New diagnostic tool in bacterial keratitis is not superior to traditional agar plates.

Abstract
Editor, B acterial keratitis is a potential eye-threatening disease. Early diagnosis is crucial to initiate correct treatment. To establish the microbiological diagnosis, direct incubation on agar plates is considered the gold standard; however, the sensitivity and specificity are low. The Copan eSwab media (Copan) is a newer liquid-based transport system that maintains viability of aerobic, anaerobic and fastidious bacteria for up to 48 hr. Furthermore, it is possible to diagnose C. trachomatis, N. gonorrhoeae and viruses by means of polymerase chain reaction (PCR). The sample is collected by the ophthalmologist, transported to the microbiology department and distributed on agar plates by a robotic plate management system. We aimed to investigate the sensitivity and specificity of Copan media compared to direct sampling on agar plates in patients suspected of having bacterial keratitis. A 1-year prospective observational study was performed at our tertiary ophthalmology department. Patients were referred by their general practitioner or an ophthalmologist. All patients had an ocular examination, and thereafter, samples from the corneal infiltrate were collected by direct smear on blood, chocolate and Sabouraud agar plate by direct incubation and on Copan. Agar plates and Copan were sent to the department of microbiology for incubation and plating. All plates were incubated in 5% CO2 for 96 hours. The growth of a bacterium was categorized as positive, unless it was coagulase-negative Staphylococcus species, Corynebacterium or viridans group Streptococci in small amounts. The sensitivity and specificity were calculated and compared by a chisquare test. A total of 114 eyes from 114 patients were examined. Retrospectively, only 45 patients had bacterial keratitis defined as an infiltrate >1 mm. We found a sensitivity of 31.1% for agar plates (14 of 45) and 24.4% on Copan (11 of 45; p = 0.5). The specificity was equal in both methods (95.7%). Positive predictive value for agar was 82.3% and 78.6% for Copan (p = 0.78). Negative predictive value was 68.0% (agar plates) and 66.0% (Copan). The overall agreement between the two methods in patients with bacterial keratitis is shown in Table 1. Two cases had more than one organism on agar plates and Copan; in these patients, organisms contiguous to the ocular surface were found. Overall, we found a trend of agar plates being more sensitive and with higher specificity than Copan for the diagnosis of bacterial keratitis; however, this was not statistically significant, possibly due to the low number of cases. Our results are similar to those of other studies on liquid media. In a clinical study (Pakzad-Vaezi et al. 2015), Copan and agar plates had comparable sensitivity. In an experimental study on pig corneas (Kaye et al. 2003), a liquid-based method (BHI) was found to be equal to direct sampling on agar plates. The advantage of the Copan is that it is a simple method and is more accessible than agar plates, and this could be relevant in the community setting. In conclusion, our results indicate that agar plates might be superior to Copan, and therefore, Copan cannot replace agar plates for diagnosing bacterial keratitis.


Letter to the Editor New diagnostic tool in bacterial keratitis is not superior to traditional agar plates Stine E. Nielsen,1 Jan Berg Gertsen,2 Mona Kjærsgaard,2 Anders Ivarsen1 and Jesper Hjortdal1 1Ophthalmology Department, University Hospital of Aarhus, Aarhus, Denmark; 2Microbiology Department, University Hospital of Aarhus, Aarhus, Denmark doi: 10.1111/aos.13051 Editor, B acterial keratitis is a potentialeye-threatening disease. Early diagnosis is crucial to initiate correct treatment. To establish the microbiological diagnosis, direct incubation on agar plates is considered the gold standard; however, the sensitivity and specificity are low. The Copan eSwabTM media (Copan) is a newer liquid-based transport system that maintains viability of aerobic, anaerobic and fastidious bacteria for up to 48 hr. Furthermore, it is possible to diagnose C. trachomatis, N. gonorrhoeae and viruses by means of polymerase chain reaction (PCR). The sample is collected by the ophthalmologist, transported to the microbiology department and distributed on agar plates by a robotic plate management system. We aimed to investigate the sensitivity and specificity of Copan media compared to direct sampling on agar plates in patients suspected of having bacterial keratitis. A 1-year prospective observational study was performed at our tertiary ophthalmology depart- ment. Patients were referred by their general practitioner or an ophthalmologist. All patients had an ocular examination, and thereafter, samples from the corneal infiltrate were collected by direct smear on blood, chocolate and Sabouraud agar plate by direct incubation and on Copan. Agar plates and Copan were sent to the department of microbiology for incubation and plating. All plates were incubated in 5% CO2 for 96 hours. The growth of a bacterium was categorized as positive, unless it was coagulase-negative Staphylococcus species, Corynebacterium or viridans group Streptococci in small amounts. The sensitivity and specificity were calculated and compared by a chisquare test. A total of 114 eyes from 114 patients were examined. Retrospectively, only 45 patients had bacterial keratitis defined as an infiltrate >1 mm. We found a sensitivity of 31.1% for agar plates (14 of 45) and 24.4% on Copan (11 of 45; p = 0.5). The specificity was equal in both methods (95.7%). Positive predictive value for agar was 82.3% and 78.6% for Copan (p = 0.78). Negative predictive value was 68.0% (agar plates) and 66.0% (Copan). The overall agreement between the two methods in patients with bacterial keratitis is shown in Table 1. Two cases had more than one organism on agar plates and Copan; in these patients, organisms contiguous to the ocular surface were found. Overall, we found a trend of agar plates being more sensitive and with higher specificity than Copan for the diagnosis of bacterial keratitis; however, this was not statistically significant, possibly due to the low number of cases. Our results are similar to those of other studies on liquid media. In a clinical study (Pakzad-Vaezi et al. 2015), Copan and agar plates had comparable sensitivity. In an experimental study on pig corneas (Kaye et al. 2003), a liquid-based method (BHI) was found to be equal to direct sampling on agar plates. The advantage of the Copan is that it is a simple method and is more accessible than agar plates, and this could be relevant in the community setting. In conclusion, our results indicate that agar plates might be superior to Copan, and therefore, Copan cannot replace agar plates for diagnosing bacterial keratitis.


Metadata
Authors
Stine E Nielsen, Jan Berg Gertsen, Mona Kjaersgaard, Anders Ivarsen, Jesper Hjortdal
Journal
Acta ophthalmologica
Publisher
Date
pm27080043
PM Id
27080043
PMC Id
Images