14.1 Antibody Identification and Rule-Out

14.1 Antibody Identification and Rule-Out

Quick Answer: An antibody identification panel is 10–11 group O reagent red cells with a known antigen profile, tested against patient plasma at the same phases as the screen. Rule-out uses nonreactive cells: if a cell is negative and homozygous for an antigen (especially Kidd, Duffy, Rh, MNS), that antibody is excluded. Phenotype the patient only if not transfused in the last 3 months (or use reticulocyte separation / genotype). A negative autocontrol supports alloantibody; a positive autocontrol means autoantibody, recent transfusion, or drug. Confirm with the rule of three and selected cells.

The June 9, 2026 BB outline parks antibody identification at IV.A.3. The exam does not want you to memorize one manufacturer's panel map. It wants you to cross out antigens on nonreactive homozygous cells, know when a heterozygous cell is not a legal rule-out, and know when the autocontrol wrecks the story.

What the panel actually is

The antibody screen is two or three group O cells chosen to carry the common clinically significant antigens. A positive screen (or an incompatible AHG crossmatch) sends you to a panel. Commercial identification panels are 10 or 11 group O cells. They are group O so anti-A and anti-B in the patient's plasma cannot agglutinate the reagent cells and fake an alloantibody. Each cell has a published antigen profile: Rh (D, C, c, E, e), Kell (K, k), Duffy (Fya, Fyb), Kidd (Jka, Jkb), MNS (M, N, S, s), Lewis, P1, Lutheran, and often Xga, Kpa, Jsa, and a few low-frequency antigens.

Run the panel at the same phases and enhancement that made the screen positive — typically immediate-spin, 37 °C, and AHG in LISS or PEG, or the gel/solid-phase method you used on the screen. Record strength, not just plus/minus. Strength is data: dosage shows up as 3+ on a double-dose cell and 1+ on a single-dose cell of the same antigen. A panel that is “all 1+” is a different problem from a panel that is 3+ / 1+ / 0.

Include an autocontrol (patient cells + patient plasma) with every panel. The panel without an AC is an unfinished test.

Phenotype the patient — but only if the cells are really the patient's

If the patient has not been transfused in the last 3 months, type the patient's own red cells for the common antigens. A person who made anti-Jka should type Jk(a−). A person who types Jk(a+) did not make anti-Jka unless you are looking at donor cells left over from a recent transfusion, a partial antigen, or a clerical disaster. Matching the phenotype to the suspected antibody is part of identification, not an optional extra.

Recently transfused (within 3 months) invalidates a routine phenotype. Circulating donor red cells produce mixed-field typing and can make a Jk(a−) patient look Jk(a+). Options the exam expects you to name:

• Use a pretransfusion specimen if one still exists.
• Reticulocyte separation (density or microhematocrit harvest of young patient cells) if the last transfusion was days to a few weeks ago and the patient is making retics.
• Molecular genotyping on white cells or buccal DNA — donor red cells do not change the patient's genotype.
• Wait 3 months if the workup is not urgent.

Do not phenotype through a positive DAT with a high-protein or AHG reagent. IgG-coated cells can false-positive some typings. Strip the IgG (chloroquine or EGA) or use IgM monoclonal reagents that agglutinate without AHG.

Rule-out: nonreactive cells do the work

Rule-out (exclusion) is the engine. You do not identify an antibody by staring at the reactive cells first. You look at every cell that is nonreactive at the phase of interest and cross out every antigen present on that cell — with one exam-critical caveat.

For antigens that show dosage — Kidd, Duffy, Rh (especially C/c and E/e), and MNS — rule out only with a homozygous (double-dose) cell. A Jk(a+b+) cell that fails to react does not exclude anti-Jka. Anti-Jka can miss a single-dose cell and still destroy a Jk(a+b−) unit. Same logic for anti-Fya on an Fy(a+b+) cell, anti-c on a C+c+ cell, anti-E on an E+e+ cell, and anti-M or anti-S on heterozygous MNS cells.

K is the usual exception. K+k− (homozygous K) cells are rare (~0.2%). Anti-K does not show clinically important dosage, so a single K+k+ nonreactive cell is accepted as a K rule-out in most SOPs.

Work the nonreactive cells first. Then look at what is left standing. If every remaining possibility is one antigen, and the reactive cells are that antigen-positive, you have a candidate — not a finished identification.

Worked mini-panel: anti-Jka with dosage

Patient: group A, D+, screen 2 of 3 cells 2+ at AHG, autocontrol 0, last transfusion 2 years ago. Phenotype on this specimen: C+c+ E−e+ K− Fy(a+b+) Jk(a−b+) S+s+.

Read the negatives first. Cells 3, 4, and 7 are 0 at AHG. Those cells are Jk(a−). They are Jk(b+) homozygous, c+, and together they carry double-dose e, Fyb, and S (cell 7 is S+s−) plus Fya (cell 4 is Fy(a+b−)). You can exclude anti-c, anti-E, anti-e, anti-K, anti-Fya, anti-Fyb, anti-Jkb, anti-S, and anti-s using homozygous (or, for K, accepted heterozygous) cells. Cell 3 is R2R2 (D+C−c+E+e−) and nonreactive, so anti-D and anti-E are out. C is not present as a double-dose antigen on any nonreactive cell — you still need a selected C+c−, Jk(a−) cell to finish that rule-out.

Now the positives. Every Jk(a+b−) cell is 3+. Every Jk(a+b+) cell is 1+. That is dosage. The pattern is anti-Jka, not anti-E (cell 3 is E+E+ and 0) and not anti-K (cells 3 and 7 are K+ and 0). The patient is Jk(a−), which fits. Issue Jk(a−) red cells, AHG-crossmatch compatible.

If you had treated cell 2 (Jk(a+b+), 1+) as if a weak reaction were a negative, you would have wrongly excluded anti-Jka. Dosage is why the homozygous rule exists. Kidd antibodies are also the ones that fade below detectability and come back as delayed hemolytic reactions — a heterozygous screen cell that is only barely positive today may be negative next month.

Autocontrol before you name anything

The autocontrol is patient cells plus patient plasma, same phases as the panel.

• AC negative, panel has a discrete pattern → alloantibody. Proceed with rule-out.
• AC positive at AHG, DAT usually positive → warm autoantibody, drug antibody, or a recently transfused patient whose plasma is coating circulating donor cells (delayed serologic or hemolytic reaction). You cannot call the panel “anti-e” until you know whose cells are reacting.
• AC positive at IS/RT only → cold autoantibody or rouleaux. Check the AHG phase after proper washing; do not let an IS auto invent an alloantibody.

A positive AC plus all panel cells positive is not an identification. It is a problem set (Section 14.3).

Selected cells and the rule of three

The first panel rarely finishes every rule-out. Selected cells are extra reagent cells you pull from another lot, a frozen rare-cell inventory, or a second panel to answer one question: “Is there still anti-C?” or “Do I have anti-E plus anti-K?” Choose cells that are negative for the antibody you already think you have and homozygous for the antigen you still need to exclude. In the mini-panel above, the missing piece is a C+c− Jk(a−) cell. If that cell is 0, anti-C is out. If it is 2+, you have a second antibody.

Rule of three (identification confidence): the plasma should react with at least three antigen-positive cells and fail to react with at least three antigen-negative cells. That distribution gives p < 0.05 that the pattern is chance. Two reactive K+ cells and a sea of K− negatives is not enough to lock anti-K on the exam. If the panel only gave you two Jk(a+b−) cells, pick a third selected Jk(a+) cell.

Identification is not complete until:

1. All common clinically significant antibodies are ruled out (homozygous where required).
2. The rule of three is met.
3. The patient phenotype (or genotype) is antigen-negative for that specificity, unless recently transfused.
4. The autocontrol fits an alloantibody story.

Exam traps

• Ruling out anti-Jka with a Jk(a+b+) nonreactive cell. That cell is not a legal rule-out.
• Phenotyping a patient transfused last week and believing the mixed-field Jk(a+) result.
• Naming an antibody from the reactive cells without crossing out on the nonreactive ones.
• Ignoring a positive autocontrol and calling the panagglutinin anti-e.
• Declaring identification complete with two positive cells. You need three and three.
• Using a group A or B “panel” cell. Identification panels are group O so ABO isoagglutinins stay out of the way.