants" have? i.e., Among hospitals where observed rates of PORADEs were significantly lower than what would be expected based on patient-mix, what were the prevailing patterns of practice as discernible using Premier's administrative-financial data? Dr. Ellis's work will set the stage for determining whether the patterns of practice at VA hospitals compare favorably with non-VA hospitals for the most common major operation performed in both systems - Joint Replacement Surgery.
Dr. Ellis will perform data management and analysis tasks and assist with dissemination of findings. Dr. Ellis will be using existing access to a multihospital database (Premier, Inc.) to help answer the following questions. These questions align with one of the two main aims of the DVAMC PSCI - which is the reduction of Perioperative Opioid Related Adverse Drug Events. The questions are: (1) What were the recent patterns of perioperative opioid use, non-opioid analgesia, and anesthesia - among patients undergoing Joint Replacement Surgery - at ~500 non-VA hospitals between Q4 2008 and Q3 2014? (2) How did institutional rates of Perioperative Opioid Related Adverse Drug Events (PORADEs) vary after adjusting for inter-facility differences in patient-mix ? (3) What patterns of perioperative opioid use, non-opioid analgesia, and anesthesia did hospitals that were "positive deviants" have? i.e., Among hospitals where observed rates of PORADEs were significantly lower than what would be expected based on patient-mix, what were the prevailing patterns of practice as discernible using Premier's administrative-financial data? Dr. Ellis's work will set the stage for determining whether the patterns of practice at VA hospitals compare favorably with non-VA hospitals for the most common major operation performed in both systems - Joint Replacement Surgery.
Engaged research activities such as patient-centered outcomes research and community-based participatory research depend on the ability of diverse stakeholders to work in partnership. The communication of stakeholder priorities is a vital element of engaged research, but evidence about the best ways to elicit stakeholder preferences is scarce. To learn about stakeholder priorities with regard to future research needs in a variety of substantive areas, the Agency for Healthcare Research and Quality (AHRQ) typically relies on ranking exercises conducted with small panels that include scientists, clinicians, informed patients, and/or advocates. Compared to this small-panel approach, large-panel surveys may allow a broader range of opinions to be expressed. Further, compared to the traditional ranking exercise, a prioritization technique known as best-worst scaling may provide richer information about stakeholder priorities. The proposed project involves the analysis of qualitative and quantitative data from small and large panels of individuals who participated in both a traditional ranking exercise and a best-worst scaling exercise. Findings will be disseminated in order to improve methods for capturing stakeholder priorities. These findings are expected to have broad implications beyond the prioritization of future research needs.