IPSS-R Calculator for Myelodysplastic Syndrome

Five weighted items to a total of 0 to 10, and the weights are FRACTIONAL: platelets and neutrophils can score 0.5 and haemoglobin 1.5, so an implementation that rounds moves a patient across a stratum boundary silently.

Last updated 8 Oct 2026

IPSS-R (Greenberg 2012)

5 items → 0–10, half points

Cytogenetic risk groupThe item that carries the most weight, and a PUBLISHED CATEGORICAL CLASSIFICATION rather than a number: five named groups, each defined by a list of specific karyotype abnormalities. The commonest are in the options; the full lists are in the source. Take the group from the cytogenetics report. The group NAMES differ between renderings — the same five appear elsewhere as very favourable, favourable, intermediate, unfavourable and very unfavourable — and one rendering read for this page adds +21 to the intermediate group where the MDS Foundation’s table does not. Neither moves a patient, because any other single or double independent clone is intermediate anyway.

Bone marrow blast percentageBands, not a number: 2% or under scores 0, over 2% to under 5% scores 1, 5% to 10% scores 2. A count of exactly 5% is the 2-point band and 4.9% is the 1-point band. Above 20% is no longer MDS in either 2022 classification.

HaemoglobinTHE 1.5 IS NOT A TYPOGRAPHICAL ERROR: severe anaemia scores one and a half points, not two. In SI units the thresholds are 100 and 80 g/L; the MDS Foundation’s calculator also prints them as 6.2 and 5.0 mmol/L.

Platelet countHalf a point for a moderate thrombocytopenia. The whole platelet item spans one point against four for cytogenetics — the proportion the derivation cohort supported, and worth knowing before weighing a falling count by eye.

Absolute neutrophil countThe smallest item: half a point, and nothing more however deep the neutropenia. The threshold is 0.8, not the 1.0 or 1.5 most neutropenia gradings use.

Age (years), for the age-adjusted score onlyAge is NOT part of the IPSS-R. It feeds only the age-adjusted figure (IPSS-RA) below, which Greenberg published as a nomogram alongside the index and described as being for survival estimation. It cannot change the headline total.

4.0pointsExample

Good cytogenetics (1); marrow blasts 3% (1); haemoglobin 9.1 g/dL (1); platelets 74 ×10⁹/L (0.5); neutrophils 0.6 ×10⁹/L (0.5); age 72

The five items, the fractional weights and the age adjustment

IPSS-R = cytogenetic group (0, 1, 2, 3 or 4) + marrow blasts (0, 1, 2 or 3) + haemoglobin (0, 1 or 1.5) + platelets (0, 0.5 or 1) + neutrophils (0 or 0.5)
Range 0 to 10

Very low ≤ 1.5 · low >1.5–3 · intermediate >3–4.5 · high >4.5–6 · very high >6

IPSS-RA = IPSS-R + (age − 70) × [0.05 − (IPSS-R × 0.005)]

the half points: 1.5 for haemoglobin under 8 g/dL, 0.5 for platelets of 50 to under 100, 0.5 for neutrophils under 0.8. Every stratum boundary is a value a half point can land on, so an implementation storing whole numbers misclassifies silently
0 to 10, and how the 10 is reached: the maximum is 4 + 3 + 1.5 + 1 + 0.5. Cytogenetics and marrow blasts supply 7 of the 10 points, so the index is driven by the karyotype and the blast count and modified by the cytopenias
the boundaries are exclusive below and inclusive above: the published bands read ≤ 1.5, >1.5–3, >3–4.5, >4.5–6 and >6, so exactly 3.0 is LOW risk and exactly 4.5 is INTERMEDIATE. The bands here own their upper bounds for that reason
what the cytogenetic groups are: a published five-group classification of specific karyotype abnormalities, not a number. The options carry the commonest defining abnormalities of each; the full lists are in Greenberg 2012 and in the MDS Foundation’s table. This page cannot classify a karyotype
age is not in the index: the IPSS-R takes no account of age. The age-adjusted form (IPSS-RA) was published alongside it as a nomogram and described as being for survival estimation; it shifts the score towards the mean for a patient away from 70 and leaves the IPSS-R band itself alone. Greenberg’s own worked example: an IPSS-R of 3.5 at age 45 adjusts by −0.81 to an IPSS-RA of 2.69
what a median is: a median is the middle of the derivation cohort and not a forecast: half of that cohort reached the figure and half did not, and a figure measured in a cohort treated decades ago is a historical measurement rather than an expectation
what has superseded it: the IPSS-M (2022) adds 31 gene mutations and produces six strata. It is not computed anywhere on this site: its model takes 43 input variables and no source carrying its beta coefficients could be read

Worked example

Good cytogenetics (1); marrow blasts 3% (1); haemoglobin 9.1 g/dL (1); platelets 74 ×10⁹/L (0.5); neutrophils 0.6 ×10⁹/L (0.5); age 72
1 + 1 + 1 + 0.5 + 0.5 = 4.0 points
4.0 sits in the >3–4.5 intermediate-risk band — median overall survival 3.0 years in the derivation cohort
Note where the last point came from. The platelet and neutrophil items contributed half a point each. An implementation rounding each item to the nearest whole number would score this patient 3 or 5 — LOW risk or HIGH risk, either side of the stratum they are actually in
Age-adjusted: 4.0 + (72 − 70) × (0.05 − 4.0 × 0.005) = 4.06. Greenberg's own published example runs the other way: an IPSS-R of 3.5 at age 45 gives −25 × 0.0325 = −0.8125 and an IPSS-RA of 2.69 — intermediate on the index, low on the age-adjusted form
Change the cytogenetic group alone from good to very poor and the total goes from 4.0 to 7.0: very high risk, and the single biggest lever in the instrument

The five items and every published weight

Three of the five items carry a half or a one-and-a-half, which is the thing most often got wrong about this score. The empty cells are empty in the published table too. Read as text out of the MDS Foundation’s own calculator page and, independently, out of a clinicaltrials.gov protocol attributing it to Greenberg 2012; the two agree cell for cell.
Item	0	0.5	1	1.5	2	3	4
Cytogenetic risk group	Very good	—	Good	—	Intermediate	Poor	Very poor
Marrow blasts	≤ 2%	—	>2% to <5%	—	5% to 10%	>10%	—
Haemoglobin (g/dL)	≥ 10	—	8 to <10	<8	—	—	—
Platelets (×10⁹/L)	≥ 100	50 to <100	<50	—	—	—	—
Neutrophils (×10⁹/L)	≥ 0.8	<0.8	—	—	—	—	—

The five strata and what they predicted in the derivation cohort

Every figure here is from Greenberg 2012’s cohort of 7,012 primary, UNTREATED adults, and that word is the important one: these are natural history figures, not outcomes on a hypomethylating agent, on luspatercept or after allogeneic transplantation. Every outcome figure here describes the cohort the index was derived in and not the patient in front of you; the spread within one stratum is wider than the gap between strata.
Risk category	IPSS-R score	Median overall survival	Median time to 25% AML evolution	Share of cohort
Very low	≤ 1.5	8.8 years	Not reached	about 1 in 5
Low	>1.5 – 3	5.3 years	10.8 years	the largest group
Intermediate	>3 – 4.5	3.0 years	3.2 years
High	>4.5 – 6	1.6 years	1.4 years
Very high	>6	0.8 years	0.73 years	about 1 in 10

Half points, and the two ways this index gets miscomputed

This is the IPSS of HAEMATOLOGY — the International Prognostic Scoring System for myelodysplastic syndrome. It is not the International Prostate Symptom Score, which shares the abbreviation, measures lower urinary tract symptoms and has nothing to do with this one: that is the International Prostate Symptom Score. The haematological IPSS was published in 1997 and revised in 2012; this page implements the revision, which separated five prognostic groups where the original had four, widened the cytogenetic classification from three groups to five, and split the marrow blast count more finely. If a report names a risk group as low, intermediate-1, intermediate-2 or high, it is the 1997 original and not this instrument.

What makes the IPSS-R different from almost every other clinical score is that its weights are not whole numbers. Haemoglobin under 8 g/dL scores 1.5; platelets of 50 to under 100 score 0.5; neutrophils under 0.8 score 0.5. The total runs from 0 to 10 in half-point steps, and four of the five stratum boundaries — 1.5, 3, 4.5 and 6 — sit exactly where a half point can land a patient. An implementation that stores points as integers, or rounds each item before summing, puts some patients one stratum out, in no fixed direction and with nothing on the screen to show it. The boundaries compound it: they are inclusive at the top and exclusive at the bottom, so a total of exactly 3.0 is low risk and 4.5 is intermediate. The cytogenetic item is a trap of a different kind — a published classification of named karyotype abnormalities into five groups, worth up to 4 of the 10 points, which has to come off the cytogenetics report. No calculator can derive it from free text and this one does not try.

What the strata mean needs stating carefully. Every outcome figure here describes the cohort the index was derived in and not the patient in front of you; the spread within one stratum is wider than the gap between strata. Greenberg’s cohort was 7,012 primary, UNTREATED adults, so the survival figures are natural history rather than outcomes on treatment, and they predate the hypomethylating agents, luspatercept and current transplant practice. The index has itself been superseded for molecularly characterised patients by the IPSS-M. Every laboratory threshold here is method- and laboratory-dependent, so the reader’s own laboratory’s reference interval takes precedence. A risk stratum is not a diagnosis and not a plan. This page computes the index, names the stratum and reports what it predicted in the derivation cohort; what follows is a decision for the treating team with the patient.

Frequently asked questions

Which IPSS is this — the blood one or the prostate one?

The blood one. This is the International Prognostic Scoring System, revised (IPSS-R), for myelodysplastic syndrome. The International Prostate Symptom Score is a seven-question symptom questionnaire for lower urinary tract symptoms and has no overlap with this instrument beyond the four letters.

Why does haemoglobin score 1.5 rather than 2?

Because that is the weight the derivation cohort supported. The weights come from a Cox model fitted in 7,012 untreated patients and rounded to the nearest HALF point rather than the nearest whole one, which is why three of the five items carry a fraction. The maximum is 4 + 3 + 1.5 + 1 + 0.5 = 10.

A total of exactly 3.0 — is that low or intermediate risk?

Low. The published bands are inclusive at the top: very low is 1.5 or less, low is above 1.5 up to and including 3, intermediate is above 3 up to and including 4.5. So 3.0 is low and 3.1 intermediate, and 4.5 is intermediate rather than high.

Can I work out the cytogenetic group from the karyotype?

Only by reading it off the published five-group table. The options here carry the commonest defining abnormalities — loss of Y and del(11q) are very good; a normal karyotype or an isolated del(5q), del(12p) or del(20q) is good; a complex karyotype with more than three abnormalities is very poor — but the full lists are longer than a calculator should reproduce, and no page can classify a karyotype from free text.

What is the age-adjusted IPSS-R, and should I use it?

A published adjustment, printed alongside the index as a nomogram and described by its authors as being for survival estimation. It adds (age − 70) × [0.05 − (IPSS-R × 0.005)], pulling a young patient’s figure down and an old patient’s up. It is shown as a second figure and deliberately not as the headline: the IPSS-R band, which guidelines and trial eligibility criteria are written in, comes from the unadjusted total.

Has the IPSS-R been replaced?

For patients who have had a myeloid gene panel, largely yes: the IPSS-M (2022) incorporates 31 gene mutations, separates six risk groups and reclassifies a substantial minority — most often out of the IPSS-R intermediate group. The IPSS-R is still what most trial eligibility criteria are written in, and the only one computable without sequencing.

Related calculators

References

Greenberg PL, et al. Revised international prognostic scoring system for myelodysplastic syndromes. Blood. 2012;120(12):2454–65. The derivation paper: 7,012 primary, UNTREATED adult patients.
MDS Foundation. IPSS-R calculator: prognostic score values, prognostic risk categories and cytogenetic risk groups.
Celgene Corporation. Protocol for NCT02631070, Table 1: IPSS-R prognostic score values, attributed in the protocol to Greenberg 2012. clinicaltrials.gov.
Harrison’s Principles of Internal Medicine, 20th ed., Table 66-4: Revised International Prognostic Scoring System — risk groups, score ranges, median overall survival and median time to 25% AML evolution.
Heidelberg University, Institute of Medical Informatics. IPSS-R data model: risk categories, survival and cytogenetic risk groups.
ipssm: JavaScript package for the Molecular International Prognostic Scoring System (IPSS-M) — 43 input variables, no stated licence and no stated source for its coefficients. The instrument this page does NOT compute.

Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/