Graefe's Archive for Clinical and Experimental Ophthalmology

, Volume 242, Issue 3, pp 204–209

Epidemiological characteristics of microbiological results on patients with infectious corneal ulcers: a 13-year survey in Paraguay

Authors

  • Florentina Laspina
    • Research Institute of Health-Science (IICS)National University of Asunción
  • Margarita Samudio
    • Research Institute of Health-Science (IICS)National University of Asunción
  • Diógenes Cibils
    • Department of OphthalmologyNational University Hospital
  • Christopher N. Ta
    • Department of OphthalmologySchool of Medicine, Stanford University
  • Norma Fariña
    • Research Institute of Health-Science (IICS)National University of Asunción
  • Ramona Sanabria
    • Research Institute of Health-Science (IICS)National University of Asunción
  • Volker Klauß
    • Department of OphthalmologyLudwig-Maximilian University
    • Department of OphthalmologySchool of Medicine, Stanford University
    • Department of OphthalmologyLudwig-Maximilian University
Clinical Investigation

DOI: 10.1007/s00417-003-0808-4

Cite this article as:
Laspina, F., Samudio, M., Cibils, D. et al. Graefe's Arch Clin Exp Ophthalmol (2004) 242: 204. doi:10.1007/s00417-003-0808-4
  • 125 Views

Abstract

Background

This is a retrospective, chart-reviewed study of patients diagnosed with infectious corneal ulcers at the Ophthalmology Department of the National University of Asunción in Paraguay. The microbiological culture results are described, as well risk factors for the development of fungal keratitis.

Methods

After obtaining approval from the Institutional Review Board, an analysis of medical charts from 1988 to 2001 was conducted and 660 patients were identified to have been diagnosed with infectious corneal ulcers due to bacteria or fungi. Demographic data were recorded, including age, gender, occupation and geographic location of their home and work (city or rural). Other information collected included the history of the presenting illness, past and current use of ocular medications and whether or not they had a history of trauma or contact lens use. Each patient had an eye examination performed by an ophthalmologist and corneal scrapings were obtained for cultures in all cases. Microbiologic culture results were analyzed.

Results

Twenty-one percent (136/660) of the specimens collected from the patients’ conjunctiva and cornea were sterile in all culture media. Of the 524 (79%) positive cultures, 267 were due to bacteria (51%), 136 to fungi (26%), and 121 (23%) cultures yielded both fungi and bacteria. Of the 430 isolated bacteria approximately 25% (103) were coagulase negative Staphylococcus, followed by 23% (94) Staphylococcus aureus, 14% (60) Pseudomonas aeruginosa and 13% (56) Streptococcus pneumoniae. Acremonium species accounted for 40% (79) of all fungi identified, followed by Fusarium species (15%) (41). Approximately two-thirds of the patients were male (n = 435). For those patients for whom a history was available, approximately half had a history of trauma. Of these, half of these again involved foreign bodies. Over-the-counter medications were used commonly, and most of those patients had a delay in diagnosis of over 1 week. Risk factors for fungal keratitis as opposed to bacterial keratitis were male gender, agricultural occupation, age between 30 and 59, history of trauma and self-medication.

Conclusions

The results of this study provide demographic data on patients with infectious corneal ulcers in Paraguay. Common causes of such ulcers are both bacteria and fungi. Most patients had self-medicated, and most had delayed seeking professional medical care.

Introduction

Infectious corneal ulcer is a sight-threatening condition that requires immediate medical treatment. In Nepal, corneal ulceration is the major cause of blindness, second only to cataracts [6]; furthermore, in 70% of child blindness in Africa, the cause is corneal opacification [7]. Thylefors et al. have estimated that up to 5% of blindness is related directly to ocular trauma and subsequent infection [19], and others have reported similar results in several population-based studies carried out in Africa and Asia [1, 2, 3, 4, 9, 10, 13, 16, 17]. However, there are few published studies evaluating the epidemiological and risk factors that predispose patients to corneal infections [5, 11, 14, 15, 20].

A report from India showed that 44% of all central corneal ulcers are caused by fungi [8]. This high prevalence of fungal pathogens in south India is significantly greater than that found in similar studies in Nepal (17%) [20], Bangladesh (36%) [6] and south Florida (35%) [19], although this may be due to differences in geographic and/or health-care delivery. Several large studies have been published about patients in North and South America, Europe and India who have corneal ulcers. However, there is a paucity of data on corneal ulcers in patients from Central America.

The etiologic factors that predispose a patient to infectious corneal ulceration and the causative pathogenic organisms must be determined in order to develop a strategy for the diagnosis, treatment and prevention of corneal infections. The goal of this study was to examine the demographics of patients with corneal ulcers and the results of cultures of corneal scrapings of these patients performed at a major hospital in Paraguay over a 13-year period.

Patients and methods

We performed a retrospective chart review of all patients who were diagnosed at the Ophthalmology Department of the National University of Asunción, Paraguay, from March 1988 to August 2001 as having a bacterial or fungal corneal ulcer. Infectious corneal ulcer was defined as a stromal infiltrate with an associated overlying epithelial defect. Corneal ulcers that were not caused by bacteria or fungi were excluded. These included those caused by herpes and interstitial keratitis, as well as neurotrophic ulcers, Mooren’s ulcers and those caused by systemic autoimmune conditions.

A total of 660 patients were diagnosed with infectious corneal ulcers most likely caused by bacteria or fungi. Data collected from chart reviews included the patient’s age, gender, occupation and geographic location (city or rural) of his/her home and work. Other information collected included a history of the presenting illness, past and current use of ocular medications and risk factors for developing cornea ulcers, specifically a history of trauma or the use of contact lenses. One ophthalmologist (DC) evaluated all the charts.

The corneal scraping for cultures was performed in all cases using a Kimura spatula under direct visualization with a slit-lamp biomicroscope after the application of topical anesthesia. A microbiologist (FL) performed direct microscopy with potassium hydroxide 10% (KOH) and Gram stains of the specimen. The samples were inoculated directly onto blood, chocolate and Saboureaud’s agar, as well as in thioglycolate broth. All culture media were incubated at 37°C for 5 to 7 days, except for Saboureaud’s agar, which was incubated at 28°C for 15 days in a humidified incubator [14]. All bacterial and fungal growth was identified by a microbiologist (FL, NF, RS).

Results

A total of 660 cases of infectious corneal ulcers were identified during the 13-year period. There were 435 (66%) men and 225 (34%) women. The age was available for 578 (86%) of the 660 cases (Fig. 1). Data on occupations were available on 480 of the 660 patients (Fig. 2). Among the 435 male patients, data on occupation were available for 335 (77%), of whom 151 were agricultural workers (45%); data were available for 145 (64%) of the 225 female patients, of whom 78 (54%) worked only within the home. Data on geographic residence were available for 544 of the 660 patients (82%). Of these, 281 (52%) resided in an urban area and 263 (48%) resided in rural areas. Further analysis of the data showed that 214 of the 362 male patients (59%) resided in a rural area (data were not available for 73 cases). In contrast, 133 of the 182 female patients (73%) resided in a city (data were not available for 43 cases).
Fig. 1

Patient age and gender distribution. The age of the patients was not available in 82 cases

Fig. 2

Patient occupations (n = 480). Data were not available in 180 cases. **Manual laborers included: mechanics, carpenters, gardeners, construction workers, butchers, aborists or metal workers; ***examples are professionals: administrative assistants, chauffeurs, salesmen, seamstresses and bakers

For 591 patients, data were available on the number days between the onset of symptoms and presentation to our clinic. Patients may have initially sought medical attention at our clinic by self-referral or may have been referred by another physician for evaluation; the ratio of self-referred patients to those referred by other physicians is unknown. Figure 3 shows the number of days from onset of symptoms to the time that the patients were examined in our eye clinic.
Fig. 3

Distribution of days between symptom onset and patient presentation at our clinic (n = 591)

History of the presenting illness was available for 530 (80%) of the 660 patients. Of these, 257 (48%) had a history of trauma. The trauma involved vegetable matter in 121 cases (47%) and metallic foreign bodies in 20 cases (8%); 10 cases (4%) involved animals, and 78 cases (30%) involved a variety of other materials. In 28 cases, the type of material associated with the trauma was not recorded. Whether or not the patients were using ocular medications was known in 555 (84%) of the 660 cases. Among these, 460 (83%) were taking medications for their infections. Figure 4 lists the medications that the patients were using at the time of presentation to our clinic.
Fig. 4

Distribution of ocular topical medication prior to clinical presentation. *No steroidal anti-inflammatory medications are one example

In a total of 136 specimens (21%) from 660 eyes, collected from both the conjuctiva and the cornea, no bacteria or fungi did grow. In the other 524 eyes (79%), bacteria, fungi or both were isolated. Of these, 267 (51%) had only bacterial growth, 136 (26%) had only fungal growth and 121 (23%) had both bacterial and fungal growth. A total of 430 bacterial organisms were isolated from the 388 patients who had either bacterial or mixed bacterial and fungal growth (Fig. 5). Two hundred and nine fungi were isolated from the 257 patients who had either fungal or mixed bacterial and fungal growth (Fig. 6). In 50 patients, fungal elements were observed on the smear samples only, and not on culture media.
Fig. 5

Isolated bacteria (n = 430). Others: Streptococcus viridans, Klebsiella pneumoniae, Moraxella nonliquefasciens, Proteus mirabilis, Escherichia coli, Neisseria gonorrhoeae and Haemophilus species

Fig. 6

Isolated fungi distribution (n = 209)

The annual number of patients who presented to the clinic appears to be the same over time, as is the distribution of percentage of patients with bacterial, fungal or mixed bacterial and fungal keratitis (Table 1). There was no trend in either the number of cases per year or in the types of keratitis.
Table 1

Distribution of patients presented to the clinic over time and culture results

Year

Bacteria

% bacteria

Fungus

% fungus

Mixed

% mixed

No growth

% no growth

Total

% of total

1988

6

21

11

38

5

17

7

24

29

4

1989

22

39

6

11

13

23

15

27

56

8

1990

14

25

15

26

23

40

5

7

57

10

1991

31

55

3

5

18

32

4

7

56

10

1992

27

60

2

4

15

33

1

2

45

8

1993

22

43

8

16

15

29

6

12

51

9

1994

14

40

8

23

8

23

5

14

35

6

1995

20

51

7

18

6

15

6

15

39

6

1996

14

36

10

26

2

5

13

33

39

5

1997

15

32

17

34

0

0

18

36

50

6

1998**

21

52

5

12

1

2

15

36

42

5

1999

14

30

12

23

3

6

18

38

47

5

2000

23

38

20

33

8

13

9

15

60

10

2001***

24

44

12

22

4

7

14

15

54

8

*Both bacteria and fungi were isolated from the same eye; **data do not include January through February; ***data do not include September through December

Using a chi-square test (Statistical Package Epi Info 6 CDC, WHO, Geneva, Switzerland), five risk factors were identified to be associated with the diagnosis of fungal rather than bacterial keratitis. Males were more likely to have fungal keratitis than were females (P = 0.002). Among the male patients, agricultural workers were at higher risk of having fungal keratitis than were urban workers (P = 0.01). The age group with the highest risk for fungal keratitis was 30 to 59 years (P < 0.001). Patients with a history of trauma were more likely to have fungal keratitis (P = 0.03), as were as those previously treated with medications (P < 0.001).

Discussion

Infectious corneal ulcer is a potentially sight-threatening serious condition. Our study of 660 patients with infectious corneal ulcers over a 13-year period at a major university hospital in Paraguay revealed several important points. Two of three patients were male, which is similar to that reported in previously published studies [14, 15], and one of three were agricultural workers. The ages of the patients were distributed from 0 to 70 years, with no particular age group more likely to be diagnosed with infectious corneal ulcers. Approximately half of the patients had a history of ocular trauma, and half of these cases involved vegetable matter. Due to the high incidence of trauma in male agricultural workers, it may be prudent to recommend the use of eye safety goggles for this subset of individuals. Given the wide availability of over-the-counter medications in Paraguay, 83% of the patients were self-medicating prior to presentation to the clinic, with the most common medications being antibiotics or antibiotic/corticosteroid combinations (36%).

In our study, multiple samples were taken from the ulcers, and enriched media were used for cultures, as in the Nepal [20] and India studies [18]. Despite the fact that 83% (460 of the 555 patients) were taking either antibiotics or antifungal agents, often in combination with other medications, 80% of the specimens obtained were culture positive. This percentage of positive cultures is similar to that reported by Upadhyay et al. in Nepal [20] and Dunlop et al. [6] in Bangladesh, but is higher than the 68% of positive cultures rates reported by Srinivasan et al. [18] in south India, and that in the study performed in Ghana (58%) [9]. The negative culture results in the present study may be attributed to the use of previously prescribed or self-administered topical medications.

Approximately one half of the positive cultures grew bacteria, one-fourth grew fungi and the remaining one-fourth grew both bacteria and fungi. Accordingly, approximately half the cultures were positive for fungi, which is similar to results in India [8], but higher than that reported from studies in Nepal (17%) [1], Bangladesh (36%) [6] and south Florida (35%) [12]. This is most likely due to multiple factors, including climate, patient population and the availability of over-the-counter medications.

The most common bacteria isolated were coagulase-negative staphylococci and Staphylococcus aureus, which accounted for one half of the bacteria cultured. The most common fungi isolated were Acremonium species (sp), Fusarium sp and Aspergillus fumigatus, in decreasing order of frequency. Risk factors that correlated statistically with a positive fungus culture rather than a positive bacterial culture were male gender, particularly those who were agricultural workers, aged between 30 and 59, had a history of trauma and were self-medicated with topical therapeutics. Given the generally poorer prognosis of patients with fungal keratitis compared to those with bacterial keratitis, it is important to recognize these risk factors, since approximately half of the culture results had growth of fungi.

We would even advocate empirically prescribing antifungal medications to patients with these risk factors who have clinical findings consistent with fungal keratitis at the initial visit, after cultures have been obtained. Due to the expansive nature of deep fungal infection and the limited penetration of topical medications, treatment as soon as infection is suspected should result in an improved prognosis. Treatment could subsequently be modified depending on the culture results.

In our study, only one in four patients was evaluated within the 1st week of onset of symptoms. This is even lower than the patients in India (44%) [8],and is most likely due to the lack of access to health care in Paraguay, most specifically, to an ophthalmology clinic; many of the patients had to travel long distances to our hospital for care. Corneal ulcers are a major health issue for the people of Paraguay. Thus, particularly in our area (i.e., middle South America), public health care policy should be directed toward patient education, prevention and prompt treatment of infectious corneal ulcers. Patients with ocular trauma, particularly those with agricultural occupations, should seek appropriate ophthalmologic care and should be discouraged from self-medication. More importantly, patients with symptoms of corneal ulcers should have prompt access to medical care and should avoid using over-the-counter medications, particularly those that contain corticosteroids.

As with all retrospective studies, there are several shortcomings of our study. Obviously, there is a selection bias, since only those patients with persistent infections are likely to seek medical care, and therefore, only the more severe cases were included in this series. The incidence or prevalence of corneal ulcers cannot be determined. Finally, our study did not include data on visual outcome, so we cannot make recommendations regarding treatment. It is difficult to follow these patients long-term since, as previously noted, many live a long distance from the clinic and were not available for follow-up examinations.

In summary, a majority of patients with infectious corneal ulcers were male and had a history of trauma. Over-the-counter medications were commonly used, and most patients had a delay in diagnosis of more than 1 week. Risk factors for fungal keratitis as opposed to bacterial keratitis were male gender, agricultural occupation, age between 30 and 59 years, history of trauma and the use of self-medication.

Acknowledgements

The authors thank the Georg-and-Hannelore Zimmermann Foundation and the Department of Ophthalmology at the Ludwig-Maximilian University, Munich, Germany, for their financial and technical support of this study.

Copyright information

© Springer-Verlag 2004