ORIGINAL ARTICLE

Pdf version

COMPARISON BETWEEN THE PANOPTIC OPHTHALMOSCOPE AND THE CONVENTIONAL DIRECT OPHTHALMOSCOPE IN THE DETECTION OF SIGHT THREATENING DIABETIC RETINOPATHY: THE KUCHING DIABETIC EYE STUDY

STATISTICAL ANALYSIS

Sensitivity and specificity of PO and CO in the detecting of STR were calculated using standard formula.

Sensitivity =

Specificity =

The 95% confidence interval of sensitivity and specificity were obtained using the biconf.exe; a stand-alone MS-DOS program which calculates exact confidence intervals for Binomial proportions and Poisson rates. The biconf.exe was downloaded free from the following Uniform Resource Locator (URL): http://www-users.york.ac.uk/~mb55/soft/soft.htm.

'Ease of use' scores were compared using chi-squared test and the method of comparison of proportion. The result was reported in 95% confidence interval for the ratio of the difference.

RESULTS

A total of 200 patients were enrolled in this study, of which 197 (98.5%) had T2DM and 3 (1.5%) had T1DM. The right eye was examined in 173 (86.5%) of patients, the left 27 (13.5%). The mean age of patients was 57.2 ± 12.1 years old (ranged from 20 to 81 years). 183 (91.5%) patients were aged 40 and above. 88 (44%) patients were male, 112 (56%) female. 105 (52.5%) were Chinese, 62 (31%) were Malay, 19 (9.5%) were Iban, and 14 (5.2%) were of other ethnic groups (Table 1).

Table 1: Age, ethnic and gender distribution of patients

Age group (years)	Malay		Chinese		Iban		Others		Total(%)
	Male	Female	Male	Female	Male	Female	Male	Female
20-30	0	2	0	1	1	1	0	0	5 (2.5)
30-39	2	5	1	1	0	1	1	1	12 (6.0)
40-49	3	8	9	1	3	0	2	2	28 (14.0)
50-59	12	13	16	17	2	3	4	2	69 (34.5)
60-69	4	7	12	19	1	4	1	1	49 (24.5)
70-79	2	4	11	15	0	3	0	0	35 (17.5)
80-89	0	0	1	1	0	0	0	0	2 (1.0)
Total	23	39	50	55	7	12	8	6	200 (100)

The duration of known diabetic status ranged from 0 (newly diagnosed DM) to 41 years. 45 (22.5%) patients were referred within one year of diagnosis. STR was detected in 41 (20.5%) cases, where 11 cases were due to DR alone, 16 due to diabetic maculopathy alone, and 14 due to the combination of both. 92.7% (38 cases) of STR, occur in patients aged 40 and above (Table 2).

Table 2: The prevalence of Sight Threatening Retinopathy (STR) according to age groups

Age group (years)	STR	No STR	Total
20-30	1	4	5
39-39	2	10	12
40-49	7	21	28
50-59	17	52	69
60-69	9	40	49
70-79	4	31	35
80-89	1	1	2
Total	41	159	200

Sensitivity and specificity
The examinations were performed using the PO as the initial investigation tool in 54% of eye, the CO in 46% of eyes (p=0.1). For CO, the overall sensitivity in detecting STR was 73.2% (95% CI: 57.1-85.8%), specificity 93.7% (95% CI: 88.7-96.9%), false negative 26.8% (95% CI: 14.2-42.9%), false positive 6.3% (95% CI: 3.0-11.3%), positive predictive value 75.0% (95% CI: 58.8-87.3%), and negative predictive value 93.1% (95% CI: 88.0-96.5%) (Table 3).

Table 3: The detection of Sight Threatening Retinopathy (STR) with Conventional Direct Ophthalmoscope (CO)

CO		SLM		Total
		STR	No STR
No cataract	STR	21	7	28
	No STR	9	88	97
Cataract	STR	9	3	12
	No STR	2	61	63
Total		41	159	200

SLM: Slit Lamp Biomicroscopy

For PO, the overall sensitivity in detecting STR was 58.5% (95% CI: 42.1-73.7%), specificity 93.7% (95% CI: 88.7-96.9%), false negative 41.4% (95% CI: 26.3-57.9%), false positive 6.3% (95% CI: 3.0-11.3%), positive predictive value 70.6% (95% CI: 52.5-84.9%), and negative predictive value 89.8% (95% CI: 84.2-93.9%) (Table 4).

Table 4: The detection of Sight Threatening Retinopathy (STR) with PanOptic Ophthalmoscope (PO)

PO		SLM		Total
		STR	No STR
No cataract	STR	18	8	26
	No STR	12	87	99
Cataract	STR	6	2	8
	No STR	5	62	67
Total		41	159	200

SLM: Slit Lamp Biomicroscopy

A total of 75 (37.5%) eyes have no cataract, while 125 (62.5%) eyes have mild cataract. Sub-group analysis revealed the sensitivity for the detection of STR using CO was 70.0% (95% CI: 50.6-85.2%) for eyes without cataract, and 81.8% (95% CI: 48.2-97.7%) for eyes with cataract. The sensitivity for the detection of STR using PO was 60.0% (95% CI: 40.6-77.3%) for eyes without cataract, and 54.5% (95% CI: 23.4-83.2%) for eyes with cataract. The differences were not statistically significant.

Ease of use
'Ease of use' for each examination with the PO and CO was scored. A subjective score of eight and above is considered easy, and a score of seven and below is considered difficult.¹³

The median score was greater when using the CO (median 9, IQR 8-9) compared to the PO (median 8, IQR 7-9) (Table 5). 82.5% of examinations using the CO were rated as eight or nine, compared to PO (60.0%) (P< 0.0001). The CO was 1.38 times (95% CI: 1.17-1.61 times) as easy to use compared to the PO.

Table 5: Frequency for “Ease of use” scores with PanOptic Ophthalmoscope (PO) and Conventional Direct Ophthalmoscope (CO)

Ease of use score	Frequency
	PO	CO
1	0	0
2	0	0
3	0	0
4	0	0
5	8	3
6	39	13
7	33	19
8	49	59
9	71	106
Total	200	200

DISCUSSION

The primary objectives of this study were to determine, the sensitivity and specificity of the CO and the PO in the detection of STR. The secondary objective was to determine the “Ease of Use” of these equipments.

The results of this study indicated that the CO and the PO are comparable in sensitivity and specificity for the detection of STR. The overall sensitivity in detecting STR using the PO was 58.5% (95% CI: 42.1-73.7%), and specificity 93.7% (95% CI: 88.7-96.9%).

Screening tools with low sensitivity or high specificity are far from ideal. Many STRs would be missed due to high false negative rate. Operator dependence and higher false negative value suggest that the PO is not a good screening tool.

This study found that the CO was 1.38 times (95% CI: 1.17-1.61 times) as easy to use compared to the PO. This is in contrast to previous study which found that the PO was easier to use.¹³ The “ease of use” of the PO and the CO were related to both the instruments and the patients. Instrument factors include intensity of illumination, working distance, as well as magnification of image and field of view. The intensity of illumination can be adjusted with the rheostat. The working distances were almost constant for each instrument. Magnification of image and the field of view achieved, depend on the refractive error of the patient. Patient factors include the type and the degree of refractive error, the degree of media opacity, photophobic response to ophthalmoscope light, the degree of patient’s cooperation during examination, the degree of pupil dilation and the amount of light entering the fundus, hence the quality of the fundus view. The cumulative effects of these factors on each examination were translated into the “ease of use” score. Hence, the “ease of use” score will vary from patient to patient. Overall, in this study, the CO was found to be easier to use than the PO in the screening for sight threatening DR.

The working distance of the CO is 2 cm, whereas the working distance of the PO is 13 cm, about six times the working distance of the CO. Greater working distance allows more comfort during examination. Although the PO provides a wider field of view, the brightness and quality of the image is poorer than that of the CO, especially in the presence of media opacities. This is undesirable in the screening of STR, as lesions such as new vessels and macular oedema could be easily missed. The brightness of the CO is about 500 lux, the PO is slightly (30%) brighter due to the halogen HPXTM lamp. Both equipments use 3.5 volt power source. The inverse square law predicts the relationship between the brightness and the distance from the light source. The brightness of the image decreases by a factor of four, when the distance between the illuminated retina and the examiner doubles. Assuming that the illuminated retina reflects light at the same luminous intensity, the brightness of the image seen by the PO is 36 times dimmer than those seen by the CO.

The CO allows greater flexibility in changing the angle of illumination. The eye-cup of the PO was not used in the study although its usage is recommended in the product usage guideline. The manufacturer claimed that the eye-cup provides a dark room effect, stabilizes the instrument and establishes a proper viewing distance while manoeuvring and focusing.

On the contrary, we found that the eye-cup failed to serve its purpose. If examination of an undilated eye is carried out in room light, the undilated pupil will still constrict due to the presence of consensual pupillary light reflex. Stimulation of the contra-lateral eye by ambient room light, will inevitably lead to constriction of both pupils. We strongly recommend examinations be performed with eyes fully dilated in a dark room.

In addition, PO with its eye-cup, allow little control over the angle of illumination, especially in eyes with undilated pupil. The eye- cup is actually a compressible rubber cup. The viewing distance provided by the eye-cup is therefore not constant. Stabilization of the PO viewing distance with the other hand, allow minor adjustment to provide the best fundus view. This simple manoeuvre also allows the examiner to monitor the patients’ eye as they gaze in the desired direction, in order to bring fundus lesions into view. If the eye-cup is used, the examiner will need to remove the PO in order to ensure proper direction of gaze. Therefore, the examination with the PO is best performed without the eye-cup.

The CO is superior in viewing the red reflex. The quality of red reflex is determined by the amount of light that enter the pupil and hence reflected back from the fundus.

The illumination system of the CO sends a diverging cone of light rays to enter the subject’s eye. On the other hand, to allow easy entry into small pupil, the Axial PointSource Optics Illumination system of the PO converge the light to a point at the cornea (1.45 inches in front of the PO). The illumination pathway then diverges widely to illuminate a very wide area of the fundus.11 At a working distance of 30 cm, the amount of light rays that enter the pupil from the PO is greatly reduced, as most are diverged away from the visual axis. Therefore, red reflex is better evaluated with the CO (Figure 1).

Figure 1: Ray diagram during red reflex examination. A: Conventional direct ophthalmoscope. B: PanOptic ophthalmoscope.

Limitations

There are several limitations that must be considered in the interpretation of the results of this study. The PO, CO and SLM are operator dependent as well as patient dependent. Inter-observer variation and intra-observer variation is bound to occur. Gill et al. showed that there is a wide range of kappa statistic, between 0.06 to 0.70, among family physicians in the screening for DR.¹² In order to avoid inter-observer variation in this study, only one investigator performed all three examinations. If the lone investigator had any preconceived bias against any one of the instruments, the result will suffer.To maintain masking, there was no history taking prior to PO and CO examination. There was only one CO, one PO and one slit lamp used in the study.

The investigator was given a PO for three months for him to familiarize himself with the equipment. The three months period may not be adequate compared to his presumed long experience with the CO.

Since DM is a systemic disease, both eyes are equally affected in most patients. The knowledge of the DR status in one eye will influence the examiner’s judgment of the DR status in the contra-lateral eye. Therefore, only one eye from the patient will be included in this study.

Patient factor include decrease level of co-operation during repeated examination. This can be due to photophobia or reluctance for repeated examination. In the Liverpool Diabetic Eye Study, 9% of patients failed to attend hospital clinic for repeat examination. Of the 91% who attended, 9% did so after four months.⁸ The defaulter rate is expected to be higher in our population. Taking account of the above difficulties, data collection for each patient was completed in one session. The investigator may remember the fundus finding during the first examination, this constitutes a limitation of the study. Coin flipping was used to mitigate this bias.

CONCLUSION

The results of this study showed that the PO is not superior to the CO for the screening of STR, and do not support the use of the PO for this purpose.

ACKNOWLEDGEMENT

This study forms part of the fulfilment for the degree of Master of Surgery (Ophthalmology) Universiti Kebangsaan Malaysia (UKM). The authors would like to thank the following parties for their generous advice and support for this study:

1. Director General of Health, Ministry of Health, Malaysia (approval for publication of this manuscript).
2. Clinical Research Centre (CRC) Kuching, Sarawak.
3. Dr Ooi Chuo Huck, Head of Department of Epidemiology, Sarawak State Health Department.
4. Professor Dr Awang Bulgiba Awang Mahmud, Professor and Head, Dept of Social and Preventive Medicine, University of Malaya.
5. Dr Aye Aye Aung, Lecturer in Community Medicine and Public Health, Faculty of Medicine and Health Sciences, UNIMAS.
6. Sekretariat Penyelidikan Perubatan dan Industri, Pusat Perubatan Universiti Kebangsaan Malaysia.

REFERENCES

1. The National Eye Survey, 1996. Disease Control Division, Ministry of Health Malaysia.
2. Zimmet P, Alberti KG, Shaw J. Global and societal implications of the diabetes epidemic. Nature. 2001;414(6865):782-7. [PubMed]
3. Malaysian Clinical Practice Guideline: Diabetic Retinopathy; 1996.
4. Photocoagulation treatment of proliferative diabetic retinopathy. Clinical application of Diabetic Retinopathy Study (DRS) findings, DRS Report Number 8. The Diabetic Retinopathy Study Research Group. Ophthalmology. 1981;88(7):583-600. [PubMed]
5. Lee CM, Olk RJ. Modified grid laser photocoagulation for diffuse diabetic macular edema. Long term visual results.  Ophthalmology. 1991;98(10):1594-602. [PubMed]
6. Photocoagulation for diabetic macular oedema. Early Treatment Diabetic Retinopathy Study Research Group. Int Ophthalmol Clin. 1987;27:265-72.
7. Hutchinson A, McIntosh A, Peters J, et al. Effectiveness of screening and monitoring tests for diabetic retinopathy--a systematic review. Diabet Med. 2000;17(7):495-506. [PubMed]
8. Harding SP, Broadbent DM, Neoh C, et al. Sensitivity and specificity of photography and direct ophthalmoscopy in screening for sight threatening eye disease: the Liverpool Diabetic Eye Study. BMJ. 1995;311(7013):1131-5. [PubMed] [Full text]
9. Bresnick GH, Mukamel DB, Dickinson JC, et al. A screening approach to the surveillance of patients with diabetes for the presence of vision-threatening retinopathy. Ophthalmology. 2000;107(1):19-24. [PubMed]
10. Diabetic Retinopathy Guidelines, 2005. Amalgamated Literature Review on the Management of Diabetic Retinopathy. Centre for Vision Research, University of Sydney (Westmead Hospital); 2005.
11. PanOptic Ophthalmoscope: A difference you can see. [Link]
12. Gill JM, Cole D, Lebowitz HM, et al. Accuracy of screening for diabetic retinopathy by family physicians. Ann Fam Med. 2004;2(3):218-20. [PubMed] [Full text]
13. McComiskie JE, Greer RM, Gole GA. Panoptic versus conventional ophthalmoscope. Clin Exp Ophthalmol. 2004;32(3):238- 42. [PubMed]
14. Klein R, Klein BE, Magli YL, et al. An alternative method of grading diabetic retinopathy. Ophthalmology. 1986;93(9):1183-7. [PubMed]
15. Moss SE, Meuer SM, Klein R, et al. Are seven standard photographic fields necessary for classification of diabetic retinopathy? Invest Ophthalmol Vis Sci. 1989;30(5):823-8. [PubMed]
16. Younis N, Broadbent DM, Vora JP, et al. Incidence of sight-threatening retinopathy in patients with type 2 diabetes in the Liverpool Diabetic Eye Study: a cohort study. Lancet. 2003;361(9353):195-200. [PubMed]
    
17. Younis N, Broadbent DM, Harding SP, et al. Incidence of sight-threatening retinopathy in Type 1 diabetes in a systemic screening programme. Diabet Med. 2003;20(9):758-65. [PubMed]
18. Bland M. An introduction to medical statistics. 3rd ed. New York: Oxford University Press; 2000.