Dear all –

Conversations, connections, questions and insights – these were the outcomes of our very first Research Colloquium hosted last week at Brandeis University.  Having set a limit of 55 participants, we exceeded that limit before early registration had even ended, and decided to accommodate 75 participants from all across the US — Northeast, Mid-Atlantic, Southeast, Midwest, Great Plains, Rocky Mountains, Southwest, West Coast and Northwest — as well as Australia, Belgium, Canada, Denmark, Netherlands, Ireland, Spain, Sweden and the UK.

Despite the number and diversity of participants, we somehow connected on a fairly intimate level.  See above for some photos!  As I always say, the best part of doing this work is the amazing people you get to meet and work with to create positive social change.

Our conversations were about research but more specifically about interventional research that enables us to learn by using rigorous measures and methods while creating positive change.  My note today is intended to continue these conversations so please feel free to respond with your own thoughts here in our blog (scroll down to the comments at the bottom of this page)— whether or not you were there!  Some thoughts to get us started:

RC Measurement and Analysis

* Participants were eager to include the patient and family in the RC network measure, along with other participants on the care team, especially as health systems are striving to foster health and wellness in the community.  The RC Survey assesses 7 dimensions – the quality of communication (frequent, timely, accurate, problem-solving) and relationships (shared goals, shared knowledge, mutual respect) among all participants – now with the option of including the patient and family as well.   In addition, the PPIC measure captures the patient’s evaluation of care team activities in a way that is potentially complementary.

* When setting the bounds of RC networks, we can remain narrow (core team only) or extend those bounds to include non-clinical participants (receptionists, housekeeping, transport workers, navigators, wellness coaches, etc.) who can play an essential role in coordination, or extend even further to include participants across multiple organizational boundaries.

* One key lesson was to include all participants who perform tasks that are needed to achieve the desired outcomes, and not be influenced by perceived status when deciding who to include.  Another key lesson was to include participants who do not currently see themselves as being on the same “team,” in order to gain a baseline for building a more inclusive team going forward, across multiple organizational boundaries.

* Participants want to measure RC at the level of coordinating care for particular patient populations, and also to measure RC at the organizational level to capture culture more broadly.  Both are possible – but the survey will be targeted differently in each case.  In the first case we ask each of the 7 questions narrowly, for example about “the care of diabetic patients” or “the transition from pediatric to adult care.”  In the second case we ask each of the 7 questions more broadly, for example about “achieving our mission as an organization.”  The 7 RC questions are always the same but the focal work process that they refer to can be targeted narrowly or broadly, depending on your improvement focus.

*  Participants also want to capture RC in a more dynamic way, with continuous updates on the 7 dimensions of RC and how they are playing out across the network, to inform both research and the process of change.  RC Analytics is working with partners to collect data in more dynamic ways than the current RC Survey allows, which is very efficient but designed for use in intervals of 3 to 6 months rather than more frequent snapshots.

* In addition to the matrices and network maps we already provide, participants want to see RC displayed in social network maps to achieve additional insights into potential intervention designs — RC Analytics is working with partners to provide these additional visualization capabilities.

RC Interventions

*  When designing relational interventions, there is much to be learned about the science of intentional change, from the human mind to the group to the organization to cross-organizational networks and beyond.  It may be useful to see these levels as connected, as argued by Wilson et al in “Evolving the Future: Toward a Science of Intentional Change,” forthcoming in Behavioral and Brain Sciences.

* While interventions are intentional efforts to change, it may be essential for their success that interventions emerge as co-constructed by participants as they seek to make sense of their situation and the opportunities/ challenges they face, guided by internal or external change agents who model the changes they are seeking to foster.

* Structural interventions — like huddles, reward systems, conflict resolution practices, performance measurement, protocols, etc. — are more likely to succeed when they are coupled with relational interventions that address how people connect with each other in terms of shared goals, shared knowledge and mutual respect — not simply imposed in a seemingly arbitrary way from above.

* Likewise with work process interventions, relational interventions can be used to build connections across standardized tasks to enable participants to coordinate “on the fly” when the situation calls for more than a standardized response.

RC Evaluation Design

* The complementary nature of relational, structural and work process interventions makes causality harder to disentangle from a research perspective.  This calls for careful evaluation design such as longitudinal designs where one intervention type is followed by the addition of a second and then a third.  For example starting with lean/work process interventions, followed by relational interventions, and then structural interventions to embed new coordination patterns into the ongoing routine of the organization.

* Alternatively we can use evaluation designs in which structural, lean/work process or relational interventions are carried out on their own in some sites, and in combination with each other in other sites.

* With these evaluation designs, we are looking to assess the impact of interventions on RC and to assess the impact of changes in RC on cost and quality performance, as well as patient/family experience, patient/family engagement, worker engagement and worker well-being, building on the cross-sectional evidence that links RC with all of these critically important outcomes.

Thanking Our Presenters and Sharing Their Presentations

Our presenters were fabulous at providing insights and launching conversations — their slides will be posted here.  We extend our sincere thanks to Kathy McDonald (Stanford), Sara Singer (Harvard), Joan Roche (U Mass Amherst), Signe Peterson Flieger (Brandeis), Ken Milne and Nancy Whitelaw (Salus Global), Peter Kreiner (Brandeis), Eric Jones (mdlogix), Jo Shapiro (Brigham and Women’s), Tony Suchman (McArdle Ramerman), Luci Leykhum and Holly Lanham (UT Health Science San Antonio), John Krueger (Iowa Health), Rob Reid (Group Health), Jose Azar (Indiana University Health), Thomas Huber and Gerard Livaudais (Quantros), Gene Beyt (RC Analytics), Farbod Hagigi (ClinicalBox), John Carroll (MIT Sloan School), Jennifer Perloff (Brandeis), Jeff Wetherhold and Gareth Parry (Institute for Healthcare Improvement).
Thanks also to everyone who came and contributed to the rich conversations. Best wishes to us all for our continued learning!


Photo credits: Gene Beyt

Stan Wallack welcomes Colloquium participants

Stan Wallack welcomes Colloquium participants