A Primer on Platelet Refractoriness for the Overworked Fellow
It's two in the morning and the platelet count didn't move. You gave a full adult dose, ABO-compatible, straight off the agitator, room temperature the whole way, and the post-transfusion count looks almost identical to the pre-transfusion count. The reflex is to order more platelets. Sometimes that's right. Often it isn't, and knowing the difference is the entire point of this post.
You cannot adjudicate refractoriness without a post-transfusion count
This sounds obvious written down, and it is routinely skipped anyway. Teams give the platelets, move on to the next task, and never draw the count that would actually tell them whether the transfusion worked. Without a post-transfusion increment, "the patient seems refractory" is a clinical impression, not a diagnosis, and it's an impression you can't act on. If refractoriness is even on your differential, the single most important order you can write is the post-transfusion count itself, timed appropriately, before you order anything else.
What actually counts as refractoriness
A single low increment isn't refractoriness either. It's a data point. The corrected count increment, or CCI, standardizes the response for body size and dose, so you're comparing apples to apples across patients:
CCI = (platelet increment) × (body surface area in m²) / (platelets transfused, ×10¹¹)
The TRAP study, the trial that anchors most of our modern definitions, called a patient refractory when the 1-hour CCI was under 5,000 on two sequential ABO-identical, fresh platelet transfusions [1]. Most clinical references use a slightly more generous cutoff, treating a 1-hour CCI under 7,500 to 10,000 as an inadequate response [2][3]. Either way, the key word is sequential. One bad number after one transfusion tells you almost nothing, and you need that number in the first place. Two bad numbers in a row is when you start believing it.
Timing matters as much as the number itself. A low 1-hour CCI points toward immune-mediated destruction, the antibody-coated platelets are being cleared almost as fast as they go in. A 1-hour count that looks fine, followed by a 24-hour count that's collapsed, points away from immune causes and toward ongoing consumption instead [2].
Non-immune causes come first, and they're the majority
This is the part that gets skipped in the rush to order an antibody workup. Sixty to eighty percent of refractory cases are non-immune [2]. Fever, sepsis, active bleeding, DIC, splenomegaly, and drugs like amphotericin B are all consuming or sequestering platelets faster than transfusion can replace them, and no amount of specially selected product will fix a problem that isn't immunological. Even in patients who do have alloimmunization, non-immune factors are frequently present alongside it, muddying the picture further [2].
Practically: before you reach for an antibody workup, ask whether the patient is febrile, septic, bleeding, DIC'ing, or splenomegalic. If the answer is yes to any of those, you may already have your explanation, and the next step is treating that, not chasing platelets.
HLA and HPA are separate workups, not one workup with two names
Once non-immune causes have been reasonably excluded, the two immune mechanisms get worked up differently, and they're often conflated when they shouldn't be. HLA alloimmunization is, by a wide margin, the most common immune cause of platelet refractoriness. HPA alloimmunization is considerably rarer, but it stays on the differential, because it won't show up on an HLA workup and a rare cause you didn't test for is still a cause you missed.
For HLA: start with an HLA antibody screen. If antibodies are detected, the next step is calculating the panel reactive antibody, or PRA, which estimates the percentage of the general donor population the patient would be incompatible with. A PRA above 20% is generally the threshold for pursuing selected platelet support; below that, enough of the standard donor pool remains compatible that special selection usually isn't worth the delay.
For HPA: this is a separate antibody, against platelet-specific antigens rather than HLA, and it requires its own test, a platelet antibody screen, distinct from the HLA workup entirely. If HLA testing comes back unrevealing and immune refractoriness is still suspected, the platelet antibody screen is the next step, not a repeat of the HLA testing.
Management, and the terminology that actually matters
Once HLA antibodies are confirmed and PRA clears that 20% threshold, the product you're ordering is HLA-compatible platelets, not HLA-matched. The distinction is real: HLA-matched means finding a donor whose HLA type mirrors the patient's, antigen for antigen, which is a narrow and often impractical ask. HLA-compatible means selecting donors who lack the specific antigens the patient has already formed antibodies against, which is a much larger and more workable donor pool. In practice, compatible is what gets ordered and what's actually achievable; matched is closer to a theoretical ideal than a standing option.
Even HLA-compatible platelets don't guarantee a good response. Compatibility improves the odds; it doesn't promise them, and suboptimal increments still happen on selected product, which brings you right back to needing a post-transfusion count to know whether this round worked either.
ABO-matched, fresh (under 48 hours) platelets tend to survive better even before you get to HLA selection, which is a low-cost first move worth trying regardless of suspected mechanism. Beyond that, options like IVIG or splenectomy get mentioned in the refractoriness conversation, but the evidence supporting either as a reliable fix for platelet-specific alloimmune refractoriness is thin, and neither should be reached for as a first-line answer.
The actual point
When the count doesn't move, or seems like it doesn't, the reflexive answer is "more platelets." But you don't actually know it didn't move unless someone drew the count. Once you have that number, often the correct next answer is "not platelets at all, fix the fever," and sometimes it's "differently selected platelets." The CCI, and knowing which antibody workup you're actually running, is how you figure out which question you're answering.
References
Trial to Reduce Alloimmunization to Platelets Study Group. Leukocyte reduction and ultraviolet B irradiation of platelets to prevent alloimmunization and refractoriness to platelet transfusions. N Engl J Med. 1997;337(26):1861-1869.
Platelet Refractoriness. Transfusion Medicine. PathologyOutlines.com. Accessed September 2026.
Evaluation of Platelet Refractoriness. UW Hematology. University of Wisconsin School of Medicine and Public Health. Accessed September 2026.
