A colleague from Texas wants to know if institutions routinely perform ABO/Rh testing at least twice as part of their pre-transfusion testing scheme, before attributing an ABO and Rhesus status to a patient. If an institution performs ABO/Rh testing at least twice as part of pre-transfusion testing, is the second typing accomplished by another technologist or by the same technologist using a second specimen? What other approaches are taken?

In reply to the above, the following responses were submitted.

1. In the experience of a Los Angeles physician who once worked as a blood bank technologist, her laboratory required each pre-transfusion sample to be tested twice for ABO/Rh, with the second testing done by a different technologist. In addition, if the patient had no ABO/Rh result in the laboratory computer, a second (separate) specimen was required to be tested for ABO/Rh typing, before releasing non-group O red blood cell units.

2. Another California blood banker reports that at her University Medical Center a second or "confirmation" blood type is required when there is no previous ABO/Rh on file. They confirm the forward type on all samples for which ABO/Rh testing is ordered, with the exception of cord blood specimens. The confirmation type is done on a new cell suspension and all patient ID is carefully rechecked. Only one patient confirmation type is done at a time (batch testing is not permitted) and they do not list more than one patient on the computer screen at a time as the results are entered. The second typing may be done by the same technologist who performed the initial type. The reason for this is that there is only 1 technologist working in the Blood Bank at night.

3. A hospital system in San Diego that includes four hospital transfusion services implemented a policy in response to last year's sentinel alert from JCAHO in which it was recommended NOT TO BATCH samples when performing crossmatching or ABO/Rh testing. This recommendation was unrealistic in a busy transfusion service, but as a compromise measure they implemented a policy that stipulated that a "first time" blood type for the purposes of compatibility testing would never be performed in a batch mode with other samples. If the patient had no historic blood type on file, the sample would be tested for ABO and Rh and reported independently of all other samples. If the patient had a historical blood type on file, then the blood type could be "batched" with other samples. The San Diego group is interested in hearing what other sites are doing.

4. Another Southern California physician reported that the AABB Standards require ABO grouping routine to be done using FOUR reagents (anti-A, anti-B, A1 cells and B cells) when determining a patient's ABO as part of pretransfusion testing. According to AABB Standards, 21st edition 5.12.1 "The ABO group shall be determined by testing the red cells with anti-A and anti-B reagents and by testing the serum or plasma for expected antibodies with A1 and B red cells. If a discrepancy is detected and transfusion is necessary before resolution, only group O Red Blood Cells shall be issued." In other words, both a forward and a back typing must be done, and the results must be compared for concordance. However, in the California physician's opinion, the AABB Standards do not go far enough to protect patients from ABO-incompatible transfusion errors, because minimum compliance with the AABB Standards only require that a patient's ABO grouping routine be done once, even if the sample tested is the first ever received in the laboratory for that patient. He suggests that it should be required to routinely validate the ABO grouping result of every new patient by testing a separately collected second sample. This approach is done at many institutions. Currently every donor unit of RBCs is required to be tested at least twice for its ABO to assure the unit is accurately labeled with the correct ABO. Why should we accept a lesser degree of accuracy when attributing ABO grouping results to patients?

5. A transfusion medicine physician in Ohio reports that at her facility they perform a repeat ABO typing on the same sample with the testing done by two different technologists. The Ohio physician admits that this approach will not find potential phlebotomy errors resulting in the wrong blood in the tube. For those institutions that do confirmations and require a second separate sample, she would be most interested in learning how this orchestrated.

ADDENDA Dec. 6, 2002

6. A Transfusion Service Technical Manager in Los Angeles reports that at his institution they perform two ABO/Rh tests on patients without a previous typing record in their facility. The tests can be performed on the same sample, by the same tech, as long as there is separation between the tests. That is, the tests are performed independently and the cell suspension used for the first test cannot be used for the second test. Parenthetically, they do not perform the second type whenever the patient has a documented previous type, including receipt (and retesting) of autologous blood.

ADDENDA Dec. 9, 2002

7. A transfusion medicine physician (currently practicing in Saudi Arabia) reports that he was "raised" in Denmark, where it was required that pre-transfusion testing include an initial sample for ABO/Rh typing, and that a second sample must be obtained by a different phlebotomist at a different time to be ABO/Rh typed. The results from the first sample are compared to the results from the second sample (to be used for wet crossmatch in case a electronic crossmatch is not sufficient). This must be in place prior to issue of blood for transfusion. In emergency situations, and in cases when it is technically impossible to obtain two independent samples in a timely fashion, only group O Rh Neg RBCs are issued until the second sample has been obtained and tested.

(Editor's query: I wonder if other e-Network forum members from Denmark might comment on current practice in their country regarding this issue.)

8. A blood banker with many years of experience in the field reports that if you look at the processes that should be controlled by requiring repeat testing of samples there are two:

• Technical or testing errors:
  • Testing the SAME sample a second time will control if the initial testing was performed correctly from scratch (ie. re-check sample ID, prepare new cell suspension, etc).
  • Testing the SAME sample 'blind' by another technologist adds another level of security.
• Sample collection errors.
  
  This can only be controlled by collection of ANOTHER specimen at a different time preferably by another person. This could also be fulfilled by a comparing a current result with a previous test result found in the records check.

She concludes that if greatest patient safety is the goal of repeat testing, it would seem that repeat testing from a DIFFERENT sample is a better choice than retesting the SAME sample over again.

ADDENDA Dec. 10, 2002

9. A transfusion medicine physician reports that at a 350-bed hospital in Los Angeles, they instituted a 'double check' system about a year and a half ago that requires confirmation of ABO group on all patients on whom they do not have an historical record of blood type, before non-group O RBC products may be released. The specimen must be from a different draw (staffing issues precluded the requirement that the phlebotomist be different, but it often is), but may also be obtained from an appropriate specimen from another area of the lab (e.g. a CBC drawn the day before), provided the sample meets blood bank criteria for proper labeling (this is convenient for the lab, and lessens second sticks on the patient). They do not require a double-check on group O patients, since if they are not group O, there is no harm in an RBC transfusion; this also helps keep the number of double-check draws down. (Editor's NOTE: While it is true that no harm is expected when inadvertently transfusing group O RBCs to a patient who is not group O, that might not be the case when inadvertently transfusing group O platelets or FFP to a patient who is not group O). The physician concludes that in emergency situations where there is not time for a double check, group O RBCs are issued pending completion of a double check.

ADDENDA Dec. 12, 2002

10. A blood banker in Ohio is finding this discussion very useful. She comments, however, that it would be extremely helpful, if possible, for the facilities that have instituted the ABO checking system by requiring a second sample (particularly for non-group O patients) prior to releasing non-O blood, to share their logistics or protocols - how is this orchestrated so that it works smoothly and efficiently, and minimizes the amount of additional group O blood dispensed?

11. A blood banker in Southern California reports that his hospital policy requires a second type to be performed on a sample drawn at a different time than the crossmatch or type and screen sample. Exceptions to this policy are:

• a historical type is on file
• the initial sample types group O
• patient has donated an autologous unit and the ABO/Rh of the autologous unit is in the blood bank record
• crossmatch sample was drawn in the operating room
  

Pre-op outpatients have their confirmatory type drawn on the day of surgery. The Transfusion Service sends a form each day to the Surgical Admit unit listing which patients need to be drawn. In many cases, there is already a sample in the laboratory which can be used (CBC, Chemistry sample, etc.) negating the need to stick the patient for a confirmatory sample. In the cases where a redraw is required, a venipuncture tech is dispatched to the patient location to collect a sample. Without exception patients have been very receptive to a second venipuncture when it is explained to them that it is for their safety. If a second sample cannot be obtained prior to the need for blood, group O red cells are issued. Confirming the blood type on a second sample has the obvious advantage detecting both testing and labeling errors. Since implementing this policy in 1996, the California hospital has detected 13 cases of incorrect ABO typing. All 13 of these errors were due to mislabeled blood samples. In 6 of the 13 cases, transfusion based on the initial blood type would have resulted in a acute hemolytic transfusion reaction due to major ABO incompatibility. Implementing the requirement for a confirmatory blood type on a second sample clearly contributes to patient safety at the reporting institution and has had a better yield than many of the other recipient safety measures that are required by regulation, e.g. NAT.

 12. A blood banker who inspects other blood banks recently performed an audit at a Southern California Transfusion Service and experienced the exact situation as described in reply #4 above. The responding blood banker agrees that repeat ABO testing is not mandated by AABB Accreditation and Standards for patients who may be first time recipients, but yet this testing is stated for all donors. The responding blood banker supports a comparable requirement for patients as well as donors. He mentioned his opinion to the transfusion service supervisor and her exact reply was, "It is not required by Standards."

ADDENDA Dec. 20, 2002

 13. A colleague in Switzerland reports that according Swiss guidelines the ABO and Rh typing must be determined on two different samples before transfusion, unless the patient has a blood type on file. In addition some hospitals carry out a bedside test immediately before transfusion. The bedside test shows only major ABO incompatibilities. The Swiss colleague adds that patients in Swiss hospitals very often do not wear an identification wristband. Usually their ID card is attached to their hospital bed. Some but not all hospitals have implemented ID wristbands .

Please submit comments to the e-Network Forum.

Ira A. Shulman, MD
CBBS e-Network Forum Editor & Moderator

W. Tait Stevens, MD
CBBS e-Network Forum Assistant Editor & Moderator