What is automated perimetry?
Here, we’ll only talk about the Humphrey visual field perimeter, which is used for 99% of visual field tests. It’s an automated, static perimeter (unlike Goldmann kinetic perimetry which requires a human operator, and uses a moving target). The Humphrey uses fixed points of light which are shown at different intensity levels. The software automatically varies the intensity of the points of lights at each location to determine the threshold – the intensity of light where the patient can see it 50% of the time.
Which subtest should I order?
HVF 24-2
This is ordered for 90% of glaucoma patients. This is your baseline exam that all glaucoma suspects and glaucoma patients need at routine intervals.
HVF 10-2
This is ordered for the 10% of glaucoma patients who are so advanced that the HVF 24-2 is mostly black, with only a central island of remaining vision. Macular diseases including plaquenil toxicity exams also need 10-2.
HVF 30-2
Generally, we order this for neuro patients. It has a wider angle and can capture peripheral field defects.
Reliability indices
- Name, demographics, etc: Make sure you are looking at the right patient!
- Fixation loss: The HVF will routinely flash dots in the patient’s physiological blind spot to check if the patient has his / her gaze fixated on the center. If the patient can actually see the spot, then it’s recorded as a fixation loss. Reliable tests have below 20% fixation loss (although many people have their own opinions about these upper limits).
- False positive: The user pressed the button when there was no stimulus. They were “trigger happy”. Reliable tests have below 33% false positives.
- False negative: The user did not see a stimulus which was brighter than one they saw earlier in the same test. Reliable tests have below 33% false negatives.
- Stimulus characteristics: 99% of visual fields (VFs) will use the size 3 white stimulus. Other sizes and colors are used for patients with late disease or retinal disease.
Which picture do I look at?
Yes, there are a lot of graphs. The two most important to look at are the Grayscale Map and the Pattern Deviation. The rest of this article will explain how to interpret these.
How do I tell if things are changing over time?
This is the million dollar question. This is what every patient will want to know and how you will decide whether to step up drop therapy, add laser, or take the patient to the operating room.
As a very, very general guideline, you can look at the density / size of the field defect, the pattern standard deviation, and the mean deviation (MD) to see if it is worsening. However, your decision should also take into account the normal variability between each visual field, the optic nerve head appearance, pressures, patient compliance, OCT, visual symptoms, etc.
This is a very complex topic and somewhat beyond the discussion of this post, so talk to your seniors and your attendings if you aren’t sure!
Top 5 most common visual field patterns
1) Nonspecific / low Reliability / inattention / patient hungry
For every interpretable, reliable visual field you get, you will also get another in which the patient thinks he should be scanning the dome for lights the whole time, is poorly positioned, is exhausted from waiting in your clinic for hours, or is too elderly and arthritic to push the button in time.
These types of inattention errors will usually register as high fixation losses, false positives, or false negatives. Or, the visual field could just be patchy all over.
If these errors are not too bad, the general gist of the field can be deduced, especially if compared to prior fields. Most often, as long as everything else is stable (IOP, ONH appearance), we just reorder these fields in a few month’s time. If you simply cannot get a visual field due to patient cooperation or attention, you can order an optic nerve OCT to follow the optic nerve head objectively (though thinning does not always necessarily correlate to field loss).
2) Superior / inferior arcuate defect
The most common early to mid stage glaucomatous field. The reason these look like arcs and come off the blind spot is that they represent the loss of bundles of nerves as they come out of the optic nerve head. The horizontal border is the horizontal raphe, which is an imaginary line dividing the upper and lower hemispheres of the retina. These are probably 25%-35% of the fields we see.
3) Blind spot enlargement
This can be seen in glaucoma, but also can occur with papilledema and optic nerve head swelling. This would likely be seen in patients with idiopathic intracranial hypertension (aka pseudotumor cerebri).
4) Severe constriction with a central island
Unfortunately, this is end stage glaucoma. At this point, many patients still have great central vision of 20/20 to 20/50, but peripheral vision is nearly gone. Here, we switch patients over to an HVF 10-2 to better follow their progression.
5) Nasal steppe
This is another common glaucomatous field. About 10% of fields show this.
Conclusion
Use this order to interpret your Humphrey visual field every time:
- Confirm it’s the right patient with name and date of birth
- Confirm it’s the right/left eye
- Look at the reliability indices
- Look at the pattern
- Look at the GHT, mean deviation, VFI, and pattern standard deviation
- Compare to the previous visual fields
Great help in understanding the VF pattern
Glad you found it helpful, Jennifer!
Highly educative
This is awesome thanks a lot for putting this together Ben! What are your thoughts on the VR based perimeters vs HFA (or as a complement)? I recently saw an ad from Olleyes (https://olleyes.com/) seems comfortable. It also looks like they are partnered with Keeler.
Hey Peter – not just me working on this, we have a big team behind the scenes that creates our content!
I’m certainly also not an expert in perimetry but have worked with Rarebit, another HVF alternative, in the past. I’m curious to see how VR based perimetry is implemented in the coming years. I can imagine there will be lots of opportunities for making large impacts in mobile settings or resource-poor areas where cost/size of HVF is a limiting factor.
Three questions: Does this chart tell me the actual degree of blindness in the visual field? The visual field efficiency score? What does a 1% VFI percentage mean?
No jargon, very nicely put and clearly explained. Thanks!!
GOOD
Very Helpful, but did not explain MD, and pattern deviation.
Great article, very helpful and educative.
Great
My dr report said 30-2 field test : scattered deflects both eyes . Can someone tell me what this means?
excelent, thanks
excellent