TIMI score for NSTEMI vs STEMI: two different scores

What the TIMI risk score for UA/NSTEMI measures

The TIMI risk score for unstable angina and non-ST elevation myocardial infarction was published by Antman and colleagues in JAMA in August 2000. It was derived in 1957 patients with UA/NSTEMI from the TIMI 11B trial and validated in 3 more trial cohorts from TIMI 11B and ESSENCE. It counts 7 predictors at 1 point each: age 65 or older, at least 3 risk factors for coronary artery disease, prior coronary stenosis of 50% or more, ST-segment deviation on the presenting ECG, at least 2 anginal events in the prior 24 hours, use of aspirin in the prior 7 days, and elevated serum cardiac markers. The total runs from 0 to 7. The outcome is a composite of all-cause death, new or recurrent myocardial infarction, or severe recurrent ischemia requiring urgent revascularization through 14 days, and its rate rose from 4.7% at a score of 0 or 1 to 40.9% at 6 or 7.

This is the score the TIMI score calculator on this site works out, with the 14-day rate for each score from the original paper.

What the TIMI risk score for STEMI measures

The TIMI risk score for ST-elevation myocardial infarction is a separate score, published by Morrow and colleagues in Circulation in October 2000. It was built from 14,114 fibrinolytic-eligible patients with STEMI in the InTIME II trial (Intravenous nPA for Treatment of Infarcting Myocardium Early II), and it predicts death within 30 days. Unlike the UA/NSTEMI score, its points are weighted: each predictor carries points according to its adjusted odds ratio from logistic regression, for a total from 0 to 14. The paper reports that its ten baseline variables accounted for 97% of the predictive capacity of the full multivariable model. Mean 30-day mortality in the trial was 6.7%. Mortality rose more than 40-fold across the score, from under 1% at 0 points to 35.9% above 8. The c statistic was 0.779, against 0.784 for the full model, performance held from 1 to 365 days, and external validation in the TIMI 9 trial gave 0.746.

30-day mortality by TIMI STEMI score in InTIME II (Morrow 2000):

AskMed does not build the STEMI score as a calculator; this page describes it from Morrow's paper so the two are not confused.

How the two TIMI scores differ

The two TIMI scores differ in almost everything except their name and their authors, the TIMI investigators at Brigham and Women's Hospital in Boston. The UA/NSTEMI score uses 7 equal yes-or-no items and a range of 0 to 7; the STEMI score uses weighted items and a range of 0 to 14. The UA/NSTEMI score predicts a composite of death, myocardial infarction or urgent revascularization over 14 days; the STEMI score predicts death alone over 30 days. Only age appears in both, and with different cut-offs: one point at 65 or older in the UA/NSTEMI score, against 2 points at 65 to 74 and 3 points at 75 or older in the STEMI score. The UA/NSTEMI score includes the coronary history, aspirin use and cardiac markers; the STEMI score instead uses hemodynamic findings, weight, infarct location and time to treatment. The STEMI score also separated outcomes better, with a c statistic of 0.779 against 0.65 for the UA/NSTEMI composite.

Which TIMI score fits which patient

Which TIMI score to use depends first on the electrocardiogram. A patient with ST-elevation MI belongs to the Morrow score, which was built in patients eligible for fibrinolysis and estimates death at 30 days. A patient with an acute coronary syndrome and no ST elevation, meaning unstable angina or NSTEMI, belongs to the Antman score, which estimates death, infarction or urgent revascularization at 14 days. ST depression on the presenting ECG counts as a point in the UA/NSTEMI score; it does not make a STEMI. Neither TIMI score was designed for every patient who walks in with chest pain: both were derived in trial populations with a confirmed diagnosis. For undifferentiated chest pain in the emergency department, Backus and colleagues found that the HEART score separated 6-week outcomes better than the UA/NSTEMI TIMI score, with c statistics of 0.83 against 0.75 in 2388 patients at 10 Dutch hospitals.

Use the HEART score calculator for that first look at chest pain, and the TIMI score calculator once UA/NSTEMI is the working diagnosis. The guide HEART vs TIMI vs GRACE covers how those three compared head to head.

Worked examples: one patient for each score

Two patients show how differently the two TIMI scores are built and read. The first is a 68-year-old man with hypertension, diabetes and current smoking, who took aspirin this week, has never had a cardiac catheterization, and had one episode of chest pain at rest today. His ECG shows ST depression and his troponin is raised, so the diagnosis is NSTEMI. On the UA/NSTEMI score he gets 1 point each for age, 3 risk factors, ST deviation, aspirin use and raised markers: 5 points. In the TIMI 11B test cohort, 26.2% of patients with a score of 5 had an event by day 14, and 7.1% died. The second is a 70-year-old woman with hypertension and an anterior STEMI, 2 hours after symptom onset, weighing 62 kg, with a heart rate of 104, a systolic pressure of 110 mmHg and clear lungs (Killip class I). Her STEMI score is 7 points.

The two totals cannot be compared with each other: a 5 on one scale and a 7 on the other measure different outcomes over different time spans. Enter the first patient in the TIMI score calculator to see the same 26.2%.