National Expert Consensus Process

Following completion of the task group work, the draft recommendations and supporting information were presented for discussion and decision-making to a broad group of stakeholders at a national expert consensus meeting held in Toronto, Ontario in April, 2010 (see Appendix 3 for a list of consensus participants). Consensus participants included the Best Practices Working Group, the chairs of each task group, and invited opinion leaders, stroke survivors, and healthcare professionals external to the development process. The selection of consensus participants was based on ensuring representation from stroke-related health professional disciplines, all sectors of the healthcare continuum, all provinces in Canada, and rural and urban work environments.

The meeting was attended by 58 of the 63 invited participants. The objectives of the 2010 consensus meeting were to:

• Discuss and where necessary modify proposed updates to existing recommendations
• Discuss potential new recommendations or other amendments
• Vote and reach consensus on all recommendations proposed for 2010
• Discuss ongoing implementation strategies
• Conduct a strategic planning exercise to explore capacity and mechanisms to monitor best practice implementation and impact.

Consensus participants received the draft recommendations, including proposed changes to existing recommendations as well as new recommendations, in advance of the consensus meeting. They were asked to review the information, provide feedback to the Working Group prior to the meeting, and indicate their degree of support for each recommendation.

For the first part of the meeting, discussion and debate took place concerning the relevance, current evidence and practice, and barriers to uptake and implementation for each proposed recommendation. Votes were taken after each topic area discussion, with the following results:

• 31 of the 35 proposed recommendations were approved as they were presented or with minor revisions and edits.
• Two recommendations were deferred pending further investigation and revision by the respective task groups (atrial fibrillation and carotid intervention). These were revised and then reviewed electronically by the consensus participants, and both were accepted.
• Two sets of recommendations from the original Stroke Care Optimization of Rehabilitation through Evidence (SCORE) guidelines pertaining to upper limb and lower limb recovery following stroke were deferred as the writing groups continued to deliberate and review the current evidence.⁵ The proposed recommendations were circulated to the consensus panel in August 2010 for review and input. They were voted on electronically and confirmed for inclusion in the 2010 stroke best practice guidelines.
• No recommendations proposed to the consensus panel were rejected or eliminated from the 2010 update of the stroke best practices.

Subsequently, a strategic planning session was held to explore ways of maximizing the impact of the Canadian Best Practice Recommendations for Stroke Care. Objectives included:

• Identifying ongoing implementation strategies
• Exploring capacity to monitor best practice implementation and impact at the local level
• Discussing mechanisms to use performance information to improve implementation of best practices, patient care delivery, and outcomes
• Gathering advice on how the CSS can continue to build its strategy for system-wide improvement across the continuum of stroke care.

Following presentations by Drs. Ian Graham, Anthony Rudd, Judith Shamian and Ben Chan, consensus participants discussed and debated a number of topics, and developed a range of broader directional statements and guidance for the CSS, as follows:

• The need to explicitly recognize the many dimensions of accountability and ensure that all parties who have accountability are held responsible for quality of care and reporting results
• The need for quality and accountability measures at the population level (surveillance), the health system level, and the local service delivery level
• The need for measures of structure, processes of care, and outcomes
• The importance of setting targets that are specific and time-limited
• The importance of disclosure and reporting
• The need for profession-specific best practice summaries
• The need to look at performance through both a patient lens and a provider lens.

Guidance that pertained to specific sections or recommendations within the Canadian Best Practice Recommendations for Stroke Care was as follows:

• Emergency medical services should be integrated more formally with regional stroke programs. Processes should be in place for emergency medical services to collect and report on their performance based on the data they collect.
• Accreditation Canada’s Stroke Services Distinction program is an opportunity to promote the routine periodic collection of standardized data aligned directly with stroke best practices for hyper-acute and acute care as well as other components.
• Hospitals and regional health authorities should be encouraged to participate in the CIHI stroke special project as a cost-effective opportunity for feasible and consistent standardized collection of stroke data.
• With respect to hospital-based rehabilitation, benchmarks and standards are required to assess whether the right people are getting access to the right service in the right setting and that resources are being used effectively.
• With regard to stroke secondary prevention services, effective measurement and evaluation mechanisms are needed to compare and contrast the myriad of service delivery models.
• A method of measuring unmet need in the community is needed. From a quality perspective, community-based service providers such as municipal recreation programs and the YM/YWCA should be engaged in monitoring and evaluating services and linkages.
• All regional stroke programs should have a region-wide telestroke strategy that ensures that all residents of the region have effective access to available stroke expertise, and data systems should be established to monitor the use and impact of telestroke programs.