In the past, the standard approach for KOL identification was to hire good Medical Science Liaisons (MSLs) from various regions and rely on their Personal Networks to identify the right KOLs in their region.
This Personal Network approach has some strengths:
- It is very fast, since the MSLs typically already know the KOLs in their personal networks
- Good chance for strong access, since the MSLs typically target KOLs they have a existing relationship, they can often rapidly gain access to those KOLs for educational discussions
- It is inexpensive, since there are no additional costs from outside vendors
The Personal Network has some significant weaknesses:
- MSLs don’t know who they don’t know, some important KOLs may simply not show up on their radar because their personal networks do not cross
- MSLs are biased toward KOLs they have strong relationships who may or may not be the KOLs that are the most effective “hubs” of their formal and informal networks
- MSLs may specifically avoid “difficult” KOLs simply because they do not want to interact with them
The Personal Network approach is most effective in a very small, highly technical specialty medical community setting. In these types of communities, since there are so few people involved, everyone knows everyone in the community and the risks of the Personal Network approach are mitigated.
The next most common approach is the use of Secondary Research networks. Secondary Research networks are developed by gathering information that is publically available about potential KOLs for secondary sources. Information that is publically available may include:
- Degrees and institutions
- Institutional affiliation
- Conference presentations
- Principle investigator
- Committee involvement
- Board membership
The list can go on. This information is collected; each piece is given a weight based on the what is perceived to be likely to contribute to their being a KOL. The weighting produces a combined value which is then used to identify an individual as a KOL.
The Secondary Research approach has strengths:
- Fact based, eliminating personal bias found in the Personal Network approach
- Broad view, looking at everyone in the therapeutic area, not limited to individuals in a specific institution or with a specific background
- Less expensive than primary research, since the information in question is available in public databases
The Secondary Research approach has some weaknesses:
- Backward looking, since it relies on a preponderance of publically available information, it tends to favor those people who have contributed for the longest period of time
- Academic oriented, since it relies on public databases of information it tends to value those who publish and speak the most and those people tend to be academics
- Subject to weighting bias, if the weighting of the information does not correspond to what is truly makes someone a KOL, the exercise can produce seemingly valid data that misses actual KOLs
The Secondary Research approach works best in therapeutic areas that are fairly static and dominated by academicians. In these types of environments, there is a strong correlation between age and importance and between academia and practice.
The third approach to KOL identification is Primary Research. The Primary Research approach identifies who is a KOL by polling the physicians treating a disease state and directly asking them who influences them. These responses are then tabulated and the names of those individuals cited the most are given the most weight and are thus considered KOLs.
The Primary Research approach as its strengths:
- Accuracy, this is the only approach that actually asks physicians who influences them and thus is free of any bias or miscalculation
- Broad view, this approach does not limit itself to only KOLs already known by the organization
- Additional insight, at the same time that physicians are polled about influence, there is an opportunity to gain additional insight into their educational needs
The Primary Research approach also has weaknesses:
- Expensive, all direct research is expensive requiring a direct interaction with many physicians and potentially the cost for honoraria
- Sample bias, the sample of physicians polled may be biased, especially if the sample size is kept down by budgetary concerns
- Time consuming, it can take a number of weeks to conduct this type of research
The Primary Research approach is the best approach for less established disease states, less academically driven disease states and for any groups looking for less well established KOLs.
REAL LIFE EXAMPLES
Any of these three approaches may work, and they are often used in combination.
For example, I had a client who was entering a new disease state and thus needed to establish their MSL group from ground up. During the initial six months, as the MSLs were coming on board and being trained, the company allowed each MSL to use their Personal Network. Once the group was fully formed, they worked with a vendor to conduct a Secondary Research analysis of the community and create the initial list of KOLs to be targeted. After the MSLs had been in the field for six more months, they identified a related set of potential KOLs that the Secondary Research had not targeted, and they commissioned a targeted Primary Research effort to identify all the KOLs in this new group.
I have had other clients that used the Primary Research approach to develop their initial list of KOLs then used the Secondary Research approach to continue to track those KOLs and ensure that their staff was fully up to speed on their activities.
Some vendors that support these types of research include:
Secondary Research: Heartbeat Experts, SteepRock, OpenQ, ThoughtLeader Select
Primary Research: AlphaDetail, Encuity Research (full disclosure - I used to be a part of Campbell Alliance, the parent company to Encuity)
In my next post, I will discuss how these KOL interactions should be managed.
As always, I am curious to know your experience with KOL identification. Please leave comments below.