Monday, December 05, 2005

HL7 Launches Marketing Effort for Katrina Relief

Thomas Beale points to this press release:

as an illustration of how HL7 people say different things to different people to suit their political/marketing purposes. Are HL7 standards making an impact on the ground in Louisiana? Or rather just members of HL7 wearing other sorts of hats?:

"This press release is hardly responsible or ethical representation of a supposed standards body.

"A similarly exaggerated claim is contained in the HIMSS announcement (see extracts below) which starts with the statement:

  • Health Level Seven’s Version 3 Normative Edition 2005 standards, dubbed The Foundation of Healthcare Interoperability, have been made available on ...

"Here, HL7 is claiming that its specifications are the basis of healthcare interoperability. That's a pretty big claim for an organisation that doesn't know much about EHRs, has only just started doing service models in the last year or so, and made its v3 message offering so complicated that almost no-one can understand it and which somehow managed to ignore all of the previous health standards work in Corbamed, CEN, UN/CEFACT, and work in GEHR, PICNIC, openEHR, and also managed to ignore basic norms of object-oriented modelling!

"Further down in the same announcement, HL7 is claimed to have provided the standards for workflow, documentation and security in health:

  • Since its inception in 1987, HL7 has worked to address the unique needs of the healthcare IT industry by developing a variety of pioneering standards ­ from messaging to documentation to security to workflow.

"None of the HL7 specifications that I know of contains anything to do with the large amount of workflow research available today; certainly there is little workflow thinking in the RIM or RMIMs (an exception being the state model of Acts). There are various doctoral theses and projects around the world which have treated the problem of workflow in healthcare directly, yet none of this work is used in HL7 to my knowledge. I guess the existence of CDA is the basis of the claim that HL7 has pioneered standards in documentation, even though projects like GEHR, CEN and openEHR have been working in this sphere for years using object-oriented models rather than just XML schemas. I don't know of any HL7 standard to do with security.

"I can only take these kinds of claims (especially intermingled with a menu of prices for buying sponsoring opportunities) as marketing, not connected with reality.

"HL7 speaks with a forked tongue also when it comes to the issue of its support for or use of specific software systems. On the one hand it maintains the official position that it is agnostic with regard to software (to preserve compatibility with different systems or when systems change). When it suits them, however, they make repeated claims to the effect that they have the standards for the "foundation of healthcare interoperability" and the like.

"They have tried very hard to make their unpopular (even within HL7) data type model an ISO standard for data types in health computing. This attempt has temporarily been blocked by members of CEN TC/251 and others in 2005, on the basis of unsuitability. I wouldn't even mind if the world standard for clinical data types came from HL7 – if it were any good – but their models are close to unimplementable, as many users of the XML-ITS have discovered. In any case, the numerous objections in principle should be enough to stop them being used for software or standards.

"To see the problems, consider this simple example: to record observations properly, you need to be able to record:

– a history of events, since observations are in past time, and multiple samples are often taken

– the data at each timepoint, which may be structurally complex, as in the case of Apgar, microbiology, BP etc

– the state of the patient at given times, which is often needed to correctly interpret the data (e.g. was the patient lying in bed or standing when the BP was taken?)

the protocol/method of performing the observation (e.g. instrumentation used).

"So, in the openEHR model of Observation you will find all of these things (see here, entry section and the data structures supporting this model here).

"None of these requirements are particular to the EHR. Whether you are making an observation in a laboratory, in a doctor's surgery, or at home, they would all apply; they also apply regardless of how and where the data are being stored. These requirements are not specific to the EHR, only to the scientific process of clinical investigation.

"I challenge anyone to show where such requirements are clearly met in the HL7 models.

"I think something has to be done about HL7's marketing arm – it seems to be getting out of control. Should standards bodies even have marketing arms?



*Since its inception in 1987, HL7 has worked to address the unique needs of the healthcare IT industry by developing a variety of pioneering standards ­ from messaging to documentation to security to workflow.*

*Today, we proudly invite you to join us in the HL7 exhibit at the Healthcare Information and Management Systems Society (HIMSS) 2006 Annual Conference and Exhibition.*

*Come be a part of this exciting opportunity as we educate the industry about HL7 and its role in the NHIN, and bring vendors and providers together to showcase how HL7 standards are used to provide real-life solutions every day in care-giving settings across the country.*

*Applications are now being accepted for participation in the Health Level Seven (HL7) HIMSS 2006 Exhibit.*

Go to to see all the ways your organization can partner with HL7 at HIMSS 2006 and tell your HL7 story to HIMSS attendees. Or contact Jonathan Himlin at 734-677-7777 x165 / to inquire further.

*Sponsorship Opportunities*

Whether or not you’re participating in the Provider Solutions Showcase, sponsoring an element of the HL7 2006 HIMSS Exhibit is a fantastic way to partner your company and its brand with the preeminent healthcare information technology standards development organization in the U.S. and abroad. Choose from among the following sponsorship opportunities:

*SOLD **HL7 Educational Theater - $10,000*

*SOLD* *HL7 Provider Solutions Showcase - $10,000*

*HL7 HIMSS Exhibit Brochure - $5,000 (ONE AVAILABLE)*

*General Exhibit Sponsorship - $3,000 (AVAILABLE ­ NO LIMIT)*

*Individual HL7 Educational Session Sponsorship - $1,000 (AVAILABLE)*


Anonymous said...

Have you found any positive points about HL7 V3. You seem to be sitting on the "anti HL7" side of the fence. Which current standards do you think would be a better choice for messaging and EHRs?

Barry Smith said...

I think we can leave it to the HL7's own marketing arm to document positive features of HL7 V3.

I think also that, in the age of internet services, it is becoming increasingly unclear whether the technology of messaging standards is needed at all. But even leaving this aside, HL7 V2 is a reasonably good messaging standard for its scope (mainly pathology results and imaging), and it is ten times simpler than HL7 V3, which is already quite widely implemented. Where, then, is the business case for Version 3?

As to the Electronic Health Record, since HL7 itself does not in fact have an EHR solution, I am tempted to say that more or less anything that has been designed in response to EHR requirements would represent an improvement on something which does not exist.