Secondary Logo

Journal Logo


Prevention of Corneal Ulceration in the Developing World

Whitcher, John P. M.D., M.P.H.; Srinivasan, M. M.D.; Upadhyay, Madan P. M.D.

International Ophthalmology Clinics: January 2002 - Volume 42 - Issue 1 - p 71-77

Corneal blindness in the developing world has traditionally been attributed to trachoma, xerophthalmia, measles, neonatal ophthalmia, and leprosy. The importance of superficial corneal trauma in agricultural work, which frequently leads to rapidly progressing corneal ulceration and visual loss, has been largely overlooked as a worldwide cause of unilateral blindness. However, Thylefors 1 estimated that up to 5% of all blinding conditions are directly related to ocular trauma and that children may be at greater risk than adults. A recent study in Uganda by Waddell 2 indicated that corneal ulceration was second only to cataract as the main cause of blindness in the younger age group. Examination of 1,135 children with subnormal vision revealed that 30.7% had visual impairment secondary to cataracts or poor surgical outcome, whereas 22.0% had visual loss after corneal ulceration.

The social and economic impact of corneal ulceration is insidious, because it causes bilateral blindness infrequently. The individuals most commonly affected are farmers or laborers; men are at slightly more risk than women; and the ulcers occur predominantly in middle-aged individuals who are in their most productive years. 3,4 Because it affects predominantly the working poor in the developing world, microbial keratitis is underreported and often neglected. A corneal ulcer can cause economic disaster for a family on a subsistence income. Farmers and laborers, who are not covered by any kind of health insurance, may have to spend all their meager wages on medical treatment and be out of work for weeks to months, ultimately driving their families into utter poverty. By conservative estimates, at least 1.5 million new cases of unilateral blindness occur worldwide every year owing to microbial keratitis. The affected individuals, who are truly victims of a “silent epidemic,”5 return to their previous jobs where they remain exposed to the same risk factors for developing ulceration in the other eye.

The medical costs for treating microbial keratitis are almost always prohibitively expensive, especially in the case of fungal ulceration, for which medications are costly and the visual outcome is usually disappointing. In South India, approximately one-half of eyes with fungal ulceration that receive appropriate treatment still develop blinding sequellae secondary to perforation, phthisis, and severe scarring. Bacterial ulcers fare somewhat better, but central corneal scarring effectively means that an individual is unilaterally blind for life because of the prohibitive cost of corneal transplantation in most developing countries, the lack of facilities and trained surgeons, and the scarcity of donor tissue for such transplantation. The true economic costs of unilateral blindness secondary to corneal ulceration are not known. Affected individuals are undoubtedly at greater risk for injury to the fellow eye and are more prone to accidents in general. In the case of bilateral blindness in children, the social and economic impact over a lifetime is incalculable.

Epidemiological Features

Most corneal ulcers in the developing world occur after relatively minor corneal trauma. 3,4 Microbial pathogens then are introduced into the corneal stroma by contamination of the object producing the abrasion, by the presence of pathogens in the environment, or by the patient's own conjunctiva or eyelids. Immediately after the corneal abrasion occurs, there is a window of opportunity during which one can interfere with microbial invasion before keratitis develops. If an infection becomes established in the corneal stroma, the rapid growth of the microbial pathogens and the resulting tissue necrosis that occurs make medical reversal of the process more difficult. The inflammatory mechanisms called into play by the human body invariably produce rapid necrosis of the corneal stroma, which then heals slowly, with the development of severe scarring over weeks to months.

Few surveys have investigated the incidence of corneal ulceration, 5 and the spectrum of microbial pathogens producing ulceration varies from population to population, as shown by large studies in Asia and Africa (table). Significantly, the most common bacterial pathogen causing corneal ulceration in the majority of these studies was Streptococcus pneumoniae. In a series from South India, half of all the ulcers were caused by fungal pathogens, with Fusarium species being the most commonly isolated organisms. 4 Only one population-based retrospective incidence survey has been reported from a developing country. Gonzales and colleagues 6 found that the incidence of corneal ulceration in Madurai District, South India, was 113 per 100,000 per year or ten times the incidence in the United States. Generalizing these findings to all of India, an estimated 840,000 people develop a corneal ulcer annually in that country.

Table 1
Table 1:
Central Corneal Ulcers in the Developing World: Geographical Comparison

A number of risk factors may predispose an individual in a developing country to microbial keratitis. Geographical environmental, social, and ethnic differences are usually of great importance. Undoubtedly, however, the most important risk factor is a history of corneal trauma. Unlike in the industrialized world, where contact lens wear is the main risk factor for developing corneal ulceration, in the developing world contact lenses are worn by a relatively small portion of the population. In contrast, the majority of the population is involved in agricultural labor, and the opportunities for repeated incidences of corneal trauma are numerous. Geographical areas that are consistently warm and humid, such as South India, tend to have a greater number of fungal ulcers, whereas regions with more temperate climates, such as the mountainous areas of Nepal, have a preponderance of bacterial ulcers. Individuals such as farmers, day laborers, and brick or stone workers are prone to frequent corneal abrasions. Often, the poorest people in any given community tend to have a higher risk for microbial keratitis because of their occupational exposure and their lack of ready access to the health care system.

Diagnosis and Treatment

Microbial keratitis, by definition, is a suppurative infection of the corneal stroma with an associated epithelial defect and signs of inflammation. The pathogens usually responsible are bacterial, fungal, or parasitic. The herpesviruses and other viral pathogens usually are excluded from this definition. Even though good epidemiological evidence does not exist, it is usually assumed that herpes simplex virus is an uncommon cause of corneal blindness in developing countries. Patients who develop a corneal ulcer usually give a history of redness, pain, and sensitivity to light, with an associated decrease in visual acuity. Although the distinction between central corneal ulcers and peripheral ulcers is somewhat arbitrary, microbial keratitis does not have to occupy the visual axis to be included in the corneal ulceration category.

Corneal ulcers are characterized by the presence of a white or yellowish stromal infiltrate or infiltrates with an associated epithelial defect. The eye exhibits signs of inflammation: injection, an anterior chamber reaction and, possibly, a hypopyon. The patient is in pain and usually presents as an acute emergency. If there is a delay in receiving treatment, as was seen frequently in a study in Nepal where patients sometimes had to walk for weeks to get to a hospital, 3 total corneal necrosis may be seen on presentation, with corneal perforation and endophthalmitis. Because of the delay in diagnosis and treatment, microbial keratitis in developing countries is often much more severe on initial presentation. Both bacterial and fungal ulcers may present with a large area of central necrosis and a significant hypopyon. Satellite lesions with soft feathery edges are a prominent sign of fungal keratitis in developing countries but, in general, the ulcers seen under these circumstances are much more aggressive than those seen in industrialized countries and quickly perforate. Likewise, bacterial ulcers progress rapidly to involve the entire cornea. Medical treatment then becomes problematic because, even if the ulcer is sterilized, necrosis is usually so extensive that surgical intervention may be necessary to maintain the integrity of the eye.

If culture facilities are not available, corneal scrapings can be of great help in making a tentative etiological diagnosis. Staining a bacterial ulcer smear with Gram's or Giemsa stain can be performed with minimal equipment. Examining corneal scrapings from fungal ulcers is greatly facilitated by using a KOH wet mount. Fortunately, Acanthamoeba is a rare cause of corneal ulceration in the developing world, but the organisms can still be identified from direct corneal scrapings stained with Giemsa stain.

In general, physicians and health care workers in developing countries are incredibly dedicated. However, they often lack training and even the simplest material resources. Equipment and instruments often are antiquated and in need of repair, and medications are limited and prohibitively expensive. Successful treatment of microbial keratitis is problematic under the best of circumstances. In the setting of a developing country where everything is in short supply, the odds are stacked against successful therapy. Because laboratory facilities often are inadequate or unavailable, the treatment of corneal ulcers is frequently empirical.

The choice of antibiotics for the treatment of bacterial ulcers must cover S. pneumoniae, as this is the most common corneal pathogen in many parts of the world. However, the gram-negative pathogens, such as pseudomonads, cannot be overlooked because of the severity of the keratitis that they produce. Any empirical therapy for a bacterial ulcer should include cefazolin or the equivalent to cover gram-positive organisms and a fluoroquinolone or an aminoglycoside to cover gram-negative pathogens. Antifungal medications are more problematic because they frequently are unavailable. Pimaricin in a 5% suspension may sometimes be in stock, and amphotericin B drops can be made up as a 0.15% topical solution. These two medications are frequently used together to treat filamentous fungi, but yeasts respond best to amphotericin B alone. Empirical antibiotic treatment of any central ulcer without first ruling out the possibility of a fungal infection invites disaster, especially in areas such as South India where fungal pathogens cause half of all the ulcers seen. Obtaining corneal scrapings and staining the smears with Gram's or Giemsa stain to make an etiological diagnosis prior to initiating therapy is critical in avoiding treatment disasters.


Because of the often prohibitive cost of treating microbial keratitis in a developing country and the invariably blinding corneal scarring that occurs, prevention is the ideal approach for reducing the epidemic numbers of corneal ulceration now occurring. Theoretically, the majority of ulcers could be prevented if a combination antibiotic-antifungal ointment were instilled into the eye immediately after an individual suffered a corneal abrasion. A large-scale program could be implemented through the existing village health care systems in many countries, and public awareness of ulcer prevention could be increased using the eye health education programs already in place. Unfortunately, a commercially available antibiotic-antifungal eye ointment does not currently exist and, until recently, no study had been done that proved that the incidence of corneal ulceration could be significantly reduced by prophylactic treatment. Upadhyay and coworkers 7 have recently reported the results of such a study in Nepal, however. They demonstrated that the application of chloramphenicol ointment three times daily for 3 days in eyes of patients who had suffered corneal abrasions dramatically reduced the rate of corneal infection after corneal abrasion. Moreover, none of the patients who presented for treatment within the first 18 hours after injury developed an ulcer but, as this time interval increased, more cases of ulceration occurred. Even though ethical considerations prevented the use of a control group, the large number of patients surveyed over the 2-year period of this prospective population-based study provides the strongest evidence thus far that antibiotic prophylaxis may be effective in preventing corneal ulceration after abrasion, if the treatment is administered in a timely fashion.


The problem of microbial keratitis in the developing world should be handled at the grass-roots level. Gonzales and associates 6 found that many of the ulcers in South India were treated by village healers and nonophthalmologists who had little knowledge of disease pathogenesis or standard methods of therapy. Remedies used by traditional healers often enhance the development of microbial keratitis after corneal abrasion by introducing bacterial or fungal pathogens in contaminated solutions or organic materials. 4 With simple training, most health care workers would be able to administer prophylactic ointment after corneal abrasions, and they could also be instructed in how to recognize corneal ulcers early during the first stages of their development. In most cases, personnel are already in place, but they require better training and specific instructions for referring suspected cases of microbial keratitis to regional health centers for treatment.

Until recently, the enormity of the problem of microbial keratitis in the developing world was not appreciated. Population-based studies sponsored by the World Health Organization are now under way in several countries in Southeast Asia to define the true incidence of corneal ulceration, to examine the importance of risk factors, and to determine the most common pathogens responsible for ulceration in each geographical area. As the epidemiology of microbial keratitis in the developing world becomes better understood, prevention and treatment programs that are evidence-based can be implemented. It is hoped that by that time new antifungal-antibiotic medications will be available for large-scale prevention programs and that more effective and affordable antimicrobial agents will be available for treating those ulcers that elude prevention.


1. Thylefors B. Epidemiological patterns of ocular trauma. Aust NZ J Ophthalmol 1992; 20: 95–98
2. Waddell K. Childhood blindness and low vision in Uganda. Eye 1998; 12: 184–192
3. Upadhyay M, Purna C, Karmacharya S, et al. Epidemiologic characteristics, predisposing factors, and etiologic diagnosis of corneal ulceration in Nepal. Am J Ophthalmol 1991; 111: 92–99
4. Srinivasan M, Gonzales C, George C, et al. Epidemiology and aetiologic diagnosis of corneal ulceration in Madurai, South India. Br J Ophthalmol 1997; 81: 965–971
5. Whitcher J, Srinivasan M. Corneal ulceration in the developing world—a silent epidemic. Br J Ophthalmol 1997; 81: 622–623
6. Gonzales C, Srinivasan M, Whitcher J, et al. Incidence of corneal ulceration in Madurai District, South India. Ophthalmic Epidemiol 1996; 3: 159–166
7. Upadhyay M, Karmacharya S, Koirala S, et al. The Bhaktapur eye study: ocular trauma and antibiotic prophylaxis for the prevention of corneal ulceration in Nepal. Br J Ophthalmol 2001; 85: 388–392
8. Carmichael TR, Wolpert M, Koornhof HJ. Corneal ulceration at an urban African hospital. Br J Ophthalmol 1985; 69: 920–926
    9. Ormerod DL. Causation and management of microbial keratitis in subtropical Africa. Ophthalmology 1987; 94: 1662–1668
      10. Dunlop AAS, Wright ED, Howlader SA, et al. Suppurative corneal ulceration in Bangladesh: a study of 142 cases examining the microbiological diagnosis, clinical and epidemiological features of bacterial and fungal keratitis. Aust NZ J Ophthalmol 1994; 22: 105–110
        11. Hagan M, Wright E, Newman M, et al. Causes of suppurative keratitis in Ghana. Br J Ophthalmol 1995; 79: 1024–1028
        © 2002 Lippincott Williams & Wilkins, Inc.