sgoertzen

When blood bank is called to release uncrossmatched blood, we specifically ask who the ordering physician is. The blood banker then places an order in EPIC called "Release of Uncrossmatched Blood" which is the statement (taken from AABB Standards) about the situation being critical enough to warrant the release of blood before compatibility testing is completed.  This order must be electronically co-signed in EPIC by the physician that we name as the ordering physician.   We have EPIC and WellSky (Mediware HCLL), but it shouldn't matter what blood bank computer system you have since this is all done (ordered/cosigned) in EPIC.  I'm attaching our uncrossmatched worksheet that includes the instructions that we follow to place this order in EPIC and assign it to the ordering physician for co-sign.  I check each day to make sure it gets co-signed. Once signed, I print the order with the co-sign tracking information, attach it to our worksheet, and file the paperwork for easy retrieval during future inspections/assessments.   
TO-381F01 Release of Uncrossmatched Blood.docx

We apply this custom label to all PR platelets and PR platelet aliquots (see attached picture).  We're a children's hospital and we've been giving PR platelets to neonates since March 2017 with no problems.
PR Platelets.docx

Our hospital went LIVE with Epic and WellSky in April 2020. The rest of the lab is Epic Beaker. We previously had MediTech house-wide. We have not experienced an interface delay between the 2 systems as David mentions above. We're a children's hospital with a level 2 trauma center.

Our hospital went LIVE with Epic and WellSky in April 2020. The rest of the lab is Epic Beaker. We previously had MediTech house-wide. We have not experienced an interface delay between the 2 systems as David mentions above. We're a children's hospital with a level 2 trauma center.

Someone above commented that a 2nd sample is only required in the U.S. for computer crossmatch (which used to be true). But with the 31st Edition of AABB Standards (effective April 1, 2018), this requirement was moved so that it now applies for all pretransfusion testing for allogeneic transfusions including all types of crossmatching (IS, AHG, and Computer crossmatching). This is more in line with CAP requirements and makes more sense in order to detect possible Wrong Blood In Tube (WBIT) events.
AABB Standards for Blood Banks and Transfusion Services, 31st Edition
5.14.5 Pretransfusion Testing for Allogeneic Transfusion  
There shall be two determinations of the recipient’s ABO group as specified in Standard 5.14.1.  The first determination shall be performed on a current sample, and the second determination by one of the following methods:
Testing a second current sample.
Comparison with previous records.
Retesting the same sample if patient identification was verified using an electronic identification system or another process validated to reduce the risk of misidentification.
Standards 5.11 and 5.27.1 apply.
 
Personal Note: If you intend to retest the same sample (by a different person or the same person), be prepared to show the AABB assessor your validation proving that your "another process" is actually validated to reduce the risk of misidentification (i.e. WBITs). 
 
CAP Checklist Requirements:
TRM.30575 Misidentification Risk
The facility has a system to reduce the risk of mistransfusion for non-emergent red cell transfusions.
NOTE:  Mistransfusion occurs from misidentification of the intended recipient at the time of collection of the pretransfusion testing sample, during laboratory testing and preparation of units to be issued, and at the time of transfusion.  Misidentification at sample collection occurs approximately once in every 1,000 samples, and in one in every 12,000 transfusions the recipient receives a unit not intended for or not properly selected for him/her.  The laboratory is expected to have implemented a plan to reduce these risks through implementation of a risk-reduction system.  Among options that might be considered are:  (1) Verifying the ABO group of the intended recipient on a second sample collected at a separate phlebotomy (including the recording of the result in the institution's historical record); (2) Utilizing a mechanical barrier system or an electronic identification verification system that ensures that the patient from whom the pretransfusion specimen was collected is the same patient who is about to be transfused.  Other approaches capable of reducing the risk of mistransfusion may be used.  The laboratory should participate in monitoring the effectiveness of the system that it implements.   The laboratory should also consider improvements in procedures and/or educational efforts as part of its program to reduce the risk of mistransfusion.
 
TRM.40670 ABO Group and Rh(D) Type Verification
The recipient's ABO group and Rh(D) type has been verified by repeat testing of the same sample, a different sample, or agreement with a historical type in the laboratory's records.
NOTE:  Repeat testing of the same sample may be inadequate unless the sample has been drawn using a mechanical barrier system or digital bedside patient identification system. For laboratories that employ computer crossmatching, serologic crossmatch techniques must be employed when ABO typing discrepancies are present (e.g. mixed field reactivity, missing serum reactivity, apparent change in blood type post hematopoietic stem cell transplant).

We're in the midst of moving to Epic. We'll have Beaker for the Lab and WellSky (formerly Mediware HCLL) for the Blood Bank.  Still too early to tell if we'll be happy with HCLL since we're still building it and will Go-Live in April 2020. 

AABB does not require or recommend that you give group O washed blood to neonates < 4 months of age.  I work at an AABB accredited children's hospital (so lots of neonate transfusions) and we have never used washed blood for them during the 28 years I've worked here. 
31st Edition of AABB Standards for Blood Banks and Transfusion Services:
5.17.2  If a non-group-O neonate is to receive non-group-O Red Blood Cells that are not compatible with the maternal ABO group, the neonate's serum or plasma shall be tested for Anti-A or Anti-B.
          5.17.2.1  Test methods shall include an antiglobulin phase using either donor or reagent A1 or B red cells. Std. 5.14.3.4 applies.
          5.17.2.2   If anti-A or anti-B is detected, Red Blood Cells lacking the corresponding ABO antigen shall be transfused.
Our policy is that we routinely give neonates fresh group O Leuko-reduced, Irradiated, CPDA RBCs, but if they must be given something other than group O (example: a directed donor unit), then we require a full AHG crossmatch be done on the unit using baby plasma. 

Someone above commented that a 2nd sample is only required in the U.S. for computer crossmatch (which used to be true). But with the 31st Edition of AABB Standards (effective April 1, 2018), this requirement was moved so that it now applies for all pretransfusion testing for allogeneic transfusions including all types of crossmatching (IS, AHG, and Computer crossmatching). This is more in line with CAP requirements and makes more sense in order to detect possible Wrong Blood In Tube (WBIT) events.
AABB Standards for Blood Banks and Transfusion Services, 31st Edition
5.14.5 Pretransfusion Testing for Allogeneic Transfusion  
There shall be two determinations of the recipient’s ABO group as specified in Standard 5.14.1.  The first determination shall be performed on a current sample, and the second determination by one of the following methods:
Testing a second current sample.
Comparison with previous records.
Retesting the same sample if patient identification was verified using an electronic identification system or another process validated to reduce the risk of misidentification.
Standards 5.11 and 5.27.1 apply.
 
Personal Note: If you intend to retest the same sample (by a different person or the same person), be prepared to show the AABB assessor your validation proving that your "another process" is actually validated to reduce the risk of misidentification (i.e. WBITs). 
 
CAP Checklist Requirements:
TRM.30575 Misidentification Risk
The facility has a system to reduce the risk of mistransfusion for non-emergent red cell transfusions.
NOTE:  Mistransfusion occurs from misidentification of the intended recipient at the time of collection of the pretransfusion testing sample, during laboratory testing and preparation of units to be issued, and at the time of transfusion.  Misidentification at sample collection occurs approximately once in every 1,000 samples, and in one in every 12,000 transfusions the recipient receives a unit not intended for or not properly selected for him/her.  The laboratory is expected to have implemented a plan to reduce these risks through implementation of a risk-reduction system.  Among options that might be considered are:  (1) Verifying the ABO group of the intended recipient on a second sample collected at a separate phlebotomy (including the recording of the result in the institution's historical record); (2) Utilizing a mechanical barrier system or an electronic identification verification system that ensures that the patient from whom the pretransfusion specimen was collected is the same patient who is about to be transfused.  Other approaches capable of reducing the risk of mistransfusion may be used.  The laboratory should participate in monitoring the effectiveness of the system that it implements.   The laboratory should also consider improvements in procedures and/or educational efforts as part of its program to reduce the risk of mistransfusion.
 
TRM.40670 ABO Group and Rh(D) Type Verification
The recipient's ABO group and Rh(D) type has been verified by repeat testing of the same sample, a different sample, or agreement with a historical type in the laboratory's records.
NOTE:  Repeat testing of the same sample may be inadequate unless the sample has been drawn using a mechanical barrier system or digital bedside patient identification system. For laboratories that employ computer crossmatching, serologic crossmatch techniques must be employed when ABO typing discrepancies are present (e.g. mixed field reactivity, missing serum reactivity, apparent change in blood type post hematopoietic stem cell transplant).

Someone above commented that a 2nd sample is only required in the U.S. for computer crossmatch (which used to be true). But with the 31st Edition of AABB Standards (effective April 1, 2018), this requirement was moved so that it now applies for all pretransfusion testing for allogeneic transfusions including all types of crossmatching (IS, AHG, and Computer crossmatching). This is more in line with CAP requirements and makes more sense in order to detect possible Wrong Blood In Tube (WBIT) events.
AABB Standards for Blood Banks and Transfusion Services, 31st Edition
5.14.5 Pretransfusion Testing for Allogeneic Transfusion  
There shall be two determinations of the recipient’s ABO group as specified in Standard 5.14.1.  The first determination shall be performed on a current sample, and the second determination by one of the following methods:
Testing a second current sample.
Comparison with previous records.
Retesting the same sample if patient identification was verified using an electronic identification system or another process validated to reduce the risk of misidentification.
Standards 5.11 and 5.27.1 apply.
 
Personal Note: If you intend to retest the same sample (by a different person or the same person), be prepared to show the AABB assessor your validation proving that your "another process" is actually validated to reduce the risk of misidentification (i.e. WBITs). 
 
CAP Checklist Requirements:
TRM.30575 Misidentification Risk
The facility has a system to reduce the risk of mistransfusion for non-emergent red cell transfusions.
NOTE:  Mistransfusion occurs from misidentification of the intended recipient at the time of collection of the pretransfusion testing sample, during laboratory testing and preparation of units to be issued, and at the time of transfusion.  Misidentification at sample collection occurs approximately once in every 1,000 samples, and in one in every 12,000 transfusions the recipient receives a unit not intended for or not properly selected for him/her.  The laboratory is expected to have implemented a plan to reduce these risks through implementation of a risk-reduction system.  Among options that might be considered are:  (1) Verifying the ABO group of the intended recipient on a second sample collected at a separate phlebotomy (including the recording of the result in the institution's historical record); (2) Utilizing a mechanical barrier system or an electronic identification verification system that ensures that the patient from whom the pretransfusion specimen was collected is the same patient who is about to be transfused.  Other approaches capable of reducing the risk of mistransfusion may be used.  The laboratory should participate in monitoring the effectiveness of the system that it implements.   The laboratory should also consider improvements in procedures and/or educational efforts as part of its program to reduce the risk of mistransfusion.
 
TRM.40670 ABO Group and Rh(D) Type Verification
The recipient's ABO group and Rh(D) type has been verified by repeat testing of the same sample, a different sample, or agreement with a historical type in the laboratory's records.
NOTE:  Repeat testing of the same sample may be inadequate unless the sample has been drawn using a mechanical barrier system or digital bedside patient identification system. For laboratories that employ computer crossmatching, serologic crossmatch techniques must be employed when ABO typing discrepancies are present (e.g. mixed field reactivity, missing serum reactivity, apparent change in blood type post hematopoietic stem cell transplant).

Someone above commented that a 2nd sample is only required in the U.S. for computer crossmatch (which used to be true). But with the 31st Edition of AABB Standards (effective April 1, 2018), this requirement was moved so that it now applies for all pretransfusion testing for allogeneic transfusions including all types of crossmatching (IS, AHG, and Computer crossmatching). This is more in line with CAP requirements and makes more sense in order to detect possible Wrong Blood In Tube (WBIT) events.
AABB Standards for Blood Banks and Transfusion Services, 31st Edition
5.14.5 Pretransfusion Testing for Allogeneic Transfusion  
There shall be two determinations of the recipient’s ABO group as specified in Standard 5.14.1.  The first determination shall be performed on a current sample, and the second determination by one of the following methods:
Testing a second current sample.
Comparison with previous records.
Retesting the same sample if patient identification was verified using an electronic identification system or another process validated to reduce the risk of misidentification.
Standards 5.11 and 5.27.1 apply.
 
Personal Note: If you intend to retest the same sample (by a different person or the same person), be prepared to show the AABB assessor your validation proving that your "another process" is actually validated to reduce the risk of misidentification (i.e. WBITs). 
 
CAP Checklist Requirements:
TRM.30575 Misidentification Risk
The facility has a system to reduce the risk of mistransfusion for non-emergent red cell transfusions.
NOTE:  Mistransfusion occurs from misidentification of the intended recipient at the time of collection of the pretransfusion testing sample, during laboratory testing and preparation of units to be issued, and at the time of transfusion.  Misidentification at sample collection occurs approximately once in every 1,000 samples, and in one in every 12,000 transfusions the recipient receives a unit not intended for or not properly selected for him/her.  The laboratory is expected to have implemented a plan to reduce these risks through implementation of a risk-reduction system.  Among options that might be considered are:  (1) Verifying the ABO group of the intended recipient on a second sample collected at a separate phlebotomy (including the recording of the result in the institution's historical record); (2) Utilizing a mechanical barrier system or an electronic identification verification system that ensures that the patient from whom the pretransfusion specimen was collected is the same patient who is about to be transfused.  Other approaches capable of reducing the risk of mistransfusion may be used.  The laboratory should participate in monitoring the effectiveness of the system that it implements.   The laboratory should also consider improvements in procedures and/or educational efforts as part of its program to reduce the risk of mistransfusion.
 
TRM.40670 ABO Group and Rh(D) Type Verification
The recipient's ABO group and Rh(D) type has been verified by repeat testing of the same sample, a different sample, or agreement with a historical type in the laboratory's records.
NOTE:  Repeat testing of the same sample may be inadequate unless the sample has been drawn using a mechanical barrier system or digital bedside patient identification system. For laboratories that employ computer crossmatching, serologic crossmatch techniques must be employed when ABO typing discrepancies are present (e.g. mixed field reactivity, missing serum reactivity, apparent change in blood type post hematopoietic stem cell transplant).

Someone above commented that a 2nd sample is only required in the U.S. for computer crossmatch (which used to be true). But with the 31st Edition of AABB Standards (effective April 1, 2018), this requirement was moved so that it now applies for all pretransfusion testing for allogeneic transfusions including all types of crossmatching (IS, AHG, and Computer crossmatching). This is more in line with CAP requirements and makes more sense in order to detect possible Wrong Blood In Tube (WBIT) events.
AABB Standards for Blood Banks and Transfusion Services, 31st Edition
5.14.5 Pretransfusion Testing for Allogeneic Transfusion  
There shall be two determinations of the recipient’s ABO group as specified in Standard 5.14.1.  The first determination shall be performed on a current sample, and the second determination by one of the following methods:
Testing a second current sample.
Comparison with previous records.
Retesting the same sample if patient identification was verified using an electronic identification system or another process validated to reduce the risk of misidentification.
Standards 5.11 and 5.27.1 apply.
 
Personal Note: If you intend to retest the same sample (by a different person or the same person), be prepared to show the AABB assessor your validation proving that your "another process" is actually validated to reduce the risk of misidentification (i.e. WBITs). 
 
CAP Checklist Requirements:
TRM.30575 Misidentification Risk
The facility has a system to reduce the risk of mistransfusion for non-emergent red cell transfusions.
NOTE:  Mistransfusion occurs from misidentification of the intended recipient at the time of collection of the pretransfusion testing sample, during laboratory testing and preparation of units to be issued, and at the time of transfusion.  Misidentification at sample collection occurs approximately once in every 1,000 samples, and in one in every 12,000 transfusions the recipient receives a unit not intended for or not properly selected for him/her.  The laboratory is expected to have implemented a plan to reduce these risks through implementation of a risk-reduction system.  Among options that might be considered are:  (1) Verifying the ABO group of the intended recipient on a second sample collected at a separate phlebotomy (including the recording of the result in the institution's historical record); (2) Utilizing a mechanical barrier system or an electronic identification verification system that ensures that the patient from whom the pretransfusion specimen was collected is the same patient who is about to be transfused.  Other approaches capable of reducing the risk of mistransfusion may be used.  The laboratory should participate in monitoring the effectiveness of the system that it implements.   The laboratory should also consider improvements in procedures and/or educational efforts as part of its program to reduce the risk of mistransfusion.
 
TRM.40670 ABO Group and Rh(D) Type Verification
The recipient's ABO group and Rh(D) type has been verified by repeat testing of the same sample, a different sample, or agreement with a historical type in the laboratory's records.
NOTE:  Repeat testing of the same sample may be inadequate unless the sample has been drawn using a mechanical barrier system or digital bedside patient identification system. For laboratories that employ computer crossmatching, serologic crossmatch techniques must be employed when ABO typing discrepancies are present (e.g. mixed field reactivity, missing serum reactivity, apparent change in blood type post hematopoietic stem cell transplant).

We designed and ordered special labels that we attach to our psoralen platelets.  The ISBT codes for our products are E8331, E8332, E8333, E8334, E8335.
Psoralen-Treated Platelets

Place this label on all Psoralen-treated (Pathogen Reduced) INTERCEPT Platelets to remind the care provider that they are equivalent to IRRADIATED and to store at room temp.

This is the way Meditech documents the splitting of products. Now with ISBT, the product code does not change when you split, so this is how the computer differentiates one split from another made from the same product.  The split number (A, B, Ba, whatever) needs to print on the tag so you can tag the product properly and the blood bank and bedside staff know which split is being issued and transfused.  I've had Meditech 27 years, we're a pediatric hospital so we make tons of aliquots, and no inspector (CAP, AABB, FDA, JC) has ever had a problem with this.  Your CAP inspector should not cite you for something they don't understand. The previous post is correct - adding an A or B to the DIN when splitting has nothing to do with a closed or open system.  I would challenge that citation.  

This is the way Meditech documents the splitting of products. Now with ISBT, the product code does not change when you split, so this is how the computer differentiates one split from another made from the same product.  The split number (A, B, Ba, whatever) needs to print on the tag so you can tag the product properly and the blood bank and bedside staff know which split is being issued and transfused.  I've had Meditech 27 years, we're a pediatric hospital so we make tons of aliquots, and no inspector (CAP, AABB, FDA, JC) has ever had a problem with this.  Your CAP inspector should not cite you for something they don't understand. The previous post is correct - adding an A or B to the DIN when splitting has nothing to do with a closed or open system.  I would challenge that citation.  

At my hospital, we split out our Blood Utilization monitoring like this (below) to cover each of these 10 AABB requirements:
1. Ordering Practices: C/T Ratios (Total and by Service), Uncrossmatched Blood Requests, Compliance with MD signature completion on "Release of Uncrossmatched Blood" forms
2. Patient Identification: Any Patient ID issues identified are reported
3. Sample Collection and Labeling: BBK Specimen acceptability report (outliers are detailed)
4. Infectious and Noninfectious Adverse Events, Incidence of Mistransfusion: Transfusion Reactions, Recalls, Notifications
5. Near-Miss Events: Anything identified would be detailed along with root-cause, corrective action
6. Usage and Discard: Units transfused, Units discarded
7. Appropriateness of Use (for all Components):
QA Audits performed by Medical Director, report pulls 100% products transfused (one type of product per month) and the most recent related lab data to when the product was issued. Outliers are investigated by the Medical Director. MDs receive follow-up letters on questionable transfusions.
Jan/Apr/Jul/Oct = RBC compared to most recent Hgb/Hct
Feb/May/Aug/Nov = PLT compared to most recent PLT count
Mar/Jun/Sep/Dec = FFP and CRYO compared to most recent PT/PTT and Fibrinogen
We also review any Autologous transfusions, Peri-Op Collection/Reinfusion (Cell Saver, Hemodilution, ECMO) and Therapeutic Apheresis and Phlebotomies performed during the quarter.
8. Blood Administration Policies and Practices: Our QA department performs random audits on 75 transfusions per quarter (coordinated with the RBC, PLT, and FFP/CRYO schedule above) and audits for MD Order to Transfuse, Issue Checks Performed, Bedside Checks Performed, Vitals Recorded. Since we are using TAR, all transfusion data is now entered into the computer so this audit can be done using computer reports. We also audit 100% of all transfusions for "Duration Less than 4 Hours" by running a computer report. We also report on Bedside Blood Administration Audits which are performed by the blood bank techs (minimum of 16 per year) and Blood Warmer QC which is performed quarterly by Biomed.
9. Ability of Services to Meet Patient Needs: We do Turnaround Time studies (Type & Screen, DAT, Crossmatch, Newborn Workup), report on any Customer Complaints and Cancelled Treatments. We also have a Bloodless Medicine and Surgery program so we report on the number of patients enrolled in the program who were treated without transfusion.
10. Compliance with Peer Review Recommendations: Letters Written/Sent to Phsyicians from the Medical Director and/or Blood Utilization Committee Chair, and Responses Received
I know, some of these do overlap somewhat, but "Ordering Practices" would be looking at how many and what kinds of products are doctors routinely ordering for various things (i.e. are they over-ordering or under-ordering for certain types of surgeries/procedures, do they order uncrossmatched too often because lack of ordering in advance) and "Appropriateness of Use" would be looking at whether they make the decision to actually transfuse those ordered products to their patients based on appropriate criteria (lab results, patient symptoms, etc.).

We prepare small aliquots into ISBT labeled and tagged bags, and nurses pull the product into a syringe at the bedside using a small Y-Set (Baxter 4C2223, 13 inches) with an in-line filter. The labeled/tagged bag is NEVER disconnected from the syringe throughout the transfusion. The syringe simply becomes a part of the line from the bag to the patient. We've been doing it this way for years and it works very well.

I've attached the form I came up with that we use.
Here is the section in my procedure that addresses method correlation:
CAP Checklist: COM.04250

If the laboratory uses more than one nonwaived instrument/method to test for a given analyte, the instruments/methods checked against each other at least twice a year for comparability of results.


 
NOTE: This requirement applies to tests performed on the same or different instrument makes/models or by different methods. This comparison must include all nonwaived instruments/methods. The laboratory must establish a protocol for this check that includes acceptance criteria. Quality control data may be used for this comparison for tests performed on the same instrument platform, with control materials of the same manufacturer and lot number. Otherwise, the use of human samples (whole blood, serum, plasma, urine, etc.) rather than stabilized commercial controls, is preferred to avoid potential matrix effects. In cases when availability or pre-analytical stability of patient/client specimens is a limiting factor, alternative protocols based on QC or reference materials may be necessary but the materials used should be validated (when applicable) to have the same response as fresh human samples for the instruments/methods involved.


 
Method Correlation is performed twice a year, comparing Echo1 vs. Echo2 vs. Manual Capture vs. Tube methods

1.       ABO/Rh – No less than 3 specimens are compared, each with different blood types, at least one should be Rh negative

2.       Antibody Screen – No less than 3 specimens are compared, at least one should be positive

3.       Antigen Typing – comparing tube to Echo, no less than 3 specimens

4.       Antibody ID – No less than 1 positive specimen is compared

Interpretation/ Acceptance Criteria:

·         Manual and Automated Capture methods are expected to correlate closely.

·         Echo1 vs. Echo2 results are expected to correlate (match) closely.

·         Capture vs. Tube methods are expected to show some variability between reactions due to the differences in the nature of the testing systems and enhancements.

·         Corrective action must be taken and documented when criteria are not met.
 
Hope this helps!       Sheri
Q0530F03 Method Correlation 8-12.doc

Here is a copy of our Level 2 Annual Blood Bank Competency form.  We also have a Level 1 (for just label checking and issuing blood), and a Level 3 for advanced Blood Bank competencies.
Q0230F06 Level2BBKComp 11-11.doc

Here is a copy of our Level 2 Annual Blood Bank Competency form.  We also have a Level 1 (for just label checking and issuing blood), and a Level 3 for advanced Blood Bank competencies.
Q0230F06 Level2BBKComp 11-11.doc

Here is a copy of our Level 2 Annual Blood Bank Competency form.  We also have a Level 1 (for just label checking and issuing blood), and a Level 3 for advanced Blood Bank competencies.
Q0230F06 Level2BBKComp 11-11.doc

AABB Std 5.25.5 requires "The records shall contain a signed statement from the requesting physician indicating that the clinical situation was sufficiently urgent to require release of blood before completion of compatibility testing or infectious disease testing".  
 
Notice that this requirement is not hinged on whether the patient actually gets transfused with the emergency released blood.  The fact that blood was requested prior to completion of testing requires the physician's signed statement, regardless of whether they end up transfusing it or not.   So yes, you need to keep the emergency release form with the physician signature.  If the form is returned with no signature, you should also chase the doctor down and get that signature, even if the blood was not given.   It can be a hassle, but those are the standards.

AABB Std 5.25.5 requires "The records shall contain a signed statement from the requesting physician indicating that the clinical situation was sufficiently urgent to require release of blood before completion of compatibility testing or infectious disease testing".  
 
Notice that this requirement is not hinged on whether the patient actually gets transfused with the emergency released blood.  The fact that blood was requested prior to completion of testing requires the physician's signed statement, regardless of whether they end up transfusing it or not.   So yes, you need to keep the emergency release form with the physician signature.  If the form is returned with no signature, you should also chase the doctor down and get that signature, even if the blood was not given.   It can be a hassle, but those are the standards.

AABB Std 5.25.5 requires "The records shall contain a signed statement from the requesting physician indicating that the clinical situation was sufficiently urgent to require release of blood before completion of compatibility testing or infectious disease testing".  
 
Notice that this requirement is not hinged on whether the patient actually gets transfused with the emergency released blood.  The fact that blood was requested prior to completion of testing requires the physician's signed statement, regardless of whether they end up transfusing it or not.   So yes, you need to keep the emergency release form with the physician signature.  If the form is returned with no signature, you should also chase the doctor down and get that signature, even if the blood was not given.   It can be a hassle, but those are the standards.