Bestimmung der Standortflora der Augenoberfläche an einem Marburger Kollektiv

Background: Ocular surface infections can cause serious complications. These include descemetoceles, perforations or secondary glaucoma up to blindness. Because of this, adequate therapy is essential. Unfortunately, the evaluation of microbiological findings in ophthalmology is often not easy, because on the one hand a normal flora exists and on the other hand the detected amounts of pathogens are very low. Therefore, the assessment whether a detected bacterium or fungus is actually causally related to the infection or originates from the normal flora is often not clear. As a consequence, the selection of an antibiotic based on resistance testing s not always possible. Therefore, our work‘s aim was to find parameters that can serve to improve the assessability of microbiological findings in ophthalmology. Material and Methods: We compared microbiological results from swabs of patients with presumed infection from 2016 to 2020 with data from swabs of patients without infection of the ocular surface from May to September 2020. The general germ detection rate, germ detection rates of individual germs, as well as the amount of germs were evaluated. In the analyses, each were divided into a group with presumed infection and a group of healthy eyes, as well as into conjunctival and corneal swabs. R.A. Fisher‘s exact χ2-test was used to determine the significance level at which germs were more likely to occur in presumed infection or in healthy subjects. We determined whether there was a significant difference in the amount of germs between the two groups using the Wilcoxon-Mann-Whitney test. Results: A total of 713 smears were analyzed. These included 106 conjunctival swabs from healthy eyes and 222 conjunctival swabs from eyes with presumed infection, and 108 corneal swabs from healthy eyes and 277 corneal swabs from eyes with presumed infection. Regarding the germ detection rates, conjunctival (p<0.001) and corneal (p<0.001) smears showed a significant higher rate in ocular healthy subjects. After exclusion of smears that were made after antibiotic pretreatment, no significan difference was found for conjunctival smears (p=0.3), but for corneal smears it was (p=0.03). For conjunctival swabs, gram-positive cocci (p<0.001) were significantly more frequent in ocular healthy patients, gram-negative rods (p<0.001) were significantly more frequent in presumed infection. No significance was shown for gram-positive rods and pleomorphs (p=0.07), gram-negative cocci (p=1) and fungi (p=0.5). For corneal smears, gram-positive cocci (p<0.001) also occurred significantly more often in ocular healthy individuals, gram-positive rods and pleomorphs (p=0.002) significantly more often in eyes with presumed infection. No significance was shown for gram-negative rods (p=0.07), gram-negative cocci (p=0.6) and fungi (p=1). In the analyses of individual germs or classes, for conjunctival swabs there was a significantly more frequent occurrence of coagulase-negative staphylococci (p<0.001) in ocular healthy individuals and of streptococci (p=0.006) in eyes with presumed infection. For Staphylococcus aureus (p=0.09), coryne- and propionibacteria (p=0.4) and Haemophilus influenzae (p=0.06) there were no significances. In corneal smears, coagulase-negative staphylococci (p<0.001) were significantly more abundant in ocular healthy subjects. Staphylococcus aureus (p=0,005), streptococci (p=0,05), coryne- and propionibacteria (p=0,008) and Pseudomonas aeruginosa (p=0,02) were significantly more frequent in presumed infections. Regarding the amount of germs, considering all germs (p=0.01), as well as Staphylococcus aureus (p=0.04) showed a significantly higher amount of germs in presumed infection. For coagulase-negative staphylococci (p=0.8), streptococci and coryne- and propionibacteria (p=1) no significance was presentable. Conclusions: We could not identify one parameter or a combination of parameters for reliable differentiation between an apathogenic normal flora and a pathogenic agent of conjunctivitis or keratitis. However, we could show that some germs are rather associated with the natural flora or with infections and that the amount of germs could be a possible indicator for a better assessability in borderline cases.