Gaze Deviation - Are We Missing This?

We know it's a sign of stroke - what I want to know is if we're missing this subtle sign and whether it matters in the end.

From “Ōkiku naru Ko”, 1959

Insert “time is brain” or whatever cliche you want here about strokes. Strokes are tough. Some are blatantly obvious, but the annoying ones are the ones that are subtle without good signs that we’re taught in medical school and residency. Adding to the anxiety, they remain one of the more litigious diagnoses in our specialty. Frightening.

One of my neurology friends wanted me to cover gaze deviation - and my first response to that is “Why should I care about gaze deviation?”, and my second response was “Wait should I care about gaze deviation?”. Let us feast on the evidence…

BUT FIRST - gaze deviation? It’s when a patient’s eyes involuntarily deviate/move to one side. Other terms that mean the same thing clinically - conjugate gaze deviation, gaze palsy, and visual gaze preference. In a real patient, this will look like the patient not being able to follow you across midline.¹ Check out this youtube video that shows a good example in about 15 seconds:

Ok now onto the feast of evidence…

Why should I care about gaze deviation?

Because it can mean, and often means stroke. Big ones (LVOs).

Gaze deviation assessed via telemedicine had a sensitivity of 66% and specificity of 92% for large vessel occlusion (LVO) strokes, which improves to 70% and 93% if you were to broaden out to all anterior circulation ischemia (not just LVO).² This association is so high that some fine people over at the Journal of Stroke and Cerebrovascular Disease put out a score (the Gaze Deviation and Paresis Score or GPS) that predicted LVOs with a sensitivity of 89%, spec. of 97%, positive predictive value of 95% and NPV of 94%.³ Those are some high numbers - though this was done in patients in Germany, it’s purpose was to identify patient prehospital, and it was a retrospective observational cohort analysis of ultimately 904. It performs better in terms of sensitivity than the G-FAST score (which is gaze palsy, facial droop, arm weakness, speech difficulty, and time) with the same specificity, otherwise GPS performs better than the other prehospital scores. It is composed of a few items from the NIHSS:

1. Gaze palsy/deviation (any abnormal gaze)

2. Motor arm weakness (some effort against gravity or worse)

3. Motor leg weakness (some effort against gravity or worse)

The score is “positive” when all three components are present, otherwise it is negative. To be honest, if I see these symptoms, I am already calling the stroke alert and rushing the patient over to CT. But I have to ask myself - how often have a spent time really teasing out gaze deviation on exam when the other signs of stroke aren’t there?

Does gaze deviation in stroke present in isolation?

No, not usually.

Gaze deviation alone performs better as a test when trying to predict for LVO compared to general ischemic stroke.⁴ Unless the patient in front of you has had longstanding chronic disease and collateral blood flow, you would suspect an LVO to present with more than just gaze deviation especially in the anterior circulation (we’re talking hemiparesis, aphasia, etc). Notably - presence of gaze deviation is associated with larger stroke lesion volumes on MRI DWI (103 mL vs. 17 mL).⁵

So… are we missing strokes because we’re missing gaze deviation?

Yes, we’re missing strokes. No, it’s not likely because of gaze deviation.

Big studies have already been published about the fact that acute ischemic strokes are being missed in the ED (14% in this 2016 study)***.⁶ There’s plenty of reasons for that - with most of these being the dreaded posterior circulation strokes that present more “I’m dizzy” and less “I can’t move my arm”, making these harder to diagnose. But it is unlikely that gaze deviation, in isolation and the lack of picking up on this sign, is causing us to miss strokes. I don’t have any data on the number of gaze deviation providers are missing however.

So what’s your take?

Gaze deviation is an important sign of LVO - look for it in your patients you’re considering a stroke in. Don’t rule out a stroke because it’s not there; do push stroke higher on your differential if it is.

---

What are you afraid of? What would you like to see on this blog?

---

---

***If you are a nice neurologist you can move on and not read this note - I only have respect for thee. If you are a nasty neurologist or want to poo-poo all over the specialty of EM, I will remind you that one time I called a stroke alert on a patient who was out walking with his mom and then all of a sudden had slurred speech and was stumbling around. After all the workup was done, something didn’t sit right with any of the EM providers so we slowed our role and decided to wait on administering thrombolytics given the negative non-con head CT, though our neurology colleagues wanted us to push the lytics. Very soon after this conversation ended, the alcohol level I ordered came back and we found out that the patient had done a few shots when mom wasn’t looking. Naughty naughty. But at least we didn’t give him the lytics.

1. Visual gaze preference (Concept Id: C4476823) - MedGen - NCBI. Nih.gov. Published 2016. https://www.ncbi.nlm.nih.gov/medgen/1388778

2. Schröter N, Weiller A, Rijntjes M, et al. Identifying large vessel occlusion at first glance in telemedicine. Journal of neurology. 2023;270(9):4318-4325. doi:https://doi.org/10.1007/s00415-023-11775-2

3. Johannes B, Habib P, Kolja Schürmann, et al. Gaze Deviation and Paresis Score (GPS) Sufficiently Predicts Emergent Large Vessel Occluding Strokes. Journal of Stroke and Cerebrovascular Diseases. 2020;30(3):105518-105518. doi:https://doi.org/10.1016/j.jstrokecerebrovasdis.2020.105518

4. Li M, Liang W, Yue P, et al. Does radiological conjugate eye deviation sign play a role in acute stroke imaging? A meta-analysis. Journal of Neurology. 2021;269(3):1142-1153. doi:https://doi.org/10.1007/s00415-021-10540-7

5. Singer OC, Humpich MC, Laufs H, Lanfermann H, Steinmetz H, Neumann-Haefelin T. Conjugate Eye Deviation in Acute Stroke. Stroke. 2006;37(11):2726-2732. doi:https://doi.org/10.1161/01.str.0000244809.67376.10

6. Madsen TE, Khoury J, Cadena R, et al. Potentially Missed Diagnosis of Ischemic Stroke in the Emergency Department in the Greater Cincinnati/Northern Kentucky Stroke Study. Panagos PD, ed. Academic Emergency Medicine. 2016;23(10):1128-1135. doi:https://doi.org/10.1111/acem.13029