2. How Good is Your Stroke Scale?
Sean Kivlehan, MD, MPH, NREMT-P
September 2013

4. Outline Stroke Review Treatment review Scale development and flaws
Rehab  tPA  recognition & time matters
Scale development and flaws
Systems issues
Future directions

5. Case #1
62 year old male, low speed MVC into pole, conscious with flaccid left arm, facial droop, and slurred speech. BP 220/100

6. Case #2
89 year old female, daughter called because she has been vomiting for past 2 hours, is dizzy, and can’t stand. No facial droop, normal speech, no focal weakness. BP 220/100. Actively vomiting.

7. Case #3
70 year old female found down in her apartment when son went to check on her. Last seen normal at dinner last night. Now semi-conscious but following commands, not moving right side. BP 200/90.

8. We’ve come far but still have a long way…
HISTORY

9. “Apoplexy” What they called it from Hippocrates on… 1950’s: TIA’s
25 years ago: Imaging
10 years ago: Imaging in community hospitals
1970 WHO definition: “rapidly developing clinical signs of focal (or global) disturbance of cerebral function, lasting more than 24 hours or leading to death, with no apparent cause other than that of vascular origin.”
Sacco 2013

10. TIA’s
1975: “episodes of temporary and focal dysfunction of vascular origin, which are variable in duration, commonly lasting from 2 to 15 minutes, but occasionally lasting as long as a day (24 hours). They leave no persistent neurological deficit.” 2002: “A TIA is a brief episode of neurologic dysfunction caused by focal brain or retinal ischemia, with clinical symptoms typically lasting less than one hour, and without evidence of acute infarction.”

11. Unclear
2009: “a transient episode of neurological dysfunction caused by focal brain, spinal cord, or retinal ischemia without acute infarction.”

12. What is a Stroke Today? Ischemic Stroke
(clot: thombotic vs embolic) – 87%
2. Hemorrhagic Stroke
(bleed: intracerebral vs subarachnoid; aneurysms & AVM’s) – 13%
3. Transient Ischemic Attack (TIA)
Caused by clot, last <5 min, no permanent injury

13. A long way to go Number 4 cause of death in US (down from #3 in 2008)
785,000 in US a year
Less than half of 911 calls made within 1 hour of onset
Newest AHA/ASA outreach: FAST
1 of 3 in 88% of strokes
Recalled well in 3 months (public)
Only 53% of strokes arrive by EMS in US

14. Jauch 2013

15. Vessels & Risk
ANATOMY OF A STROKE

21. Where is the Stroke? Anterior Circulation Posterior Circulation
Weakness
Aphasia
Sensory deficit
Posterior Circulation
Dizzy
Vomiting
Ataxia

22. Now that we have the vessels down, how do they get messed up?

23. Source: Arizona.edu

24. Ischemic

25. Hemorrhagic
40% 30 days!
Sacco 2013

29. What Causes These Things?
Embolic
Bacterial emboli from valvular vegetations
Blood clots from cardiac dysrhythmias
Platelet-fibrin complex from atherosclerotic plaque disruption
Septic emboli from infection
Fat emboli from traumatic injury
Particulate emboli from IV drug injection
Thrombotic
Secondary to intrinsic vascular disease in the occluded vessel
Atherosclerotic disease most common cause in U.S.
Other causes include
Vasculitis
Hypercoagulable states
Polycythemia
Vascular dissection

30. Risk CAD: Double HTN: Triple Cardiac Failure: Quadruple
A-Fib: Quintuple
Wolf, Stroke, 1991

32. And what about those TIA’s?
4 to 20% will progress to stroke within 90 days
Half of those in first 24 hours
ABCD2 score
Age (>60)
Blood Pressure (>140/90)
Clinical Features (weakness/speech)
Duration (>60 minutes)
Diabetes
Koenig, NEJM, 2007

33. From Rehab to Cure
TREATMENT

34. Treatment
Old teaching – 5 pages out of 1100 Focused on recognition, managing airway, manage herniation symptoms

35. Suddenly, a cure!

36. What we KNOW Earlier recognition  Earlier EMS
Earlier EMS  Earlier ED notification
Earlier ED notification  Earlier imaging
Earlier imaging  Earlier tPa
Earlier tPa  Better outcomes

37. AHA Implementation Strategy for EMS within Stoke Systems of Care (1997)

38. Key Things Check a sugar And the pupils! Prior stroke/seizures?
Time of onset!
Wake up strokes
Stroke mimickers

40. EMS Interventions ABC’s etc Oxygen to maintain SpO2 >94%
Head of bed flat for SBP <120 and give isotonic saline
Antihypertensive treatment controversial (>220)
Glucose if FS <60
Establish PIV
Labs if build into system
None of this should delay transport
Jauch 2013

41. The most important piece of info….
Time patient last seen normal
EMS often the only people that can get this info on scene (family, bystanders)
Also:
Signs of hx of seizures or trauma on scene
PMHx (prior stroke, diabetes, HTN, AF
Meds on scene (anticoagulants)
Recent illness, surgery, trauma
Collect phone numbers, code status, family members
Bring family to ED
EMS are the eyes & ears of the ED

42. tPA (Fibrinolytic Therapy)
FDA approved in 1996 for 3 hour window
Showed symptom improvement at 24 hours and 1 year (OR 1.9; CI )
ICH: 6.4% in tPA and 0.6% in placebo
3 to 4.5 hour window w/ extra exclusions
4.5 to 6 hour window did not help
NINDS, NEJM, 1995

43. Jauch 2013

44. Mechanical Retrieval & IA tPA
Merci
Penumbra
Solitaire
Trevo
IA tPA ok for up to 6 hours
Good for patients with IV tPA contraindications
Can give in addition to IV tPA

53. So Now, TIME MATTERS

55. And we need to Recognize
Can we? 1995: Reading EMS (Ohio) 58 months, 4,413 patients 96 (2%) dx with stroke or TIA by EMS 10 excluded (lost data) 86  62 (72%) dx with stroke or TIA in ED So, 24 patients (28%) were misdiagnosed by EMS (false positive) Dispatch: 30 of 62 dispatched as something other than stroke BUT WE DON’T KNOW: how many people were having strokes & EMS called it something else?
Kothari, 1995

56. Maybe it’s because of the education?

57. The NIH Stroke Scale
WHO IS HAVING A STROKE?

63. What they did in Cincinatti
NIH stroke scale – huge & takes a while
Watered it down to what matters:
Facial palsy
Asymmetric arm weakness
Speech abnormalities

64. 100% Sensitivity 88% Specificity Kothari, 1997
They took 299 patients (74 known stroke from previous study compared to 225 randoms in ED
Kothari, 1997

65. What does that mean? Sensitivity: True Positive
(you found all strokes)
Specificity: True Negative
(you found all non strokes)
So this test (in theory):
Catches every stroke
1 in 9 aren’t really having a stroke (false positive)

66. Who was falsely labeled? (also, our differential)
Seizure (3)
Alcohol Intox (3)
Cardiac Arrest (2)
Abdominal Pain (2)
Dizziness/Vertigo (2)
Sepsis/Pneumonia (2)

67. Limitations Done by ED Physicians (not EMS)
Stroke patients were sick – enrolled in stroke trial
Done in hospital (not in field)
Misses posterior circulation strokes

68. Next: Test on EMS 24 EMS Providers (17 medics, 7 EMTs)
4 EMS providers & 1 ED physician did CPSS on each patient
860 scales on 171 patients (38 with stroke, 11 with TIA)
Scoring for each component same between ED physician, Medic, EMT
CPSS had sensitivity 59% and specificity 89%
Missed: 10 posterior circ & 3 anterior circ (atypical)
Kothari 1999

69. Paramedics vs Physicians
Nor 2004

70. Kothari 1999

71. What about this posterior circ thing?
10/14 were missed!
Testing ataxia would have caught 6/10
Downplayed: ataxia difficult to reproduce, least reproducible item on NIHSS, rarely treated w/ tPa
Kothari 1999

72. What happened?
1997: all stroke dx patients were in a thrombolytic trial, so they had more severe symptoms and all received tPa
1999: neurologists were asked to find stroke mimickers, and stroke patients were just “discharge dx” of stroke, didn’t necessarily get tPa
1997
1999
Sensitivity
100%
59%
Specificity
88%
89%

73. Did it get better? San Diego, 2005, 12 month period, 86,989 calls
Matched CAD (stroke), EMS report w/ CPSS done & stroke dx, & hospital ICD-9’s by billing info
Dispatch as stroke: 2,140 (2%)  882 to study hospital  367 (42%) hospital stroke dx
EMS dx as stroke: 995 (1.3%)  493 to study hospital  193 (40%) hospital stroke dx
Missed: Dispatch: 73, EMS: 247
Ramanujam 2008

74. Verdict? EMS Sensitivity 44%, PPV 40%
Flaws: didn’t follow all patients in SD, so don’t really know true neg.
Possible data is missing, or labeled “AMS etc”
Others?
Bray – 78% sensitivity
No scale (Smith) – 61% sensitive, PPV 77%

75. Nail in the coffin? Durham, 2004 They use CPSS
All transports are discretely coded (so no missed EMS dx’s)
Hospital (Duke) has stroke registry, compared this to EMS transports’ dx’s
Frendl 2009

76. Durham 184 EMS stroke/TIA charts 30 excluded (uncon), leaving 154
CPSS only used on 37% of stroke patients!
70% of patients w/ actual stroke had at least 1 CPSS finding (Sensitivity)
52% of patients w/ CPSS finding had a stroke (Specificity)
No effect of scene time or L&S use to ED
Frendl 2009

77. What are you goals?
IF: identify greatest # of stroke patients, then add more elements (11% of missed strokes had grip strength/visual deficit/numbness or tingling)
BUT: if you are going to active stroke teams off of this, this will put false positives through the roof
IF: justify transport diversion, then the specificity NEEDS to increase.
Is it better to bypass a patient’s primary hospital when we are only right half of the time?
PRO
CON
No missed strokes
Loss of Continuity
Best care for stroke patients
Increased cost (system & patient)
EMS unit OOS time
Accident Risk

78. Another option? Yes – the competitor! LAPSS Starts with exclusions:
>12 hour onset
Age <45 yo
Hx seizures
Wheelchair/bedbound
Finger stick <50 or >400
Then does:
Arm weakness
Grip Testing
Facial droop
Kidwell 1998

79. LAPSS Plans to miss: This is in the name of increasing specificity
Isolated visual field cuts
Isolated aphasias
Brainstem & Cerebellar infarcts
This is in the name of increasing specificity
Focuses on motor deficits
Tested by ED physicians retrospectively on known stroke patients in thrombolytic trial (similar to CPSS), 93% sensitive
Held up in testing – 89% sensitive & 99% specific
Kidwell 2000
Kidwell 1998

80. The Australians Studied CPSS vs LAPSS
…and threw in their own, the MASS
(Melbourne Ambulance Stroke Screen)
Supposed to be a combo of the 2
Showed better sens/spec but the approach was flawed – didn’t track stroke discharge from hospital (so we have no idea how many were missed)
“Faster Access to Stroke Therapies (FAST)” study
Bray 2005

81. MASS? Facial Droop Unilateral hand grip/weakness Unilateral arm drift
Slurred/abnormal speech
ADDS: rule out mimics
<45 yo
Hx seizures/epilepsy
Abnormal BSL
Bray 2005

82. Bray 2010

83. Bray 2005

84. The British don’t want to be left out
Face Arm Speech Test (FAST)
Same as CPSS but you don’t need the patient to repeat the sentence
Validated in Newcastle, UK
EMS caught 144/183 (79%) of strokes
Still high levels of missed posterior strokes
Can’t measure the false negatives (a theme)
Harbison 2003

85. Scale Summary CPSS: based off of NIHSS, misses post circ
LAPSS: Longer, but rules out non-tpa candidates and increases specificity
FAST: Even quicker than CPSS, basically the same thing, from the UK
MASS: Combo of the CPSS & LAPSS, building on strengths, from Australia

86. Wait – There’s More! Miami Emergency Neurological Deficit (MEND)
Precursor to new Advanced Stroke Life Support (ASLS) course
CPSS + 6 more NIHSS components
level of consciousness, orientation, commands, visual fields, gaze, leg motor, limb ataxia, sensation
Brotons 2012

87. Ontario
CPSS w/ more exclusions
Chenkin 2009

88. Case #1
62 year old male, low speed MVC into pole, conscious with flaccid left arm, facial droop, and slurred speech. BP 220/100 Right MCA (anterior circulation), positive on all scales

89. Case #2
89 year old female, daughter called because she has been vomiting for past 2 hours, is dizzy, and can’t stand. No facial droop, normal speech, no focal weakness. BP 220/100. Actively vomiting. Posterior circulation stroke. Abbreviated scales negative, NIHSS positive.

90. Case #3
70 year old female found down in her apartment when son went to check on her. Last seen normal at dinner last night. Now semi-conscious but following commands, not moving right side. BP 200/90. Left MCA, large stroke, positive on CPSS but out of window.

91. Now, transport
Where, how
ALS v BLS
Air?
Stoke centers
Retriage

92. Field tPA, telemedicine, & ultrasound
FUTURE

93. PHANTOM-S Pre-Hospital Acute Neurologic Therapy and Optimization of Medical Care in Stroke
The ambulance:
Neurologist
Paramedic
Radiographer
CT Scanner
POC Lab
Teleradiology
The dispatch:
For any suspected stroke
Weber 2013

95. The Results 23 patients got tPA in 3 months
Door-to-needle 62 minutes (vs 98 minutes)
2 patients bled, 1 died
Weber 2013

96. Source: Mayfield Clinic.com & Radiographics

98. Does EMS notification reduce door to needle?
Yes!
And door to imaging
And pretty much every other metric
Bae 2010, Mosley 2007

99. Summary
Stoke management
AHA
Scales
Transport

100. Why does this matter?
EMS are the first medical contact for 38-70% of stroke patients
EMS recognition is key to hospital notification
Notification shortens door-to-needle time
That improves outcomes for stroke patients
BUT, we are missing a LOT of stroke patients (up to 40%)
They go to non-stroke center or wait in triage
Bray 2005