Introduction

Infectious keratitis is a leading cause of monocular blindness worldwide.1 Furthermore, in developing countries, it has been recognized by the WHO as a ‘silent epidemic’ with an incidence rate up to 113 per 100 000 (0.11%).2 Corneal blindness is the end result in the majority of these infections, and more disastrous outcomes such as corneal perforation, endophthalmitis, or phthisis may occur.

Extensive literature describes a standard technique for evaluating clinically suspected infectious keratitis, which includes obtaining corneal scrapings for smears and cultures prior to initiating antibiotic treatment.3, 4 The traditional evaluation of clinically suspected infectious keratitis includes Gram stains, as well as multiple cultures. Culture media commonly used include blood agar, chocolate agar, and a fungal medium such as Sabouraud agar.5 Nevertheless, it was reported in Southern California that as many as 48% of patients with clinically suspected infectious keratitis received antibiotic therapy without prior culture, and only 22% of the general ophthalmologists were found to maintain culture media supplies in their offices.6

Several reasons have been offered as explanations for the disparity between the traditionally recommended evaluation of infectious keratitis and actual practice. These explanations include laboratory costs, time constraints in busy office settings, the cost of maintaining fresh media in an office setting, and the difficulty of transporting the inoculated plates, as they require immediate transportation in special conditions. Other reasons can be the sense that empiric treatment is adequate for the majority of cases and a past experience of low yield from the cultures.

Contrary to the traditional culture media, swab transport media are potentially less expensive and simpler to maintain, inoculate, and transport. It has also been reported that, in certain conditions, transport media can yield as high a growth percentage of bacteria as blood agar plates.7 Unfortunately, the yield of the transport medium is dramatically reduced with time from the primary inoculation.

BACTEC Peds Plus F (Becton Dickinson, Sparks, MD, USA) (Figure 1) is a specialized broth medium that accommodates small-volume samples (<3 ml). The medium is an enriched soybean-casein digest broth, and contains resins for antibiotic neutralization. The BACTEC Peds Plus F is a well-known and widely used broth culture medium for blood and other body fluids in paediatric practice. It is suitable for aerobic and anaerobic bacteria as well as fungi.8 These characteristics of the BACTEC Peds Plus F broth, along with no need for immediate incubation and easy transportation, makes it potentially advantageous as the swab transport medium in ophthalmologic settings.

Figure 1
figure 1

(a) BACTEC Peds Plus F (Becton Dickinson, Sparks, MD, USA), (b) Eppi 40 tube (Eppendorf International), (c) Regular 2.5 ml syringe.

Full size image

A literature search revealed that the use of commercial broth media in the work-up of infectious keratitis has not been reported previously. The primary goal of this study was to elucidate whether the BACTEC Peds Plus F could be used for corneal cultures from clinically suspected infectious keratitis. Secondary goals were to compare the yield between this method and the traditional methods, and to review the relevant literature.

Methods

All cases with newly diagnosed, nonviral, clinically suspected infectious keratitis presented at Soroka University Medical Center, Beer-Sheva, Israel, between June 2003 and June 2004 were included in this prospective, comparative study. Cases with any prior antibiotic treatment, or a clinically suspected viral keratitis, were excluded. Corneal scrapings were obtained by a single ophthalmologist (AK) from all patients complying with the above criteria. A common protocol allowing the diagnosis of bacterial and fungal keratitis was used in all cases. Under topical anaesthesia (Benoxinate HCl 0.4%) and slit-lamp magnification, corneal scrapings were obtained from the base and edge of the ulcer using three different sterile surgical blades (no. 15). The samples were inoculated directly onto blood agar, chocolate agar, and Sabouraud agar. Another scraping was carried out, using a different blade, and the specimens were inserted to a regular sterile Eppi 40 tube (Eppendorf International) previously filled with 1.5 ml of sterile saline 0.9%. The fluid from the Eppi 40 tube was then drawn using a regular 2.5 ml syringe, and inserted into regular BACTEC Peds Plus F broth (Becton Dickinson, Sparks, MD, USA) (Figure 1). For each case, an additional sample was taken, using a regular amine agar gel transport swab without charcoal (COPAN Venturi Transystem) for secondary inoculation at the bacteriology laboratory in our medical centre. There was no policy regarding the order in which these two scrapings and swab were taken. The blood and chocolate agar plates were incubated at 37°C. Blood agar plates were incubated under both aerobic and anaerobic conditions, chocolate agar plates were incubated in 5% carbon dioxide, and Sabouraud agar plates were incubated at 25°C to enhance the growth of fungi. The BACTEC broths were evaluated in the bacteriology department in our medical centre using the BACTEC 9240 system (Becton Dickinson Diagnostic Instrument Systems, Towson, MD, USA). The blood and chocolate agar plates, the BACTEC broths and the Sabouraud agar plates, were incubated for 3, 7, and 14 days, respectively, before being reported as ‘no growth.’ Any isolation of a definitive, noncontaminant keratitis-causing pathogen was defined as ‘positive growth.’ For each case, the growth outcomes of the blood, chocolate, and Sabouraud agar were summed and were called the ‘Traditional method.’ These results were added to the secondary growth from the swab, and were compared with the growth outcome from the BACTEC alone. McNemar's test was used for pairwise comparisons of the rates of positive growth between the two groups. P<0.05 was used to denote a significant statistical difference.

Results

In total, 30 eyes (17 right and 13 left) of 30 patients (19 males and 11 females) with clinically suspected infectious keratitis were included in this study. The growth results in the BACTEC group and the ‘Traditional method’+swab transport media group is shown in Table 1.

Table 1 The growth outcome of the culture methods
Full size table

The yield of the two culture methods is outlined in Figure 2. The BACTEC method yielded 53.33% growth (16 cases), while the yield in the Traditional method+swab transport media group was 50.0% (15 cases). There was no statistically significant difference between the two groups (P=1.0, χ2=0.0). As can be seen in Table 1, in some cases both methods were able to identify an organism, while in other cases only one method succeeded, or even both methods failed.

Figure 2
figure 2

The distribution of growth results in the Traditional method+swab transport medium group and the BACTEC group.

Full size image

The combined positive growth of both BACTEC and Traditional Method+swab transport media is 73.33% (22/30 cases), and is statistically significant higher than the yield of the Traditional method+swab transport medium alone (P=0.024, χ2=5.14), and also statistically significant higher than the yield of the BACTEC alone (P=0.042, χ2=4.17).

The type and rate of identified organisms is outlined in Table 2. The most common identified organism in our series was coagulase-negative Staphylococcus, which was discovered in 50.0% (11/22) of all positive identifications.

Table 2 Type and number (%) of identified organisms from positive cases
Full size table

Discussion

The traditional academic approach for the initial diagnosis of clinically suspected infectious keratitis considers diagnostic scraping of corneal ulcers mandatory.5, 9 In most cases, standard media to be inoculated include blood and chocolate agar for recovery of aerobic and facultatively anaerobic bacteria and fungi, and Sabouraud agar for recovery of fungi.5 Nevertheless, several recent reports describe a decrease in the frequency with which ophthalmologists perform standard cultures, and an increasing use of transport media for submitting samples to microbiology laboratories for evaluation.6, 10, 11

The disparity between traditional academic and actual practice can be explained by laboratory costs, time constraints in busy office settings, the cost in maintaining fresh media in an office setting, and the difficulty in transporting the inoculated plates, as they require immediate transportation in a special conditions. Another reason can be the sense that empiric treatment is adequate for the majority of cases.

Unfortunately, even when standard cultures were obtained, the growth yield were disappointing. It is interesting that even with direct inoculation of a known organism (Streptococcus pneumonia) onto rabbit cornea,7 the growth yield of the blood and chocolate agar reached only 69%. Table 3 lists several reports of infectious keratitis studies in humans, in which the data concerning the growth yield were represented or could be calculated. A summary of 7888 cases reported in the literature reveals that the average growth yield of the Traditional culture medium is only 50.47%. The yield of the Traditional method+swab transport medium in our study (50.0%) is similar to the average value found in the literature, but when pooling the results from the Traditional+swab transport medium with the yield of the BACTEC in our study, the outcome is a high rate of 73.33%.

Table 3 Reported growth yield of the Traditional methoda
Full size table

The use of a special broth-based culture medium to improve the growth yield in infectious keratitis was previously described in two reports (Table 4).18, 19 During a 5-year study period, Simcock et al18 reported that the use of a broth increased the growth yield from an average of 41.5% (81 of 195 cases) when using Traditional methods alone, to 79.4% (54 of 68 cases) when adding the use of the broth. Schonheyder et al19 found a yield of 85.2% when using a broth in a 1-year study (23 of 27 cases), but they did not report the yield in cases without broth use. In both studies, the broths were prepared locally in the same institute in which the studies were conducted, and were not available commercially.

Table 4 High growth yield when using a broth medium
Full size table

As seen in Table 2, the most commonly identified organism in our study, both in the traditional method and BACTEC broth, was coagulase-negative Staphylococcus. In the past, coagulase-negative Staphylococcus was often disregarded as a significant cause of keratitis because of its presence on the skin and eyelids as normal flora.20, 21 However, it is now clear that coagulase-negative Staphylococcus is an important causative organism in infectious keratitis, as shown in many studies worldwide,12, 13, 16, 19, 22, 23 as well as in another study from our country.24 Our results are consistent with these findings. Measures should be taken to decrease the probability of contaminations during specimen collection. These measures include preventing the patient from blinking during corneal scraping and disinfecting the broth rubber head with 70% alcohol (or equivalent solution) prior to specimen injection.

The BACTEC Peds Plus F is a well-known and widely used broth culture medium for blood and other body fluids in paediatric practice. It is suitable for aerobic and anaerobic bacteria, as well as fungi,8 as proven by the growth of Candida parapsilosis in one of our cases.

The BACTEC method has previously been shown to be superior to Traditional methods when culturing blood and synovial fluids.25, 26 Moreover, its sensitivity relative to the Traditional method led to the recognition of formerly rare pathogens as relatively common causative pathogens in various paediatric diseases.27, 28

The approximate cost (in US$) of both methods, as was collected from several on-line sources, is ∼7.5$ for the traditional method and ∼10$ for the BACTEC broth.

Our results show the efficacy of the BACTEC broth in the workup of clinically suspected infectious keratitis. When using BACTEC along with the Traditional method+swab transport medium, there is a statistically significant higher yield than using either methods alone. Adding this to the other advantages of this method, namely a single culture medium, time-savings, easy transportation, no need to keep and maintain a supply of fresh agar media, and no need for immediate incubation, the BACTEC broth, though slightly more expensive, is a worthy adjunct to the Traditional method+swab transport media in the workup of clinically suspected infectious keratitis. This method is especially suitable for office settings and remote clinics, but also can be used in hospital setting, as an adjunct, when available, as it will probably increase the total growth yield.