Description of the CLS Accession (OBR-20) and Filler Order (OBR-3)

From: Deane Atkinson Deane.Atkinson@albertahealthservices.ca
Date: 2 April, 2012 9:17:45 AM MDT
To: ”'holly.corneil@imaginehealthcentres.ca'” holly.corneil@imaginehealthcentres.ca
Cc: Millennium eLab MillenniumeLab@cls.ab.ca
Subject: FW: CLS eLab - use of Accession vs. Filler order number

Hi Holly,

Here is a description of the CLS Accession (OBR-20) and Filler Order (OBR-3).

CLS generally treats the Accession (OBR-20) as a number that ties together orders made on a patient visit. Still, it is possible to get multiple accessions per visit. The important thing to note is that results under a single accession can go back within multiple messages.

When it comes to message corrections, or updates in general, the Filler Order Number (OBR-3), in conjunction with the Accession, helps identify messages that are ok to overwrite.

I've included a few examples taken from our previous validation file (please see attached):

Typical lab order. One common accession, but it's transmitted in two messages. The results contained within are complementary, and are not meant to overwrite. Hence, notice that the Filler Order Number is different in both messages.

If an update for the Urinalysis portion were to come in, its filler order number would be 0067880709^101LA.

Typical lab correction. A surgical report arrives in a single message. Note the 'F' for 'Final' flags in the OBX segments. Later, a second message arrives with 'C' for 'Corrected' flag in the OBX segments. Because the Filler Order Number and Accessions match, you can now display the second message instead of the first one.

If you take a look at the body of the report, you'll notice that the corrected report is composed of 1) a small description explaining the correction, and 2) the body of the original report. So the 'overwrite' mechanic is assumed here, since all the previous info is contained in the second message.

A compound example. A set of results under a single accession arrives. They are contained in 3 messages (1-3). Later, updates for each of those messages arrive (4-6). Note that accessions are the same across the board, but Filler Order Numbers indicate which message can overwrite which.

Interesting things to note: The 'Differential part1' has no changes at all, so message 4 has no correction comments - it's effectively a 'resend message'.

The CBC has a single corrected value, so message 6 has a correction comment. 'Differential part2' had almost every value corrected, so message 5 is filled with correction comments. In every case, the update message carries the old values, as well as the corrected values, so you're not losing data by overwriting messages 1 and 3.

Occult Blood. A result arrives, and gets two consecutive updates. Vendors make two common mistakes here: 1) they display *all* three messages side by side, which is a lot of duplication, or 2) they only display the first message for some reason. Since the Filler Order Number and Accessions are the same, displaying message #3 would suffice.

A trickier example: Glucose Tolerance. The initial result is supplemented with a 1 hour and a 2 hour result later. This test is a bit of an oddity - the Filler order number matches, and the contents have redundancy, like in Example 2. However, the accessions are different, and the collection times for all 3 of these results will be different.

Technically, if only message 3 were retained, the only thing that would be lost is the exact collection time. It is debatable whether the doctor is interested in the collection time here, so long as they know which is the 0.5h, 1h, and 2h result.

For completeness, it's probably better to process all three messages as complementary, based on the different accession numbers. I believe I've passed validations in the past that used either of these options. In my personal opinion, this procedure would be better designed if the collection time were included in the contents, and they all had the same accession - although that's unlikely to happen soon.

• example1-5.zip