Posted by Erin Holt · Feb 06, 2017 at 4:20 PM

NEED HELP: ELR Specimen question from HL7

Hello,

As you know HL7 is currently going through ballot reconciliation for the next release of the Laboratory Reporting Implementation Guide, and specifically the Lab Reporting to Public Health Profile, otherwise known as ELR.

There was a comment included in the ballot regarding specimen that sparked discussion on whether or not the SPM group should be required, thus making the SPM required. Currently in the guide, the Specimen Group is RE [0..*] (required but may be empty, 0 to many repetitions), where if the Specimen group is present, the SPM is required. However, for the public health profile, there is an additional requirement that at least one SPM segment be present in each message. We are unsure of the intent of this additional requirement and so we are looking to gather some information from you on the importance of the SPM segment. See screenshot.

Please send us your feedback on the questions in bold below by COB Feb 15th or join us on Thursday 2/16 for the weekly HL7 OO/PHER call dedicated to the ELR ballot recon where we will discuss this topic.

From your public health prospective, how important is it that you get specimen information?

What would happen if you didn’t get an SPM? How does that affect your usage of the ELR message or affect your practice? Is it important to get specimen information about every result? Is it important to get at least one specimen per message?

Additionally, there are questions about whether or not the data element usages for SPM data elements are appropriate based on what information the lab will always have, sometimes have, or may not routinely have in order to require an SPM. For example, a lab may have a specimen ID (SPM2) but may not have Specimen type (SPM4). See screen shot below.

If the SPM segment is R (required), and a lab sends you an ELR message with only a Specimen ID and no Specimen Type, how would that affect your work?

…

All other data elements in the SPM have a usage of Varies (have to look at the profile constraints in the description) other than the Specimen Reject Reason and the Specimen Condition which are both RE.

Erin Holt Coyne, MPH|CPHIO

Office of Informatics and Analytics

TN Department of Health

Comments & Events

Julie Luepke

Hi Erin,
I am checking with our Cloverleaf team about the SPM questions. We have a meeting on Friday so hope to be able to address these questions then concerning the capture and reporting of the specimen/source on reports.
Julie

Julie Luepke
Coding/Billing Compliance Analyst
Laboratory Compliance Office
Integrity and Compliance Office

Feb 06, 2017 at 5:33 PM Notified 38 people

Marion Kainer

For CRE, getting specimen type is really important in order to subcategorize as colonization or infection.
We would still want to get the report if it were missing, but would need to follow-up with the reporter to get clarification.
With the ARLN and ability to now access a lot more colonization testing, this will become even more important.

[cid:image001.png@01D0AAAF.AF3502B0]
Marion A. Kainer MD, MPH, FRACP, FSHEA | Director
Healthcare Associated Infections and Antimicrobial Resistance Program
Andrew Johnson Tower, 3.403
710 James Robertson Parkway, Nashville, TN 37243
p. 615-741-7247; fax 615-741-3857
marion.kainer@tn.gov<mailto:marion.kainer@tn.gov>
tn.gov/health
Connect with TDH on Facebook<http://www.facebook.com/TNDeptofHealth> and Twitter<http://twitter.com/#!/TNDeptofHealth> @TNDeptofHealth!

Our Mission – To protect, promote and improve the health and prosperity of people in Tennessee.

Feb 07, 2017 at 6:34 PM Notified 41 people

Daniel Golson

As for the Division of Lab Services at the Tennessee Department of Health, I do not see any situation where we would not include an SPM segment in a result message either an LRI or an ELR message. We also would not accept an order without an SPM segment. As you know the major advantage to the SPM segment over the use of other message fields to contain pertinent specimen details is that it provides better detail in a more structured format. From the SPM, at present the most important piece of information we use from an electronic order is the specimen type (SPM-4). We use this in our matching process for the electronic order and the actual specimen (i.e. it is part of the information our laboratorians use to validate the order is correct to the specimen received. We in turn, use that same specimen type information in returning an electronic lab result. We are also now exploring the use of the Specimen Reject Reason field (SPM-21).

Feb 13, 2017 at 4:01 PM Notified 42 people

Scott Troppy, Surveillance Epidemiologist at MA DPH

Erin - sorry for my tardy response - was on a vacation.

As for Massachusetts - We definitely want the specimen ID. That is critical. The specimen type (or specimen source, I’m assuming) is equally important. Even though all sources are reportable, something like an outpatient urine doesn’t really have a public health intervention compared to an inpatient blood. We also might be limiting confirmatory testing at MA-SPHL to invasive specimens, and it would be helpful to match with ELR data. I agree with Marion/TN that with the ARLN and ability to now access a lot more colonization testing, this will become even more important going forward.

From Sita (which I think she sent already to you in an email but wanted to share with the group)
From your public health prospective, how important is it that you get specimen information?

We would consider it extremely important. MA has several constraints on the HL7 251 implementation guide, and requiring specimen source in SPM.4.1 is one of them. We would welcome any changes that negated the need for additional state-based constraints.

For several reportable diseases (Streptococcus pneumoniae, H. influenzae, MRSA, VRSA, Group A streptococcus, Group B streptococcus) we only require tests on sterile specimens (blood, CSF etc) because we are primarily interested in invasive infection – the only way we can ensure that ELR we receive is what we require is if we get specimen source information in the report.

For STDs in particular, it has become very important to know specimen source because the PH epidemiologists triage their outreach work based on specimen source, among other factors (anal, pharynx, genital).

Our childhood blood lead program is insistent that they should know whether the blood is venous or capillary.

We have a minimal dataset that will make a paper lab report acceptable, and specimen source, specimen ID (or accession number) and date of collection are part of that dataset.

One specimen per message is sufficient as long as all the tests/results in that message are from the same specimen.

Specimen collection date is also critical, given that epidemiologists routinely look at the difference between acute and convalescent titers for arboviruses; or consecutive tests over a period of time to determine if a chlamydia infection, for example is a new infection or a re-infection.

If the SPM segment is R (required), and a lab sends you an ELR message with only a Specimen ID and no Specimen Type, how would that affect your work?

If the result was one of the reportables described above, then someone from the PHA would call the ordering facility to find out specimen details – an increase in labor for both PHA and labs, which is entirely contrary to the intent behind ELR.

Finally, we have not heard any reasons why specimen information should not be made R.

Feb 15, 2017 at 4:50 PM Notified 43 people

Steve Pon

Regarding California, specimen information is important to public health intervention. If the sending facility has valid information about the specimen, it should be programmed and sent appropriately. Some sending organizations simply insert "unknown" for specimen info, which is built into a properly structured (technical) SPM segment, but obviously this is not epidemiologically useful. I figure that the cases in which a specimen is truly unknown are rather rare. So, it takes both technical support of SPM, plus content/vocab management, to get info that is useful.

Of course, we want all the necessary information, accurately, completely, and immediately. The downside of making anything (R)equired is a technical (automated) rejection upon lack of that segment or field. This can have the unintended effect of preventing public health from receiving actionable information sooner, rather than later, even if it is less than ideal (incomplete).
~Steve

Feb 22, 2017 at 5:49 PM Notified 44 people