New Injury Severity Score (NISS) Calculator
New Injury Severity Score (NISS) Calculator
Sum the squares of the three highest AIS severities anywhere in the body, ignoring region. The NISS is never lower than the ISS, and the gap is widest in multiple severe injury to one region.
New Injury Severity Score
Three worst injuries to 0–75Two severe chest injuries (AIS 4 and AIS 4) and a serious extremity injury (AIS 3)
Formula
where A, B and C are the three highest AIS severities anywhere in the body, irrespective of region · any AIS 6 sets NISS = 75
- the one rule that differs
- the NISS is “the sum of the squares of the Abbreviated Injury Scale scores of each of a patient’s three most severe” injuries, with no region constraint. The ISS requires the three codes to come from three different body regions. Everything else about the two scores is identical
- NISS is never below ISS
- the ISS selection is a restriction of the NISS selection, so the NISS is greater than or equal to the ISS for every patient. Equality happens when the three worst injuries already lie in three different regions, which is why a single-injury patient is useless as a test case
- where the gap opens
- multiple severe injury concentrated in one region. Three critical chest injuries: ISS 25, NISS 75. Two severe injuries and a serious one in one region: ISS 16, NISS 41. These are the patients the score was modified for
- reported performance
- in the 1997 paper’s two retrospective validation sets — 3,136 patients over four years at a level I centre in Albuquerque and 3,449 over four years at Emanuel Hospital in Portland — the receiver operating characteristic area for predicting mortality was 0.869 for ISS against 0.896 for NISS in the first and 0.896 against 0.907 in the second, with Hosmer-Lemeshow calibration statistics of 29.12 against 8.88 and 83.48 against 19.86. The authors concluded that the NISS “should replace ISS as the standard summary measure of human trauma”
- which score a registry uses
- both are in use and this page does not claim a global answer, because the one international source read for this page reports centres supplying ISS and centres supplying NISS in the same study. What that source does flag is sharper: “AIS dictionary differences could affect the discrimination between severely and less severely injured patients across national and international registries”, so the dictionary version matters as well as the formula
- where the codes come from
- The Abbreviated Injury Scale codes themselves come from a copyrighted dictionary licensed by the Association for the Advancement of Automotive Medicine, which licenses it for software only through an API and forbids its reproduction. This page therefore does not and cannot assign an AIS code to an injury: it takes the severities the reader already holds from a licensed coder or a trauma registry and applies the published formula over them.
Worked example
Two severe chest injuries (AIS 4 and AIS 4) and a serious extremity injury (AIS 3)
The three worst injuries are 4, 4 and 3, and the NISS does not care that two of them are in the same region
4² + 4² + 3² = 16 + 16 + 9 = 41 NISS points
The ISS for the same patient takes one chest code: 4² + 3² = 25. A 16-point gap on three injuries
Make all three injuries critical and confined to the chest: NISS 5² × 3 = 75, ISS 5² = 25. The widest gap the two scores can show, and it appears in the most severely injured
Now enter 4, 0, 0 — a single severe injury. NISS 16, ISS 16, identical. An implementation that sorted or selected wrongly would pass this case, which is why it is not the one to test on
Enter 6 anywhere and the answer is 75. Without that convention three 6s would give 108, outside the shared scale
ISS and NISS on the same patients
| Injuries | ISS | NISS | Gap |
|---|---|---|---|
| One severe injury (AIS 4) | 16 | 16 | 0 — the degenerate case |
| AIS 4 head, AIS 3 chest, AIS 3 extremity | 34 | 34 | 0 — already three different regions |
| AIS 4 chest, AIS 4 chest, AIS 3 extremity | 25 | 41 | 16 |
| AIS 4 chest, AIS 4 chest, AIS 3 chest | 16 | 41 | 25 |
| AIS 5 chest × 3 | 25 | 75 | 50 |
| Any AIS 6 | 75 | 75 | 0 — both forced by convention |
Reported discrimination in the 1997 validation sets
| Cohort | ROC area, ISS | ROC area, NISS | Hosmer-Lemeshow, ISS vs NISS |
|---|---|---|---|
| 3,136 patients, level I centre, Albuquerque, four years | 0.869 | 0.896 (p less than 0.001) | 29.12 vs 8.88 |
| 3,449 patients, Emanuel Hospital, Portland, four years | 0.896 | 0.907 (p less than 0.004) | 83.48 vs 19.86 |
One rule changed, and why it matters most in the worst patients
The New Injury Severity Score differs from the Injury Severity Score in exactly one rule. The ISS takes the highest Abbreviated Injury Scale severity from each of three different body regions; the NISS takes the three highest severities anywhere. Everything else — the squaring, the ceiling of 75, the convention that an AIS 6 forces 75 — is the same. That single change was proposed in 1997, and its consequence is arithmetic rather than opinion: because the ISS selection is a restricted version of the NISS selection, the NISS can never be lower than the ISS for the same patient.
Where the two scores agree is in the majority of patients, whose three worst injuries happen to lie in three different regions. Where they diverge is in multiple severe injury concentrated in one part of the body — a chest crushed in three places, a head with three separate critical lesions, a limb with several severe injuries. Three critical chest injuries give an ISS of 25 and a NISS of 75. Two severe chest injuries and a serious one give an ISS of 16 and a NISS of 41. The gap is not evenly distributed: it opens widest in exactly the group whose outcomes the score is used to predict, which is why mixing ISS and NISS in one dataset is a problem and not a rounding detail.
The 1997 paper tested the change on two retrospective single-centre series, 3,136 patients in Albuquerque and 3,449 in Portland. The receiver operating characteristic area for mortality rose from 0.869 to 0.896 in the first and from 0.896 to 0.907 in the second — small gains. The calibration improved far more, with Hosmer-Lemeshow statistics falling from 29.12 to 8.88 and from 83.48 to 19.86, meaning the NISS predicted better across the whole range rather than only on average. The authors concluded that the NISS should replace the ISS as the standard summary measure.
It has not entirely done so. Both scores remain in use, and one international study of a consensus trauma template collected ISS from some centres and NISS from others. That same study raises the point that matters more than the choice: differences in the AIS dictionary version between registries affect how severely and less severely injured patients are separated, so comparing two registries requires knowing the dictionary as well as the formula. This page applies the published formula to severities the reader supplies; the AIS dictionary that produces them is licensed and is not reproduced here.
Frequently asked questions
What is the difference between ISS and NISS?
The ISS uses the highest AIS severity in each of three different body regions. The NISS uses the three highest severities anywhere, so two or three severe injuries to the same region all count. That is the only difference, and it means the NISS is never lower than the ISS for a given patient.
Can the NISS be lower than the ISS?
No. The three codes the ISS selects are always available to the NISS selection, so the NISS is greater than or equal to the ISS for every patient. They are equal when the three worst injuries already lie in three different regions, which covers most patients.
Does the NISS predict mortality better than the ISS?
In the 1997 derivation paper’s two validation cohorts it did, modestly on discrimination and substantially on calibration: ROC area 0.869 against 0.896 in Albuquerque and 0.896 against 0.907 in Portland, with Hosmer-Lemeshow statistics of 29.12 against 8.88 and 83.48 against 19.86. Both were retrospective single-centre series reported by the score’s own authors.
Can I apply the usual ISS strata to a NISS?
Not safely. There is no published NISS stratification, and the ISS bands reclassify patients when applied to a NISS because the NISS runs higher for the same patient. The one threshold this page can point to from a source it read is the consensus trauma template study’s use of NISS 16 or above to define the severely injured.
Does the AIS 6 rule apply to the NISS?
Registry implementations carry it across, and this page does the same: an AIS 6 anywhere gives 75. The 1997 definition as published is just the sum of squares of the three most severe injuries, which for three AIS 6 injuries would be 108 — outside the 0 to 75 scale the two scores share.
Related calculators
References
- Abstract of A modification of the injury severity score that both improves accuracy and simplifies scoring (Osler, Baker and Long, 1997), PMID 9420106, read at qxmd.com/r/9420106.
- Agency for Clinical Innovation, Institute of Trauma and Injury Management. Injury scoring. NSW Health. aci.health.nsw.gov.au/networks/trauma/data/injury-scoring
- Texas Department of State Health Services, Injury Prevention Unit. Texas EMS and Trauma Registries — 2021 trauma data, 18 August 2023.
- Association for the Advancement of Automotive Medicine. AIS License — commercial (software) and healthcare system licensing terms for the Abbreviated Injury Scale. aaam.org/ais/ais-products/ais-license
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/