Allocation Rules for Platelets and Plasma

Medinfo Hematos IIG software is rules-based so the institution may set its own custom rules for all processes.  One chooses a framework and then adds any additional rules it needs for optimization.  Turnkey systems do not offer this flexibility.

The rules for platelet and plasma components are much simpler than those for RBCs since usually we only consider ABO type.  There are two modes:  regular and emergency, the latter applying if not all the patient testing (including historical checking) is available.  The components, on the other hand, must meet all criteria before being considered for patient use.

Please note that any donor with antibodies is automatically excluded from plasma and platelet production based on our donor testing criteria.

Example rules for plasma follow:

For platelets, note that for adults and anyone else >= 20 kg, I gave any type of platelet pool or plateletpheresis component without regard to ABO matching.  With our production method, I did not give Rh immunoprophylaxis to females of child-bearing age receiving platelets from D-positive donors based on our clean (essentially RBC-free) Reveos automated production process.

For platelets, there were also different allocation rules for regular and emergency mode:

Similarly, allocation rules for granulocytes, etc. can be made and enforced by the software.  Low-B-titer group A universal plasma would also be easy to implement.

Are These Garbage Antibodies?

In my Middle Eastern experience, I have found antibody panels from a certain manufacturer show a lot of weak, nonspecific reactivity which disappears with another manufacturer’s panels, particularly when gamma-heavy-chain-specific reagents are used:

  • In the following example, 4/15 cells show weak or 1+ reactivity at AHG phase only using the manufacturer’s whole molecule IgG antiglobulin reagent—no reactions at enzyme phase.  This reagent detects both heavy and light chains.
  • When the panel from another manufacturer is used with a gamma-heavy-chain-specific reagent, there is no reactivity at either AHG or enzyme phases.

Normally, in my protocols, the first encounter in this situation, I would do extended RBC phenotyping beyond Rh and Kell systems as well.

In my experience, the nonspecific reactions detected are not clinically significant and they may be ignored if the conditions of the following algorithm is used:

  1. If nonspecific, enzyme-labile reactions found in the first manufacturer’s panels and AHG, then repeat with the second manufacturer’s panels using gamma-heavy-chain-specific reagents.
  2. If the second manufacturer’s panels are negative, use the latter for full AHG crossmatch and release crossmatch-compatible components if the panel is negative.

What do you do when you encounter this is in your laboratory?  Please share your thoughts.

Manufacturer 1’s Panel:

Manufacturer 2’s Panel:

Regular and Emergency Mode RBC Antigen Matching

Note:  This is an update of a previous post.

In the previous posts I outlined how Medinfo handled antibody screening and identification.  This post reviews how antigen matching is used based on these results.

There are two modes, regular and emergency.  If the patient has not had at least two ABO/D determinations and/or does not have a recent antibody screen within the past 3 days, then emergency mode must be selected with its own rules.  Otherwise, the regular mode applies.

Regular Mode:

In general, if there is a clinically significant antibody, an RBC unit which has not been matched for the corresponding antigen or has the corresponding antigen cannot be routinely selected.  However there is a hierarchy here also:

  1. Absolutely prohibited release—no one can override the logic (e.g. giving group O to a patient with anti-H)—not even the transfusion medicine physician can override this
  2. Restricted release—only transfusion medicine physician can authorize the release of the incompatible or untested unit (e.g. giving C-positive unit to someone with anti-C)
  3. Least-incompatible for warm autoimmune hemolytic anemia WAIHA:  requires transfusion medicine physician approval
  4. Informational release:  authorized staff may release antigen-incompatible or untested unit but a pop-up menu appears and asks them to accept (e.g. Lewis untested unit in a patient with anti-Lea).
  5. Antigen-specificity matched—the usual mode for patients with antibodies

Examples of Regular Mode rules follow (these are not the complete lists but just provided to show the complexity of the process).


Emergency Mode:

This is much more restricted for selection of ABO/D and other antigen typings.  An example follows:

If low-titer group O whole blood is available, then a specific rule must be added to both regular and emergency mode to allow this to be given to any ABO type except Bombay.

Inter-Depot Transfer, Blood Delivery, Type and Antigen Matching

This is an update of a previous post.

The final components from the component preparation center may be sent to various depots (freestanding location and/or hospital blood banks.  There should be complete traceability for every step (from donor reception, collection, testing, and processing) transport between locations, and finally the exact storage site, which might include which refrigerator/freezer/incubator and even shelf/position number for each component is stored.  The end of that document showed rules for type/antigen matching.

For disaster planning, rapid inventory enumeration by type is very important.  This can be very time-consuming manually.  With our Medinfo Hematos blood bank system, we could quickly get total inventory across the Qatar or by hospital in less than one minute.  We could also quickly find antigen-matched units across the system and reserve it at any one site for another if necessary.

Smart blood bank dispensing refrigerators, as offered by Haemonetics and Angelatoni, may also serve as depots and take the place of a hospital blood bank for some dispensing.  These solutions can also capture vital information about the storage conditions of the components and prevent release if the storage criteria are not met.  They can also interface with blood bank computer systems and use the main system’s logic for the dispensation rules.

Upon receipt at the hospitals from the blood processing center, the forward ABO and D typing must be confirmed.  We used D reagents which detected partial D so we would call such donor units as D-positive.  However, if a patient type reagent insensitive to partial D types is used, it is possible for a unit to be typed as D-negative whereas in the donor center it might be D-positive.  Sometimes, nothing types consistently as D-positive:  all you can say is that with a particular reagent and lot number, there is or isn’t reactivity.

The greatest complexity is for RBCs since potentially so many antigens exist.  Criteria for matching/ignoring certain antigens must be made.  Critically significant antibodies such as the Kell, Duffy, Kidd, and certain Rh (D and c) must be antigen matched.  A robust blood bank computer system can enforce these rules.

For other components, antigen/typing may be less important.  In fact, in most situations, any type of platelets can be given to anyone (except neonates).  Despite the potentially incompatible plasma, there is rarely significant hemolysis.  In fact, if pooling platelets without regard to blood types is done, a platelet transfusion is a common cause of a positive direct antiglobulin test DAT—something that is not clinically significant.  No one died of a positive DAT by itself for this reason.

Specific rules for compatible plasma types are important, but nowadays, low-anti-B-titer group A plasma may be used like universal AB plasma.  The challenge is to be able to perform the ABO titration (specifically anti-B) quickly—titration can be a slow process, even with automated equipment.  A similar situation for low-titer, universal group O whole blood requires both anti-A and anti-B titration (I will return to this topic in a future post).