Evaluation and Interpretation

Not every reaction in the tube matters to the patient. This area is about deciding which serological findings could genuinely shorten the survival of transfused red cells and which are laboratory noise. A technologist weighs the phase, thermal range and strength of reactivity, checks the autocontrol, and considers the antibody's known behaviour. Lewis, M and P1 antibodies reacting only below 37 °C rarely justify antigen-negative units, whereas Kidd, Duffy, Kell and Rh specificities do, even when the current reaction is weak or the antibody has since become undetectable. That judgement drives unit selection and how the finding is recorded on the patient's file.

Some results are not merely abnormal, they are impossible. A group that contradicts the patient's historical record, a forward and reverse type that disagree, a panel pattern no single specificity explains, or an incompatible crossmatch against a negative screen all signal that something outside the patient's biology has gone wrong. The technologist stops, decides whether the fault sits in the sample, the reagent, the technique or the patient, and resolves it before anything is reported. Collecting a second, independently drawn sample is often the fastest way to separate a labelling error from a real serological puzzle.

When a patient reacts during or after a transfusion, the laboratory owns a defined part of the investigation. The opening steps are the same regardless of cause: confirm the transfusion has been stopped, repeat the clerical check of patient identifiers against the unit, inspect the post-transfusion plasma for the pink or red colour of haemolysis, and perform a direct antiglobulin test on the post specimen with ABO regrouping of both patient and unit. The pattern of findings then separates a febrile non-haemolytic reaction from acute haemolysis, bacterial contamination, circulatory overload, an allergic response or transfusion-related acute lung injury.

This is the analytical core of the blood bank: reading grouping, antibody screening, identification and crossmatch results as one picture and deciding what a patient may safely receive. It covers interpreting reaction grades and phases, applying the rule of three when a specificity is claimed, excluding antigens on non-reactive cells, recognising dosage, resolving mixtures of antibodies, and knowing when an incompatible crossmatch against a negative screen points to an antibody aimed at a low-incidence antigen carried by that donor. The output is always a decision: which units are compatible, which antigen-negative units must be ordered, and whether an electronic crossmatch is permitted at all.

Choosing the right component is as much a laboratory judgement as choosing compatible units. This area covers matching the product to the clinical defect, such as red cells for oxygen-carrying capacity, platelets for thrombocytopenia or platelet dysfunction, frozen plasma for multiple factor deficiencies, and cryoprecipitate when fibrinogen falls below roughly 1.5 g per litre. It also covers the modifications a specific patient needs. Irradiation prevents transfusion-associated graft-versus-host disease in directed donations from relatives, HLA-matched platelets and severely immunocompromised recipients, while washing suits recurrent severe allergic reactions and IgA deficiency. Volume, rate and paediatric aliquoting belong here too.

Blood bank work leans heavily on packaged commercial systems: gel and solid-phase platforms, foetomaternal haemorrhage screens, molecular typing kits, and the humble vial of IgG-sensitised check cells. Each carries its own controls, validity criteria and list of interferences, and a result is only reportable when those internal controls have behaved. The technologist reads the result against the manufacturer's stated interpretation, confirms lot number and expiry, recognises when a control failure invalidates a whole run rather than a single well, and knows which confirmatory test must follow a positive screening result.