Start Your Search
e-Health 2018 Virtual Meeting
Celebrate, Grow & Inspire Bold Action in Digital Health - Vancouver, BC
This product offers access to the e-Health 2018 Keynote / Plenary Presentation Live Webcasts, the recording of these 4 sessions and access to all PDF/Presentation Slides of each conference presentation.
Group Discounts Available for 5+ Purchases. Contact us to request group pricing.
PDF's of presentation PowerPoints are now online!Presentation Date(s):
Non-Member Price: C$120+tax Digital Health Canada Member Price: C$100+tax
- May 27 - 30, 2018
- Total Presentations: 240
OS27 - Disrupting Technology into the Next Decade
- Type: Oral Session
- Track: Technical/Interoperability
- Presentations: 6
- Coordinates: 5/30/2018, 10:30 - 12:00, Fairview V Room, Conference Level
OS27.01 - Distributed Consent Management by Blockchain
Purpose/Objectives: Different medical software applications are adopted proliferate for different purposes and organizations, including clinics, hospitals and pharmacies. The difficulty of enforcing consistent privacy and consent rules in multiple systems is one factor encouraging centralized data management. Centralized data repositories may then attempt to leverage technologies such as portals to mimic interoperability instead of actually supporting data exchange protocols. This paper presents an alternative approach which separates consent management from data records. Consent directives are maintained in a blockchains ledger, which is distributed and publicly accessible. As a consequence of blockchain technology, this consent ledger can be copied and re-distributed, is inherently consistent and reliable across different applications, is immutably secure with respect to consent history, and can be referenced by any records system, including paper records. Since the consent ledger is public, the resulting infrastructure allows any application or user to verify consent without requiring special authorization to access the ledger. Future designs, technology or records systems will also have access to the ledger without requireing re-engineering or revision to the blockchain. Additional implementation to enforce privacy and consent rules for different systems is not necessary since all systems can access the same distributed consent ledger. Access is unconstrained since the consent ledger is public and replicable. Since health record data is not stored with the consent ledger, public accessibility of the ledger does not increase the risk of privacy breach.
Methodology/Approach: A demonstration implementation is provided using Solidity e-contract language under the Ethereum blockchain technology.
Finding/Results: We illustrate a consent directive model which supports authorization, delegation and revocation of consent. It also supports configurable data specification within consent directives, so any data storage technology or data type can be referenced. The semantics for describing record data can be revised, without modifying the model, so changes in regulations or health policy can be reflected without requiring software revision.
Conclusion/Implications/Recommendations: We argue that the flexibility of an open consent model and reference implementation for any record and data technology can encompass existing privacy enforcement mechanisms, such as role based access control, since such mechanisms can be mimicked by determining or redefining consent directive semantics. In addition, there is deceased risk of technology or regulation lock-in, as technology, legislation and social policy of today are less prone to become part of the the out-of-date and too-expensive-to-replace legacy systems of tomorrow.
140 Character Summary: A blockchain health consent ledger increases flexibility in privacy protection, avoids technology lock-in and allows health information policy to evolve.
OS27.02 - Deep Learning Techniques to Improve Patient Care with Neural Networks
Purpose/Objectives: Neural Networks are biologically inspired learning algorithms. Artificial networks have come a long way and are now considered some of the most powerful and robust learning algorithms deployed in numerous emerging software-based innovations. The goal behind neural networks (deep learning) is to construct a large hypothesis space of functions that contains a good approximation to the underlying function that represents the deterministic behavior of a process that generates the data. In the case of healthcare, the are definitive use cases of applying advanced predictive models using similar patient data to train a neural network. With improved care opportunities that are guided by predictive systems, comes the added potential of better care management including case management, disease mangagement, and high-risk case identification. An example of a neural network is shown below. The talk will demonstrate the results of applying deep learning training techniques to patient electronic health data.
Methodology/Approach: TensorFlowTMis an open source software library for numerical computation using data flow graphs. Nodes in the graph represent mathematical operations, while the graph edges represent the multidimensional data arrays (tensors) communicated be- tween them. The flexible architecture allows you to deploy computation to one or more computer systems with a single API. TensorFlow was originally developed by researchers and engineers working on the Google Brain Team within Googles Machine Intelligence research organization for the purposes of conducting machine learning and deep neural networks research, but the system is general enough to be applicable in a wide variety of other domains. In the context of this presentation, TensorFlow was applied to patient records that were formulated by the Johns Hopkins Adjusted Clinical Groups (ACG) model. The ACG system measures health status by grouping diagnoses into clinically cogent groups. The goal of the ACG system is to assign each individual a single, mutually exclusive ACG value, which is a relative measure of the individual's expected or actual consumption of health services. ACGs are closely related to many health trends.
Finding/Results: Results show nearly 50% reduction of error in cost prediction.
Conclusion/Implications/Recommendations: Results show promise of neural networks to reduce the mean cost prediction error of patients by over 50%.
140 Character Summary: In healthcare, there are definitive use cases of applying advanced predictive models using similar patient data to train neural networks.
OS27.03 - The Power of Patient/Provider Messaging: From Human to AI
Purpose/Objectives: Patients love being able to message with their providers, and messaging between in-clinic visits can improve adherence and outcomes. However, many providers are concerned about the extra time that may be associated with patient messaging. This session explores best practices for managing remote patients and driving adherence based two clinical studies of interactive care plans, with people with Parkinson's disease and for seniors at risk of falls. In both studies, patients were accessed in person, and then assigned a personalized mobile care plan, that included the ability to message remotely with a care provider. Providers used this communication and also analysis of patient-reported outcomes to advance the care plans without in-person visits, and improve patient adherence remotely. We will also explore insights from machine learning classification applied to over 80,000 messages between patients and providers, and discuss how artificial intelligence when combined with caregiver interactions can be used to scale care. The session will debunk common myths about patient messaging, and show an opportunity that combines both technology and a human touch.
Methodology/Approach: Study protocol for Falls Study: https://bmcgeriatr.biomedcentral.com/articles/10.1186/s12877-017-0618-x Study Protocol for Parkinson's Study: Study goals -Provide safe and challenging exercise intervention via mobile -Enable patient and provider messaging -Decrease in-person patient visits without impacting outcomes Methodology for ML message classifer: -Trained on 80,000 patient provider messages and keywords that may indicate adverse event -Initial input, manual classification of messages into categories of "informational, questions, urgent"
Finding/Results: Findings are still being collected for REACH study and will be presented at ACRM at end of October 2017. Initial indicators are positive. Patients in the Parkinson's study were 81% adherent to their care plan and reported 9/10 patient satisfaction with the program. Lower activated patients saw 2 times greater increase in activity with mobile intervention than those in control group. Analysis of 80,000 patient messages, 70% informational, 28% questions, and 2% urgents. Benefits of patient messaging can outweigh the perceived costs. Note that we expect to have further analysis of messaging, and the outcomes from the REACH study available to present by the conference in May 2018.
Conclusion/Implications/Recommendations: This session explores both qualitative feedback from remote messaging, and compare messaging styles to the theoretical framework for behavior change, and identify how messaging and adherence to interactive care plans fulfill the targets of self-efficacy, outcome expectations, motivation, knowledge, and social persuasion. Digital interventions that connect patients with providers outside the clinic can have positive impact on patient outcomes, without increasing the costs of care. Patients can benefit from remote interactions with providers, and these can either replace other forms of communication.
140 Character Summary: Patient and provider communication backed by machine learning can improve patient outcomes and satisfaction, without increasing costs of care.
OS27.04 - Iris Scanners as an Identification Tool for Individuals Experiencing Homelessness
Purpose/Objectives: The rationale for this research is the disadvantage that is inherent to the frequent loss of ID cards and the subsequent limited access to non-emergency healthcare services. The overall objectives of this study were to assess the functionality of iris scanning technology in a community setting and to evaluate the acceptability of iris scanning for client identification among participants.
Methodology/Approach: Participants were recruited from a Salvation Army homeless shelter who were either staying there for the month or accessing its services such as the food bank. As participants checked-in to the shelter, they were asked to have a scan of their iris which generated a unique identifying number for each participant. In addition, 50 participants were then asked to perform a second iris scan to verify that the technology could accurately identify them and match their identifying number with their previous scan. Participants answered a short questionnaire focusing on client acceptability and feasibility of the iris scanning identification system. Iris scanning was performed with iris recognition equipment called Seek Avenger, developed by Crossmatch Technologies. Quantitative and qualitative analyses were conducted to generate descriptive statistics and to determine a thematic grouping of responses, respectively.
Finding/Results: The research team recruited 200 participants over the course of three visits. A total of 191 participants agreed to an iris scan. Qualitative findings revealed three themes for agreeing to a scan; safe and fast identification, incentive for participation ($2 coffee card), and to help a good cause. Of the 167 participants who answered the question on identification preference, 146 (87%) stated they preferred an iris scan over a health card. Reasons for this preference included simplicity, and losing ID cards/not required to carry an ID card. The 21 participants (13%) that preferred an ID card to iris scanning was either for fear of being tracked by government agencies, information falling into criminal hands for misuse, or the lack of system capacity to manage the data. The iris scan was successful in scanning the eye(s) of 182 participants and unsuccessful for six participants, resulting in a success rate of 97%. Furthermore, 50 participants agreed to a second iris scan of which 49 were accurately matched with their previous scan, resulting in an accuracy rate of 98%. The remaining 2% was due to the participant not being able to stay still long enough for the scanner to focus on the eye.
Conclusion/Implications/Recommendations: Based on the current qualitative and descriptive statistics, iris scanning is a safe and reliable form of identification that can withstand fraudulent activity. If implemented, it can be cost-effective in this population given the frequency of ID card replacement, and the lack of access to non-emergency health care services, the latter of which leads to an increase in emergency department usage and therefore funding. This study recommends the establishment of a quality-assured iris recognition program for identification of individuals experiencing homelessness in order to gain access to services. Concerns and perceptions of surveillance, inabilities of the system, and safety of information must be addressed before implementation.
140 Character Summary: Iris recognition was found to be an acceptable, reliable and feasible method of identification among individuals experiencing homelessness.
OS27.05 - Using Blockchain to Enable Informed Patient Consent for Research
Purpose/Objectives: The PROOF Center (Vancouver), St. Pauls Hospital Vancouver, University of Nebraska Medical Centre and Deloitte are collaborating to reimagine how clinical and genomics data is shared in a secure and transparent manner, and setting up the foundations for the next evolution of research analytics. Currently, the process of enrolling a new patient in a research program is time intensive and paper based. The enrolled patient has little control over their own data and the process for providing their consent to share and receive their data. The objective is to prove that a blockchain-powered solution will enable the process to identify the right patient and enroll them in the research program. The solution will make it easy for the patient to own and provide consent to share their data, enabling easy access for researchers.
Methodology/Approach: Developed a solution with consent workflows that are powered by blockchain. For the PoC we went with a web-based UI, but we envision building a mobile app in future. The process leveraged a hybrid agile approach that provides for an up-front framing and stakeholder alignment phase followed by iterating prototype sprints. The key stakeholder for this POC were: PROOF (Prevention of Organ Failure) Center, St. Pauls Hospital Vancouver, Nebraska Medical Centre, and Deloitte.
Finding/Results: Through this POC, we were able to collect patient consent and store it in a transparent, secure and a verifiable manner. Timestamps corresponding to consent statuses provided an audit trail for audit purposes. The key benefit for the health / research community is that blockchain technology can be leveraged for a more efficient, secure, and reliable process of accessing recorded patient consent before releasing the patients data. Blockchain technology also allows for secure audit trail of the shared patient data.
Conclusion/Implications/Recommendations: We have seen in this proof-of-concept study that all consent-related data can leave an unfalsifiable and verifiable fingerprint on the Blockchain. This is important both on the stakeholders side, letting them prove the existence and the consistency of the data, and on the patients side, giving them more visibility, transparency, and hence control over their consent. Tracking the complex data flow with numerous diverse stakeholders, and documenting it in real-time through a timestamping workflow, is a key step towards proving data transparency and inviolability, and could improve clinical trial process. The application of Blockchain technologies in the context of clinical research is broad and promising. The decentralized nature of blockchain technology helps introduce communities to contemporary clinical research, thus allowing researchers to enroll patients using a more targeted approach.
140 Character Summary: Blockchain based solution captures and tracks patient consent in secure and verifiable manner while enabling targeted enrolling of patients in research programs
OS27.06 - Improving Management of Long Stay Patients with Machine Learning Prediction
Purpose/Objectives: A grand challenge that hospitals face is managing facilities and manpower under the uncertainty of how long an admitted patient will stay. This uncertainty can restrict efficient utilization of hospital resources by prohibiting effective scheduling and coordination of early discharge preparations. Long stay patients (?30 days) are particularly impacted, where prolonged hospitalization is frequently associated with significant social, economic, physical and psychological burden. At Fraser Health, the year-to-date average number of long stay patients is 17% higher than the fiscal year-end target. Early identification of long stay patients and effective care planning to reduce long stay can lead to better health outcomes, greater patient and provider satisfaction, reduced risk of adverse incidents and complications, and contribute to overall healthcare system sustainability.
Methodology/Approach: Health and Business Analytics (HBA) has designed a predictive model called long stay predictor (LSP) that can predict whether an admitted patient will be a long stay patient or a short stay patient with 75% accuracy. Through artificial intelligence trained on over 10,000 historical long stay cases, the model is capable of identifying new long stay patients within 48 hours of admission. Patients flagged by the predictor are notified through daily email, upon which an early escalation of discharge planning is triggered to identify discharge barriers and establish a monitoring process on the patient. Staff can follow-up on individual patients real-time through an in-house web tool. Figure 1. Models inputs & outputs Figure 2. Workflow
Finding/Results: At present, the tool has been deployed across White Rock/South-Surrey, Delta, Burnaby and Chilliwack. A 5-month post-implementation analysis at Peace Arch Hospital reveals a 16% reduction in long stay patients, and 25% reduction in bed days. Operational lessons learned from this project, along with novel insights that reveal why some patients are long stay, will be shared at the conference.
Conclusion/Implications/Recommendations: Fraser Health has created a tool to predict patients likely to become long stay, and have operationalized this information to help clinicians engage in early care and discharge planning, leading to a significant reduction in bed days. Next steps include expanding adoption across all Fraser Health sites and adapting model to predict readmissions.
140 Character Summary: Predictive modelling provides timely alerts to improve care planning for long stay patients at Fraser Health, contributing to a 25% reduction in bed days.
OS28 - Bringing Mental Health to the Forefront
- Type: Oral Session
- Track: Clinical Delivery
- Presentations: 6
- Coordinates: 5/30/2018, 10:30 - 12:00, Granville I Room, Conference Level
OS28.05 - A Smart Homes Concept for an Inpatient Psychiatric Population
Purpose/Objectives: The objective of this exploratory pilot study is to gain insight into the use of smart technologies for individuals with mental illness living in transitional hospital apartments. This pilot study, planned after a successful previous trial by Corring, Campbell and Rudnick (2012), is currently examining the feasibility of implementing a smart homes intervention for participants with mental illness and to determine if further modifications are needed prior to wide-scale community deployment in the homes of discharged participants. The innovative use of technology may be one such strategy in increasing the communication between clients and health care providers. In addition, it was hypothesized that the intervention will assist participants nearing discharge into the community with greater support for successful integration by facilitating independence and improving self-reported health outcomes.
Methodology/Approach: Up to 20 participants (aged 18-85) who are inpatients at either Parkwood Institute or Southwest Centre for Forensic Mental Health Care and meet inclusion criteria will be recruited. Participants will be provided with screen devices such as smartphones, tablets and smart mirrors that can provide video-conferencing capabilities and send questionnaires to their health care providers through the Collaborative Health Record program. These smart technologies will be linked with the Lawson Integrated DataBase which is programmed to transmit prompts and reminders to coincide with the participants care plans. Health adjunct Bluetooth-enabled monitors such as blood pressure monitors, heart rate monitors, weigh scales, glucometers, medication dispensers, and sensor floor mats will also be available. Participants are allowed to select their preferred screen devices and health adjunct monitors based on their health needs. The research team will use a mixed-methods (quantitative & qualitative) design to assess the feasibility of the technology, as well as the perceptions and health of the participants. Upon a one week minimum stay in the apartment, participants will complete a semi-structured interview with research staff. This will be followed up with a 6-month interview post-discharge. Furthermore, focus groups will be conducted with hospital staff to further evaluate the feasibility of the smart technologies.
Finding/Results: The quantitative results from the participant interviews and the qualitative findings from staff focus groups will be discussed. Quantitative findings will include an analysis of the Perception of Smart Technology tool, the Housing History Survey Form, the Community Integration Questionnaire, the EQ-5D Health Utilities Index, Short Form-36, and the Health, Social, Justice Service Use Questionnaire. Thematic analysis will be used for open-ended questions and focus group discussions.
Conclusion/Implications/Recommendations: This pilot study is being conducted to assess the feasibility of using smart technology within an inpatient hospital setting for individuals with mental illness. It is envisaged that this pilot study will provide information to enhance the intervention before wider-scale adoption of the technology in the community. The implications of this study could inform health policy and decision makers to adopt more smart technologies into mental health care and/or treatment plans. Long term implications could include being able to effectively serve more individuals with mental illness, and prevent homelessness and criminalization of the population under study.
140 Character Summary: A pilot study investigating the use of smart technologies in providing support to individuals with mental illness living in transitional hospital apartments.
OS28.02 - Physician Adoption of Standardized Order Sets and Electronic Order Entry
Purpose/Objectives: Prescription errors and unexplained prescriber variation within hospital practice greatly contributes to adverse drug events that lead to potential safety incidents. Patient safety initiatives at other institutions prioritized reducing these occurrences and out of these initiatives, electronic-order entry has demonstrated a significantly decrease adverse drug events. The multi-site structure of Trillium Health Partners (THP) fosters different physician cultures, practices and protocols between sites. Standardization of physician order sets across the three locations will decrease prescriber variation and provide a standardized patient experience. The Clinical Order Sets, Technology and Standardization (COTS) project aims to harmonize order sets and embed these in an electronic-order application while promoting physician adoption of the new electronic system.
Methodology/Approach: Standardized order sets were developed from best-practice literature/guidelines and follow Quality Based Procedures (QBPs) and Quality Standards where applicable. Approval was obtained from a multidisciplinary team of physicians, clinical educators, pharmacists, and nurses to ensure applicability. All order sets were approved at an organized Order Set Committee. The Mental Health Program was identified as an early adopter group to implement the use of the order set technology platform on their units. Forty-six physician pre- and post-implementation surveys were completed to identify demographics and any barriers related to technology uptake. Parameters such as user demographics, attitudes towards harmonized orders and towards electronic entry were assessed. Technology uptake and application usage data were also analyzed. Findings were used to identify barriers and enablers in adopting standardized electronic order entry.
Finding/Results: Gender, awareness of QBPs, and computer proficiency impacted the uptake of standardized electronic order entry, while a physicians duration within their current position had little effect. Clinicians recognized that harmonized orders improved decision-making whereas their attitudes regarding harmonized orders improving patient safety were unclear. EntryPoint uptake and patient safety incidents may be monitored long term.
Conclusion/Implications/Recommendations: Findings offered insight into perspectives around change management and standardized electronic order entry. A robust change management plan has been developed to promote clinician utilization of only order entry thought user-acceptance testing, hands-on demonstration, in-time training, and peer-to-peer education. Process mapping sessions are being completed to embed newly acquired hardware into clinician practice. Attitudes from physicians in different programs will be aggregated over the next seven months as the system becomes available throughout THP.
140 Character Summary: Physician usage of standardized electronic-orders and attitudes towards their implementation was analyzed to promote uptake across multiple sites.
OS28.03 - Global Experiences in Building Cognitive Digital Health Systems
Purpose/Objectives: Digital and cognitive care has the potential to contribute to better and safer care of patients, more effective health providers as well as to the sustainability of health systems world-wide. With increasing digitization of the healthcare world, cognitive systems help organizations unlock new opportunities. In a recent Institute of Business Value study, 4 out of 5 healthcare executives believe that cognitive computing will have a critical impact on the future of their healthcare business and even more intend to invest in cognitive capabilities (IBM IBV, 2016). But what is a Cognitive Digital Health System and how are organizations getting started on these journeys? We will define Cognitive digital healthcare with an example of how cognitive care got started in the mental health domain and provide guidance for organizations wanting to begin their own cognitive journey.
Methodology/Approach: Cognitive platforms are designed to ingest vast quantities of structured and unstructured information from numbers and text to audio, video, images and other data. With an increasing volume and velocity of data, cognitive solutions focus on assembling new kinds of data in machine readable forms and on curating information identifying new patterns and insights to accelerate discoveries, treatments and insight. Various cognitive tools and methodologies are proving useful in digital healthcare, including virtual agents, natural language understanding, advanced analytics for risk scoring and predictive modelling and Internet of Things functionality. For example, the IBM #Here4U team with mental health as focus area developed the #Here4U virtual agent for youth in need of support, aiming to reduce bullying, anxiety and addiction with preventative analysis and intervention through digital personal connection thats available anytime, anywhere. The virtual agent is a set of cognitive technologies, including a conversation agent that responds to text-chats, natural language processing, machine learning and training by identifing presenting issues such as stress, anxiety and depression. After this successful experiment, the team applied the same foundation to build virtual agents for PTSD and for employee mental health support.
Finding/Results: The team successfully built a mental health virtual agent proof of concept utilizing a cognitively trained chatbot to detect presenting issues. The team then reused the underlying cognitive components and data structures to build other mental health virtual agent solutions An effective approach to understanding cognitive in digital health systems is to create a cognitive program and then fund a series of use cases in a learning environment.
Conclusion/Implications/Recommendations: Cognitive technologies are so new that cognitive strategies are journeys into the unknown. Unlike traditional technologies, cognitive programs must be designed to explore different linked use cases, iterate, experiment with sets of technologies and pivot in different directions as circumstances, benefits and adoption becomes clearer. Digital Hospitals are complex ecosystems with hundreds of clinical and business processes containing thousands of sub-processes. Cognitive solutions will not be standalone, identifiable applications, but will be capabilities that are embedded through the digital hospital fabric to seamlessly unite patients, clinicians, staff, assets and information throughout the hospital, delivering the right information and resources at point of care at the right time.
140 Character Summary: An effective approach to Cognitive digital health systems is to create a cognitive learning program and then experiment with use cases built on a common foundation.
OS28.04 - Telemedicine and Its Growing Usage
Purpose/Objectives: To demonstate the adoption, applicability and benefits of using telemedicine.The intended outcomes were to prove an improvement in the accessibility and timelines of specialized mental health services for patients at the receiving institution, improving clinical outcomes for the recepients of the service and to reduce costs associated with the provision of mental health service.
Methodology/Approach: We piloted the use of telemedine in a correctional setting over a 12 month period. The clinic was held once weekly with a mean of four clinics or sixteen hours per month. Pre and post case discussions were facilitated with staff in addition to seeing the patient. Inclusion and exclusion criteria were applied. Patients were seen following screening done by in reach staff at the correctional institute. Patients referred included, male or female, adult or Youth offenders with documented or suspected serious mental disorders in need of assessment and/or treatment. Treatment was provided in the form of Pharmacological and Psychological interventions.
Finding/Results: We found that there was an increased level of understanding of the criminal justice system and the inmate population. Appropriate and efficient use of the inpatient hospital beds with some referrals being initiated, some being averted and others being expedited for clinical reasons. From the receiving institution point of view, patients expressed satisfaction and comfort using Telemdicine. Enhanced access to specialist service was reported through decreased wait time, timely follow up, reduction in wait time for urgent assessments for those on suicide watch.
Conclusion/Implications/Recommendations: The improvement in the services and mutual satsfaction between the service provider and the receiving institution in addition to improved outcomes have resulted in more contracts being awarded. We are now a dedicated service in the province of Ontario who are providing service to at least six correctional institutions many of which are remote. Some of these institutions have had very little Mental Health Support over the years. The team has a team of Physicians who are supported by excellent staff from the Learning and Development team who coordinate and ensure the smooth running of the clinics.
140 Character Summary: The rise of the machines: Telemedicine allows timely access to care with reduction in costs and inconvenience associated in providing in person care.
OS28.01 - Development of Mental Health Reporting Framework
Purpose/Objectives: This presentation will focus on the development of a robust enterprise wide reporting framework to support organizations strategic, operational and quality improvement plan.
Methodology/Approach: A meaningful reporting project was launched with a goal of streamlining reporting; establish measurement standards and definitions; reducing number of and variation in reports and facilitating single source of truth for enterprise data. This multi-phased project included conducting an in-depth current state analysis exercise to understand the measurement, reporting and analytics needs and activities of CAMH. Following the current state analysis, a gap analysis exercise was conducted to identify reporting efficiencies, opportunities for product alignment and unmet reporting needs. Recommendations from the gap analysis exercise resulted in formation of an enterprise reporting framework comprising of report mapping by audience, frequency and business intelligence product alignment.
Finding/Results: First year of reporting framework implementation resulted in the following: Governance structure: A Data and Reporting Governance Committee was established to provide CAMH-wide leadership and oversight for data quality assurance, measurement and reporting priorities, and alignment of reporting across CAMH. Two sub-committees were also created to specifically focus on internal measurement and reporting; enterprise data quality. Centralized Intake process: A centralized intake process was established to ensure a standardized method is followed for anyone requesting reports from the Reporting and Analytics department Figure1. Centralized Intake Process of Reports Annual measurement plan: Annual measurement & reporting plans are being developed for priority areas as per the operational and quality improvement plan Reporting products to aid in data driven decision making: The Corporate Balance Scorecard, Clinical Programs and Key Priorities Dashboard were developed to measure organization wide performance against mission, vision and quality improvement priorities. Noticeable improvements post balanced scorecard implementation included: Significant increase in medication reconciliation completion rate at discharge. In 17-18-Q1 the rate was 4% above target at 77%. 17% increase from16-17-Q4. Figure 2. Medication reconciliation discharge rate over time
Conclusion/Implications/Recommendations: Development of the centralized intake process for reporting, enterprise reporting framework has improved customer service by increasing quality, efficiency and efficacy of reports generated at CAMH. It has enables the organization to move towards data driven decision making and quality improvement initiatives.
140 Character Summary: This abstract outlines the development and implementation of a Mental Health Reporting Framework that supports the organizations strategic goals.
OS28.06 - Insights to Impact: Helping Organizations Use Data for Business Improvements
Purpose/Objectives: The Centre for Addiction and Mental Health (CAMH) provides high quality and client-centered care to meet the needs of people facing addiction and mental health challenges. To support these diverse clinical services, CAMH established a Reporting Framework to create a single source of truth for clinical, finance and workforce reporting. iManage is the self-serve management reporting tool that informs and enhances planning, decision making and performance improvement. The purpose of this abstract is to showcase how the development of Key Performance Indicators (KPIs), through stakeholder engagement, led to the active use of iManage and the ability to transform lives through data.
Methodology/Approach: The CAMH Business Intelligence (BI) team leveraged iManage, an enterprise reporing portal, in order to develop a Suicide Risk Assessment (SRA) Dashboard that includes a series of KPIs to provide managers and clinical teams with important information related to suicide risk assessment and clinical best practices. There were two critical steps used to implement KPIs: stakeholder mapping and business requirements gathering. The first step to building effective KPIs is to understand the targeted audience. The BI team, employing the knowledge of data usage in CAMH by clinical staff, mapped all potential users of an SRA dashboard. All users were later refined into a single user group. The second step in this approach was to understand which KPIs would provide the most value and benefit to the user group. With the support of the applications and clinical teams, the BI team was able to map out key suicide risk pathways and responsibilities. For example, patients who are assessed as high risk should be monitored more closely than those identified as moderate risk. The user group was leveraged to identify the step by step patient journey and the areas along the pathway that should be measured and monitored. These focus points were then prioritized and appropriately developed into KPIs.
Finding/Results: Significant stakeholder engagement allowed the SRA dashboard to be developed with a number of distinct views, each representing a different focus area from the business focus group. The dashboard series contained an overarching performance summary dashboard, as well as requirement specific dashboards focusing on; assessment compliance, care plan compliance, continuous observation orders and incidents. The SRA dashboards are actively used by CAMH stakeholders to provide a weekly lens to monitor improvements and observe care planning on current clients, as well as noticing any trends from previous weeks data. The development of the SRA Dashboard and the availability of rich data drove SRA Completion rates up from 82% in August 2015 to 95% by August 2016.
Conclusion/Implications/Recommendations: Through effective stakeholder mapping and business requirement gathering, the BI team was able to build the correct KPIs for an enterprise SRA dashboard. This in turn led to the use of KPIs in iManage to create business improvement initiatives. The principles applied in this model should be considered for future data projects, in order to enhance engagement and improve clinical outcomes.
140 Character Summary: Effective development and evaluation of key performance indicators allows for optimal data adoption and application for the benefit of improved patient outcomes.
OS29 - e-Health Solutions for Patient Self Management
- Type: Oral Session
- Track: Clinical Delivery
- Presentations: 6
- Coordinates: 5/30/2018, 10:30 - 12:00, Granville II Room, Conference Level
OS29.01 - Change Before You Have to…Emerging Models of Care
Purpose/Objectives: Healthcare is in a perpetual state of change with new treatments, models of care and technologies being implemented daily. What used to be the traditional doctors bag now includes handheld devices facilitating the care connection with patients and other care providers outside of the hospital and doctors office. At the same time, patients are demanding not only prompt service, but also convenience of access for their health care. This workshop will explore emerging models of care enabled by virtual technology. This includes the growing expectation that a patients primary care provider has immediate access to the information they need to provide a continuum of care no matter where the patient is in a rural setting or an urban location.
Methodology/Approach: The panel will review new evolving models of care enabled by technology at Providence Health Care in Vancouver and the BC Interior led by the Kootenay Boundary Division of Family Practice from the provider and patient experience. Rapid Access to Consultative Expertise is a telephone advice line where family physicians (FPs) can call one number, choose from a selection of specialty services and be routed directly to the specialists cell phone for advice usually within a few minutes. eCASE, electronic consultation to specialists expertise is a non-urgent model of e consultation where FPs can access specialists advice through email. Other models of virtual health including secure messaging, enotification and video conferencing in traditional and not-so-traditional settings (ICU, Mobile Maternity, PreSurgical) will be discussed.
Finding/Results: Evaluation metrics, based on the Triple Aim Framework show promising results in relation to - Experience of Care, Population Health, and Per Capita Cost. Provider and user satisfaction is high at 95%, 60% of interactions avoid a face-to-face consultation, 32% of interactions avoid an emergency department visit and patients have improved access to speciality care.
Conclusion/Implications/Recommendations: These models are revolutionizing communications between specialists and family physicians - building relationships, providing clinical decision support, and opening access to different modalities of care - so patients have a continuum of care in their own community. By exploring future models of care through technology we will: Present information from the patient and provider perspective on remote consultation services Delve into new models of care eCase, RACE, secure messaging, Mobile Maternity, non-traditional telehealth Investigate the shifting value base related to virtual care do our values match those of our patients?
140 Character Summary: This workshop will explore emerging models of care enabled by virtual technology from the provider, user and patient perspective.
OS29.02 - Feasibility of Remote Patient Monitoring for COPD and Heart Failure
Purpose/Objectives: 1. Describe strategies used for initiation and modification of remote patient monitoring/telehomecare. 2. Describe lessons learned from feasibility period to sustainability and scalabilty. 3. Promote patient success stories and outcomes. The Remote Patient Monitoring Program focused initially on Congestive Heart Failure (CHF) and Chronic Obstructive Pulmonary Disease (COPD) patient populations, whose chronic conditions lead them to frequent emergency departments and multiple hospital admissions. Management of Diabetes Type II was included in recent months. Technology is used as a platform to deliver healthcare outside the conventional setting in the patient's own home. Patient data is electronically transmitted (e.g., symptoms, vital signs, outcomes) from the home to the clinical team with a goal to identify evidence-based care interventions, provide education, support, and health and wellness coaching that improves patient self-management. Early evidence shows that expected benefits of: reduced length of stay, emergency department visit and acute admission reductions will be realized. High satisfaction rates with the patient care experience with this healthcare delivery model have been demonstrated. Over 700 patients have been enrolled in the program to date with numbers expected to double in the next 12 months - improving access. Patient success stories and outcomes will be shared to demonstrate the impact remote patient monitoring/telehomecare can have - especially in remote locations.
Methodology/Approach: Patient identification was competed using electronic data examination and direct patient referral. Health session kit (iPad and peripheral devices) were delivered using in-home assisted installation and through a self-install (direct ship to patient) model. These methods were analyzed for efficiency and cost. Referrals received via fax, email or by telephone Eligible patients are contacted either by phone and/or letter Patients who accept are consented and enrolled for a 4 - 6 month period iPad mini, BP cuff, pulse oximeter, and weigh scale delivered to the home Biometric data and symptom question responses are delivered remotely to the RPM dashboard daily to be monitored by RN. RNs contact the patient when pre-established individualized thresholds are surpassed. Advice, coaching and intervention is provided as required. Pre-scheduled coaching calls are completed for goal setting, action planning, self-management support and behaviour modification
Finding/Results: Direct patient referral is more efficient but requires increased focus on clinician engagement. Self-install model (direct ship) is more cost efficient but requires more technical support via phone and clinician. Technology is the enabler. Clinical self-management support, education, and intervention are considered most valuable to patients. ER visits and acute admissions will be examined in the 12 month period before enrollment in the RPM program and 12 months after enrollment Patient surveys are administered on enrollment, completion of the program and 4 months after disenrollment A benefits analysis report will be completed by December 2017
Conclusion/Implications/Recommendations: Audience will hear results of the program: - enrollment and equipment feasibility anlysis - findings in efficiencies and cost reductions - learnings for scalability and expansion - outcomes for patients and efficiencies for organizations - patients related stories of success
140 Character Summary: RPM empowers and engages patients to be experts in their own care. Self-management support and education using technology is key to success in achieving outcomes.
OS29.03 - Personalized Chronic Disease Management: Balancing Apps and Health Coaching
Purpose/Objectives: Medically complex patients with co-existing conditions often receive conflicting clinical advice, hindering their abilities to appropriately self-manage multiple chronic conditions (MCC), resulting in frequent hospitalizations, premature admissions to long-term care, and a decreased quality of life. The aim of this demonstration project was to iteratively evaluate the impact of a patient-centered mobile remote patient monitoring (RPM) system, coupled with health coaching, on the experience of care, health outcomes, frequency of hospitalizations, and overall well-being of patients with MCC. The RPM system specifically enabled patients with MCCs (i.e., heart failure, chronic obtrusive pulmonary disease, chronic kidney disease, and/or diabetes) to track and monitor their own biometric data and patient reported outcomes, triggering appropriate self-care instructions and social supports. The alerts generated were also remotely monitored by a nurse and health coach; the nurse monitored intervened when appropriate, and the health coach promoted self-management and healthy behaviour change through the empowerment approach. Together, the nurse and health coach collaborated and facilitated social support for the patients. The three key priorities of this project were to 1) improve the patient experience, 2) improve health outcomes, and 3) reduce avoidable hospitalizations and emergency room visits.
Methodology/Approach: The demonstration project had a rolling recruitment of patients who were diagnosed with heart failure, chronic obtrusive pulmonary disease, and/or diabetes. Once enrolled, patients visited the health coach at the clinical site to be on-boarded into the study, receive training on the RPM platform, and complete the study activities, such as study questionnaires and blood tests. The patients were enrolled in the program for a duration of 6 months. The project team developed a health coaching protocol that was adapted based on the needs expressed by the patient, moving towards a highly personalized health coaching protocol.
Finding/Results: Over the 6-month period, we recruited 40 patients who had one or more of the following: heart failure, chronic obtrusive pulmonary disease, and/or diabetes. In addition to clinical outcomes, we collected analytics around both the usage of the mobile apps and the individualized health coaching delivered to each patient, and conducted 31 semi-structured interviews with patients. Preliminary analysis suggests that the preferred ratio of technology to health coaching was highly variable among participants, however technology assisted delivery of health coaching had high acceptability and perceived effectiveness among patients.
Conclusion/Implications/Recommendations: While the benefits of remote patient monitoring for complex patients in specialty clinics have been demonstrated, there remains a gap in the support these patients receive prior to the escalation of their condition. This project demonstrated that these patients valued the ongoing support of a health coach in conjunction with remote patient monitoring, and that the ratio of the technology and coaching vary greatly on the individuals needs and their unique psychosocial complexities. The next generation of mobile platforms need to consider that chronic disease management requires a blended approach, that moves away from one-size-fits-all concept, and towards a model that is able to titrate the ratio of technology and customized support based on individual needs.
140 Character Summary: Complex patients in primary care valued the combination of digital remote patient monitoring and highly individualized health coaching for the management of MCCs.
OS29.04 - End-User Engagement: Electronic Self-Care Application for Patients with Heart Failure
Purpose/Objectives: Heart failure (HF) is a chronic disease that affects over 1% of Canadians and is associated with a significant economic burden (2.8 billion/year). Self-care is key to the management of HF and can potentially lead to better clinical outcomes. Proper HF self-care includes tasks such as daily weight and symptom monitoring, as well as adjusting diuretics based on weight. Nevertheless, patients find HF self-care challenging, with less than 50% of patients regularly weighing themselves. Mobile applications can support self-care but barriers such as literacy, numeracy and mild cognitive impairment can lead to challenges in adopting technology. Our previous work supports the use of a paper-based standardized diuretic decision support tool (SDDST) promoting self-care in older individuals with HF to manage their daily weight and adjust their diuretic dose accordingly. The primary objective of this study was to use participant (HF patients, informal care-providers [CPs]) input to convert our paper-based SDDST into a user-centered electronic mobile application.
Methodology/Approach: We recruited patients (male and female, age > 60) with a confirmed diagnosis of HF, and their CPs from the Heart Function Clinic at the Hamilton Health Sciences General site. HF patients were categorized into three groups, 1) adequate self-care patients, 2) inadequate self-care patients without a CP or 3) inadequate self-care patients with a CP, based on their self-care abilities measured with the Self-Care Heart Failure Index (SCHFI) where a score of > 70 is considered self-care adequate. We are conducting semi-structured interviews with HF patients and CPs using Persona-Scenarios. Interviews are transcribed verbatim and analyzed using NVivo, version 10, for emerging themes regarding self-care. This study has received ethics approval from the Hamilton Integrated Research Ethics Board.
Finding/Results: Thus far, we have interviewed 6 patients (4 male, 2 female, mean age: 74) and 3 CPs. We have identified 3 major themes which include 1) Challenges with technology, 2) Communication and assistance with circle of care, and 3) App customization. Many of the challenges patients and CPs mentioned involved their unfamiliarity with technology. However, participants were supportive and more likely to actively use the HFApp when informed of the interventions inclusion of volunteer and nurse assistance. Data collection and analysis is still in-progress and will be completed by the end of December.
Conclusion/Implications/Recommendations: Many mHealth apps fail to maximize their potential due to the lack of end-user engagement. To our knowledge, this is the first the study that includes patient and CP feedback throughout the design process. Our solution focuses on the patients needs, where the app must be easy to operate, robust and effective. Today, value-based outcomes and patient engagement are key to managing costs, which self-care can help achieve. We expect that simpler more user-friendly apps will result from our study supported by a patient-centric model of feedback.
140 Character Summary: Self-care is key for HF management. We have designed a simple, patient-centered mHealth app to improve HF-self-care in the home setting.
OS29.05 - A Business Case for Artificial Intelligence in Remote Patient Monitoring
Purpose/Objectives: The use of Remote Patient Monitoring (RPM) technology to manage chronic disease patients is gaining popularity. Dealing with chronic conditions only at the acute state in an episodic manner has led to high demand and inefficiencies in most health care delivery systems (Thorpe et al. 2004). It has been suggested that early diagnosis, continuous care, and coordinated health care would reduce system burden and make chronic disease management more effective (Rajan et al. 2013). However inefficient use of technology can in fact lead to less efficiency itself. In our recent CPRPM study paramedics using RPM technology observed that close to 50% of the patient interactions were non-coachable moments. The introduction of Artificial Intelligence (AI) in the RPM system not only increases the efficiency of the system but also increases the engagement of patients for coaching purposes and to keep patients engaged. Preliminary research has differentiated between three roles: 1) patient monitoring provides patients with access to their daily readings, 2) provider feedback helps patients understand and learn how to self-manage their condition, and 3) digital leverages artificial intelligence to predict health risks and exacerbations. The objective of this study is to examines RPM technologies used by patients not in isolation, but rather within an ecosystem that includes learning technology, providers, patients, and feedback that supports and gently pushes or nudges positive health behaviors and outcomes.
Methodology/Approach: Observational study that utilizes a prospective longitudinal cohort design to compare the role of RPM across two use cases. The first use case is human-centric as its design is highly dependent on the active role of the provider to respond to systematic alerts and provide feedback. The second use case is technology-centric as its design incorporates data analytics and artificial intelligence to generate more customized alerts and predict risks and exacerbations. The outcomes evaluated are the impact of both use cases on healthcare system utilization (i.e., reduction in 911 calls, reduction in ED visits). EMS Service Rurality Target # Patients Use Case 1 Target # Patients Use Case 2 Total Sample EMS 1 Rural 10 10 20 EMS 2 Rural 15 15 30 EMS 3 Urban 10 10 20 EMS 4 Urban 15 15 30 Total 50 50 100
Finding/Results: A previous study of 212 participants involved in the first use case resulted in a 42% reduction in 911 calls and 40% reduction in ED visits. Although results were extremely promising, it is difficult to scale results to a larger population due to high demand on provider influence. By comparing the original use case with the technology-centric alternative, benefits and challenges for both use cases will be identified.
Conclusion/Implications/Recommendations: The implications of our study are relevant to organizations and government bodies involved in connecting the business of health and technology to transform health care into a more sustainable delivery model.
140 Character Summary: The introduction of smart home AI in the remote monitoring system not only increases the efficiency of the system but also increases the engagement of patients.
OS29.06 - Scaling up Home Telemonitoring: Insights and Lessons from TEC4Home
Purpose/Objectives: TEC4Home Heart Failure (HF) examines using home tele-monitoring to support the safe transition of care from hospital to home. Home monitoring equipment is used to collect biometric measurements (weight, blood pressure, pulse, and oxygen saturation) from patients daily to detect deterioration and implement early interventions, thereby avoiding unnecessary ED readmissions and hospitalizations. This project is led by a Canadian university in partnership with the provincial health ministry, regional health authorities, and a health technology partner. The four year initiative is designed to evaluate the efficacy of home tele-monitoring and to scale up and spread the practice provincially. The Phase 1 feasibility study at 3 sites demonstrated a reduction in re-hospitalizations and improvement in quality of life, resulting in the expansion to the Phase 2 randomized controlled trial with 22 community sites across the province, starting in 2018. This abstract explores insights and lessons learned while scaling up across health authorities in a real-world research trial implementation.
Methodology/Approach: In Phase 1, the feasibility study tested and refined procedures in three domains: model of care (i.e. how monitoring nurses work with patients and health professionals); model of technology (i.e. how the equipment and data will be used); and model of research (i.e. what metrics to use to track outcomes). This approach proved effective. When scaling up from three sites to multiple sites across health authorities in Phase 2, a fourth domain, model of management, was added as another stream. Topics requiring consideration within each domain are listed: Model of Care: How to account for and accommodate variations in standards of care and resourcing between sites, and identify areas of harmonization? How can monitoring nurses develop and share best practices provincially, while also retaining practices relevant to community-specific needs? Model of Technology: How will the monitoring equipment be delivered to patients homes in various communities, including ones with limited connectivity and those in remote areas? How will data be captured and shared across jurisdictions using different electronic health systems? Model of Research How to develop capacity for research on sites and across health authorities to set up support for future trials? How to develop a common evaluation framework, identifying key provincial metrics with health authority and site specificity? How to pool data to facilitate analysis and prospective tracking for future quality improvement? Model of Management How to create new job descriptions for monitoring nurses in keeping with health system and professional standards? How to establish policies and practices across health authorities to enable sharing of monitoring nursing capacity and data? How to transition from project to sustainable program with appropriate governance?
Finding/Results: We will discuss how the above framework guided the project to scale up and roll out the trial.
Conclusion/Implications/Recommendations: TEC4Home aspires to generate high quality evidence to define future standards of care in home tele-monitoring. This will contribute to the evidence-based policy translation needed to guide the implementation and sustainability of home tele-monitoring in the province. This empirical approach can help harmonize administrative practices and create new channels of collaboration to introduce innovations into routine health practices, establish sustainability, and increase research capacity for an enduring legacy in digital health research. This work is submitted on behalf of the TEC4Home Healthcare Innovation Community.
140 Character Summary: Scaling up TEC4Home Heart Failure: Insights for collaborative "real-world" large scale trials and implementations.
OS30 - Enabling High Value Analytics
- Type: Oral Session
- Track: Clinical Delivery
- Presentations: 6
- Coordinates: 5/30/2018, 10:30 - 12:00, Cambie Room, Conference Level
OS30.01 - Using Business Intelligence for Insights into Healthcare Data
Purpose/Objectives: Healthcare organizations gather large volumes of data which continues to exist in legacy formats making it difficult to analyze or use effectively. Over the last few years, we have engaged in collaborative research projects to apply business intelligence (BI) techniques for integrating, analyzing and reporting on such data. The projects have spanned areas of Critical Care, Services Availability, Ambulatory Care, Patient Transfer, and Trauma. Services availability is guided by population, patient and case mix group profiles. While population profiles focus on demographics, the patient profile provides an overview of health-related metrics within selected regions by showing information such as births, commonality of chronic conditions, and prevalent or vaccine preventable diseases. Comparative reports are generated at various levels of hierarchy ranging from health service delivery areas to individual facilities. Adequate analysis of ambulatory care sensitive conditions results in preventable hospitalizations and enhanced patient care. In addition, predictive analytics models were built to guide resource allocation based on the forecasted trends. The inter-facility patient transfer dashboard uses BI techniques to analyze data related to healthcare infrastructure and services, and provides a web-based system to quickly identify optimal destinations for inter-facility transfers. The solution is now being extended for province-wide adoption. Finally, the trauma project is intended to provide data-driven perspective of incidences, mortality and transportation along several dimensions.
Methodology/Approach: Our agile methodology consists of building multi-dimensional online analytical processing (OLAP) cubes and render reports using business intelligence tools. Intuitive navigation eliminates the need for training or user manuals; this is further enhanced by mapping tools, customized shape files and embedded objects. Data visualization, adhoc reporting, and ease of use has been the key factors in rapid adoption and deployment of these solutions. Aesthetically pleasant and interactive dashboards display KPIs with ability to navigate at finer granularity using multi-level drill-down and drill-through reports. Parameterized reports allow selection of multiple dimensions simultaneously and are rendered in a matter of seconds while sifting through years of data. The performance is further enhanced by connecting selected reports directly to optimized backend data warehouse. User-friendly web forms safely constrain future data entry and ensure consistency. For existing repositories, integration modules are developed to cleanse and upload data from disparate sources. Data anonymization and aggregation is used where warranted. For privileged information, access controls have been implemented. Our designs are modular and allow for incremental development.
Finding/Results: The benefits are multi-faceted with the audience ranging from managers and physicians to strategic decision makers. In some cases, this work has also yielded unintended benefits. For instance, the critical care and services availability dashboards have been used for orientation of newly recruited physicians.
Conclusion/Implications/Recommendations: Our objective is to share our findings from several years of demonstrated success with peer groups and demonstrate the effectiveness of our approach. Most development occurs in an academic setting working closely with partners in the healthcare organizations. The presentation will consist of numerous dashboards, web forms and reports. The challenges encountered will also be discussed.
140 Character Summary: Analytical processing techniques can immensely accentuate healthcare data making it more valuable. This will be demonstrated via dashboards, forms and reports.
OS30.02 - Using Advanced Analytics to Predict Clinical Aggression in Mental Health
Purpose/Objectives: Leveraging the advanced analytics capabilities inherent in a fully integrated EHR, the hospital is piloting an analytics solution with the goal of improving capabilities to predict and clearly identify clinical aggression (CA) and enhance staff capacity to respond effectively to risks of future CA from a diverse patient cohort. Objectives: -Reducing incidents of CA, harm to others, and restraint/seclusion use while increasing therapeutic engagement -Increase confidence and competence of staff in identifying and managing the antecedents of CA -Improve safety for both clinicians and patients
Methodology/Approach: A number of risk assessment tools for imminent aggression and discrete data are captured in the EHR; however, they have limited predictive power with respect to indicating when a patient could become aggressive. Additionally, the predictive utility of these scales are in the very near term; and the ideal predictive model would be able to predict CA risk days prior to its occurrence to have time to implement robust interventions. To address this, we have partnered with an expert to create a neural network predictive model that will leverage data found within unstructured and structured data sources in the EHR. If the model is successful at predicting CA, the implications for next steps are vast, including ability of clinicians to be alerted to become proactive to improve patient outcomes. Validation of the predictive model was completed over 2 months with a pilot focusing on the predictive models ability to accurately predict CA incidents with current data over a four month period. Pre-implementation data will be utilized to compare prediction performance rates with implementation groups. A comparison of expected verses actual number of CA incidents will be conducted utilizing reports which identify the total number of CA incidents against the expected number of CA incidents. Control charts reviewing trends for the CA incidents, harm to others, and restraint & seclusion use (incidents and duration) will be utilized to evaluate. A chart audit of CA occurances will be conducted to identify what the care planning intervention.
Finding/Results: Implementation of the predictive model is still in progress. An evaluation framework has been developed and the predictive model will be evaluated along the following domains: -Model Performance- How accurate is the model in predicting CA? -Outcomes-Are there any improvements in key outcome and balancing indicators: Incidents of CA, harm to others (incident data harm to staff or patients), and restraint & seclusion use All findings from the pilot will be presented at the 2018 eHealth conference.
Conclusion/Implications/Recommendations: The implications of creating a predictive tool for CA are vast. This approach will allow staff to identify dynamic risk factors that contribute to the risk for CA and subsequently facilitate proactive refinement of care plans to mitigate or address the risk and potentially reducing aggressive incidents and restraint and seclusion use. Moreover, from a risk management and operational perspective, implementation of a predictive alerting tool for CA has the potential to positively impact patient incidents as well as staff injury rates related to patient aggression, and the subsequent sick time and costs to the healthcare system associated with these events.
140 Character Summary: This predictive solution has the potential to reduce clinical aggression and restraint/seclusion use which would improve patient outcomes and staff safety.
OS30.03 - The Impact of Leveraging Analytics to Drive Adoption and Engagement
Purpose/Objectives: Three organizations are working together to support the implementation of a province-wide service to provide clinicians with faster and better access to specialists via eConsult. The complex, multi-faceted initiative involves technology implementation and evaluation, business process design and the establishment of a program governance framework. This oral presentation will showcase the important role that analytics has played in the successful adoption of this provincial digital health service, and in driving effective engagement with the initiatives stakeholders. Presenters will highlight key elements of the analytical framework (methodology, tools), and share lessons learned which can be leveraged for other digital health initiatives.
Methodology/Approach: Dashboards were developed to support project team members, field staff and key stakeholders through an evolutionary journey that focused on clinician recruitment, target setting, adoption and process optimization. Since launching as an Excel pivot table-based solution offering static single-view content, the dashboards have evolved to adopt a dynamic, multi-view format that aligns with the changing needs of stakeholders and evolving priorities of the provincial initiative.
Finding/Results: The dashboards provide key stakeholders and project staff with the ability to analyze, drill down and observe patterns in different geographic areas, to develop individualized change management and practice improvement plans for clinicians participating in the initiative. The dashboard provides timely access to stakeholder-specific research and reporting and on important topics including: Usage patterns: Analysis of past eConsult usage data, and the ability to sort users into various adoption categories. User interviews: Based on adoption categories, users are identified and interviewed to build user and practice profiles. Registration processes: Leveraging the usage pattern analysis, populations of inactive users can be identified. When engaged, these populations often cite registration delays as a cause for their low/non-existent usage patterns. Targeted recruitment: Monitoring the flow of eConsults helps partners to identify specialty-based priorities for service recruitment. The ability to quickly view analytics in the dashboards has allowed field teams to localize and target their change management and adoption strategies, increasing efficiency and outcomes. Timely access to region-specific data has provided regional partners with the information they need to become effective champions for the provincial initiative. In turn, regional partner resources have provided invaluable insight and engagement opportunities for the initiatives field staff to consider while recruitment and adoption activities continue. Aided by the dashboard-driven analytics, the provincial initiative has successfully recruited 10,500+ primary care providers, and 550+ specialists spanning 100+ specialties, resulting in more than 56,000 eConsults sent since the initiatives launch in January 2015.
Conclusion/Implications/Recommendations: The data provided by these dashboards has enabled eConsult service partners to tailor adoption approaches, and has helped foster a collaborative approach to deployment. Decisions are made based on usage data, and the opportunity to aggregate eConsult data with data on other digital health initiatives has provided a deeper understanding of primary care providers needs and usage patterns. Through analytics, this initiative has been able to monitor progress, dynamically course correct, increase user adoption, and measure success.
140 Character Summary: A provincial eConsult initiative is demonstrating how analytics can effectively support localized change, planning and adoption strategies.
OS30.04 - Predicting Personalized Surgical Outcome Through Analytics & Machine Learning
Purpose/Objectives: Up to one quarter of all patients who undergo knee arthroplasty surgery are not satisfied with the operation results and can continue to experience pain and suboptimal joint function. Despite being a common procedure, no algorithms which bring together predictors in a unified model to be used at the point of care exist. Prediction algorithms provide an opportunity to identify the patients who are at risk of poor surgical response. Having these analytical models available at the point of care to predict personalized risk before the procedure enables better case selection, patient preparation and improved clinical outcomes. The arthritis data science team at the University Health Network (UHN) has developed analytical models to predict post surgical pain, function and satisfaction. Each patient assessed is provided with a personalized risk report prior to having the procedure. Machine learning allows for the identification of predictors and enables the data models developed to continually evolve.
Methodology/Approach: The UHN arthritis data science team leveraged Microsofts Cortana Intelligence Suite to ingest, process and transform relevant data for the creation of predictive models and to visualize the data insights. The team leveraged 4 existing clinical databases with clinical and patient outcome data to derive predictive models for personalized risk, post surgical pain, function and satisfaction. Patient reports were created and provide a real-time personalized visualization of predictive outcomes, risk and contributing factors at the point of care.
Finding/Results: Over 9,500 records from four data sets were included in the data analysis to create the predictive models. Data included in the analysis was collected between January 2011 and August 2017. Microsofts Azure and Cortana Intelligence platforms were leveraged to identify the predictors of success or failure of a surgery. In addition, the probability of surgery success, likelihood of post surgical pain, function, patient satisfaction and 30-day hospital re-admission was also derived through the creation of multiple experiments and the development of models. All of the models developed were trained and scored to measure and compare accuracy. The most accurate models were visualized at the point of care by providing patients with a personalized probability of surgical success based on the derived model predictors.
Conclusion/Implications/Recommendations: Machine learning and predictive analytics provides an opportunity to leverage large data from multiple siloed datasets and focus on outcomes in healthcare. Data scientists require keen business and data understanding to develop and train the most precise predictive models. Data Preparation and quality is essential for accurate processing and transformation.
140 Character Summary: The arthritis data science team at UHN has developed analytical models to predict the probability of surgery success by leveraging machine learning technologies.
OS30.05 - Data Mining Twitter to Detect Prescribing Cascades: A New Concept
Purpose/Objectives: A prescribing cascade occurs when the adverse event from a drug therapy is misdiagnosed as a new medical condition, resulting in a subsequent drug therapy, medical devices, or diagnostic tests (Figure 1). Traditionally, prescribing cascades have been evaluated using administrative data to assess recorded drug therapies. However, these data do not capture information such as the use of over-the-counter drugs, devices or tests, thus limiting the scope of our understanding of the prescribing cascade. Over 11 million people in the United States have used social media to post information about health and treatment issues. Over 600 million active users are registered on Twitter. Recently, an approach to use Twitter to detect signals of potential drug-to-drug prescribing cascades was described by Hoang et al. We explore the feasibility of using Twitter to identify expanded prescribing cascades, using dementia as an example. Figure 1: The Expanded Prescribing Cascade (adapted from Rochon & Gurwitz, 2017.)
Methodology/Approach: A challenge with searching social media for clinical health information is the wide variety of synonyms, colloquial terms, and informal language used to describe conditions, medications and symptoms. Clinical data dictionaries, including the Consumer Health Vocabulary, were identified, which links lay speech about health to technical terms used by healthcare professionals. Using the Twitter Application Programming Interface, a preliminary search was run to identify the level of tweeting relating to dementia. The search terms used were Dementia, Alzheimer, and Lewy body. The collected tweets were then manually explored for general sentiments and user demographics.
Finding/Results: Feasibility testing within a five-hour window revealed 872 potentially relevant tweets, suggesting that Twitter users tweet about dementia every 20 seconds on average. Manual exploration of these tweets showed that the majority were posted by caregivers of people with dementia, or healthcare professionals. Of these tweets, 6 pertained to dementia drug therapies. We expect that other medical conditions that are more prevalent in the general population will have more relevant tweets.
Conclusion/Implications/Recommendations: Twitter is a tool for patients, caregivers or providers to post information relevant to prescribing cascades, and drug therapies in general. This platform is an unexplored resource for identifying potential prescribing cascades, which may allow opportunities to collect previously unavailable data on over-the-counter drugs, devices and tests. These data can also inform future population-level exploratory studies about the consequences of prescribing cascades.
140 Character Summary: We explored the feasibility of using social media meta data to identify prescribing cascades, using drug therapies for dementia as an example.
OS30.06 - From the Boardroom to the Bedside: Transforming Care Through Analytics
Purpose/Objectives: This presentation will highlight how the implementation of a collaborative approach to performance measurement and management improved outcomes at one hospital. Specific areas of interest include the implementation of electronic unit level scorecards with cascading Key Performance Indicators, automated workflows and the roll out of a consultancy model. The objective was to enable the achievement of strategic and operational goals through timely information and knowledge.
Methodology/Approach: Our business intelligence portal was upgraded and re-designed to provide a real-time data on how our key performance indicators (KPIs) are doing against a set of established targets. KPIs were cascaded down to program and unit levels from the corporate scorecard. To enhance transparency and accountability, all clinical and non-clinical leaders have access to the business intelligence (BI) portal and data is available at any time through the hospitals intranet page and on mobile devices. Electronic workflows were designed and implemented. These workflows facilitate tiered leadership review of the performance of all KPIs; enable the entering of comments and action plans and approval of these by the appropriate leaders. Workflows are enabled through e-mail notifications with links to the BI portal. A consultancy model of support was implemented to compliment the unit based scorecards. The consultancy model is an organizational structure where analysts from decision support, finance, human resources and quality work together to support the customer with consultation and advice to improve performance. The overall goal of the consultancy model is to ensure the leaders have access to information they need to provide quality patient care, plan the right programs and services to meet the needs of the patients and clients in an efficient and effective manner. The quality improvement team is linked in when it is identified that quality improvement initiatives are required to improve performance in a specific area.
Finding/Results: Since the implementation of the new system, a number of corporate indicators have shown marked improvement indicating the new model is having positive effect. Two specific examples: Weight Loss is one of the KPIs monitored through the unit based scorecards. A target of 9.8 % was set. Performance of this KPI showed 13.5% of patients having weight loss. Detailed analysis by the consultants demonstrated areas of focus in order to improve performance. Quality improvement initiatives were implemented which resulted in a significant decrease in the percentage of patients who experience weight loss. Most current performance at Q2 17/18 is 2%. Weighted patient days is another one of the KPIs monitored through the unit based scorecards. Targets for this KPI were not being met. Not meeting the target for this KPI can have a significant negative financial impact to the hospital. Deeper analysis and interprofessional collaboration including process and quality improvement initiatives have significantly improved performance in this indicator.
Conclusion/Implications/Recommendations: The roll out of electronic unit based scorecards coupled with a consultancy model significantly increased uptake and use of data and have driven up performancre across the board.
140 Character Summary: A collaborative approach to analytics enables the translation of strategic objectives from the boardroom to the bedside and improves outcomes.
PS06 - Digital Health Engagement Through Benefits and Data
- Type: Panel Session
- Track: Executive
- Presentations: 3
- Coordinates: 5/30/2018, 10:30 - 12:00, Fairview III Room, Conference Level
PS06.01 - Connected Health Information in Canada: A Benefits Evaluation Study
S. Hagens, Evaluation Services, Canada Health Infoway; Toronto/CA
A. Motulsky, Centre de Recherche du Centre Hospitalier de l'Université de Montréal, University of Montreal; Montreal/CA
J. Wilson, Gartner; Greely/CA
Purpose/Objectives: Effective, timely sharing of patient health information among care providers is central to quality of care. Important progress has been made in the implementation, adoption, and use of electronic systems that are connecting health care providers with the information they need to provide care. In Canada, jurisdictions have created connected health information through interoperable electronic health records (iEHRs) which provide access to drug, lab, diagnostic imaging, immunization information, as well as clinical reports from multiple care settings. The objective of this panel is to discuss how connected health information is transforming health care in Canada. More specifically panel members will discuss: 1) the current landscape of connected health information; 2) impact of connected health information against expected benefits; 3) jurisdictional experiences with use and optimization of systems; and 4) how to sustain and spread the value of connected health information to transform healthcare.
Methodology/Approach: In 2017, Canada Health Infoway undertook a study to understand the adoption, use and impact of connected health information across Canada. This study, based on national clinician studies, project research and evaluation, valued the effects of the foundational iEHR infrastructure as accessed through different point of care solutions in different care settings. Panelists will address a number of key activities that were undertaken as part of the study: measurement of the adoption of the iEHR; assessment of the benefits accruing from connected health information through review of available evidence; and key informant interviews with provincial/territorial stakeholders to understand the current and future use of connected health information nationwide.
Finding/Results: The pan-Canadian Study on Connected Health Information calculated benefits accruing to health care system stakeholders (patients, providers, health system). Benefits substantiated by evidence include reducing duplication of lab and diagnostic imaging tests, enhancing timeliness of care, more effective ambulatory care and emergency department interactions, optimizing scope of practice for clinicians and improving equity in care through the availability of health information. Financially quantifiable benefits were driven by improvements in clinician and clinical practice productivity; avoided health system utilization due to improved patient safety; reduced patient time and expense, and reduced duplication of diagnostic tests. Equally as important but difficult to quantify are benefits related to improved access to information for clinicians. A majority of clinicians now have access to connected patient information either through integrated point-of-care systems or through separate web-based viewers. In some jurisdictions integrated iEHR viewers are available where clinicians are accessing connected patient information through a single solution such as their main clinical record system. Over 300,000 health care providers across Canada are currently accessing the iEHR through one of these methods, compared to 170,000 two years ago.
Conclusion/Implications/Recommendations: Panelists will provide clinical and health system leadership perspectives on what has contributed to the impact seen to date and how health leaders can move forward to sustain, spread and achieve further value through the use of iEHRs. Discussion will focus on priority focus areas such as interoperability, advanced functionalities that enable e-referral, e-consults, and the use of the foundational iEHR infrastructure to support patient portals and analytics.
140 Character Summary: Connected health information is transforming care in Canada. More than 300,000 health care professionals and their patients are realizing widespread benefits.
PS06.02 - Txting for #youthmentalhealth: Provider and #ptexp
Purpose/Objectives: It is estimated that 1.2 million Canadian youth suffer from mental health issues, yet only 20% receive treatment. Yet delivering services to young people can be difficult. There are many reasons young people in need of mental health services may not wish to, or be able to, access traditionally-delivered mental health services. These include the desire for privacy, lack of local services or personal transportation to them, lack of acceptability of using cell phones for talking, limited data plans or bandwidth, etc. Digital health can play a key role in providing accessible and appropriate services to this population. The objective of this session is to provide an overview of the current state of digital mental health in Canada, present the experience of one provider offering a new 24/7 crisis texting line to youth in Canada, and amplify the patient point of view by presenting the patient experience in their own voice. These presentations will inform discussion during the question period around best practices for digital mental health solutions, engaging youth, and adopting a patient-centred approach.
Methodology/Approach: Youth are increasingly less likely to use telephone or online chat when seeking help. Yet cellphones are nearly ubiquitous, and the use of texting is more acceptable to young people. For this reason, a bilingual, 24-hour, anonymous crisis text service, launched in 2017, provides an opportunity to address this accessibility gap. The crisis line provides service to youth without significant wait times. Using texting eliminates the need to download an app or use a cellphone data plan. The service is available in remote areas and areas that are typically underserved. A texting service allows an easier entry point for a youth in crisis, allowing them to be connected to on-the-ground resources including emergency services and child welfare after engaging with the texting service. The technology platform can be continually improved and lends itself to rigorous monitoring and evaluation. The solution can be integrated into other platforms, such as social media (Facebook, Snapchat, etc.).
Finding/Results: Data from an initial release across one province will be presented. The data will inform local and national actors on trends and patterns, as well as add to the knowledge base of how young people experience mental health issues, and which words or concepts may be leading indicator for crisis. The data will also show usage patterns and areas of demand, which will provide lessons for scale-up to the national level. The effectiveness of outreach and engagement activities associated with the limited production release will also be assessed. Data from this implementation in Canada will be compared to the experience of other countries.
Conclusion/Implications/Recommendations: The use of texting is an innovative way to provide services for youth in crisis. Its success has implications for how health services are delivered, whether to youth or others. It is a tool with the potential to improve patient experience, the health of the youth population, and contribute to health care system efficiencies.
140 Character Summary: Digital mental health can be more accessible, acceptable and effective for youth. 24/7 texting crisis service is an opportunity for kids to get help the way they want.
PS06.03 - Command Centres: Shining the Light Between the Seams
Purpose/Objectives: NASA, municipalities, airports and the military all operate command centres, in which people, analytics and actionable insight come together in one physical space with the goal to routinize very high levels of performance. Large scale command centres are now emerging in healthcare at a time when the system is precariously congested and resource-constrained. These centres of gravity play mission critical roles in the orchestration of patient care activities by shining a light into the dark corners of operations to de-risk, prioritize and synchronize patient care delivery.
Methodology/Approach: The panelists will describe the healthcare command centre journeys they have been on, including the problems they are tackling, their implementation approach, results achieved and lessons learned. One panelist will describe how a hospital turned vision into reality when it leveraged its vast digital infrastructure and opened Canadas first hospital-wide command centre in 2017 as part of its quest to become a high reliability organization. Today, the command centre is ingesting data elements in real-time from multiple source systems across the hospital to trigger actions in the moment that reduce length of stay, mitigate delays in care and expedite patient flow. Another panelist will discuss its organizations goals and approach for implementing a regional command centre that optimizes how capacity is utilized across multiple locations in its network. A third panelist, that is partnering with hospitals and health systems around the world to design, build and activate healthcare command centres with deep capability, will provide a compelling look at the scalability of command centres across different problem spaces from patient flow, length of stay and access to quality, safety and patient experience.
Finding/Results: The vast amount of data that is being generated at every turn in a healthcare organization is overwhelming for most caregivers. Command centres are demonstrating this data can be put to good use and results in significant benefit when it is carefully mixed, filtered, processed and presented to the end user. That benefit comes in the form of reducing patient waits, de-stressing caregivers on the front lines and optimizing the use of limited resources such as staff, beds, equipment, etc. The panelists will share numerous real-world instances of applying real-time and predictive analytics strategically to solve specific challenges and drive outcomes that were otherwise not attainable.
Conclusion/Implications/Recommendations: Hospital and health system command centres are challenging the status quo such that functions across the enterprise no longer need to work in silos, decision-makers no longer need to act without good information and caregivers can spend more time delivering care instead of coordinating care. All of this translates into better health service delivery and better quality of care which all patients deserve.
140 Character Summary: Like NASA mission control, large-scale hospital command centres enable real time situational awareness and prompt action in the moment for direct patient benefit.
PL04 - Lunch Closing Keynote Address
- Type: Plenary Session
- Presentations: 1
- Coordinates: 5/30/2018, 12:00 - 14:00, Kitsilano Ballroom, Conference Level
PL04 - Workplace Wellness For A Better Bottom Line
Abstract not provided
Login or Purchase: Please login to view this content if you have already purchased this Virtual Meeting. To purchase, please select the corresponding button above. Click here to login.
Conference Attendees & Access Code holders: Click here to enter your Access Code. Already entered your Access Code? Please login.