Purpose/Objectives: CCO collects and manages one of the most comprehensive healthcare data sets in Canada. The data is used to support research and inform decision-making in Ontario's health system. Data and analytics are integral to meeting the growing demand for greater health system accountability, better health outcomes and improved patient experience. One of CCO’s strategic priorities is to support evidence and knowledge generation. To support the increased demand for data to support research, CCO’s Data Disclosure Team worked to increase the capacity to support researchers both internal and external to CCO, while still meeting CCO’s privacy and legal responsibilities.

Methodology/Approach: Supporting researchers in their data needs is a resource-intensive process. Support for a typical research data request is required from coordinators, research associates, privacy specialists, as well as senior management across the organization. In order to continue supporting evidence and knowledge generation and ensure the delivery of quality service for data requests, the Data Disclosure Team focused on improvements in four areas: Engagement with stakeholders; Governance; Streamlined processes; and Pilot of a cost recovery model.

Finding/Results: The team began by focusing on quick wins by improving stakeholder communications at first contact, improving online information, and clarifying roles and responsibilities in the process. Subsequently, a multi-phased improvement plan was implemented, including: Improvement of communication tools (online; email; forms) to ensure researchers were equipped with the information necessary to request data, as well as to ensure the best quality data was disclosed; Launch of the Data Disclosure Subcommittee to oversee the decision-making and issues in a complex process; Process improvements, beginning with detailed process mapping to identify streamlining opportunities, monitored through weekly and monthly performance reports; and Launch of the pilot year of the cost recovery model to support the increasing resource requirements of data disclosure at CCO.

Conclusion/Implications/Recommendations: In the past year, CCO’s Data Disclosure Team has worked with researchers across the province to help address their research questions. The most frequent research topics include: Cancer Prevention; Cancer Diagnosis; Cancer Treatments; and Access to Care. A quarterly report to CCO’s executive team has demonstrated sustained improvements over 6 quarters. This translates to an improved experience for researchers through: Better understanding of the steps and requirements to access data for research purposes; Improved quality of data disclosed; and More timely access to data.

140 Character Summary: To meet the increased demand for data for research purposes, CCO’s Data Disclosure Team worked to streamline and improve processes to support researchers across ON.

Purpose/Objectives: Ontario is a very diverse province; while Northern Ontario contains 88% of the province’s landmass, 94% of the province’s population resides in Southern Ontario. This creates very different contexts for health care service provision. The Ontario Telemedicine Network (OTN) uses technology to connect patients and health care providers across the province. Although there has been research on women’s health, aging and health, and the implications of rurality on access to health care services, there is little research to date showing the connections between these factors and telemedicine utilization in Ontario.

Methodology/Approach: We obtained OTN medical service utilization data collected through the Ontario Health Insurance Plan (OHIP) and provided by the Ministry of Health and Long Term Care (MOHLTC). Initial investigations used data that was aggregated by census subdivision, billing code, and month and year of visit. The data did not contain information on age and sex of patients, but we were able to identify visits with gynecological, obstetrics and geriatric billing codes to infer age and gender for preliminary analyses.

Finding/Results: A total of 22% of clinical telemedicine visits occur with patients from rural Northern Ontario; an area with only 2% of Ontario’s population. For telemedicine patient visits using obstetrics or gynecology codes, this figure rises to 72%. In comparison, 14% of visits with obstetrics and gynecology codes occur with patients residing in the urban north, 10% in the urban south and 3.5% in the rural south. Similarly, at 42%, a disproportionate number of visits using geriatric codes take place with patients from rural Northern Ontario, compared with 30% from urban Southern Ontario, 16% rural Southern Ontario and 12% urban Northern Ontario.

Conclusion/Implication/Recommendations: Northern Ontario’s sparse population across a large geographic area creates unique barriers to health care services, and much of the north is medically underserved. Our data suggest that telemedicine is being used differently by men and women and by older adults in rural Northern Ontario than in the rest of the province. The findings demonstrate the potential to increase access to medical services in medically underserved areas of rural Northern Ontario for women and older adults and reduce the need for medically related travel.

140 Character Summary: This research examines selected age- and sex-specific telemedicine use in Northern and Southern Ontario by rurality.

Purpose/Objectives: A community-based hospital participated in a pilot project to demonstrate the benefits of using a standardized terminology language (International Classification for Nursing Practice (ICNP)) embedded within their electronic medical record (EMR) system to increase data quality and facilitate data extraction for e-clinical analytics to support continuous quality improvement. This presentation reviews the methodology used to build organizational capacity to facilitate the use of advanced clinical e-functions such as ICNP and e-clinical analytics to standardize data collection and employ technology-enabled data extraction for outcome evaluation.

Methodology/Approach: The hospital was required to implement ICNP-encoded nursing order sets to optimize wound care management. External consultants were contracted for the project to build organizational capacity by providing high level training, mentoring and support to: • The Project Lead who was tasked with leading the integration of ICNP-encoded nursing order sets within their MEDITECH information system to standardize and automate data collection for outcome evaluation; • The Information Technology (IT) staff to facilitate data extraction for e-clinical analytics using ICNP codes. The implementation strategy was divided into four phases: Phase I – Project Kick Off & Information Gathering Session - The project sponsor provided a project overview to all key stakeholders, validated the background information obtained and gathered more detailed information from the key informants to share with the external consultants. Phase II – Training & Technical Build Support - The consultants provided training on order set integration within the EMR. This training was divided into modules that coincided with the various stages of the technical build. This approach provided opportunities for direct application and ongoing support throughout the build facilitated by remote system access and WebEx. Phase III - Data Extraction - The consultants worked with the organization’s IT staff to extract the required data using the ICNP codes to support e-clinical analytics. Working with the site, testing and user acceptance were achieved.Phase IV – Support Migration to Production, Go Live & Post-Go Live - During this phase the consultants provided go-live and remote support to ensure adoption, data extraction and reporting functionality during the first 1 month post Go Live.

Finding/Results: The hospital successfully integrated the ICNP-encoded nursing order sets within their EMR. Data were collected seamlessly as the clinicians used the order sets to facilitate evidence-based wound care. The IT staff were successful in their efforts to extract the data required to evaluate key outcome indicators using e-clinical analytics. This methodology was welcomed by the staff who previously relied on a time-consuming manual data collection process.

Conclusion/Implication/Recommendations: This project demonstrates the benefits of technology-enabled outcome evaluation facilitated through the use of a standardized terminology language (i.e. ICNP). It greatly reduces the time required for data collection to support outcome evaluation for quality improvement. It also offers hope for others who rely on clinical data for research, benchmarking, trending and policy decisions, to name a few.

140 Character Summary: This presentation reviews an approach to build capacity to use ICNP to automate data collection and extraction for outcome evaluation.

Purpose/Objectives: Audits are important to monitor the ongoing quality and efficiency of transfusion service processes. Typically, audits combine manual observation and laboratory information system generated data to derive conclusions of how a process is performing, and suggest recommendations on its improvement. The centralization of transfusion services introduces more challenges in auditing, especially if transfusion services are distributed over significant geography and if there are limited human resources available. We describe a novel use of process mining, recently adapted for use in financial auditing, to further characterize an automated red cell unit dispensing fridge which was being suboptimally used at our institution.

Methodology/Approach: Red blood cell unit transaction data and attributes from October 1, 2013 to September 30, 2015 from the laboratory information system of a large distributed multisite transfusion service in Eastern Canada (Nova Scotia Health Authority, Central Zone) were queried, preprocessed, and process maps were created. Subject matter experts were asked to create a process map of their perception of the blood fridge process. Subsequently, blood fridge maps were displayed for real-time exploration, and comments were collected and thematically analysed.

Finding/Results: There were 2302 unique red cell units which encountered the VG-Hemosafe inventory destination. There were 1597 different variants, each representing a different unique sequence of pathway activities that a red cell unit could take from receipt to final disposition. The manually created process map at simple unidirectional flows, while the process mining derived map demonstrated complex looping, though the inventory receipt process was common between both types of maps. The process maps were relevant and valid, and comments were grounded on insight and process confirmation. The exercise also resulted in policy changes and gave us insight and impetus to re-initiate discussion with the manufacturer.

Conclusion/Implication/Recommendations: We demonstrate a novel use of process mining to enhance our transfusion services understanding of how our institution was using an automated remote blood dispensing fridge without the need for any manual auditing. We demonstrate that process mining allows for significant intelligent discussion and reflection on this process, and this technique could be applied to other processes..

140 Character Summary: Process mining can be applied to further understand the functioning of transfusion services and enhance inventory management.

Purpose/Objectives: The Canadian Institute for Health Information links data from health care organizations and health ministries across Canada to create new knowledge. The new knowledge supports system and policy-decision makers in answering questions about the health or healthcare of population groups or healthcare sectors. This presentation provides an overview of recent analytical products that incorporate a variety of approaches to linking data.

Methodology/Approach: Key success factors in data linkages are the use of standards and robust linkage methodologies. Standardized data can be linked at the individual level across hospital, long-term care, home care and community settings. Expanding knowledge about a specific sector can be achieved by linking physician billing, drug, and financial data. Linking clinical and financial data at the person and/or organization level provides insight into the cost of providing services, laying the foundation for examining efficiency and value for money. Data may also be used in non-traditional ways, for example, using payment data linked to clinical data to better understand how services are provided in various settings and to identify best practices.

Finding/Results: Results from these studies demonstrate how linked data creates new knowledge: 1) Hospital, emergency, and drug data were linked to show the number of emergency visits, hospitalizations, and the use of psychotropic medications by children and youth with mental health disorders are increasing, especially for those with mood and anxiety disorders living in urban areas. 2) Physician billing data was linked with hospital and emergency data to show that increased continuity of care with a family physician is associated with reductions in hospitalizations for people with chronic diseases and reductions in ED visits for people with conditions better managed in primary health care settings. 3) Drug data was linked with clinical assessment (InterRai) data to show that a large proportion of seniors living in long-term care exhibiting severe aggression were not being treated with antipsychotics, suggesting non-drug alternatives were often considered. 4) Financial data was linked with hospital clinical data to calculate an efficiency indicator that can be used at a system level to examine variation, and by individual managers to identify areas for further investigation and manage trends in spending.

Conclusion/Implications/Recommendations: This poster demonstrates how new knowledge created using linked data can inform decision and policy makers, and lead to action within the health system. Several examples are presented to illustrate the approach and the types of data being linked.

140 Character Summary: See how CIHI's linked data brings results and contributes to improvement of the health system, using standard data and linkage methodologies

Purpose/Objectives: One in five Canadians suffer from chronic pain, and given the long wait time for specialty pain programs, many are cared for by their family doctor. However, family doctors only receive a limited amount of pain training during medical school. Project ECHO (Extension for Community Healthcare Outcomes) is a telemedicine initiative that teaches primary health care providers (HCPs) how better to manage their patients’ chronic medical conditions using a hub (experts) and spoke (learners) model involving case-based discussions and didactics. Project ECHO Ontario Chronic Pain teaches HCPs best practices in the management of chronic pain, and responsible opioid stewardship. Research Questions: A) Do ECHO’s HCPs increase their skills and competence levels (in pain management and opioid stewardship) after attending Project ECHO? B) How is the knowledge diffusing within or outside the online community created by ECHO and is it leading to improved care for patients whose HCP attended Project ECHO? C) In what ways, if any,have HCPs gained insights into their own motivations, confidence levels and processes for managing their chronic pain patients after attending Project ECHO?

Methodology/Approach: This study is based on in-depth interviews conducted with HCPs who are either participating in Project ECHO or have completed their training with ECHO. Interviews will be conducted with 12-15 HCPs to evaluate their experience with ECHO, the impact of ECHO on their own clinical practice, how ECHO spreads to other HCPs, and how technology facilitates medical education. Maximum Variation sampling is being used to select the participants to ensure a representative provider population. The HCPs will include a mix of: a) those from rural and urban Ontario, b) presenters or non-presenters of ECHO cases, and c) a variety of HCP disciplines (e.g., physicians, nurse practitioners, physical therapists, occupational therapists, and social workers). Data analysis will use thematic analysis, which includes inductive and deductive coding strategies. The Diffusion of Innovations framework will guide inductive coding, while deductive coding will create codes and themes that fall outside of the framework. The codes will be generated on an ongoing basis with regular feedback from members of the research team. Multiple coders will take part in checking and interpreting the data.

Finding/Results: We have interviewed three HCPs to date. Preliminary results suggest that participation in Project ECHO increases HCPs’ a) confidence in managing patients, b) knowledge of specific pain resources, and c) skills for managing chronic pain in the community using a multi-modal approach. In addition, video-conferencing has been a successful continuing education method for HCPs to share their knowledge in a no-shame community and this online learning platform has increased the interactions of HCPs.

Conclusion/Implication/Recommendations: Preliminary results suggest that ECHO is increasing HCP knowledge, skill and confidence in caring for chronic pain patients in the community, and that this teleconferencing modality is a viable approach to educating primary care HCPs in managing chronic pain patients. This research can be used to guide further evaluations of ECHO programs and help researchers understand how knowledge can be shared innovatively through the use of online communities.

140 Character Summary: Understanding how Project ECHO is an innovative method to learn and share knowledge about chronic pain management with healthcare providers in online communities.

Purpose/Objectives: Increasingly, traditional models of in-office, appointment based care are being questioned as optimal to provide care for homebound and complex patients in our community. The pillars of a patient's medical home include timely access to care, coordination, team based approach to care and continuity of care for a population of patients. Achieving these goals can be difficult for patients who are homebound; use of various digital health technologies, interoperability of clinical systems and remote communication can optimize care delivery and experience of care for this population of patients and their caregivers. Further, use of data from exisiting electronic systems allows for measurement of patient outcomes and use of data for population identification. This presentation will describe a model of care that addresses the needs of this population, the role of technology in improving care and experience for these patients and where the gaps remain for seamless integration of electronic systems for home based primary care. Objectives: (1) Describe a model of care for complex and homebound patients, in keeping with the goals of the patient's medical home. (2) Identify opportunities for use of exisiting electronic systems for measurement of care delivery and improvements in clinical outcomes (3) Describe use of exisiting electronic systems for communication, care coordination and interprofessional team based care for homebound and complex patients. (4) Spark innovative thinking and problem solving to address existing barriers to communication and outcome measurement.

Methodology/Approach: Review of exisiting care model will be described and how various exisiting clinical systems (i.e. clinic electronic medical record, home care MDS and hospital EDIS) are used for data collection, patient identification and measurement of outcomes. Further, use of these systems for communication and information sharing amongst team members will be detailed. The role of the provincial electronic health record in care delivery for homebound and complex patients will be explored with examples given. Gaps within exisiting systems and opportunities for improvement will be described.

Finding/Results: Home based primary care, provided by an interprofessional team with intensive case management, to a population of high risk patients or patients who are high system users, has demonstrated a reduction in emergency room use and hospital bed days for this population.

Conclusion/Implication/Recommendations: Recognizing those in our community who are at high risk and in need of intensive primary care, requires us to look outside the walls of the clinic and change our approach to care delivery. Measurement of these innovative models of care is imperative to demonstrate success but also for appropriate identification of populations in need. Existing electronic systems can provide opportunites for communication, information sharing and data collection to improve care for this population. Ongoing barriers exisit to further optimization of care delivery for this population of patients and their care givers.

140 Character Summary: Care for homebound patients requires optimal and innovative use of exisiting electronic sytems to improve their health and healthcare experience.

Purpose/Objectives: To provide "virtual consultations" in the community facilitated by a "community health worker". This will enable clients and providers to make timely and accessible health care connections, conducive to producing better health outcomes.

Methodology/Approach: Virtual care technology and processes will be implemented in a community care setting. Community Health Workers (CHW) will be connecting to primary care and/or specialized care providers from the patient's home. This is in support of Primary Care Home initiative to increase continuity, and improve care and coordination between clients and clinicians. Initially the technology will be tested to ensure technical, security, and privacy needs are met. Because this is a community setting, broadband connectivity is not a given, thus hotspots will be used. Virtual models supporting primary care home is anticipated to have the following benefits: increase access to primary care services for remote communities, enhance continuity and coordinated care particularly for outreach services, patient-centered care, support patients with chronic diseases, patients requiring mental health and substance use services; access to specialists; and acute care services in remote service areas, enable the primary care home to provide continuity of care to community’s patients and, achieve the triple aim (e.g., improve provider and patient experience – decrease provider isolation; improve population health; and improve sustainable cost).

Finding/Results: Expected results will include: New virtual model to support “Primary Care Homes” linkages to patient/families/communities and family physicians, specialist and community health services. Strengthen community health services into local virtual teams. Change Island Health processes and staff models to provide people with better access to these services, especially when it is remote urgent (Rapid Response, 7 days a week). Educate staff to support patients in setting their own health goals in a proactive care plan. Support people to live at home as long as possible by promoting a ‘Home is Best’ approach. Develop methods for finding and monitoring people at risk of deteriorating health. Build and expand upon partnerships with community health providers, such as First Nations Health Authority, Divisions of Family Practice, local government and not for profit agencies.

Conclusion/Implication/Recommendations: Virtual care is one of the identified elements of the Primary Care home and provides a seamless approach for Community Health and Care to link frail and elderly and complex care clients to their physicians and clinicians directly from home, contributing to rapid response and early intervention goals. Virtual care through home support is a strong fit that will enable a shift in the point of care directly to the home, furthering Island Health's goal of 'right care, right time, right place'.

140 Character Summary: Harnessing existing home support capacity, virtual care accessed in the home represents a proactive paradigm shift in healthcare.

Purpose/Objectives: To explore adult children caregivers' (ACCs) experiences with online and in-person peer support exchange while caring for an elderly parent. Specifically, we aimed to answer two questions: 1) How do ACCs use online and in-person modalities to obtain support? 2) What type of support is exchanged within each modality?

Methodology/Approach: *Design: We employed a qualitative descriptive approach. Qualitative description entails a concise and descriptively rich analysis that remains true to the data. Recruitment: Brochures distributed through several in-person support groups and the social media channels of national organizations. The first author also used her Twitter account to tweet the study link. Procedure: Participants accessed the online survey where they were first asked about their eligibility. Consent was then sought. At the end of the survey, participants could volunteer for an in-depth qualitative interview. Data Collection: Each caregiver participated in an in-depth semi-structured interview that was conducted over the phone. Each interview was transcribed verbatim by a professional transcriptionist. The authors then checked for accuracy by cross-referencing the transcripts with the original audio files. Data Analysis:* Thematic analysis was performed to identify themes from the caregivers’ narratives. Line-by-line coding informed the development of a coding framework which was applied to all transcripts. NVivo version 10 qualitative data analysis software was used. Analyses included comparing and contrasting the coded data and categorizing similar ideas. All authors participated in the final phase of the thematic analysis which entailed constant comparison until categories could be grouped into ‘themes’ that were distinct from one another.

Finding/Results: In total, 15 adult children caregivers (ACCs) participated in an interview. The average age of ACCs was 51 years old (Range: 41-65 years old). The majority of ACCs (80%) indicated that their peer was a family member, long-time friend or co-worker– suggesting that this population mobilizes their existing network for peer support. Theme 1: ACCs take a pragmatic approach to peer support exchange. This was characterized by ACCs' blended use of communication modalities and mention of telpehone calls and tetxing as supplementary modes of communication. ACCs interacted online with peers to meet their practical needs (e.g. efficient and fast communication). Conversely, they interacted in-person to meet relational needs (e.g. desire for high quality relationships). Theme 2: The nature of peer support that ACCs received transcended the interaction modality. Regardless of whether the ACCs interacted with peers online or in person, they consistently received emotional, informational and appraisal suppport across modalities.

Conclusion/Implication/Recommendations: Dichotomizing support as either ‘online’ or ‘in-person’ may detract from our ability to understand how ACCs use multiple modalities to achieve their support goals. ACCs’ approach to peer support was complex. This highlights the need for future interventions to emulate their naturally pragmatic and flexible support-seeking style.

140 Character Summary: Adult children #caregivers use a blend of communication modalities to obtain #support from #peers. Type of support received transcends online/in-person modality

Purpose/Objectives: We are still in the early days of a clinical data revolution. While medical devices generate an enormous amount of clinical data, most of this data goes unused, siloed in proprietary device formats and systems that are nearly impossible to access. However, this data is essential to realizing the promise of improved healthcare by the emerging advances in predictive analytics and retroactive data analysis systems. Exactly what does it take to get this data and then what can you do with it? What does it take to get a data acquisition project rolling and how can you empower your clinicians, researchers, and innovators to make breakthrough discoveries?

Methodology/Approach: Using a variety of systems and methods, we have developed techniques and methods for collecting large amounts of data for customers such as Sick Kids in Toronto. In addition, we have 12 years of experience connecting medical devices for companies such as Hospira, Baxter, and Smiths Medical. This data is sometimes placed in the medical record, but is now being placed into systems from IBM, Google, Hitachi, GE, and others to perform real time predictive analytics.

Finding/Results: Having implemented/integrated and developed software for device integration at hospitals across the US and Canada, we have taken years of expertise and put it into our own products and solutions for data acquisition and storage. This has resulted in some of the most reliable, high quality data that researchers have been able to use from devices that generate huge amounts of data such as patient monitoring systems. Medical record systems are not device data warehouses. Data is generated at a high rate from some devices, both numeric and waveform data, that cannot be stored at the resolution necessary for research, or simply not at all.

Conclusion/Implications/Recommendations: Although standards exist and are being further developed, there are still a large number of devices that need connectivity that will be around for years and don't utilize standards. We have helped customers sort through these systems to provide data that can be used for more than just the medical record, it is stored and can be used for research for as long as is needed. When most people think of "big data" in healthcare, they think of the analysis of the data and systems that perform it (such as IBM Watson). What they miss is that the most difficult part is still the link from the device to those systems.

140 Character Summary: We will bring our expertise of the task of connecting medical devices and device data for hospitals, researchers, and technologists alike.

Purpose/Objectives: The medical device industry is an innovative health care marketplace that is regulated less stringently than drugs under federal legislation in Canada and internationally. Currently there is a large information gap in terms of which devices provide the best outcomes with good value in terms of device pricing as well as the need to perform longer-term surveillance of medical device safety. The Canadian Institute for Health Information (CIHI) is expanding the capability of its primary hospitalization database, the Discharge Abstract Database (DAD), to capture medical device product information, beginning with hip and knee replacements, in order to support these patient safety and procurement agendas.

Methodology/Approach: CIHI has managed the Canadian Joint Replacement Registry (CJRR), Canada’s only national medical device registry, since 2001. This Registry captures hip and knee replacement product (barcode) information to support the monitoring of device performance and related outcomes. CIHI is working to integrate the capture of the CJRR information via the DAD, which will directly associate the product information with the associated medical-device related hospitalization. Activities undertaken as part of this initiative included an assessment of the CJRR data elements vis-à-vis the DAD, comparisons with international arthroplasty registries, and consultations with clinical and classifications experts, as part of a multi-disciplinary team that also included IT business analyst, architecture, and developer staff.

Finding/Results: As of April 2018, the DAD will include additional data elements to capture detailed manufacturer and product number information associated with each hip and knee replacement device. The DAD will also have the enhanced ability to received scanned barcode information and parse out requisite information, a key functionality given that the current lack of standard barcode formats in the industry. These additions lay the groundwork for future medical device product capture. Several provinces in Canada have mandated CJRR collection and it is expected that this integration will increase the reporting to support device surveillance and procurement decisions. For instance, product characteristics, such as bearing surface, are known to influence the revision rates of joint replacement surgeries. Early revisions need to be reduced, as such surgeries are significant procedures for the patient, take longer for recovery, and are costly to the health care system. These medical device products have varying costs negotiated through private procurement contracts; under the current pressures to reduce health care costs, it is important that their relative outcomes be considered during purchasing arrangements.

Conclusion/Implications/Recommendations: An estimated half a million medical-device related hospitalizations and procedures occur in Canada annually, incurring significant costs in the health care system. By building in the capability for medical product device capture directly linked with the patient hospitalization record, CIHI is enhancing its ability to provide outcomes data in a manner that supports the work of regulatory and supply chain/procurement agencies.

140 Character Summary: CIHI will be capturing medical device product information in its primary hospitalization database to support improvements in outcomes data.