• Users Online: 54229
  • Home
  • Print this page
  • Email this page

   Table of Contents      
LETTER TO EDITOR
Year : 2001  |  Volume : 49  |  Issue : 4  |  Page : 279

In reply


Correspondence Address:
M R Nair


Login to access the Email id

Source of Support: None, Conflict of Interest: None


PMID: 12930128

Rights and PermissionsRights and Permissions

Keywords: Automation, Humans, Perimetry,


How to cite this article:
Nair M R. In reply. Indian J Ophthalmol 2001;49:279

How to cite this URL:
Nair M R. In reply. Indian J Ophthalmol [serial online] 2001 [cited 2024 Mar 29];49:279. Available from: https://journals.lww.com/ijo/pages/default.aspx/text.asp?2001/49/4/279/14682


  Dear Editor, Top


Having grown up on a diet of Goldmann field charting I always found it difficult to interpret automated perimetry. I must commend Ravi Thomas et al for a most elegant presentation of automated perimetry (IJO, Vol. 49:June 2001).

However, I would like ask the authors that in the light of automated perimetry what is the present status of early glaucoma field as elaborated by Traquair and others; Baring of the Blind Spot, Isolated scotomas in the Bjerumm's area, Seidel's scotoma & Roenne's step? Are we still looking for them?

Also, has automated perimetry thrown up newer field defects which we need to be aware of?


  In Reply Top


Peer appreciation is the ultimate reward. I thank Dr. Shukla for that as well as his infinitely kind comments.

As far as the queries are considered:

1. Adjusting for low sensitivity areas caused by retinal hemorrhages, exudates, etc:

The conditions referred to would usually be expected to produce localized defects. Small, but localized defects. In general if we adjust for something, we loose the ability to detect THAT something. For example, in pattern deviation plot we adjust for generalized depression; therefore generalized depression DISSAPEARS from the pattern deviation plot. If we adjusted for localized scotomas, localized scotomas would then become undetectable. Regrettably it isn't possible to selectively adjust for those consequent to a particular cause and differentiate them from lesions we may be interested in.

Also, we can think of adjusting if we have the ability to detect such small defects. The 30-2 programme has a resolution of 6 degrees. It can miss the blind spot. If we combine it with a 30-1 program the resolution becomes 4.2 degrees. The 10-2 has a resolution of 2 degrees. It is therefore unlikely that these conditions will cause a detectable defect and therefore become a diagnostic problem. Especially since the fields will never be interpreted in isolation. That was ONE major message in the article.

2. The article was intended to be as simple an approach to interpretation as we could possibly make it. We wrote up our clinical approach, avoiding anything (like the numbers component of the total and pattern deviation plots) that we don't use for interpretation. We would have elaborated on the numerical component of those printouts here, but that would take more space than the editor would probably allow. For a detailed, but simple explanation I would therefore recommend Anderson and Patella's book on Automated Perimetry.

3. The age issue is very important and pertinent. I especially thank you for asking this as I learnt something: I did not have the answer and had to consult Dr. Patella. His response is copied below:
"We did not include any juveniles in the normative database. The minimum age was either 18 or 21, depending upon what was the local legal age for adulthood in each venue where normals were collected. Then we come to the implicit question, which is whether the normative database may be applied to kids. Certainly the 'normal' value may be applied, since the age correction is less than a decibel per decade.

What we don't know is whether the range of normal sensitivity is different in kids, simply because they are less disciplined test-takers. I wouldn't be surprised if it turned out that, e.g. the 1% limits for TD or PD or MD were larger for 10-year-olds than for age 25 or 35, but the limits probably are not smaller in kids than they are in adults. Thus, I think we can say that if a field looks normal on Statpac it probably is normal. If a child's results don't fall within the Statpac normal limits, then it's more likely that it has to do with patient test-taking performance than would be the case in an adult.

A number of paediatric ophthalmologists have been interested in compiling a normative database for kids, but none were interested enough to actually undertake the task. From my point of view the problem is not so interesting, simply because the juvenile patient population is such a small fraction of the perimetric testing load, and because I don't actually believe that using the adult limits imposes any significant problem. Hope that helps."

Hope that really helps!

I would like to thank Dr. Nair for his very kind comments on our article.

The defects we are looking for in automated perimetry are really quite similar to what we looked at in Goldmann perimetry. It is the format that makes some of them seem different.

Let's take isolated scotomas in Bjerrums area for example. Compare it to the first of Anderson and Patella's criteria; for the pattern deviation plot: "three or more non-edge points that are depressed to an extent expected in less than 5% of the normal population. These points should be clustered in the arcuate area." That sounds suspiciously like an isolated scotoma in Bjerrums area, doesn't it? And if such a cluster (especially if larger) were to occur in the upper or lower vicinity of the blind spot, wouldn't that qualify as a Siedel's scotoma? Baring of the blind spot is a similar non-specific depression in the same region. And while not mentioned in the article, a difference in sensitivity of 5-10 decibels across the nasal horizontal meridian would fit a nasal step. This can occur in the central or peripheral field. In fact there is a nasal step program that I used to be fond of. I must check out its validity.

Newer defects. The only thing I can think of is that large variation of a point's sensitivity could be the earliest sign of pathology. The short term fluctuation (SF) was an indicator of this variability. However as discussed this in the article SF is also an indicator of reliability. We have also discussed why SF was abandoned in the newer SITA program. I for one was sorry to see it pass away, but the gain was the shortened testing time.

May I deviate from older terminology to older techniques? The older techniques are still not "out". They are very useful. The Bjerrum's screen is an excellent and very versatile tool. Though it lacks the sensitivity of the automated perimeter and requires a trained person, it can provide useful diagnostic and gross follow up information. And the Goldman Perimeter continuous to be a gold standard for perimetry. A beautifully designed and calibrated instrument that regrettably requires a well trained and versatile technician. The older techniques are still very useful; but a calibrated standard automated perimeter is just as valid, more reliable and more convenient to use.




 

Top
 
 
  Search
 
    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

 
  In this article
Dear Editor,
In Reply

 Article Access Statistics
    Viewed2975    
    Printed113    
    Emailed0    
    PDF Downloaded0    
    Comments [Add]    

Recommend this journal