Fall term update: visit our COVID-19 Information website for more information.
We have an extensive and broad program looking at human factors in health care
Last Updated: April 24, 2020
Duration: May 2020 – April 2021
Artificial Intelligence (AI) systems have been continuously developed with time for various applications in the healthcare domain. Despite having made promising contributions like disease diagnosis, patient risk stratification, drug discovery, treatment selection e.t.c., research has shown that clinicians find it difficult to interact with these AI systems making adoption and implementation very slow. The concern of the clinicians is usually regarding the black box models and how they do not find them understandable, reliable, trustworthy, ethical or fair. We are working on research into providing interpretable and explainable AI systems in healthcare. We are also investigating the interaction of both clinicians and patients with these AI systems.
Duration: 2019 – 2023
Physiological monitoring in critical care is of utmost importance when examining critically ill patients’ health. Multiple interfaces exist in the ICU and often cause higher cognitive workload as data trending and retrieval can be challenging. Within this project, we use Cognitive Work Analysis models to understand clinicians and aim to support diagnosis and thus better outcomes of patient care.
Related Publications:
Sponsors and Partners:
Duration: September 2019 - Present
This research project is lead by the University of Waterloo School of Pharmacy, Dr. Tejal Patel and her research team, in collaboration with the AIDL.It is estimated by the World Health Organization that only 50% of people who live with a chronic illness in developed countries adhere to their medication regimen. Non-adherence can result in poor disease outcomes, increased re-hospitalization, and significant costs to our health care system. The objectives of this research are to examine the acceptance of a smart blister packaging system by patients, understand the impact of the system on patient behaviour, as well as the impact of adherence data visualizations on health care providers.
Sponsors and Partners:
Duration: September 2019 - Present
Children with medical complexities require 24-hour care and often receive home care nursing to temporarily relieve family caregivers of their responsibilities. However, there is no standard for the handoff of care between these individuals. When a breakdown in communication occurs between caregivers, these children are put at risk for the incorrect administration of a medication and errors in care performance that often involve the use of medical technology. A lack of understanding the best practices for a specific child’s medical condition can result in further variability of the child’s treatment and comfort in their home. Stellar Care is a mobile application that was developed by 4th year Biomedical Engineering students (University of Waterloo, Class of 2019) to facilitate effective communication between family caregivers and home care nurses for continuity of care.This project is currently exploring the design of a conversational agent to support families and nurses in their respective caregiver roles in the home.
Video: https://www.youtube.com/watch?v=o3uPSR0nkOc
Related Publications:
Duration: Sept 2019 – Present
The objective of this research is to investigate how stress management training affects performance in the workplace. To do this, participants experienced a stress-inducing virtual reality simulation made to emulate an industrial setting. Measurements of heart rate variability were collected during the simulation to be used as an indicator of stress. With the use of HRV and other qualitative measurements, we hope to identify the effect stress management training has on an individual’s overall performance and stress levels.
Sponsors and Partners:
Duration: 2018
The high number of ventilated patients in the ICU are often a significant group affected by the outcome of human errors. This pilot study aimed to detect usability errors with different ventilator interfaces, i.e. transitioning from a familiar to an unfamiliar ventilator as well as their equivalent interfaces when standardized nomenclature is incorporated as proposed in the ISO 19223.
Related Publications:
Duration: 2017
The purpose of this study was to investigate how people interact with external information environments for health information gathering and how the exposure to health information actually impacts their behaviour.
Related Publications:
Duration: January 2014-December 2020
This multiyear project has had several stages. First we looked at the data quality tradeoffs that primary care physicians make. Then we implemented a persuasive solution to improve their data entry experience. Now we are looking at data quality in other contexts.
Duration: January 2015-December 2020.
This multiyear project has been looking at the human side of medication management. What decision support do pharmacists need? How do patients manage their medications? How do we improve adherence to medication based treatment? This project is funded by TELUS Health and through NSERC and CIHR grants.
Chin*, J., Üreten*, E., Burns, C. (2018). Search for health information across the lifespan: Validation, confirmation or learning. Poster presented at the Cognitive Aging Conference 2018, Atlanta, GA.
Duration: January 2011 - March 2014
Critically ill patients in the intensive care unit (ICU) require constant specialized observation by a number of clinicians (nurses, respiratory therapists, doctors) at a given time. To make decisions and plan treatments, these clinicians need to observe, combine and understand a wide range of information from various devices located at the bedsides of multiple patients. The problems with existing technology are: a) fixed location of monitoring devices at the bedside; b) information overload from too many devices, each with its own display; c) no integration and interaction between devices; d) displays which do not show information in a way that is easy to understand; and e) lack of historical information to support understanding of data and clinical decision making.
To develop the Monitoring Messenger device we have completed a detailed analysis of the work-flow in a pediatric ICU. Based on the results of this analysis we have developed mock-ups of the device. Now we will improve and finalize the device with the direct involvement of healthcare providers.
The novelty of our project is in its use of a formal design approach and transferring technology developed for the OR into the ICU.
Related Publications:
Sponsors and Partners:
Duration: November 2012 - Present
With the rising number of hip and knee replacement surgeries, more patients are requiring rehabilitation to accelerate the healing process. Rehabilitation clinics are led by physiotherapists who prescribe and customize workout programs based on a clinical assessment of the patient’s progress. These assessments are performed by visual observation, patient questionnaires and goniometry, which can measure range of motion of only a single joint while the patient is at rest. However, even when carefully performed, human errors and biases are likely to occur due to the inherent element of subjectivity in these methods.
To address this concern, an Automatic Rehabilitation System (ARS) is being developed to provide the physiotherapist with quantitative patient data using human motion tracking technologies. Our lab tackles the issue of how to display this information effectively while minimizing the added complexity so that physiotherapists can better assess patient recovery. To this end, evaluations of early prototypes have been conducted with patients and physiotherapists at St. Joseph’s Health Centre (Guelph, ON). This project is in conjunction with the Adaptive Systems Laboratory (Department of Electrical and Computer Engineering) and Cardon Rehabilitation & Medical Equipment Ltd.
Related Publications:
Sponsors and Partners:
Duration: Completed February 2014
Research indicates that behavior change interventions can assist a patient's performance with respect to their health, and empower them to take a proactive role in their well-being. The objective of this research was to identify features that make remote health-monitoring systems more effective. To that end, a persuasive design approach was used to design a prototype application that encourages patients to consult their home monitoring devices appropriately. An exploratory user study was conducted to evaluate the participants’ opinions about the developed prototype, their preferences, as well as their concerns regarding the different components of the mockup.
Related Publications:
Sponsors and Partners:
Duration: Completed September 2013
This research was about developing an effective user interface for a decision support tool for mobile devices, aimed at presenting emergency department waiting-time information to the patient population. The prototype received very positive feedback from a variety of stakeholders, the head of the Emergency Department, and the trained nurses. A user study was also conducted to evaluate the usability of the system.
Sponsors and Partners:
Duration: 2006 - 2008
The project provided guidance to improve the labelling of injectable pharmaceutical containers. Existing labels for injectable pharmaceutical containers, specifically ampoules and vials, were evaluated against the Health Canada regulations and the standard developed by the CSA International. A FMEA was performed on the label reading process to identify potential human errors in the process and prioritise their criticality. Based on the results of FMEA, improved label designs will be developed and evaluated via user testing.
Related Publications:
Sponsors and Partners:
Duration: 2004 - 2006
This project developed a mobile decision support tool for cardiac care nursing coordinators doing telephone triage. The support included patient care algorithms, research evidence, electronic record keeping, advanced display ideas, and a mobile drug database. A two month field trial of the support tool showed it was readily adopted, gave useful advice, and supported the triage task well.
Related Publications:
Sponsors and Partners:
Duration: 2002 - 2005
The mobile EID (mEID) diabetes management project involved the design and development of a mobile application to help diabetics better visualize elements of their condition, thus improving understanding and management. Variables such as blood glucose and plasma were identified at each level of the Abstraction Hierarchy in the Work Domain Analysis (WDA). The information requirements extracted from the WDA were subsequently used to design the mobile information displays for the application. A usability evaluation of the application revealed that users performed more efficiently with modified task-based mEID displays vs. mEID displays alone.
Related Publications:
Sponsors and Partners:
