Comparison of Prehospital Stroke Severity Scales

Scale Description and Items Sensitivity Specificity Positive Negative Accuracy Additional Additional Notes for
Predictive Predictive Validity & Consideration
Year Published Value Value Reliability
(PPV) (NPV) Measures
Cincinnati Stroke 3-item, 0- to 5- point scale Equiv. to Equiv. to ID an LVO ID an LVO ID an LVO Positive • Validated in prehospital setting
Triage Assessment NIHSS ≥ 15 NIHSS ≥ 15 (score ≥ 2): (score ≥ 2): (score ≥ 2): likelihood and with external data sets
Tool (C-STAT) • Gaze (0/2) (severe (severe 0.653 0.783 0.753 ratio (LR+) • Middle-of-road for time to
• Arm weakness (0/1) stroke): stroke): 4.096 complete this scale compared to
[formerly • Level of consciousness (0/1) 0.771 0.841 Alternative Alternative others on this list (5/7)7
Cincinnati Equiv. to Equiv. to ID an LVO ID an LVO Negative • Fails to recognize the
Prehospital Stroke NIHSS ≥ 10 NIHSS ≥ 10 (score ≥ 2): (score ≥ 2): likelihood importance of cortical signs,
Severity Scale (moderate (moderate 0.696 0.796 ratio (LR-) such as aphasia and particularly
(CPSSS)] stroke): stroke): 0.486 neglect, which are highly
0.641 0.911 associated with large cortical
20151,2 ID an LVO: ID an LVO: infarcts3
0.711 0.701 • CPSS most commonly cited in
CSC need: CSC need: EMS statewide protocols as an
0.571 0.791 example or the recommended
ID an LVO ID an LVO Area under the receiver operating characteristic curve scale to use (see bottom of grid,
(score ≥ 2): (score ≥ 2): (AUC) values: page 3 for how this differs from
0.563 0.853 C-STAT/CPSSS)3
Alternative Alternative Severe stroke2: 0.89
ID an LVO ID an LVO Moderate stroke2: 0.90
(score ≥ 2): (score ≥ 2): (original CPSS similar, severe: 0.83 and moderate 0.95)2
0.596 0.866
Alternative Alternative ID and LVO5: 0.72
ID an LVO ID an LVO
(score ≥ 2): (score ≥ 2):
0.832,7 0.402,7
Facial palsy, Arm 5-item, 0- to 9- point scale ID LVO ID LVO ID LVO ID LVO AUC values3: • Validated in prehospital setting
weakness, Speech • Facial palsy (0/1) score ≥ 3: score ≥ 3: score ≥ 3: score ≥ 3: FAST-ED=0.81 as reference but no external data sets (yet)
changes, Time, Eye • Arm weakness (0/1/2) 0.713 0.783 0.843 0.763 NIHSS=0.80, P=0.28 • Easy to learn and remember
deviation, Denial / • Speech changes (0/1/2) RACE=0.77, P=0.02 given that many EMS agencies
neglect (FAST-ED) • Eye devitation (0/1/2) ID LVO ID LVO ID LVO ID LVO CPSS=0.75, P=0.002 already are familiar with FAST3
scale • Denial / Neglect (0/1/2) score ≥ 4: score ≥ 4: score ≥ 4: score ≥ 4: • FAST-ED had comparable
0.613 0.893 0.823 0.793 accuracy to predict
20163 Two thresholds of ≥3 and ≥4 • LVO to the NIHSS and higher
were used because of high accuracy than RACE and CPSS3
Youden Index values (0.490 and
0.491, respectively) for
identifying LVOs.
Scale Description and Items Sensitivity Specificity Positive Negative Accuracy Additional Additional Notes for
Predictive Predictive Validity & Consideration
Year Published Value Value Reliability
(PPV) (NPV) Measures
Rapid Arterial 5 of 6-item scale (last item is score ≥ 5: score ≥ 5: score ≥ 5: score ≥ 5: score ≥ 5: Positive • Validated in prehospital setting
Occlusion based on which side pt has 0.553 0.873 0.683 0.793 0.773,4 likelihood and with external data sets
Evaluation Scale deficits on); 0- to 9- point scale ratio (LR+) • A more time-consuming scale to
(RACE) • Facial palsy (0/1/2) ID LVO: ID LVO: ID LVO: ID LVO: 4.176 complete compared to others
• Arm motor function (0/1/2) 0.596 0.866 0.706 0.796 on this list (7/7)7
20144 • Leg motor function (0/1/2) Negative
• Head & gaze deviation (0/1) Alternative Alternative Alternative Alternative likelihood
• Based on side, do only one: ID LVO: ID LVO: ID LVO: ID LVO: ratio (LR-)
• R side: Aphasia (0/1/2) 0.854,7 0.684,7 0.424,7 0.944,7 0.486
• L side: Agnosia (0/1/2)
Any score > 4 considered highly
likely an LVO
Los Angeles 3-item, 0- to 5-point scale ID LVO: ID LVO: ID LVO: ID LVO: Positive • Validated in prehospital setting
Prehospital Stroke developed for prehospital and 0.576 0.846 0.666 0.786 likelihood and with external data sets
Screen (LAPSS) emergency department (ED) use ratio (LR+) • Middle-of-road for time to
---
Motor Scale Alternative Alternative 3.506 complete this scale compared to
(LAMS) • Facial droop (0/1) ID LVO: ID LVO: others on this list (4/7)7
• Arm drift (0/1/2) 0.817 0.897 Negative • Official scale used in statewide
20015 • Grip strength (0/1/2) Additional validity measure (available for this scale only): likelihood EMS protocol for Rhode Island
Convergent validity with NIHSS (early-post arrival): ratio (LR-)
• 0.49 (LAMS prehospital) 5 0.516
• 0.89 (early-post arrival) 5
Prehospital Acute 3-item scale, 0- to 3-point scale ID LVO ID LVO ID LVO ID LVO Positive • Validated in prehospital setting
Stroke Severity to identify emergent large vessel (score≥ 2): (score≥ 2): (score≥ 2): (score≥ 2): likelihood but no external data sets (yet)
(PASS) scale occlusion (ELVO) in patients with 0.666 0.836 0.686 0.816 ratio (LR+) • Designed for simplicity and
acute ischemic stroke 3.846 rapid application
20176 • Level of consciousness (0/1) • PASS validity scores similar to or
• Gaze palsy/deviation (0/1) --- Negative better than CPSSS, LAMS, and
• Arm weakness (0/1) likelihood RACE values in detecting verified
ratio (LR-) LVOs in its original design and
0.476 validation study6
AUC for ID and LVO : 0.744
 Scale Description and Items Sensitivity Specificity Positive Negative Accuracy Additional Additional Notes for
Predictive Predictive Validity & Consideration
Year Published Value Value Reliability
(PPV) (NPV) Measures
Vision, Aphasia, 4-item scale; designed to quickly 1.00 0.90 0.74 1.00 • Not yet validated in prehospital
Neglect (VAN) assess functional neurovascular (compared (compared (compared (compared setting or with external data
anatomy. Patient is considered to NIHSS ≥ to NIHSS ≥6: to NIHSS ≥6: to NIHSS ≥6, sets (yet)
20167 either VAN positive (ELVO 6, also 0.74)7 0.58)7 also 1.00)7 • Performed by nurses at early
7
present) or VAN negative. 1.00) hospital arrival in original study7
• Weakness of Arms • Shortest time needed to
---
• Visual field complete this scale compared to
• Aphasia others on this list (2/7)7
• Neglect • Strong sensitivity (1.00) and NPV
Patient must have weakness plus (1.00) values in original study
one or all of the V, A, or N to be • Easy to remember
VAN positive
Cincinnati Included to differentiate from CPSSS. CPSS does not look at gaze or level of consciousness (in CSTAT/CPSSS), but rather Most commonly cited in
(Prehospital) facial droop and speech. Cincinnati Prehospital Stroke Scale: statewide EMS protocols
Stroke Scale i. Speech: Have patient state “You can’t teach an old dog new tricks” (Abnormal = wrong word, slurred, or absent speech)
(CPSS), for stroke ii. Facial droop when asked to show teeth or smile (Abnormal = one side does not move as well as other)
recognition only iii. Motor: Have patient close eyes and hold out both arms (Abnormal = arm cannot move or drifts down when held out)
Patients with 1 of these 3 findings as a new event have a 72% probability of an ischemic stroke.
If all 3 findings are present the probability of an acute stroke is more than 85%
NIHSS (for Included in this listing as a score ≥ 6: score ≥ 6: score ≥ 6: score ≥ 6: score ≥ 6:
comparison only) comparison to measures above 0.763 0.703 0.553 0.853 0.723
--- ---
score ≥ 10: score ≥ 10: score ≥ 10: score ≥ 10: score ≥ 10:
0.643 0.853 0.683 0.833 0.783
Sources

1. McMullan, J. T., Katz, B., Broderick, J., Schmit, P., Sucharew, H., & Adeoye, O. (2017). Prospective Prehospital Evaluation of the Cincinnati Stroke Triage Assessment Tool.
Prehospital Emergency Care, 1-8.
2. Katz, B. S., McMullan, J. T., Sucharew, H., Adeoye, O., & Broderick, J. P. (2015). Design and validation of a prehospital scale to predict stroke severity. Stroke, 46(6), 1508-1512.
3. Lima, F. O., Silva, G. S., Furie, K. L., Frankel, M. R., Lev, M. H., Camargo, É. C., ... & Nogueira, R. G. (2016). Field Assessment Stroke Triage for Emergency Destination. Stroke,
47(8), 1997-2002.
4. de la Ossa, N. P., Carrera, D., Gorchs, M., Querol, M., Millán, M., Gomis, M., ... & Escalada, X. (2014). Design and Validation of a Prehospital Stroke Scale to Predict Large Arterial
Occlusion. Stroke, 45(1), 87-91.
5. Kim, J. T., Chung, P. W., Starkman, S., Sanossian, N., Stratton, S. J., Eckstein, M., ... & Restrepo, L. (2017). Field Validation of the Los Angeles Motor Scale as a Tool for
Paramedic Assessment of Stroke Severity. Stroke, 48(2), 298-306.
6. Hastrup, S., Damgaard, D., Johnsen, S. P., & Andersen, G. (2016). Prehospital Acute Stroke Severity scale to predict large artery occlusion. Stroke, 47(7), 1772-1776.
7. Teleb, M. S., Ver Hage, A., Carter, J., Jayaraman, M. V., & McTaggart, R. A. (2016). Stroke vision, aphasia, neglect (VAN) assessment—a novel emergent large vessel occlusion
screening tool: pilot study and comparison with current clinical severity indices. Journal of neurointerventional surgery, neurintsurg-2015.