TMGal

In emergencies, we always accept verbal orders for transfusion.  These should be followed up by a request documented in our electronic medical record, but that's after the fact.  If you have a paper system, then the followup order is documented that way.  There is a regulatory/accreditation requirement, which I consider bureaucratic, obstructive and useless,  that these emergency requests require a signed release from the ordering practitioner, if the transfusion is not fully tested for the recipient.  

In our 25 bed critical access hospital, uncrossmatched blood orders are always verbal to the Blood Bank and all LIS entries are made by the Blood Bank Tech thereafter.  If the patient is registered (true name or John Doe), we perform the Emergency Release Routine in Meditech (otherwise we resort to paper back up procedures).  This routine assigns the unit to the patient, and allows us to choose whether or not to add to a specimen in use or create a new specimen order.  This same screen documents the issue of the unit to the patient.  An Issue/Transfusion card prints and is attached to the unit. A Practitioner's request for uncrossmatched blood form is filled out.  The unit is delivered to the care area by the Tech but not handed to the nurse until a provider signs the request form.  A nurse can sign but the provider must countersign when time permits. This form is immediately brought back to the Lab and additional units can be added to it need be.  The completed form, after review, is scanned into the EMR. The form has space for communicating any antibody history or other special transfusion requirements.

We accept verbal requests but should be followed by an order for emergency released (RBC, WB or plasma) placed in Epic. We issue blood with a triplicate form and require a physician's signature and a copy returned to the Blood Bank.  The form documents the units that were issued, product type, attestation statement, diagnosis, and issue information including a visual inspection of the product, person picking up the product, date/time for issue and the employee ID of the person issuing the product(s)

MTP and Emergency release is a verbal phone call to the blood bank with a signed form during the event.  Routines need an order faxed from the provider or orders placed into Epic.  Happy New Year! 

We have a form the Dr. must sign.  It includes the unit number/numbers, the Dr.s signature time and date, the tech involved signature time and date.  if the unit is O negative or type specific.  (We try to give type specific if possible and limit uncrossmatched blood to 2 units.. try)  We are a small rural hospital.  The ER calls the request to us.  on a rare occasion surgery will request emergency release blood.  same thing they call us.

We have been using an Erytra Eflexis (Grifols) for almost 2 years now and we are very happy with it and the service team.
It is interfaced to Sunquest.

We have been using a Bio-Rad IH-500 (and a SAXO as backup) for about a year now.  Very happy with them.  They work with the IH-Com that serves as a command centre for both and as middleware.  We also use the antibody software and upload our QC to Unity.  (We were manual gel/tube prior).
sandra 

I am getting ready to install an ECHO Lumina.   Just had our initial set up meeting yesterday.  Switching from gel to solid phase.
Looking forward to a bit more standardization in this department.  We use DI but the ECHO also comes with its own middleware (or so I believe).

We have two IH 1000.  BioRad has been great to work with.
They are workhorses. 
They are complicated and tend to have a fair amount of downtime.
Many years ago when we first introduced automation it was solid phase.  A lot of our patients were then coming up D positive (D neg history) as the methodology was a lot more sensitive than tube.  The same thing happened a few years ago when we switched to gel.
You'll also pick up more colds and junk with gel.
Overall we're happy.  We might consider an IH 500 someday for the titers.

Great feedback. Thank you.

Great feedback. Thank you.

I appreciate in a perfect world the error rate would be zero; however, this is not feasible as humans are entering data and humans will make mistakes. I am not looking for someone to tell me my error rate is OK because it compares to someone else - I am looking for some benchmarks. For example, if I am at an error rate of 6% and folks in similar sized facilities accessioning similar numbers are around 3% then I know I need to seriously look into our processes and can work towards that goal. If I happen to be around the same error rate as other comparable sites then I will not expect to see as large a decrease in the error rate but will continue to ensure it trends downward.
thanks for the discussion...
 

I appreciate in a perfect world the error rate would be zero; however, this is not feasible as humans are entering data and humans will make mistakes. I am not looking for someone to tell me my error rate is OK because it compares to someone else - I am looking for some benchmarks. For example, if I am at an error rate of 6% and folks in similar sized facilities accessioning similar numbers are around 3% then I know I need to seriously look into our processes and can work towards that goal. If I happen to be around the same error rate as other comparable sites then I will not expect to see as large a decrease in the error rate but will continue to ensure it trends downward.
thanks for the discussion...