Wednesday, December 29, 2010

How should we pay for standards development in HealthIT

I know of a few different models:
  1. Organizations pay $$$ to be members.  Then the SDO gives away the standards for free.  This includes organizations like OASIS, IETF and the W3C.
  2. Organizations pay $$-$$$ to be members, and the SDO makes the material available to members for free, and non-members for a fee $/$$ (often the cost of membership).  ASTM and HL7 are like this.
  3. Organizations that use the standards pay $-$$$ for them.  CPT is like this.
  4. Governments whose populations use the standards pay $$$$ for them, and the standards are available to those populations for free.  IHTSDO (SNOMED CT) is like this.
  5. Governments develop them $$$-$$$$ and given them away for free.  RxNORM and UMLS are like this.
  6. Some organizations pay $$$ to participate in related testing events, and the guides are given away for free.  Other development costs are paid for by benfactors.  IHE is like this.
$ - tens to hundreds
$$ - hundreds to thousands
$$$  - tens of thousands or more
$$$$ - Millions or more

There are challenges to each model. 

Model #1 doesn't seem to work for vertical markets like healthcare because the $$ available are simply not enough to support the work enough to give away the IP.

Model #2 can cut out "little guys" and open source initiatives who don't have resources to purchase the standard or become members.

Model #3 annoys the people who have to pay, but at least puts the onus on the organization benefitting from their use.  Collecting payment is a challenge.  It's not clear how this funding is distributed when payment is for an individual product/standard.

Model #4 is interesting, except when it means that governments exert too much control over what they are paying for.

Model #5 lacks balance in the development side.  I like it for dealing with large aggregations of data like RxNORM or UMLS, but not so good when it's other content that needs consensus development.

Model #6 doesn't seem to scale up and relies on largess from organizations that want to influence what is going on.

Models 1, 4, 5 and 6 rely the largess from somewhere to make standards freely available.  Models 2 and 3 make users of the standards pay for them.  In all cases, the payers recieve some benefit.  In case 1, it seems to be prestige and influence in the development.  Case 2 is sort of an amalgamation of case 1 and 3.  Case 3 is pretty clear that the users benefit, and pay for the standard.  But I know of very few cases where this model is used elsewhere.  Case 4 and 5 seem to argue for public good as the reason for investing by governments.  Case 6 is interesting, as the benefit that comes from testing is certainly worth the investment.

The reason this is interesting to me has to do with my post on HL7 Strategic Initiatives.  It has the most to do with HL7's business model.  It should be no secret that HL7 is considering a new model.  One of the strategic initiatives is "Align HL7's business and revenue models to be responsive to national bodies while supporting global standards development" and the two measures of sucess are:
  • A new HL7 International business/revenue model has been developed
  • A revised HL7 International membership/governance has been developed
HL7 would love to be able to make its IP more accessible to the users.  The key question is what is the best way to do that.  I have my own opinions, and would love to hear yours.


  1. Observation: Standards are often like healthcare, i.e., those who pay and those who do the work are not necessarily those who primarily benefit.

    Fix healthcare funding and you may find a model that can help fix standards development.

  2. Keith...

    I'd like your opinion about which category the ANSI healthcare EDI standards "published" by the Washington Publishing Company based on ASC X12 committee meetings fit.

    I don't know enough about how publication fees are earned or shared...but I know that WPC charges a HUGE amount for them.

    I am guessing category #3?

    Steve S.
    Gilbert, AZ

  3. Good question. It looks sort of half-way between #2 and #3. Even though members recieve significant discounts, this is mostly an IP sales funding model. The table data cost is amazing at +$13K, but it's the entire set of X12 standards if I read it correctly. The cost of other publications (a couple hundred dollars) is better than what other SDOs charge in my experience.

  4. Keith, DICOM fits your #1. The fee to participate is sliding based on market size.

  5. CDISC seems to be pretty successful with basically #1