|LETTER TO EDITOR
|Year : 1997 | Volume
| Issue : 4 | Page : 259
Estimating cataract blindness in India
R Thomas, J Muliyil
Source of Support: None, Conflict of Interest: None
|How to cite this article:|
Thomas R, Muliyil J. Estimating cataract blindness in India. Indian J Ophthalmol 1997;45:259
| Editor:|| |
A landmark study that used distant direct ophthalmoscopy (DDO) to detect cataract estimated the annual incidence of cataract blindness in India at 3. 8 million. Another survey using a retinoscope for DDO estimated that prevalence of cataract blindness had increased by 90% (Limburg H, personal communication).
DDO has good interobserver variability, but its use in cataract detection has not been validated. In a study (unpublished data) we determined the sensitivity and specificity of DDO to be 96% and 87%, respectively. After adjusting for this, the true probability of cataract was determined using the following formula:
where PD is the probability of disease, and PT+ is the probability of being tested positive.
If 10% of subjects test positive (cataract on DDO) and even if sensitivity is 100% and specificity 96%:
This is a 60% overestimate. We used a higher figure for specificity than our study as vision is usually tested during surveys and increases the specificity. The true specificity in field conditions is unlikely to be this high.
When the number of cataract operations is the yardstick by which ophthalmologists are judged, the situation becomes hauntingly reminiscent of the target oriented family planning drives in the 1970s; we must be reasonably sure that the estimates are indeed correct. As seen above, the specificity of the diagnostic test may seriously distort the actual prevalence of the problem and explain the alarming incidence rate and secular trends in the frequency of cataract. Till accurate data gathered in an appropriate manner (vision with pinhole; slit lamp and DDO to make the diagnosis) is available, it would seem reasonable to use the available data after adjusting for the validity (sensitivity and specificity) of DDO.
We would be happy to provide the full text of the intended article and formula derivation to interested readers, especially those who may share similar concerns.
| References|| |
Minassian DC, Mehra V. 3. 8 million blinded by cataract each year: projections from the first epidemiological study of incidence of cataract blindness in India. Br J Ophthalmol
Mehra V, Minassian DC. A rapid method of grading cataract in epidemiological studies and eye surveys. Br J Ophthalmol