ACTA 0 P H T H A L M 0 L O G I CA 68 (1990) 445-449

Screening for glaucoma with a non-mydriatic fundus camera

Anja Tuulonen, P. Juhani Airaksinen, Antonio Montagna and Heikki Nieminen

Department of Ophthalmology (Head: Leila Laatikainen), University of Oulu, Finland.

Abstract. 183 first-degree relatives of glaucoma patients were photographed by a technician with non-mydriatic fundus camera in order to study the suitability of wide angle black-and-white fundus photog-aphs in screening for glaucoma. The success rate of photography was 92%. The optic disc and retinal nerve fiber layer abnormalities were evaluated from the photographs by an ophthalmo- logist. 31 subjects (17%) were referred to further ophthal- mological examinations. We found 6 (3%) new glauco- mas. In addition, in 6 patients (3%) retinal nerve fiber layer defect was the only abnormality, 5 subjects (3%) showed a hemorrhage and 2 eyes had collateral vessels as a sign of asymptomatic venous stasis change at the optic disc. Only 1 of the 6 (17%) patients with glaucoma would have been found with tonometry alone. The results of this study indicate that non-mydriatic retinal camera is a useful tool in screening for glaucoma.

Key words: optic disc - retinal nerve fiber layer - photo- graphy - infrared camera.

Open angle glaucoma is insidious in onset and pa- tients with early stages of disease are often asymp- tomatic. However, patients with glaucoma should be found as early as possible in order to prevent, or at least retard, further visual loss.

Tonometry, visual field testing and ophthalmo- scopy have been used to screen for glaucoma (Becker et al. 1960; Levi & Schwartz 1983; Rosen- thal & Perkins 1985). Tonometry, which is simple and rapid, finds patients with abnormally high in-

traocular pressure. However, all patients with in- creased intraocular pressure do not have glaucoma and all patients with glaucoma do not have raised intraocular pressure (Bengtsson 1981; Shiose et al. 1981; Eddy et al. 1983). Testing of visual fields is more time consuming and definite standards for mass visual field screening with automated peri- metry have not been fully provided (Bengtsson & Krakau 1979; Keltner &Johnson 1983). Ophthal- moscopy has been under-utilized for glaucoma screening (Levi & Schwartz 1983), although with skillful direct ophthalmoscopy early glaucomatous changes can be detected at the optic disc. Photo- graphy has been used to solve the problem of a lack of enough qualified personnel to perform ophthalmoscopy during mass screening. Optic disc viewers or grids on the photographs have been used with success to identify glaucomatous disc changes also by non-medically qualified observers (Shiose et al. 1981; Hitchings et al. 1983).

The purpose of this study was to evaluate the suitability and usefulness of wide angle black-and- white fundus photographs in screening for glau- coma. The photographs were taken by a technician with a non-mydriatic fundus camera. The photo- graphs were evaluated by an ophthalmologist. In addition to optic disc observation, special atten- tion was paid to retinal nerve fiber layer analysis which may reveal the very first structural glauco- matous abnormalities.

445

Fig. 1. Fig. 2. 51-year old patient with a wedge-shaped defect in the in- ferior arcuate nerve fiber layer. Humphrey 30-2 visual field was abnormal and intraocular pressure was 20 mmHg. The patient’s mother has glaucoma and his

brother’s right eye is shown in Fig. 2.

43-year old patient with an optic disc hemorrhage at the the inferior border of the optic disc and a wedge shaped retinal nerve fiber layer defect at corresponding location. Humphrey 30-3 visual field was abnormal and intraocu- lar pressure was 19 mmHg. The patient’s mother has

glaucoma and his brother’s left eye is shown in Fig. 1.

Material and Methods

A Canon CR3-45NM non-mydriatic 45 degree reti- nal camera was used to photograph the subjects’ eyes. With this picture angle one can observe the optic disc and obtain a good overview of the retinal nerve fiber layer (Fig. 1 and 2). The eye is observed in a dark room (less than 5 lux) on a monitor with infrared light and focused by aligning the split lines on the retinal image. Therefore, no mydria- tics were required but the pupil should dilate to 4 mm or more in dark. The photographs were taken on black-and-white Kodak Panatomic-X film with a green filter (Wratten No. 58). The photographs were developed using Agfa Refinal for 7 min in 20°C and enlarged paper-prints were made by the eye clinic’s photographer (HN).

The photography took place in 3 health care centres in northern Finland. A technician from each health care centre was trained to take the photographs during 2 days at the clinic. The camera was transported to the health care centres

446

where both eyes of a total of 183 first-degree rela- tives of glaucoma patients were photographed. The relatives, found according to the registry of glaucoma patients at the eye clinic, were informed either directly or in the local newspaper. The rela- tives of glaucoma patients were chosen because positive family history of glaucoma may increase the risk of developing glaucoma (Becker et al. 1960; Levi & Schwartz 1983; Rosenthal & Perkins 1985) and the family members of glaucoma pa- tients were thought to be easily motivated for photography.

The number of glaucoma patients in the registry was 208 in the 3 communities. At least one relative of 119 of 208 (57%) glaucoma patients of the regis- try was photographed. In addition, many subjects with unchecked positive family history volun- teered for photography. This was not an epidemio- logic study and therefore we did not try to trace all relatives and the diagnosis of glaucoma patients outside of the registry was not confirmed.

The photographs were independently evaluated

by two observers (PJA and AT) and all patients with suspicious or obvious findings were invited to the eye clinic for ophthalmological examination and visual field testing with Octopus G1 program or Humphrey 30-2 program. Due to cooperation problems, in two cases the central visual field was tested with a Friedmann visual field analyzer. Pe- ripheral visual fields were tested with a Goldmann perimeter (II/4 object).

Results

Threehundredfortysix of 366 photographs (95%) were assessable. The quality was good or excellent in 328 of 366 photographs (900/0), fair or poor but assessable in 18 of 366 (5%) photographs. 20 of 366 photographs (5%) were unsuccessful; in 15 of 183 subjects (8%) photography of one or both eyes was unsuccessful. In the younger age groups closing of the lid was the major cause for failure while cata- ract caused problems in the older age groups. There tended to be more failures in the beginning of study in each centre.

It took about 30 min for the ophthalmologist to analyze the photographs of 183 patients (100 photographs / 10 minutes). A total of 31 subjects (17%) were referred to further ophthalmological examinations. The agreement of the referred cases between the 2 observers was 92%. In 19 of 31 (61%) referred cases an abnormal finding typical to glau- coma could be confrmed at the first clinic visit. 2 of the remaining 12 referred subjects had other eye disease than glaucoma; pale optic disc after retro- bulbar neuritis and anterior ischemic optic neuro- pathy.

Six of 183 subjects (3%) had typical glaucoma- tous changes in the optic disc, retinal nerve fiber layer and automated visual fields. In addition, 9 patients (s0/o) had nerve fiber layer abnormalities with normal or suspicious looking optic disc cup- ping. Three of these 9 patients also had abnormal changes in automated visual fields. There were 4 pairs of siblings among the 15 patients with glau- coma andlor retinal nerve fiber layer abnor- malities (Fig. 1 and 2).

A total of 6 subjects were referred because of a hemorrhage or suspicion of a bleeding at the optic disc. Hemorrhage was confirmed in 4 cases. In one eye a vessel loop and in another eye a deposition of pigment was misintepreted as a hemorrhage. In

Table 1. The age distribution of the total material, the referred

group and the group with unsuccessful photography.

Number of patients

Total Age/years

21-30 31-40 41-50 51-60 61-70 71-80 8 1-90

Total

2 (1%) 7 (4%) 117 (140/0)

68 (37%) 9/68 (13%) 7/68 (10Ya) 60 (330/0) 13/60 (22%) 3/60 (5%) 27 (15%) 3/27 (11%) 3/27 (11Oh) 16 (9%) 6/16 (38%) 3 (1%) 113 (33%)

183 31/183 (17%) 15/183 (8%)

addition, one patient who was referred because of suspicious looking optic disc showed a hemor- rhage at the first clinic visit. Two of 5 patients with a hemorrhage showed a localized nerve fiber layer defect corresponding to the location of the bleed- ing (Fig. 2).

Two eyes had collateral vessels as a sign of asymptomatic venous stasis change at the optic disc (Tuulonen 1989). One of them had diffuse reti- nal nerve fiber layer atrophy and normal Hum- phrey 30-2 visual field. In the other the collateral vessels loop was the only abnormal finding.

The age distribution of total material, referred group and subjects with unsuccessful photography is presented in the Table. The mean age ( k SD) of the 6 glaucoma patients was 63 f 13 (range 43-79) years, 60 f 9 (range 48-73) years in the 9 patients with retinal nerve fiber layer abnormalities, and 50 f 5 (range 43-55) in the 5 patients with an optic disc hemorrhage. Among the 31 referred cases 3 sub- jects had intraocular pressure >20 mmHg: one of them had glaucoma, and 2 subjects had suspicious looking optic discs but normal nerve fiber layer and normal visual fields.

Among the 19 patients in whom structural and/ or functional abnormality was confirmed at the first clinic visit, 13 (68%) would have been found by estimating optic disc changes alone, that is, glauco- matous optic disc cupping, optic disc hemorrhage and/or asymptomatic venous stasis changes. In 6 of the 19 (32%) subjects the retinal nerve fiber layer

447

defect was the only detectable abnormality. In 9 of the 19 (47%) patients the automated visual fields were abnormal. Only 1 of the 19 (5%) subjects would have been found with tonometry alone.

Discussion

The results of this study agree with previous studies (Shiose et al. 1981), indicating that non-my- driatic retinal camera with a 45 degree picture angle is a useful tool in screening for glaucoma. In the study of Shiose et al. (1981) trained paramedi- cals photographed the left eye of 11 660 subjects using non-mydriatic fundus camera as a part of automated multiphasic health testing program. The colour slides were then screened by the para- medicals in an image contrasting projector to quantitate abnormalities in pallor, cupping, and retinal nerve fiber layer and to detect splinter he- morrhages. Optic disc and nerve fiber layer abnor- malities were the most specific parameters for de- tection of glaucoma and correctly predicted visual field loss in 68% of the glaucomatous eyes (Shiose et al. 1981).

In the present study we used black-and-white paper prints. Green filter was chosen because it in- creases the contrast at the optic disc and enhances the visibility of the nerve fibers (Airaksinen et al. 1982). The 92% success rate of photography per- formed by a technician after a 2-day trainiig period, and 92% agreement of the findings be- tweens the 2 observers was good.

Among 11 660 subjects screened for glaucoma Shiose et al. (1981) found 39 new glaucomas (0.33%) with non-mydriatic camera alone, a much lower percentage than the 3.3% (6 of 183 subjects) of the present selected study material. We want to" emphasize, however, that this study was not planned as an epidemiological analysis but merely to test the suitability of non-mydriatic black-and- white photography as a method of screening for glaucoma.

With this technique new patients with glaucoma were found, some of them in a very early phase of the disease. There were 6 patients with normal optic discs and normal visual fields but in whom retinal nerve fiber layer analysis showed definite glaucomatous abnormalities.

Clinical information of intraocular pressure and

448

visual fields may significantly bias the evaluation of optic disc photographs (Shrader et al. 1989). Half of the optic discs, considered non-glaucoma- tous with a history of normal intraocular pressure and normal visual fields, were regarded as glauco- matous with a history of increased intraocular pressure and abnormal visual fields when evalu- ated a second time (Shrader et al. 1989). In the present study the analysis of photographs was not biased by intraocular pressure since tonometry was not part of the screening procedure. Only 3 of 31 referred subjects, or 1 of 6 newly detected glau- comas, had increased intraocular pressure at the first clinic visit.

By combining retinal nerve fiber layer analysis and optic disc estimation one can find new glau- coma patients in very early stages of the disease. On the other hand, patients who have suspicious looking optic disc cupping may avoid unnecessary referrals if they have normal retinal nerve fiber layer. Need of no dilation saves time, photography is independent of the subject's response and the analysis of one pair of photographs takes but a few seconds of the ophthalmologist's time.

Acknowledgments

This study was supported by the University Central Hos- pital of Oulu and the Sigrid Juselius Foundation.

References

Airaksinen P J, Nieminen H & Mustonen E (1982): Reti- nal nerve fibre layer photography with a wide angle fundus camera. Acta Ophthalmol (Copenh) 60:

Becker B, Kolker A E & Roth F D (1960): Glaucoma family study. Am J Ophthalmol50: 557-567.

Bengtsson B & Krakau C E T (1979): Automatic perim- etry in a population survey. Acta Ophthalmol (Co- penh) 57: 929-937.

Bengtsson B (1981): The prevalence of glaucoma. Br J Ophthalmol65: 46-49.

Eddy D M, Sanders L E, Eddy J F (1983): The value of screening for glaucoma with tonometry. Surv Ophthal- mol 28: 194-205.

Etchings R A, Brown D B & Anderton S A (1983): Glau- coma screening by means of an optic grid. Br J Oph- thalmol67: 352-355.

362-368.

Keltner J L &Johnson C A (1983): Screening for visual field abnormalities with automated perimetry. Surv Ophthalmol28: 175-183.

Levi L & Schwartz B (1983): Glaucoma screening in the health care setting. Surv Ophthalmol 28: 164- 174.

Rosenthal A R & Perkins E S (1985): Family studies in glaucoma. Br J Ophthalmol69: 664-667.

Shiose Y, Komuro K, Itoh T, Amano M & Kawase Y (1981): New system of mass screening of glaucoma as part of automated multiphasic health testing services. Jpn J Ophthalmol35: 160-177.

Shrader R R, Pistoia 0 A, Nicholl J E & Steinmann W C (1989): The effect of clinical information on optic disc assessment. 1989 Annual Meeting of the American Academy of Ophthalmology. Ophthalmology (Suppl), p 125.

Tuulonen A (1989): Asymptomatic miniocclusions of the optic disc veins in glaucoma. Arch Ophthalmol 107: 1475- 1480.

Received on February 22nd, 1990.

Author's address:

Anja Tuulonen, MD, Department of Ophthalmology, University of Oulu,

Finland. SF-90220 Oulu,

29 Acta Ophthal. 68.4 449