Jump to content
PathLabTalk

validating patient histories in new LIS


Recommended Posts

We are importing all of our historical blood bank data from Cerner into SCC SoftBSA/SoftBank, most importantly our patient blood types, antibodies, antigens, comments and special requirements.  We are trying to figure out what number or percentage of patient histories to validate.  Any input?  I can't find a CAP reg related to this.

Link to post
Share on other sites


The IT department and software company should be jointly setting up and validating. You explain in excruciating detail to your IT people what you need and they make sure it can / will be done. Find out how imported history will be displayed and will there be a difference between imported and entered data. For example will special needs be imported as a comment or as a hard stop. Will comments have a maximum number of characters per comment or in total. ABO subgroups etc.

What I am saying is; more important than percentages (as most are straight forward) find a variety of unusual patient histories and check with IT and the software rep to ensure their history will be imported to your satisfaction. Once imported you are stuck. 

There is a strong argument not to import. This will allow a clean database that you know is consistent and accurate. When you do a history search you search the old system and put a comment in the new system saying search done, information transcribed, date and initials of who did the search. If you are cautious you cold set up a policy of two history searches and the old system then need not be searched.

Link to post
Share on other sites
14 hours ago, Ensis01 said:

The IT department and software company should be jointly setting up and validating. You explain in excruciating detail to your IT people what you need and they make sure it can / will be done. Find out how imported history will be displayed and will there be a difference between imported and entered data. For example will special needs be imported as a comment or as a hard stop. Will comments have a maximum number of characters per comment or in total. ABO subgroups etc.

What I am saying is; more important than percentages (as most are straight forward) find a variety of unusual patient histories and check with IT and the software rep to ensure their history will be imported to your satisfaction. Once imported you are stuck. 

There is a strong argument not to import. This will allow a clean database that you know is consistent and accurate. When you do a history search you search the old system and put a comment in the new system saying search done, information transcribed, date and initials of who did the search. If you are cautious you cold set up a policy of two history searches and the old system then need not be searched.

I should have specified that the historical data was entered into our archive to my exacting specifics and I have been mini-validating all along because it was an extremely complicated import that took our reporting team months to iron out all the information.  All of the historical data will live in SoftBSA with the patient data eventually importing in to SoftBank.  Unfortunately, my generalist staff only works in BB a couple of days per month and we have already had a near miss due to the history not being in our live system so import of patient data is crucial.  They had trouble transcribing a complicated patient from our old Cerner system into SoftBank while the historical data was getting sorted out.  Right now I'm about to officially validate BSA prior to them copying our live database and loading the patient information in the copy of the SoftBank database which will also have to be validated.  Right now I'm inclined to check the ABO/Rh history on 500 patients which is about 1%, check 100 patients with antibody or antigen histories and all patients with special instructions such as irradiation or comments.

 

We are importing unit and rhogam data as well and I think it would be a good idea to check at least a few of every product type we have ever used in the past decade with a heavy emphasis on our common products.  All of this is necessary because we don't have a contract to maintain access to our old system and the files that IT will have as raw data from our old system will be inaccessible and not interpret-able by lab staff.  Tests will also be available in my archive but we can also get those on the archived patient chart in an easier format so I don't see the point in spending a ton of time validating this other than making sure a handful match for each test.

Link to post
Share on other sites

We just did a history conversion from Meditech 5.67 to Meditech Expanse.  I chose 10% as my target and used a random number generator to blindly select the records to validate.  I also made sure to add a few patients that I knew had antibodies and special requirements, but found several just using the generated numbers.  Have fun!

Link to post
Share on other sites

If you have the IT department do the validation make sure they are blood bankers, or at least understand exactly what you are validating.  I've done this twice in my career.  I took several of my frequent flyers that had multiple antibodies, some of our patients that had previous antibodies but currently showed negative screens and those patients with "odd special needs" i.e. things outside the usual antibody problems.

Like Banker Girl I did 10% of the total converted records, yes that is a lot of pts in a big data base, but as you do your validation you will get comfortable with the converted data in the new system.

We also did more than one pass of converted data.  Checked the first vendor extract, found some "misses", they re-coded the extract, checked the second pass, etc. By the time we made it to go live the data was clean.  We did have to have some "comments" in the new system as it's hard to map like for like especially with special instruction.

Key point is to make sure you have good documentation for when your assessors and inspectors show up.  I've been asked on more than one occasion to provide the data conversion validation documentation.

Good luck and have fun :)

 

 

 

Link to post
Share on other sites

If you are importing into SCC, are you also importing into your EHR? We are a Epic facility and at go live missed the opportunity before go live to import into Soft and Epic from Sunquest and Meditech.  We have now imported into Soft but wish we had done it at go live into Epic. I was not involve in the validation.

Link to post
Share on other sites
41 minutes ago, PBM RN said:

If you are importing into SCC, are you also importing into your EHR? We are a Epic facility and at go live missed the opportunity before go live to import into Soft and Epic from Sunquest and Meditech.  We have now imported into Soft but wish we had done it at go live into Epic. I was not involve in the validation.

I have not heard of importing into EPIC.  What sorts of data would get imported into EPIC?  We have an archive patient chart that everything from the Cerner chart went in to so I can see the old transfusion records and any test results on the patients but that is it.  Our reporting team dropped the ball on getting the files tested to see if the format was correct for SoftBSA in a timely manner so we have been working on getting the file formats correct since July.  They are finally correct after about a thousand iterations.  I've been checking pieces and examples the entire time but we are starting the official validation of BSA this week.  This includes all BB tests, specimen IDs, patient demographics, blood types, antigens, antibodies, special requirements, comments, plus all RhoGAM and unit histories.  If that goes well we copy our patient database from SoftBank and will overlay BSA with the copy, revalidate the patients or a portion of the patients.  Then when we are satisfied this is correct we will take SoftBank LIVE down to import the BSA data and spot check a portion of the patients to make sure there wasn't a snafu in the import process, not so much the data itself.

Link to post
Share on other sites
2 hours ago, Likewine99 said:

If you have the IT department do the validation make sure they are blood bankers, or at least understand exactly what you are validating.  I've done this twice in my career.  I took several of my frequent flyers that had multiple antibodies, some of our patients that had previous antibodies but currently showed negative screens and those patients with "odd special needs" i.e. things outside the usual antibody problems.

Like Banker Girl I did 10% of the total converted records, yes that is a lot of pts in a big data base, but as you do your validation you will get comfortable with the converted data in the new system.

We also did more than one pass of converted data.  Checked the first vendor extract, found some "misses", they re-coded the extract, checked the second pass, etc. By the time we made it to go live the data was clean.  We did have to have some "comments" in the new system as it's hard to map like for like especially with special instruction.

Key point is to make sure you have good documentation for when your assessors and inspectors show up.  I've been asked on more than one occasion to provide the data conversion validation documentation.

Good luck and have fun :)

 

 

 

The IT department is only responsible for creating the exports (50+ different export files), it is 100% my job as the supervisor to make sure the format is working in our archive.  When it didn't work I reported the issue then the IT folks worked with the vendor to figure out how to format the files to get them to display correctly.  I'm extremely particular and wouldn't have trusted any one else to understand or get it right anyway.  It sounds like a small part but we have decided we are doing about 1% of our 52000+ patients.  Plus a percent of unit and RhoGAM histories as well.  But we are hand picking to make sure all patients with special requirements are checked and at least 1 example of every possible antigen and antibody is checked to make sure the translation is correct, plus a large portion of patient with comments.

Link to post
Share on other sites
1 hour ago, Jessica A said:

The IT department is only responsible for creating the exports (50+ different export files), it is 100% my job as the supervisor to make sure the format is working in our archive.  When it didn't work I reported the issue then the IT folks worked with the vendor to figure out how to format the files to get them to display correctly.  I'm extremely particular and wouldn't have trusted any one else to understand or get it right anyway.  It sounds like a small part but we have decided we are doing about 1% of our 52000+ patients.  Plus a percent of unit and RhoGAM histories as well.  But we are hand picking to make sure all patients with special requirements are checked and at least 1 example of every possible antigen and antibody is checked to make sure the translation is correct, plus a large portion of patient with comments.

We had our historical data imported from our legacy system into Softbank, however we did not import it into Epic. When we are looking at lab results, or using the order logic, to determine if a second verification blood type is needed, it can now only look back to our Epic go live date and not through our entire history of results. It is manual work to determine if a verification specimen is needed.

Link to post
Share on other sites
23 minutes ago, PBM RN said:

We had our historical data imported from our legacy system into Softbank, however we did not import it into Epic. When we are looking at lab results, or using the order logic, to determine if a second verification blood type is needed, it can now only look back to our Epic go live date and not through our entire history of results. It is manual work to determine if a verification specimen is needed.

EPIC should not have any impact on needing a 2nd specimen.  That's actually a setting in the hosparams of Softbank as well as a crucial piece of information in 1 of the data files.  If SoftBank is set to allow for a blood type from an outside import as an initial type AND the correct file has the DATE of last ABO/Rh test included, the system will recognize the old blood type AND that the old type counts toward electronic crossmatch (and type specific RBCs if your hosparam is set that way).  Initially, the data files were missing the date for the last blood type and I made them import that as well.  Now, assuming the validation goes well, we won't need to do retypes for patients with imported histories.

Link to post
Share on other sites
  • 2 weeks later...

I agree with Jessica A,  With a downstream BB system (SoftBank, Cerner, WellSky etc) you shouldn't need to check EPIC for previous pt history, that should be in your converted data from your old system. That you would have validated prior to your SoftBank go live.

I worked with 14 Blood Banks with converted data from 3 legacy systems.  None of them used EPIC as a source for previous patient history.  Aren't your patient historical results in SoftBank?

Just curious why you would import historical BB data into EPIC?  Our Medical Records department was responsible for patient data converted into EPIC but BB data source of truth was always the BB system.

 

 

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Advertisement

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.