Currently during lab result upload in OpenELIS, there is a restriction that accession number used in the input CSV should not be already created/used in the system. In one of our implementations, we have a scenario where lab order is placed from EMR for a particular test and sample is collected in ELIS. Since this test is performed by a machine in groups, the hospital waits for 93 samples to be collected to perform this test. Once the test is performed, entering the results for 93 patients manually is a tedious task and error prone. Hence they want to leverage on the bulk result upload feature in ELIS. However, ELIS doesn’t accept these results because the accession number has already been generated by the system during sample collection. Is there a way we can bypass this restriction so that it should accept accession numbers which are already generated but the results are not yet entered.
The CSV file upload for sample (actually results), must have the accession number specified. I believe this was done for scenarios where the accession was not generated through ELIS but handled offline and eventually uploaded - during the course of which both sample with valid accession is created and results recorded. I don’t think there is any other way, other than to create another option for file import where the accession is already created. Check here
In our use case, some samples will have accession number which is already generated and some samples will have new accession numbers. So we are planning to enhance the existing feature. The solution that we came up with:
- If Accession number is not present in the system, upload the lab result without any further checks
- If Accession number is already present in the system, check if there is data against the test that you are trying to upload
- If data(result) is not present against the lab test, upload the result for the test
- If the data(result) is already present, do not upload and recorded it in error file
- If there are 2 tests in CSV upload with the same accession number and if data is already present for one test in ELIS and not present for the other, upload the result for the test against which data is not already present. Failed event is recorded in error file for the test for which data is already present.
Does this seem viable and will product be interested in merging this solution?
I think so.
Note for case 2 and specifically 2.3, you might have to really figure out what really happens on the EMR side. Things to note of
- What happens if the Order was created in EMR?
- Also if multiple tests were done off the same sample, and one of the results were alread published, what happens on the EMR front when the result syncs.
- I remember a case, where if orders came from EMR order, and there were multiple tests for the same sample and accession, then if you publish one result without others, the other tests were automatically considered as cancelled! @arjun can you confirm this?
- What if there is no associated order in the EMR?
- I think this is an easier one, but good to check if for the same accession number another test result were published - what happens.