Indian Journal of Ophthalmology

: 1990  |  Volume : 38  |  Issue : 1  |  Page : 10--13

Blue field entoptoscopy in diabetic retinopathy

Rainer N Mittl, Ram Tiwari, Emanuel Wilkes 
 Edward S. Harkness Eye Institute, Columbia, New York, USA

Correspondence Address:
Rainer N Mittl
The Edward S. Harkness Eye Institute, Columbia-Presbyterian Medical Center, 635 West 165th Street, Suite 314, New York 10032


Blue field entoptoscopy was performed in 50 consecutive patients with diabetic retinopathy. Several statistically significant correlations were obtained. The number of «DQ»flying corpuscles«DQ» was reduced in eyes harboring a vitreous hemorrhage, and in severe and proliferative diabetic retinopathy. On the other hand, patients with a history of focal or panretinal photocoagulation had better corpuscle counts than patients who had never received any treatment. It appears therefore, that argon laser photocoagulation is beneficial in preserving the blue field entoptic phenomenon.

How to cite this article:
Mittl RN, Tiwari R, Wilkes E. Blue field entoptoscopy in diabetic retinopathy.Indian J Ophthalmol 1990;38:10-13

How to cite this URL:
Mittl RN, Tiwari R, Wilkes E. Blue field entoptoscopy in diabetic retinopathy. Indian J Ophthalmol [serial online] 1990 [cited 2021 Jan 19 ];38:10-13
Available from:

Full Text


Blue field entoptoscopy has been used to detect macular pathology in the. presence of cataracts and in ocular trauma. [1],[ 2 ] It has not been shown to be useful in the presence of a vitreous hemorrhage. [3],[4] Current tests for pre-vitrectomy macular evaluation include two-light dis­crimination, Maddox rod orientation and colour percep­tion. The Purkinje phenomenon, an entoptic response elicited by transillumination of the globe, is also valuable, but considered specific only if the capillary-free zone of the retinal vascular tree is visualized [5] , It is easier to perceive the entoptic image in the blue field.

in our study, blue field entoptoscopy was performed in a series of consecutive patients with diabetic retinopathy. In most cases, Fundus visualization was adequate for direct correlation of macular pathology and entoptic findings. The purpose of the study was to establish the usefulness of blue field entoptoscopy in the presence of diabetic retinopathy. The nature, type and extent of dia­betic retinopathy responsible for reduction or elimination of the blue field phenomenon was assessed.


Fifty consecutive diabetic patients of the Retinal Clinic at Columbia-Presbyterian Medical Center were evaluated with the Blue Field Entoptoscope BFE 100 (Medical In­strument Research Associates, Inc., Waltham, Massa­chusetts). Thirty-five patients were female, the remain­der were male. The ages ranged from 22 to 82. Five patients had diabetes mellitus for less than 5 years, six between 6 and 10 years, twenty-one between 11 and 20 years, fourteen between 21 and 30 years, and four over 31 years.

The subjects were asked to estimate the number of 'flying corpuscles' in each of the four quadrants seen through the eye piece of the instrument. Initial observa­tions were recorded in the patient's own words. The dis­trilaution and number of 'flying corpuscles' was recorded and all responses were listed separately for each quad­rant. In some cases, the patients did not recognise the blue field entoptic phenomenon, even though there was reason to believe they should. Answers were then solicited by suggesting the presence of 'dancing spots', `moving stars', etc. This was usually followed by an af­firmative response. In a sample group of 10 % of the patients, the test was repeated on three different occa­sions. The answers were quite consistent for each patient, apparently, in part based on memory.

The number of 'flying corpuscles' were recorded faith­fully, even though not infrequently, the estimates were obviously too high. No attempt was made to influence the patients' responses. Since separate answers were obtained for each quadrant rather than for the total, the results were sometimes grossly inaccurate. The correct number of 'flying corpuscles' is known to be between 20 and 40. The estimates given by our patients even for each quadrant were sometimes multiples of the *ex­pected figures. Only when the number of 'flying corpuscles' were significantly reduced did they seem to correlate better. This may have in part been due to the fact that it is much easier to count a small than a large number of particles in motion.

Since there was much tainting of the results by differ­ences in personality (e.g, euphoric vs. depressed), comparisons of the results between individual patients were not attempted. Rather, the patients were placed in large groups on the basis of a variety of clinical findings.

All measurements were first obtained at low illumination. If the patient had significant media opacities, the test was repeated at medium and high levels. The stronger light was used to assure optimal test conditions for cases with cataract or vitreous hemorrhage.

In addition, slit lamp examination, applanation tonom­etry, indirect ophthalmoscopy and three-mirror fundus contact lens examination was performed. The visual acuity was recorded. Retinal pathology was evaluated using fundus sketches, colour fundus photographs, and, if indicated, fluorescein angiograms. For each patient, the eye with the lower visual acuity was entered into the study. In cases of equal vision, the right eye was used.

Statistical tests were based on estimates of the numbers of 'flying corpuscles' in four quadrants. The individual readings were then added, totalled and averaged. Evalu­ation of the various comparisons under consideration was accomplished via nonparametric tests. Such tests were chosen over those based on the normal distribution (as the within-group variation for all comparisons was quite large and nonhomogeneous.

The Friedman Two-Way ANOVA, Wilcoxon Matched Pairs Tests and the Mann-Whitney U test were used [6].All tests were two-tailed with the rejection region set at alpha = 0.05.

Selected parameters of the individual patient's systemic or ocular disease state were compared to the findings on blue field entoptoscopy

1. Diabetes mellitus was classified as Type I (Juvenile) or Type II (Maturity-onset).

2. Patients with a past or present history of labile or mani­fest, controlled or uncontrolled hypertension were iden­tified as 'hypertensive'.

3. Lens opacities regardless of type and distribution were graded from 0 to III on the basis of density. Grade 0 included aphakia.

4. Visual acuity was obtained.

5. Any significant vitreous hemorrhage was recorded, including time of onset and state of organisation.

6. Diabetic retinopathy was classified as mild, moderate or severe on the basis of fundus findings. This grading was based on the estimated number, intensity and type of the individual fundus lesions. Macular pathology was given special consideration. The groups of 'mild and moderate' diabetic retinopathy were subsequently combined for statistical purposes, because it was diffi­cult to draw a distinct line between 'mild' and 'moderate' in many cases. The group as a whole was then con­trasted to the 'severe' group in all statistical evaluations. 7. Cases with proliferative and non-proliferative diabetic retinopathy were compared. 'Proliferative' constituted any type of epiretinal tissue, active or inactive, fibrous, vascular or fibrovascular. Eyes that merely contained a few isolated patches of proliferation were not included in this group. The change had to be clinically and morpho­logically significant.

8. Eyes with evidence of previous argon laser or xenon arc treatment were classified by the term 'photocoagu­lation' provided the treatment extended over at least one quadrant or the sum equivalent thereof.


Statistical analysis revealed no significant differences in blue field entoptic examination in the following groups of patients

1. Type I (Juvenile) versus Type II (Maturity-onset) diabetes mellitus.

2. Diabetes mellitus and hypertension versus diabetes mellitus alone.

3. Grade 0 versus Grade I versus Grade II versus Grade III cataracts.

4. Visual acuity of 20/20 to 20/80 versus visual acuity of 20/100 to 20/200. In patients with visual acuities of less than 20/400, only one was able to visualise a small number of 'flying corpuscles'.

On the other hand, blue field entoptoscopy was success­ful in detecting statistically significant differences in : 1. Eyes with vitreous hemorrhage versus eyes without it. 2. Severe versus mild to moderate diabetic retinopathy [Table 1].

3. Proliferative versus non-proliferative diabetic retino­pathy [Table 2].

4. Eyes with a history of photocoagulation versus eyes that had never received photocoagulation [Table 3].

The number of 'flying corpuscles' was lower in vitreous hemorrhage and in severe and proliferative diabetic retinopathy. It was higher in eyes that had received photocoagulation.


Which are the controlling factors in blue field entop­toscopy ? It is necessary to have a sufficient, but as of yet undetermined number of viable photoreceptors and a largely intact perifoveal capillary network. Characteris­tically, in diabetic retinopathy, both, photoreceptors and capillaries may have suffered damage directly or may have been impaired sequentially. Unfortunately, blue field entoptoscopy does not provide the means to sepa­rate these factors, rather, it provides an all-or-none re­sponse. If negative, macular function may be considered abolished, if positive, macular function is retained. These two statements have to be qualified, however.

If a patient does not see any 'flying corpuscles', addi­tional factors have to be taken into consideration. The patient may have been unable to express the observa­tions accurately, or the ocular media may have been ob­scured to such an extent that blue light was not transmit­ted, even at the highest level of illuminatipn.

On the other hand, if a patient spontaneously and correctly describes the 'flying corpuscles' in the blue field, macular function has to be present. The extent of function remains unknown, however. This represents one of the drawbacks of blue field entoptoscopy, it is largely a qualitative examination. Even though the number of 'flying corpuscles' in the central 20° field is known to be between 20 and 40, this range is far too wide for any quantitative predictions. Only reduction to numbers lower than 10 or 5 is suggestive of macular impairment. This vague relationship between macular pathology and perception of 'flying corpuscles' is also the reason that any prediction of visual acuity on the basis of corpuscle numbers is purely speculative'.

In this study, initially, it was expected that the patients with mild to moderate retinopathy would observe more 'flying corpuscles' than those with severe diabetic retino­pathy. The data presented in [Table 1] indeed supported this hypothesis. Further investigation, however, indi­cated that our positive results might have been the result of confounding. Since thirteen (40.6 %) of the thirty-two patients with severe diabetic retinopathy had prolifera­tive disease, while only two (11.1 %) of the eighteen with mild to moderate retinopathy had proliferative disease, we were concerned that the significant difference ob­served between the two types of diabetic patients analy­sed in [Table 1] was actually a result of the larger proportion of patients with proliferative disease in the group with severe retinopathy. In order to address this concern, the fifteen patients with proliferative disease were removed and the data of patients with mild to moderate versus severe retinopathy re-analysed. The results are shown in [Table 4]. As anticipated, a significant result was not obtained, thus confirming our suspicion, that the initial positive results shown in [Table 1] were due to an excess of patients with proliferative disease in the severe retinopathy group. Therefore, blue field entop­toscopy was a valuable tool in differentiating patients with and without proliferative disease. However, it was not adequate in distinguishing patients with mild to moderate retinopathy from those with severe retinopa­thy when proliferative disease was absent. Due to the small sample size the power necessary to detect such a difference was relatively low.

These findings support the conclusion that proliferative diabetic retinopathy is more likely than the non-prolifera­tive variant to reduce or eliminate perception of the 'flying corpuscles' in the blue field. While macular edema is the leading cause of blindness in diabetes mellitus, it is compatible with retention of (an entoptic response though reduced) for prolonged periods of time. Prolifera­tive diabetic retinopathy, on the other hand, appears to result in a more dramatic and thorough destruction of

cones and/or perimacular capillaries, more likely associ­ated with extinction of the blue field entoptic phenome­non.

This concept of the concurrence of proliferative disease and loss of the blue field entoptic phenomenon was also suggested by the clinical course in at least one of the cases under study. In this particular patient, moderate proliferative diabetic retinopathy with seemingly intact macular morphology concurred with almost total ab­sence of a response in blue field entoptoscopy. Within a short time, in spite of therapeutic intervention, increas­ingly florid diabetic retinopathy resulted in pthysis bulbi. In this case, early loss of the blue field entoptic phenome­non may have been an indicator of future severe prolif­erative disease.

An apparent beneficial effect on the number of 'flying corpuscles' was observed in eyes with a history of photocoagulation. The statistical significance of this finding was only borderline, however, and a larger and more differentiated patient sample will be necessary to clarify this result [Table 3]. Severity of diabetic retinopa­thy had no direct influence on this result. 33 percent of patients who received photocoagulation (7 of 21) had severe proliferative disease versus 27.6 percent who did not (8 of 29). Eyes that had undergone photocoagulation seemed to be better off, regardless how many treat­ments had been performed and if they had been ablative or focal in nature. This is even more surprising because loss of one or two lines on the Snellen chart is not uncommon following completion of panretinal photoco­agulation[7].


1Sinclair SH, Loebl M, Riva CE : Blue field entoptic phenomenon in cataract patients. Arch Ophthalmol 97 : 1092 - 1095, 1979.
2Sinclair SH, Loebl M, Riva CE : Blue field entoptic test in patients with ocular trauma. Arch Ophthalmol 99 : 464 - 467, 1981.
3Fuller Go, Hutton WL: Presurgical evaluation of eyes with opaque media. New York, Grune and Stratton 19, 1982.
4Mittl RN, Green DH : Blue field entoptoscopy in vitrectomy candidates. Ophthalmic Surgery 18 : 357 - 358, 1987.
5Michels R : Vitreous Surgery : St. Louis, The C.V. Mosby Company 104, 1981.
6Siegel S : Nonparametric statistics for the behavioral sciences. New York, McGraw Hill Book Company 1956.
7Murphy GE : Limitations of blue field entoptoscopy in evaluating macular function. Ophthalmic Surgery 14 : 1033 - 1036, 1983.