Platelet Refractoriness Interpreter
Platelet Refractoriness Interpreter
A poor platelet increment is usually consumption, not antibody. Enter the corrected count increment at one hour and at 20–24 hours with the clinical context, and see whether this is immune refractoriness or something being done to the platelets after they arrive — because the timing is what separates them.
Immune or non-immune refractoriness
CCI at 1 h and 24 hCCI 11,400 at 45 minutes and 2,800 at 20 hours, on ABO-identical platelets, after two consecutive poor increments, in a patient who is febrile and septic
The definition, taken apart into the inputs above
Four clauses, four inputs on this page — and the immune verdict is returned only when all four are satisfied.
- the corrected count increment
- computed on the corrected count increment calculator, which handles the ×10⁹/L to /µL conversion, the body surface area correction and the dose in ×10¹¹. This page takes the answer as an input and does not repeat the formula
- 7,500 at one hour
- below this at 10–60 minutes is a poor recovery. The platelets did not arrive
- 4,500 at 20–24 hours
- below this at 18–24 hours is poor survival. The platelets arrived and then went. Matsui and colleagues state the pairing explicitly: a 24-hour increment below 4,500 with a one-hour increment above 7,500 suggests non-immunological refractoriness, while both below their thresholds indicates suspected immunological refractoriness
- two consecutive occasions
- a single poor increment is not refractoriness. Too many innocent explanations — a low-dose component, a late sample, a fever, a transfusion given during a bleed
- ABO-identical
- a major ABO mismatch clears platelets immediately and mimics an antibody exactly. The definition excludes it by construction
- no dominant non-immune factor
- part of the definition, not a caveat. Non-immune causes are roughly 80% of refractoriness, which is why this page reaches the HLA screen last
Worked example
CCI 11,400 at 45 minutes and 2,800 at 20 hours, on ABO-identical platelets, after two consecutive poor increments, in a patient who is febrile and septic
11,400 at 45 minutes is above 7,500 → the platelets were recovered into the circulation normally
2,800 at 20 hours is below 4,500 → they did not survive there
An HLA or HPA antibody destroys transfused platelets within minutes, so it cannot produce a good one-hour increment. The shape of the curve excludes it
Two poor increments and ABO-identical components would have satisfied two of the four clauses of the BSH definition — but the one-hour increment is not poor, so the first clause fails
Sepsis is on the list of consumptive causes, and it is sitting in the inputs
Verdict: non-immune. Treat the sepsis, time the next transfusion closer to when the platelets are needed, and re-measure
Now change one number. Set the one-hour increment to 4,200 and everything else stays the same — the verdict becomes a non-immune cause still, because sepsis is present and BSH's definition requires its absence. Clear the sepsis selector as well and only then does the page return immune refractoriness and an HLA screen. Three inputs have to line up before that test is the right test
Reading the two increments together
| CCI at 1 hour | CCI at 20–24 hours | What it means |
|---|---|---|
| Above 7,500 | Above 4,500 | The transfusion worked. If the count is still low, look at the dose and the marrow, not the response |
| Above 7,500 | Below 4,500 | Recovery good, survival poor — consumption. Sepsis, fever, splenomegaly, bleeding, DIC, amphotericin B, graft-versus-host or veno-occlusive disease. An antibody cannot produce this shape |
| Below 7,500 | Below 4,500 | Both poor — immune refractoriness is possible, but only after ABO-identical components, two consecutive episodes and the absence of a dominant non-immune factor. Then screen for HLA class I and HPA antibodies |
| Below 7,500 | Above 4,500 | Physiologically unlikely. Check the sample timings and recalculate — increments fall with time, they do not rise |
Causes, and how common they are
| Non-immune — about 80% | Immune — about 20% | |
|---|---|---|
| Mechanism | Platelets are recovered and then consumed or sequestered | Platelets are destroyed on arrival by antibody |
| Timing signature | One-hour increment adequate, 20–24 hour increment poor | Both increments poor |
| Causes | Sepsis and fever, splenomegaly, active bleeding, disseminated intravascular coagulation, amphotericin B and other drugs, graft-versus-host disease and veno-occlusive disease after transplant | HLA class I antibodies, from pregnancy or previous transfusion; HPA antibodies, much less often |
| What to do | Treat the cause and re-measure. BSH: patients with non-immune causes should not receive HLA-selected platelets | HLA class I antibody screen; HLA-selected components if positive (2C); HPA-selected if HPA antibodies and continuing refractoriness (2C) |
| Why the order matters | This is four cases in five, and the treatment is of the underlying illness | HLA-matched components are scarce and slow to source; asking for them in the wrong patient costs somebody else the unit |
Timing is the discriminator, and the antibody screen is not the first test
A patient with a marrow that cannot make platelets gets a transfusion, and the count barely moves. The reflex in that situation is to suspect antibodies and send an HLA screen, and in about four cases out of five that is the wrong test. The British Society for Haematology puts non-immune causes at approximately 80% of platelet refractoriness, and a 2025 systematic review reaches the same figure independently — non-immune factors up to 80%, immune-mediated refractoriness “the remaining 20%”. HLA-matched components are scarce, take days to source, and are needed by somebody. So the practical question is not “is this refractoriness?” but “which kind?”, and there is a clean way to answer it that costs one extra blood count.
The discriminator is timing. An HLA or HPA antibody destroys transfused platelets essentially on contact, within minutes, so an immune cause produces a poor increment at one hour and a poor increment at twenty-four. Consumption does not: the platelets are recovered normally into the circulation and are then used up or sequestered over the following day, so the one-hour increment is fine and the twenty-four hour increment is poor. Measure the corrected count increment at both timepoints and the shape of the curve tells you which mechanism you are looking at before any antibody test is sent. Matsui and colleagues state the rule directly: a twenty-four hour increment below 4,500 with a one-hour increment above 7,500 suggests non-immunological refractoriness, while both values below their thresholds indicates suspected immunological refractoriness. Those are the same two thresholds the site’s corrected count increment calculator bands on, and that page — not this one — is where the increment is worked out, with the body surface area correction and the platelet dose in ×10¹¹ that the formula needs.
The BSH definition of alloimmune refractoriness is worth reading one clause at a time, because every clause is doing work: “a 10 minute to one hour increment of less than 5 × 10⁹/l on 2 consecutive occasions, using ABO-identical platelets and in the absence of predominantly non-immunological factors”. The poor early increment is the immune signature. Two consecutive occasions exists because a single disappointing count has too many innocent explanations — a component at the low end of its dose range, a sample taken an hour late, a fever that day, a transfusion given in the middle of a bleed. ABO-identical exists because a major ABO mismatch clears platelets immediately through a completely different mechanism and is indistinguishable from an antibody on a one-hour increment; it is also the reason BSH recommends ABO-matched platelets when available to maximise increments. And in the absence of predominantly non-immunological factors is not a caveat tacked on the end — it is the clause that stops four patients in five being investigated for something they do not have. This page returns an immune verdict only when all four are satisfied, which is why the antibody screen sits at the bottom of its rule chain rather than the top.
When the non-immune causes are the answer, they are worth naming, because most of them are treatable and none of them is treated with different platelets. Sepsis and fever consume platelets; splenomegaly sequesters up to a third of the circulating mass; active bleeding and disseminated intravascular coagulation use them; amphotericin B is the drug most consistently implicated; and graft-versus-host disease and veno-occlusive disease do it after transplant. Treat what is there, consider timing the transfusion closer to when the platelets are actually needed rather than far in advance, and re-measure the increment on an ABO-identical component. BSH is explicit that patients whose refractoriness is non-immune should not receive HLA-selected platelets. When the four clauses are satisfied and the screen is right, send HLA class I antibodies first and HPA antibodies if those are negative or if refractoriness continues despite matched components; BSH recommends class I HLA-selected components for patients with class I antibodies and HPA-selected components for those with HPA antibodies and continuing refractoriness, both at 2C. Tell the blood service early, because matched components are not sitting on a shelf. And keep re-checking the non-immune list, because a patient with an HLA antibody can become septic on Tuesday like anyone else.
Frequently asked questions
What is platelet refractoriness?
A repeated failure of the platelet count to rise as expected after transfusion. The British Society for Haematology defines the alloimmune form specifically: a 10 minute to one hour increment of less than 5 × 10⁹/L on two consecutive occasions, using ABO-identical platelets and in the absence of predominantly non-immunological factors. Expressed as a corrected count increment, the commonly applied thresholds are below 7,500 at 10–60 minutes and below 4,500 at 18–24 hours. Note that the definition requires a repeat: a single poor increment is not refractoriness, and labelling it as such starts an investigation and a supply problem that may not be needed.
How do you tell immune from non-immune platelet refractoriness?
By when the platelets disappear. HLA and HPA antibodies destroy transfused platelets within minutes, so an immune cause gives a poor corrected count increment at one hour and a poor one at twenty-four. Consumption does not touch the platelets on arrival — they are recovered normally and then used up over the following day — so the one-hour increment is adequate and only the twenty-four hour value is poor. That is why the second measurement is worth taking: a service that only ever measures the one-hour increment cannot distinguish the two mechanisms and will send antibody screens on patients who are simply septic.
Should an HLA antibody screen be sent for every poor platelet increment?
No. Non-immune causes account for roughly 80% of platelet refractoriness — the figure appears in the BSH guideline and independently in a 2025 systematic review — so an antibody screen is the wrong first move in most patients. Check the other three clauses of the definition first: were the platelets ABO-identical, has the poor increment happened on two consecutive occasions, and is there a dominant non-immune factor such as sepsis, splenomegaly, bleeding, disseminated intravascular coagulation or amphotericin B? BSH is explicit that patients with non-immune causes should not receive HLA-selected platelets — those components are scarce and slow to source, and asking for them in the wrong patient denies them to someone who needs them.
What are the non-immune causes of a poor platelet increment?
Sepsis and fever, splenomegaly, active bleeding, disseminated intravascular coagulation, amphotericin B and other implicated drugs, and graft-versus-host disease or veno-occlusive disease after stem cell transplantation. A major ABO mismatch belongs on the list too, although it behaves differently — it clears platelets immediately and therefore mimics an immune cause on a one-hour increment, which is exactly why the BSH definition specifies ABO-identical components. Most of these are treatable, and none is treated with different platelets.
Where does the corrected count increment come from?
It is the platelet increment corrected for the size of the recipient and the dose transfused, and it is calculated on the site’s corrected count increment calculator, which handles the ×10⁹/L to /µL conversion, the body surface area term and the platelet content in ×10¹¹ printed on the component label. This page deliberately does not repeat that formula; it takes the two results — at 10–60 minutes and at 18–24 hours — as inputs and reasons from the pair. The commonest arithmetic error is dividing by the number of packs rather than by the platelet content in ×10¹¹, which inflates the result roughly threefold.
Related calculators
References
- Estcourt LJ, Birchall J, Allard S, et al. Guidelines for the use of platelet transfusions. British Society for Haematology; Br J Haematol. 2017;176(3):365–394. Alloimmune refractoriness defined as “a 10 minute to one hour increment of less than 5 x 10⁹/l on 2 consecutive occasions, using ABO-identical platelets and in the absence of predominantly non-immunological factors”; non-immune causes account for approximately 80% of cases; “ABO matched platelets should be used when available to maximise increments (2C)”; patients with non-immune causes should not receive HLA-selected platelets; class I HLA antibodies indicate class I HLA-selected platelet transfusion (2C), and continuing refractoriness with HPA antibodies indicates HPA-selected platelet transfusion (2C).
- Matsui R, Hagino T, Tsuno NH, et al. Does time of CCI measurement affect the evaluation of platelet transfusion effectiveness? Transfus Apher Sci. 2021;60(3):103123. CCI at one hour: “values lower than 7500/μL considered refractory”. CCI at 24 hours: “CCI-24 h values lower than 4500/μL indicating platelet transfusion refractoriness (PTR)”. “CCI-24 h 7500/μL” suggests non-immunological refractoriness; “CCI-24 h <4500/μL and CCI-1 h <7500/μL" indicates suspected immunological refractoriness.
- Platelet antibody screening for preventing post-transfusion platelet refractoriness: a systematic review and meta-analysis. Transfus Clin Biol. 2025. Refractoriness “operationally defined by an inadequate post-transfusion platelet count increment, often measured by a Corrected Count Increment (CCI) below 5000 or 7500 within 1–24 h post-transfusion”; non-immune factors account for “up to 80%” of cases, with immune-mediated refractoriness, usually HLA alloimmunisation, accounting for “the remaining 20%”.
- Hod E, Schwartz J. Platelet transfusion refractoriness. Br J Haematol. 2008;142(3):348–360.
- Slichter SJ, Kaufman RM, Assmann SF, et al. Dose of prophylactic platelet transfusions and prevention of hemorrhage. N Engl J Med. 2010;362(7):600–613.
Medical Disclaimer: The tools and content provided here are for educational and reference purposes only. They are not intended to substitute for professional medical advice, diagnosis, or treatment. Clinical decisions should always be based on the comprehensive assessment of a qualified healthcare professional.