Healthy Heart Kit

Cardiovascular Disease Online

Centre for Chronic Disease Prevention and Control
Cardiovascular Disease

Coordinated Myocardial Infarction Registry Workshop

February 25, 1996
Westin Harbour Castle Hotel, Toronto, Ontario
Summary of Workshop Proceedings

Participants

Agenda

Executive Summary

Summary of Workshop Proceedings

Session 1
Session 2
Session 3
Session 4

Participants

Present:

Regrets:

Agenda

Session 1

Chair: Dr. Taylor

Co-Chair: Dr. Johansen

Overview- Dr. Davies
HSFC Perspective- Dr. Armstrong
LCDC Perspective- Dr. Taylor
MACSTRAC- Dr. Daly
Nova Scotia Acute MI Outcomes Monitoring Project- Dr. Cox

Session 2

Chair: Dr. Naylor

Outline of Draft Proposal- Dr. Davies
Roundtable Discussion: Consensus Issues- All
Break-Out Groups- All

Session 3

Chair: Dr. Armstrong

Co-Chair: Dr. Taylor

MI Registries:Why your investment is money in the bank- Mr. Tholl
Neurological Registry- Dr. Penn
Roundtable Discussion: Synthesis of Break-Out Groups- All

Session 4

Chair: Dr. Dagenais

Co-Chair: Dr. Wasi

Action Plan- Dr. Davies
Data Management Issues- Mr. Burnett
Closing Remarks- Dr. Davies

Executive Summary

Purpose. Health Canada, the Ottawa Heart Institute, the Heart and Stroke Foundation of Canada (HSFC) and Statistics Canada brought together key cardio-vascular disease professionals from across Canada to achieve consensus on mechanisms to create a national coordinated myocardial infarction (MI) registry. Workshop goals were to determine if partnerships with clinical academic centres, the health care industry, government agencies, and private agencies are possible and, if so, on what terms.

Advantages of an MI Registry. The advantages of an MI registry include the provision of relatively inexpensive, population-based data that are missing from current registries. These data would have sufficient clinical detail to allow the analysis of such things as risk factors compared to outcomes and the cost-effectiveness of different treatment strategies.

Other Registry Projects. The following registries currently in operation or planned were described for participants:

MACSTRAC/FASTRAK project, Toronto/Hamilton area. MACSTRAC is primarily a quality improvement project with registry possibilities. It is currently operating in 45 cardiac care units (CCU's). A new program FASTRAK II will expand the MACSTRAC database to include the emergency and ward phase of treatment in addition to the CCU phase.
Canadian Registry of Myocardial Infarction and Stroke (CROMIS), Nova Scotia. CROMIS is intended to provide precise numbers on MI and stroke cases in hospitals. It is a registry nested in an administrative-type database.
Nova Scotia Acute MI Outcomes Monitoring project, Nova Scotia. This project was launched in 1994 in collaboration with all but one hospital in Nova Scotia and NHRDP seed funding for the collection of clinical information not provided by routine hospital discharge data. MI was selected for initial study. Future plans include adding pre- and post-hospital phases to the database.
STROKENET, planned cross-Canada. This neurological registry is intended to provide on-line data collection and information access to medical decision-makers via internet links in hospitals across Canada. Dr. Penn (University of Alberta) suggested that an MI registry could be implemented in tandem with STROKENET without incurring extra infrastructure costs.

Draft Proposal. Dr. Davies (Ottawa Heart Institute) proposed that a representative subset of hospital centres be selected for in-depth data collection according to the following modular plan: (a) standard in-hospital data collection using FASTRAK or equivalent program (Module 1), (b) additional in-hospital data collection (Module 2), (c) follow-up data collection three months after discharge (Module 3), (d) follow-up data collection one year after discharge through linkages with other databases (Module 4). Partners in the MI registry would include clinical and academic centres, government agencies, private agencies, and the health care industry. Contributions from partners would differ by group.

Action Plan. After obtaining consensus from participants on the possibility of a partnership, Dr. Davies proposed that three committees or working groups be formed to arrive at solutions to (a) partnership and funding issues, (b) data issues, and (c) technical issues. Dr. Naylor (Institute for Clinical Evaluative Sciences) recommended partnerships with FASTRAK and their sponsor, Hoffman-LaRoche pharmaceutical company.

Data Management. Mr. Burnett (President, Enhance Management Perspective Inc.) reviewed various solutions for handling the new registry. He concluded that a digital system called a "Personal Digital Assistant" was the best option for data management compared to paper, e-mail and fax alternatives.

Session 1

Chair: Dr. Taylor, Health Canada
Co-Chair: Dr. Johansen, Ottawa Heart Institute

Introduction and Concepts- Dr. Greg Taylor, Chair

Dr. Taylor opened the proceedings by stating that the purpose of the workshop was to bring together key cardio-vascular disease professionals from across Canada to achieve consensus on the creation of a national coordinated myocardial infarction (MI) registry.

Overview- Dr. Rick Davies, Ottawa Heart Institute

Dr. Davies highlighted the various reasons for developing an MI registry. The most important reason for the registry is to address the so-called crisis in the health care system by doing a better job of measuring the benefits of the system against expenses. A coordinated MI registry will supply those data on cardiovascular disease (CVD) populations that are missing from current registries. Registry data have the advantage of being relatively inexpensive compared with data from randomized controlled trials, and of providing representative, population-based data with sufficient clinical detail. Registry data also have the potential to generate new knowledge by identifying hypotheses for future randomized control trials and allowing the analysis of such things as risks factors compared to outcomes and the cost effectiveness of different treatment strategies.

Dr. Davies emphasized that building an MI registry will require a coordinated approach through partnerships with clinical academic centres, the health care industry, government agencies, and private agencies. The goals of the workshop are to:

Health Science Policy Council of the Heart and Stroke Foundation of Canada (HSFC) perspective- Dr. Paul Armstrong, Past HSFC Chair, current Chair of Dept. of Medicine University of Alberta Hospital

Dr. Armstrong explained that five years ago an HSFC task force in conjunction with its principal professional partner, the Canadian Cardiovascular Society, met with groups from volunteer organizations, government, industry and academic medicine to develop a vision for enhancing cardiovascular health and research and reducing the economic and social burden of cardiovascular disease. The task force recommendations included the promotion and development of partnerships and the development of a registry for cardiovascular disease. A pilot registry project, the Canadian Registry of Myocardial Infarction and Stroke (CROMIS), was an attempt to develop some national registry standards. Dr. Armstrong concluded that the initiative of this workshop carries forward and broadens the scope of these endeavors.

Laboratory Centre for Disease Control Perspective (LCDC)- Dr. Greg Taylor, Chief of Cardiovascular Diseases Division, LCDC, Health Canada

Dr. Taylor stated that the LCDC, as the national public health agency, is committed to its core functions of surveillance, risk assessment and disease prevention and control in the area of cardiovascular disease. This MI registry project will produce information that will enhance cardiovascular disease surveillance as well as improve cardiovascular disease control and risk factor assessment. Dr. Taylor concluded that LCDC brings a unique resource to the table as the only national public health agency.

MACSTRAC Program- Dr. Paul Daly, Director of MACSTRAC project, Director of Cardiac Care Unit (CCU) Toronto Hospital

Dr. Daly stated that the MACSTRAC project was launched in September, 1994. This project was designed as a Continuous Quality Improvement (CQI) project from which a registry project could be built. The project currently has the participation of 45 cardiac care units (CCU's) in the Toronto/ Hamilton area, including urban and rural hospitals of varying sizes. These hospitals comprise a representative sample of a specific geographic area.

Dr. Daly described the program as the collection of a clinical summary triggered by a CCU admission, usually to the emergency room. The two basic data collection tools are (a) a one-page form filled out by the primary caregiver at the bedside, containing core content in addition to secondary information revisable every three to six months and (b) a log book standardized across all CCU's, allowing MACSTRAC to keep track of the percentage of CCU admissions being reported. MACSTRAC reports back to the centres with one-page summaries of information relevant to a particular clinical sub-group, such as the infarct, acute coronary syndrome, and non-acute coronary syndrome populations. These reports to the centres provide information such as (a) basic demographics, (b) length of stay, (c) past medical history, (d) nursing workload, (e) secondary intervention rates, (f) mortality and (g) discharge.

The down-side of the MACSTRAC database, according to Dr. Daly, is that it is limited to the CCU phase. A new program, FASTRAK II, will expand the database to include the emergency room and ward phases of treatment. By the end of 1996, approximately 170 hospitals should be participating in FASTRAK II, resulting in clinical summaries for about 200,000 patients per year with acute coronary syndrome. It is expected that 40,000 to 50,000 of these patients will be admitted to hospital with acute MI.

Preliminary Results from Phase 1 of the Nova Scotia Acute MI Outcomes Monitoring Project- Dr. Jafna Cox, Dept. of Medicine and Dept. of Community Health and Epidemiology, Dalhousie University; Principal Investigator of the Canadian Registry of Myocardial Infarction and Stroke (CROMIS) Dr. Cox divided his talk into two parts. The first part of his talk dealt with the Canadian Registry of MI and Stroke (CROMIS) and the second with the Nova Scotia Acute MI Monitoring project.

CROMIS. Dr. Cox explained that the objective of CROMIS was to provide accurate information on all cases of MI and stroke in their hospital as an aid to research and policy development. They chose to nest their registry within an existing administrative-type data base. Dr. Cox noted that the over-reporting of MI, typically a misdiagnosis of unstable angina, can result in a false positive coding rate of from 20 - 57%. Their solution to this problem was to improve as much as possible the accuracy of MI and stroke diagnosis by adapting MONIKA and Oxfordshire project guidelines respectively. The new case definitions in the form of checklists for medical staff were piloted and the data from the project are currently being analyzed. Dr. Cox stated that although the CROMIS database is expected to provide relatively precise information on the true numbers of MI and stroke across Canada, the database will not provide clinical detail around individual cases or insight into case management issues.

Nova Scotia Acute MI Outcomes Monitoring project. In order to address the need for clinical information, the Nova Scotia Acute MI Outcomes Monitoring project was launched in 1994 as part of a collaborative venture with all but one hospital in Nova Scotia and seed funding from NHRDP. The goals of the project were to develop a template to analyze the process of disease in the province and then match the process to outcomes. MI was selected for initial study.

The project was split into three phases. Phase I involved collection of baseline MI management data for a 6-month period. The data was analyzed and then presented to health opinion leaders in the province for interpretation and feedback regarding changes to data processing. Phase II involved another six-month period of data collection. Dr. Cox illustrated the type of data accessible from this registry, such as (a) risk factor information, (b) distribution of cases across hospitals, (c) treatment rates, (d) mortality, (e) medical costs, and (f) length of hospital stay.

Dr. Cox concluded that they have established an outcomes monitoring and feedback process through a fairly unique tripartite relationship with government, health care providers and industry. Future plans include developing a SWAT- type unit to respond to particular issues as needed and building pre- and post- hospital phases to the database.

Session 2

Chair: Dr. Naylor, ICES

Dr. Naylor began the second session by noting that there are major questions around (a) medical costs, (b) quality of medical care, given the evidence that mortality rates from acute MI are higher than trials would predict, and (c) appropriate pharmacotherapy. He added that there is a contrast between simple data collection methods based on refinements of administrative data that potentially lack clinical detail and those methods that may provide a more detailed look at the process and outcomes of care including the post-hospital phase. Dr. Naylor concluded that one of the questions facing participants is how those two methodologies can complement each other to give a true population-based perspective on care.

Outline of Draft Proposal- Dr. Rick Davies, Ottawa Heart Institute

Dr. Davies stated that he would be speaking about the collection of data not currently available for most registries and an outline for a partnership to accomplish this.

Data Collection. Dr. Davies proposed that a representative sample of a subset of hospital centres be selected for more in-depth data collection. He presented four data collection modules. Module 1 would involve the collection of in-hospital data, using FASTRAK or the equivalent. Module 2 would be the collection of additional in-hospital data such as (a) severity in disease, (b) co-morbidity, (c) results of any diagnostic testing, (d) treatment received, and (e) name and dose of medication prescribed at discharge. Module 3 would consist of follow-up data three months after discharge. Module 4 would involve follow-up data one year after discharge. This last module would consist of linkages with other databases. Dr. Davies stated that the methodology for data collection should (a) be simple, including an unambiguous case definition, (b) incorporate a mechanism to change data elements as new questions arise and new knowledge accumulates, (c) be based on planned analyses, and (d) be implemented on a manageable subset of patients before making plans to expand the dataset.

Partnership. Dr. Davies next proposed an outline for collaboration between partners whereby partners would share data ownership. Partners would be drawn from four broad groups: (a) clinical and academic centres, (b) government agencies, (c) private agencies, and (d) health care industry.

Contributions from partners would depend on the group. For example, clinical academic centres would (a) collect accurate and complete data on all eligible patients, (b) give input as to important clinical questions to ask, and (c) contribute to data publication. Government agencies would (a) support registry workshops and meetings, (b) provide analytic expertise and resources, (c) contribute to reports and publications, (d) assist with links to other databases and (e) be committed to data dissemination. Non-governmental agencies would (a) promote the registry to constituents, (b) participate in workshops and meetings, (c) give access to peer review, (d) disseminate data, and (e) use registry results in health policy advocacy and health promotion. Industrial partners would contribute (a) financial support, (b) intellectual input on drugs, tests, devices and health economics, and (c) assist with data publication.

Access to Data. Dr. Davies proposed a policy on data access whereby each partner has access to the registry data such that the confidentiality of patients and the interests and intellectual property of other partners are protected.

Data Analysis. Resources, particularly from experts in statistics and epidemiology at LCDC and other partner government agencies, would need to be put aside to support high quality data analysis and writing for group publication.

Data Dissemination. Final reports would include the names of all partners. The protection of the intellectual property rights and legitimate interests of all partners would predominant.

Data Collection. Data collection should be simple and minimize the work load at the clinical sites. Centres would need to be reimbursed monetarily in the range of $10 per module multiplied by three modules for each patient.

Roundtable Discussion: Issues Involved in Achieving Consensus

The main issues raised during the discussion revolved around (a) the need for clear rules governing data access and data analysis, (b) the possibility of multiple data collection modules of an optional nature, (c) the start-up as well as ongoing financial costs of this ambitious project, and (d) the need to strengthen the risk factor profile.

Session 3

Chair: Dr. Armstrong, University of Alberta Hospital
Co-Chair: Dr. Greg Taylor

Myocardial Infarction Registries: Why Your Investment in Money in the Bank- Mr. William Tholl, National Executive Director of the Heart and Stroke Foundation of Canada; Adjunct Associate Professor at Faculty of Medicine, Queen's University; Economist

Mr. Tholl proposed that the time is right for a coordinated rather than disparate approach to the evolution of MI registries from three main perspectives: (a) a health policy perspective, (b) an economic perspective, and (c) a Canadian perspective.

From a health policy perspective, keeping track of people going through the system is important for ensuring that politicians who are in the midst of health reform are not cutting in areas of need and knowing the implications of some of these cuts in terms of health care. From an economic perspective, a coordinated approach has the advantage of risk-sharing and leveraging time, effort and money particularly with regard to maximizing the use of limited research talent in Canada. From a Canadian perspective, there is an advantage to linking datasets nationally, especially since there is evidence that the social system as well as the health system is contracting. MI registries and other tracking systems are ways of keeping tabs on what may be happening in a deleterious way as a consequence of system change and pointing to positive approaches to changing the system.

Neurological Registry- Dr. Andrew Penn, Associate Professor of Neurology and Medicine, University of Alberta

Dr. Penn described the STROKENET project of the Synapse Publishing Group at the University of Alberta. STROKENET's goal is to provide an on-line mechanism for getting information out from central sources to medical decision- makers, such as clinicians, and bringing information back again. The STROKENET project is funded by a joint government-industry venture called the Canadian Network for the Advancement of Research, Industry and Education (CANARIE) and is now part of the Network of Centres of Excellence project, HEALNET. STROKENET also has a role in a new G-7 Global Health Applications Initiative.

Dr. Penn proposed that an MI registry could be implemented in tandem with STROKENET's personal computer and internet links planned for hospitals across Canada without incurring extra infrastructure costs. The necessary tools to do this would be an MI guidance system consisting of the best practice for management of MI and the creation of some data collection forms, for example by FASTRAK. Dr. Penn noted that the University of Alberta would like to take a lead in generating an MI guidance system.

Dr. Naylor voiced two concerns regarding the above proposal. These were that data capture may not be complete due to possible non-compliance by physicians, and necessary data fields may not be captured using standardized forms.

Break-Out Groups

Three break-out groups were formed and met for an hour and a half before lunch. The groups were (a) a Partnerships group (led by Dr. Naylor), (b) a Data and Technical Issues group (led by Dr. Cox), and (c) a Funding group (led by Dr. Armstrong). Group leaders presented a synthesis of their group's discussions after Dr. Penn's talk.

1. Partnership Group- Dr. David Naylor

Dr. Naylor summarized four key issues: (a) networking, (b) follow-up data, (c) FAST TRACK partnership, and (d) data access. Regarding the importance of networking, Dr. Naylor noted that to determine the representativeness of a sub- population it will be necessary to link up with other databases, such as by comparing FASTRAK II hospital data with CIHI data on each hospitalization. The second issue, obtaining follow-up data at three months and one year after discharge, will require funding through partnerships with industry, agencies and FASTRAK. With reference to the third issue, Dr. Naylor stated that he felt they were tied into a partnership with FASTRAK which already has excellent buy-in with the provider hospitals and with the Hoffman-LaRoche pharmaceutical company given that Hoffman-LaRoche is spending 1.4 million a year on this project. On the fourth issue, Nalyor pointed out that legal issues around data access need to be hammered out.

2. Data and Technical Issues Group- Dr. Jafna Cox

Dr. Cox stated that by using FASTRAK for the registry it will be possible to get information on acute coronary syndromes, but for the pilot project it would be best to stick to MI. He also suggested they start with a regional population base because the way to arrive at a national vision is through an understanding of regional constraints. In addition, the regional centres should be empowered to do some analysis and contribute non-core questions of local importance to ensure buy-in to the project. Dr. Cox stated that the group tried to determine what central issues would be addressed with this database. The following core issues were raised by the group: (a) incidence and prevalence for acute MI, (b) access or need issues such as access to treatment, (c) risk factors and co-morbidity, (d) management of thrombolytic therapy and other major interventions, and (e) core outcomes including mortality and readmissions to hospital.

In terms of technologies, the group advised exploring linkages to existing databases wherever possible. Where databases do not exist, electronic data capture was preferred because it is cheaper and more accurate. Regional resources would also need to be explored which would probably involve paper forms, possibly carbon copy tear-offs to make it easy for nurses. The process should (a) be prospective where possible but will inevitably involve some retrospective search through medical records, (b) follow the nursing process, (c) be patient-specific, not event-specific, and (d) have data quality control, for example by a regional project nurse coordinator.

Dr. Cox noted that a clear definition of an MI was necessary. The group proposed a dual category consisting of definite or typical symptoms and probable categories.

3. Funding Group- Dr. Armstrong

Dr. Armstrong stated that the principal issues for the group were (a) what are the real costs of the project, (b) what are the benefits to justify costs, and (c) who should pay. In terms of project costs, the group calculated that it would cost $2-3 million over 5 years or about $110 per patient per year. The benefits of the project would be (a) a better understanding of the disease, (b) a modification of the behaviour of health care providers and the public, and (c) modification of therapies of interest to the health industry, for example the pharmaceutical industry. Regarding who pays project costs, Dr. Armstrong said that industry and the government are the groups with money to contribute.

Session 4

Chair: Dr. Gilles Dagenais, Universite de Montreal
Co-Chair: Dr. Safia Wasi, HSFC

Action Plan- Dr. Rick Davies, Ottawa Heart Institute

Dr. Davies initially sought and obtained consensus from participants on the possibility of a partnership. Dr. Davies then proposed that committees or working groups be formed to arrive at solutions to the issues around partnership. These groups would cover one of the following issues: (a) partnership and funding, (b) data issues and (c) technical issues.

Dr. Naylor added that before progressing further with sponsorship and funding, he felt that it was important to have an initial meeting with FASTRAK and their sponsor, Hoffman-LaRoche, to work out the practicalities of (a) the inclusion of other industrial partners, (b) data linkages, and (c) data sharing.

Data Management Issues- Mr. Ted Burnett, President of Enhance Management Perspective Inc.

Mr. Burnett provided a technological review of various solutions for handling the new registry. He considered four key factors: (a) data collection, (b) data transmission to a central registry, (c) the procedure for collating information at a central registry and entering it into a common database, and (d) a tool to deal with that information through linkages to other databases, data analysis and reporting. Mr. Burnett concluded that a digital system, called a "Personal Digital Assistant" was the best value compared with paper, e-mail and fax options in terms of (a) flexibility, (b) quality assurance, (c) cost-effectiveness, and (d) portability. The disadvantages of a digital system, however, include the need for training and a power source.


Last Updated: 2003-12-29	Important Notices

Centre for Chronic Disease Prevention and Control Cardiovascular Disease