Current Funded Research Projects

Model-Driven Process Guidance to Improve the Safety and Efficiency of Human-Intensive Healthcare Processes, Joint funding initiative between the National Science Foundation and the Agency for Healthcare Research and Quality (Award # 1234070), $326,336, 9/12 – 8/15, PI, with Co-PIs Lori Clarke and Elizabeth Henneman.

The research objective of this Advancing Health Services through System Modeling Research project is to develop and evaluate health information technology (IT) interfaces that can be used to guide healthcare providers in real-time as they complete the complex, error-prone blood transfusion process. The researchers will determine how the interfaces should be designed to best help individuals complete the process safely, and to effectively alert and guide individuals when process failures or exceptional situations arise. The researchers will develop and evaluate these interfaces using laboratory-based evaluations and experiments in a realistic clinical setting. This project will establish foundational principles for the design of real-time health IT-based process guidance systems that improve healthcare process safety and efficiency. 

While health IT has the potential to guide individuals completing complex healthcare processes, it is often inflexible, forcing individuals to perform non-ideal, inefficient, and potentially unsafe processes. This project has the potential to dramatically reduce medical errors by designing health IT interfaces that allow individuals flexibility in the way they complete processes, yet provide enough guidance to maintain a high level of process safety. While this project focuses on the blood transfusion process, the fundamental findings will be generalizable to other complex healthcare processes. Through this project, graduate and undergraduate students will be trained by an interdisciplinary team of engineering, computer science, and nursing faculty members. The process guidance system and interfaces developed and evaluated through this project will be integrated into clinical simulation laboratory exercises used in the university?s nursing education curriculum.

CAREER: Computational Approaches to Model Physicians’ and Patients’ Interactions with Health Information Technology, National Science Foundation (Award # 1150057), $400,008, 7/12 – 6/17, Sole PI

The research objective of this Faculty Early Career Development (CAREER) project is to model physicians' and patients' cognitive and behavioral interactions with health information technology (IT) as these individuals deal with two costly chronic diseases - diabetes and hypertension. The researcher will model these interactions using methods from applied psychology, visualization, and computer science. Using these models, the researcher will assess the quality of physicians' and patients' cognitive and behavioral interactions across existing commercial health IT systems, and will design new computer interfaces that improve how physicians and patients interact with health IT.

By providing guidance for the effective design of health IT, this research will directly support current government initiatives aimed at stimulating the adoption of health IT. There is overwhelming evidence that health IT failures often result from a faulty understanding of how individuals use information while completing tasks and making decisions. This research aims to increase physicians' and patients' adoption of health IT by making it easier for them to find and use the right information at the right time. As chronic disease management poses a significant burden on the healthcare system, the specific design guidance identified via the proposed research has the potential to improve clinical outcomes and reduce the costs of managing these patients. Through this research, PhD students will be able to work closely with healthcare provider mentors and participate in interdisciplinary research teams with physicians, nurses, and computer scientists. Additionally, the researcher will develop realistic case studies to expose undergraduate and graduate students to the growing field of health systems engineering.

BRIGE: Quantitative Model-Based Visualizations of Complex Health Care Processes, National Science Foundation (Award # 1032574), $174,336, 9/10 – 8/13, Sole PI

This Broadening Participation Research Initiation Grants in Engineering (BRIGE) grant provides funding to design and evaluate visualizations that will allow health care decision makers, including policy makers and hospital administrators, to understand how health care workers and patients complete health care processes. Decision makers will be able to use the visualizations to make better choices about how to improve the health care processes, thereby increasing process efficiency and reducing process-related medical errors. The researcher will design visualizations using existing process data from two studies. In the first study, researchers used eye-tracking technology to document how health care workers verified a patient's identity, with and without barcoding technology. In the second study, researchers observed the surgery process, wherein some surgeons were sleep deprived. The researcher will evaluate decision makers' attention allocation, interpretation of information, and use of information when making process improvement choices based on 1) summary statistics and charts, 2) complex numerical information, and 3) varying forms of the visualizations.

If successful, the results of this research will improve health care decision makers' abilities to redesign health care processes. By fine-tuning the design of the process visualizations, they will be able to analyze and improve more health care processes in less time. By empirically evaluating how they use the visualizations, the research will ensure that the visualizations are easy to understand and useful to the decision makers. This approach is scalable and lends itself to visualizing processes in non-health care domains.

Controlling Disease Using Inexpensive Information Technology – Hypertension in Diabetes (CONDUIT-HID), Agency for Healthcare Research and Quality, Subcontract to UMass Medical School (Award # R18 HS18461-01A1), Barry Saver (PI), $1,975,613, 7/10 -  6/14

This project will develop and test a low-cost approach to using health information technology (HIT), aimed at improving the effectiveness and cost-effectiveness of care for chronic health conditions that are amenable to home self-monitoring, that is easy to disseminate. In contrast to many other HIT-based interventions, we will utilize commercial, off-the-shelf technology rather than custom, expensive HIT. We are using hypertension control among persons with diabetes as our test case because there is documented need for improving control of hypertension in this high-risk population and studies estimate that improving hypertension control in diabetes is more cost-effective than most other medical interventions and possibly even cost-saving in direct health care dollars. This high-value return on investment is important for encouraging adoption, expansion, and dissemination of HIT innovations. Our intervention will involve recruiting 400 persons with diabetes and uncontrolled hypertension receiving care through Fallon Clinic. Half of them will be randomly assigned to receive an automated blood pressure cuff capable of uploading readings through a computer, plus instruction on how to connect their cuffs at home or in the clinic to upload their information into a popular and free commercial personal health record (PHR) system. These blood pressure data from the PHR will be transferred automatically into Fallon Clinic's electronic health record system and will alert Fallon Clinic's existing team of diabetes care management nurses. Subjects whose blood pressure is uncontrolled will have their medication regimens intensified by these nurses according to protocols. Intervention subjects will receive regular outreach calls from the diabetes care nurses if their blood pressure remains uncontrolled or they are not periodically uploading blood pressure readings. After one year, we will compare outcomes between control and intervention subjects. Our primary outcomes will be change in mean blood pressure and proportion of subjects with controlled blood pressure. We will also measure a range of secondary outcomes including costs of the intervention, medication utilization, and a variety of patient-reported outcomes. Furthermore, we will interview and observe study subjects and care providers to gain a better understanding of factors affecting uptake and use of the intervention. We will examine continued use of the intervention after formal study participation ends and will also measure uptake of the intervention by control subjects offered delayed entry at the end of the formal study. Lastly, we will compare our study and findings to other, recent studies using HIT to improve hypertension control and develop a set of best practices and recommendations for future efforts in this area.

Collaborative Development of Climate Information for the Connecticut River Basin using Shared Vision Forecasting, National Oceanic and Atmospheric Administration (Award # NA10OAR4310182), with Casey Brown (PI), Richard Palmer, and Erin Baker, $299,835, 9/10 – 8/12, Co-PI

Water resources management has been identified as a sector where climate information on seasonal to interannual timescales has significant potential benefits. Yet the documented use of climate information for improved water management decision making is surprisingly rare. Previous research has identified a variety of causes including the difficulties associated with interpreting probabilistic forecasts and the institutional challenges to changing standard procedures. In addition to these established impediments, we propose that lessons from the study of the adoption of new technologies in society are underutilized in the design of climate information pilot projects and experiments. Further, we hypothesize that the diffusion-of-innovations model of technology adoption provides guidance that will significantly influence the uptake of climate information by water managers who do not currently use it. Motivated by this model, we will evaluate water managers' use of climate information and experimentally test a proposed facilitation method (Shared Vision Forecasting) that is drawn from the lessons of computer-assisted negotiation in water resources. The setting is the Connecticut River Basin, the largest river in New England and the source of water supply for the metropolitan Boston area. This effort consists of three components. The first is to establish a climate information knowledge network for the water managers of the Connecticut River Basin. Here we leverage an existing project focused on ecosystem health in the basin that convenes water managers to promote awareness, interest and evaluation of climate information. The second component characterizes the group's use of climate information and climate information needs through survey and discussion during annual workshops. Evaluation will be conducted prior to and following the outreach efforts. The third component is the experimental evaluation of the effect of collaborative co-production of tailored climate information by the water managers with forecasters. The Shared Vision Forecasting (SVF) approach will be applied to climate information development, evaluation and trial with a subset of a Connecticut River climate knowledge network. It is expected that the SVF-based approach will promote forecast adoption by increasing the compatibility and decreasing the complexity of climate information and by enhancing the positive interpersonal interaction effects. We test these expectations through a mixed-method approach, combining survey techniques with the water managers and experiments in a controlled laboratory with student subjects. With this approach we evaluate the applicability of diffusion-of-innovations theory to the use of climate information by water managers and evaluate SVF as a method to overcome previously identified impediments to adoption. Anticipated benefits include a better understanding of the adoption process for water managers and the identification of a potentially powerful methodology for enhancing the use and utility of climate information by water managers.