Core Topics …

Network Visualisation & Analysis, Knowledge Transfer, Influence, Opinion Leader, Early Adopter, Peer-evaluated Expertise, Knowledge Strategy, UCINET^TM, KNETMAP^TM

Key Issue …

Network Visualisation & Analysis to Identify Subject Matter Experts

CASE 4: Using Network Visualisation & Analysis (NVA) to Identify Influential Up-And-Coming Opinion Leaders in A Customer Community & Generate A Knowledge Location Map Shared With The Community

The North American pharmaceutical company Pharmco presented a new paradigm to its customer community of physician and specialist stakeholders: the informal network as the new currency of information exchange and value creation. A map showing peer-evaluated ‘hubs’ and ‘authorities’ was generated by NVA and then shared with all contributors.

Innovative pharmaceuticals such as Pharmco recognise physicians are the most important stakeholder in their value chain and that interdependence is the modus operandi of the knowledge era. Physicians and specialists want different ways to access information, and network maps afford a unique representation and new perspective of the structure in which they operate.

Issues Addressed:

Pharmco is one of the world’s top ten pharmaceutical groups focusing on four therapeutic areas which include internal medicine and oncology. Continuing medical education is an important enterprise activity that serves all stakeholders. In an effort to identify a more diverse group of subject matter experts (SMEs), it elected to pilot an information exchange mapping initiative. This was a win-win initiative for both the sponsor and stakeholders.

Top SMEs are frequently not accessible for consultation. They generally have demanding schedules where their time is divided between practice, research and speaking engagements. When the top SME is not around, who is the next in line who can give an opinion on a course of treatment? And when the next in line is not available, what is the pool of up-and-coming SMEs available for consultation?

Objectives:

· Create a network map where all stakeholders have access to the best sources of information;

· Invite the up-and-coming experts to take on key roles for developing better continuing medical education sprogrammes.

Approach:

· Two queries were delivered by telephone survey or via email:

First Query: When you discuss new treatments and practices with other physicians, whose opinion, recommendation, or advice do you seek when discussing cardiology?

· Physicians respond to each query by selecting individuals from a pick list of names;

· New names (including external contacts) are added to the list;

· The data is displayed in a dynamic knowledge network map that updates after each submission.

We used the data-gathering tool KNETMAP^TM for this pilot. This application generates dynamic Web-based knowledge network maps in real-time after each submission by a participant.

Figure 1: NVA map based on Query 1; Case 4

With the cooperation of 791 physicians who were requested to provide data in response to two queries, two “knowledge network maps” were generated.

Maps were made available for viewing by participants after each submission and were archived for retrieval on the Web site www.sun-project.ca, either for decision support, for location of expertise, or for monitoring changes in the knowledge networks. Specific attributes designating category (general practitioner/family medicine, specialist etc.) were assigned to all participants; these attributes showed up as colour coded nodes on the map.

Results:

· Lists of known subject matter experts;

· Lists of up-and-coming subject matter experts.

____________________________________________________________________

Nodes are color coded by the following location attributes. All names are pseudonyms. This is sample data only.

n Specialist n General Practitioner/Family Medicine

n X -- all others, including external contacts

Node Attribute

Michael Johnson n # of incoming links (7); # outgoing links (2)

Alex Mathers n # of incoming links (11); # outgoing links (2)

Jeremy Heston n # of incoming links (11); # outgoing links (1)

Sam Henderson n # of incoming links (7); # outgoing links (0)

Laurence Clement n # of incoming links (5); # outgoing links (0)

John Honeywell n # of incoming links (6); # outgoing links (1)

John Whelan n # of incoming links (8); # outgoing links (2)

Thomas Martin n # of incoming links (11); # outgoing links (1)

Alan Jamieson n # of incoming links (7); # outgoing links (1)

Jeffrey Neilsen n # of incoming links (20); # outgoing links (1)

John Barker n # of incoming links (5); # outgoing links (2)

John Trotter n # of incoming links (13); # outgoing links (2)

Neil Fleischer n # of incoming links (4); # outgoing links (2)

*Definition of REACH: Reach-In measures how influential a node is. The metric looks at both direct and indirect ties. By calculating how many unique nodes seek the advice/expertise/opinion of node X, the influence of node X can be determined. The influence of node X goes up if other influential nodes seek its advice/expertise/opinion. The sphere of influence for node X can be determined by viewing both direct and indirect in/out links surrounding node X -- incoming links show who seeks out node X, while outgoing links reveal who, if anyone, node X seeks for advice/expertise/opinion.

Node Specific Data

Michael Johnson: outgoing links are John Trotter, Neil Fleischer
Alex Mather: outgoing link is Thomas Martin

Thomas Martin: incoming link is Alex Mathers; outgoing link is Alex Mather

Alan Jamieson: incoming link is Jeffrey Neilsen; outgoing link is Jeffrey Neilsen

Jeffrey Neilsen: outgoing link is Alan Jamieson

John Trotter: incoming link is Neil Fleischer; outgoing link is Neil Fleischer

Neil Fleischer: incoming links are Michael Johnson, John Trotter; outgoing links are Michael Johnson, John Trotter

Figure 2: Analysed Data

_________________________________________________________________________________________________

Benefits

· Reduced subjectivity in identifying SMEs due to the peer evaluation approach

· Identification of individuals with deep subject matter knowledge

· Decision support for targeted CME development

Future Considerations

· Generating and archiving knowledge network maps of subject matter experts and making such maps and/or lists available to all staff;

· Using subject matter expert network maps as orientation tools for medical schools;

· Updating the maps and securing permission to make the names public.

Conclusion

The data gathered in this pilot revealed many of the 'lynchpins' in the flows of knowledge instrumental to sourcing information in the urology knowledge domain. Such patterns are generally only manifest in informal networks and particularly significant for the healthcare community because there are no managerial lines for information to flow. These informal links help circulate information and are responsible for significant activity that sustains the effective functioning of the structure. Of significant interest were the “surprise” SMEs who surfaced as key resources.

Feedback from the participants was increasingly enthusiastic from the start of this initiative. Now that the first two maps have been generated, it is hoped that participants will update and maintain this network via an email reminder every six months. Network maps show the relationships based on information exchange between colleagues. These maps have significant potential for decision support related to making advances in healthcare delivery.