Secondary Logo

Journal Logo

Original Study

Topical Steroids as Adjunctive Therapy for Bacterial Keratitis: Evidence From a Retrospective Case Series of 313 Cases

Khoo, Pauline BSc (Hons); Cabrera-Aguas, Maria MIPH, MBBS; Watson, Stephanie L. PhD, FRANZCO∗,†

Author Information
Asia-Pacific Journal of Ophthalmology: September-October 2020 - Volume 9 - Issue 5 - p 398-403
doi: 10.1097/APO.0000000000000320
  • Open


Bacterial keratitis is a sight-threatening condition and an important cause of corneal inflammation.1 Conventional management consists of corneal culture and empirical broad-spectrum topical antibiotics.2 The use of topical steroids as adjunctive to antibiotic therapy for the treatment of bacterial keratitis remains controversial. The drug's anti-inflammatory properties may help control the host response and reduce corneal neovascularisation and scarring, leading to improved clinical outcomes. However, the immunosuppressive effect of steroids may promote bacterial replication and slow the recovery from the infection.3

Four randomized controlled trials (RCT) have evaluated antibiotic and topical steroid use in the treatment of bacterial keratitis.4–7 They found topical steroid treatment in bacterial keratitis to have nondetrimental effects, the Steroid Corneal Ulcer Trial (SCUT) also found a beneficial effect on visual outcomes.7 Additional subgroup analyses were performed on specific causative organisms isolated from the SCUT patients; topical steroids were not associated with better outcome with Pseudomonas (P.) keratitis and patients with Nocardia keratitis had poorer outcomes.8,9 These trials were conducted in India and the United States; however, patients were mainly recruited from India. Causative organisms and hence outcomes vary with geographic region in bacterial keratitis.1,10,11 Current literature either have a small sample size,4–6 not applicable to the Australian population due to the low numbers of contact lens wearers and high trauma rates,7 or main causative organism differ to the spectrum of disease seen in Sydney, Australia.8,9 To inform practice and understand global trends, there is a need for robust contemporary local data.

This study reports the use of topical steroids as an adjunctive treatment for bacterial keratitis in a large quaternary hospital in Sydney, Australia. In particular, it focuses on the clinical characteristics and effects of topical steroids on the most common organisms isolated in bacterial keratitis: coagulase-negative Staphylococcus (CoNS), P. aeruginosa, and S. aureus in these patients.


A retrospective case series study was conducted including all adult patients with microbial keratitis and a positive bacterial corneal scrape between January 1, 2012 and December 31, 2016, presenting to the Sydney Eye Hospital, Sydney, Australia. The Sydney Eye Hospital is a quaternary referral unit for eye disease and is the major public hospital in Sydney and also serves the rural area of New South Wales.10 Patients were identified from hospital coding and pathology data. The study adhered to the tenets of Declaration of Helsinki and was approved by the South Eastern Sydney Local Health District Human Research Ethics Committee (HREC ref 14/282). Patient consent was not needed due to the retrospective nature of the study.

Patients were excluded if: they were younger than 18 years, had no record of microbial keratitis, no corneal scrape performed, had previous use of topical steroids before corneal scraping, there was a possible herpetic infection or a history of herpetic keratitis, their vision was 6/9 or better at initial presentation, corneal culture result included fungal or protozoal species, they presented with a corneal perforation, or their cornea perforated or their eye was eviscerated or enucleated during treatment. The exclusion criteria were modeled on the SCUT12 and Green et al.13

Medical records were reviewed, and the following data were collected: socio-demographic information, ocular and systemic history, clinical presentation [eg, visual acuity (VA), epithelial defect, hypopyon, and corneal thinning], pathology, management, and outcomes (eg, epithelial defect healed, VA, corneal scarring, and complications). Baseline visit was based on the patient's initial visit to the Sydney Eye Hospital and therefore, the epithelial defect might not have been present at their initial visit. Patient outcome was classified according to Khoo et al.14 Study data were collected and managed using REDCap (Research Electronic Data capture, Nashville, TN) hosted at the University of Sydney. Ulcer size was defined as the geometric mean of the longest diameter of the epithelial defect and infiltrate, and its perpendicular diameter.10


Corneal scrapes were taken in accord with local protocols from patients who had a clinical diagnosis of keratitis at presentation to the Sydney Eye Hospital.15 Briefly, samples were collected using sterile surgical blades or disposables needles. Superficial corneal samples were directly inoculated onto 2 glass slides for microscopy and culture media (2 blood agars, chocolate agar, Sabouraud's agar slope, and cooked meat medium) in the clinic. Corneal swabs were also taken for polymerase chain reaction testing for detection of herpes simplex virus and acanthamoeba DNA. Samples were then delivered to New South Wales Health Pathology (NSWHP).

Topical Steroid Use

Steroid treatment classification was based on Green et al.13 Patients were considered to have received high-dose steroid whether they commenced treatment within the first 7 days after the corneal scrape procedure and their daily treatment was ≥6 drops of prednisolone acetate 1%, phenylephrine hydrochloride 0.12% (Prednefrin forte Allergan, New Jersey), dexamethasone 0.1% (dexamethasone, Alcon, Fort Worth, TX), or prednisolone sodium phosphate 0.5% (prednisolone Minims, Bausch and Lomb, Rochester, NY). If patients were prescribed fluorometholone 0.1% (FML, Allergan) or any other steroid drops once or twice a day or if they started >30 days after corneal scrape procedure, they were considered “low dose.” All other patients’ treatment was classified as “regular dose.”

Statistical Analyses

Statistical analyses were performed by using SPSS software (version 24 for Mac, IBM). Sociodemographics and clinical features were compared between organism groups and statistical significance assessed using chi-square test for proportions and Kruskal-Wallis test for nonparametrically distributed variables (age, visual acuity, epithelial and infiltrate size, admission days, and healing time). Post hoc tests were performed on data where statistical significance was shown. Additional analysis using chi-square test for proportions and Kruskal-Wallis test were performed on patients who had worn contact lenses and had a CoNS or P. aeruginosa infection. Statistical significance was defined as P < 0.05. Co-infections were excluded from the main causative organism subgroup analyses.


Patient Demographics

Over a 5-year period, a total of 1052 microbial keratitis cases were identified. Patient demographics, risk factors, microbiological results, and patient outcomes of the entire case series have been previously reported.14,16

A total of 313 cases from 308 patients met the criteria and were included in this case series. Three patients had a re-infection in the same eye, 1 patient had a separate episode of infection in the other eye and another patient had a bilateral infection during the 5-year study period. The most common reasons for exclusion were patients with culture-negative keratitis, previous use of topical steroids before presentation to the hospital, or an initial vision at presentation of 6/9 or better.

The median age of patients included in the case-series was 51 years [interquartile range (IQR) 36–72], of whom 147 (48%) were females. The main causative organisms identified were CoNS, P. aeruginosa and S. aureus.17 There were significant differences between organisms for age (P = 0.004, Kruskal-Wallis test). Post hoc test analysis revealed patients with S. aureus infections to be significantly older compared with patients with CoNS (P = 0.008) and P. aeruginosa (P = 0.001) infections (Table 1).

Baseline Clinical Features of Patients With CoNS, P. aeruginosa, and S. aureus Infections

The most common risk factors identified were contact lens wear (CLW), ocular surface disease (OSD), ocular trauma, and history of corneal transplantation (Table 1). P. aeruginosa was the most common organism in CLW (P < 0.005, chi-square test). S. aureus infections occurred more frequently in patients with an OSD (P < 0.05) and/or history of corneal transplant (P < 0.05).

Clinical Features at Presentation

The median VA at initial presentation was 1 logarithm of the minimum angle of resolution (logMAR) (IQR 0.5–1.8), with an epithelial defect present in 81% (n = 255) of cases, hypopyon in 25% (n = 78) and corneal thinning in 9% (n = 27). The median epithelial defect size was 4 mm2 (IQR 1–10). Patients with P. aeruginosa and S. aureus infections presented with more severe keratitis. Post hoc analysis showed the epithelial size to be significantly larger in P. aeruginosa than CoNS (P = 0.002) at presentation. A hypopyon occurred more frequently in P. aeruginosa and S. aureus infections than CoNS (P < 0.001, P = 0.049, respectively). A higher proportion of S. aureus infections had corneal thinning compared with CoNS (P = 0.004) and P. aeruginosa (P = 0.043) (Table 1).

Topical Steroid Usage Patterns

Topical steroids were used as part of management in 192 (61%) cases. High-dose steroids were prescribed in 22 (11%) cases, regular-dose in 88 (46%), and low-dose in 82 (43%). The median time until the implementation of steroid use was 4 days (IQR 3–7). The median duration of steroid use was 26 days (IQR 12–55). The main topical steroids prescribed were prednisolone sodium phosphate minims (95, 49%), dexamethasone minims (28, 15%), fluorometholone (27, 14%), and dexamethasone (23, 12%). The most common initial frequency of topical steroid use was 3 to 4 times a day (86, 45%), followed by twice a day in (47, 24%) (Table 1).

Patient Outcomes

At the final visit, no significant differences were observed between the steroid and no steroid group, except for the duration of epithelial healing time. Significantly longer durations were observed in the steroid group (11 days) compared with the “no steroid” group (6.5 days) (P < 0.001) (Table 2). On subgroup analysis, similar results were seen with patients with CoNS (P = 0.014). In P. aeruginosa group, a higher proportion of patients on topical steroids had a scar present at final visit compared with the no steroid group (88% vs 33%, P = 0.019), they also had lengthier days of admission in hospital (8 vs 5 days, P = 0.033) and longer healing time (19 vs 4 days, P = 0.006). There was no difference in patient outcome with or without topical steroid use for patients with a S. aureus infection.

Clinical Outcomes of Patients With CoNS, P. aeruginosa, and S. aureus infections

The effects of topical steroids on patient outcomes were compared between CoNS, P. aeruginosa, and S. aureus infections. There were no significant differences in patient outcome for final VA, change in VA, healing of the epithelial defect, and duration of hospital admission. However, significant differences were seen in the P. aeruginosa infection steroid group, with a higher proportion presenting with corneal scarring at final visit (P = 0.02) and longer healing times (P = 0.04) compared with CoNS and S. aureus.

For patients with contact lens associated bacterial keratitis, larger epithelial defects and the presence of a hypopyon were observed at baseline visit in steroid compared with the no steroid group (P = 0.005 and P = 0.001, respectively). At final visit, the no steroid group showed greater overall improvement; with better VA (P = 0.012), shorter healing times (P = 0.005), and reduced likelihood of scarring (P = 0.034) (Table 3). For CLW with CoNS keratitis, there were insignificant differences with the use of topical steroids at baseline and final visit. Whereas in patients with a P. aeruginosa infection associated with CLW, longer hospital admissions (P = 0.041) and healing times (P = 0.011) were observed with topical steroids. However, it should be noted at baseline visit, these patients with CLW associated P. aeruginosa had worse initial VA (P = 0.072) and larger epithelial defect size (P = 0.05) with steroid use than no steroid use (Table 3).

Clinical Outcomes of Contact Lens Wearers With Bacterial Keratitis

Adverse effects were compared between the topical steroid and no steroid group. There was no significant difference observed for corneal thinning or increased intraocular pressure, across all groups (P > 0.05). More patients in the steroid group had an unhealed epithelial defect by 21 days compared to no steroid group (P = 0.004) (Table 4).

Adverse Effects of Patients With CoNS, P. aeruginosa, and S. aureus Infections


The use of adjunctive topical steroid therapy for bacterial keratitis remains controversial. In this retrospective study, we report the usage patterns, clinical characteristics, and outcomes of patients with bacterial keratitis on topical steroids. At our center, 58% (n = 606) of cases of culture-positive bacterial keratitis were prescribed topical steroids, of which 30% (n = 313) met our inclusion criteria. We found no overall significant difference at baseline and final visit with and without steroid use; however, the use of topical steroids did result in longer healing times (11 vs 6.5 days).

Existing studies evaluating antibiotic and topical steroid use were conducted in the United States, Canada, India, and South Africa. Due to the small sample size in 3 of the studies, the effectiveness of topical steroid use was inconclusive. The SCUT was the largest RCT conducted, Streptococcus pneumoniae, Nocardia species, and P. aeruginosa were the most common causative organisms identified in the study. Although P. aeruginosa was the most causative organism, a low proportion of the total number of cases occurred in CLW as contact lens use is less common in India.9,18 The microbiology from the SCUT did not represent the spectrum of disease in Australia. In Australia, CLW is the most common risk factor.10

The SCUT study was a high-quality RCT; however, the therapeutic effect of topical steroids in the SCUT may not represent real-world outcomes. Treatment protocols in the SCUT trial were fixed, whereas in routine clinical practice, reported in our study, treatment protocols were altered depending on the patient's progress. Furthermore, subanalysis from the SCUT found that topical steroids may be associated with worse outcomes when administered for specific organisms, for instance, Nocardia species.8 In Australia, Staphylococcus species and P. aeruginosa19 infections were the most common causative organisms and our study focuses on these species.19,17

Delayed corneal wound healing is recognized as a potential adverse effect of topical steroid therapy.20,21 Previous studies have shown topical steroids to significantly delay corneal healing22,23; however, the delay does not affect the final visual outcome.24 Our study found adjunctive topical steroid therapy to be associated with longer duration of reepithelialization, specifically in CoNS and P. aeruginosa infections. In the SCUT study,7 no difference in epithelial defect healing was observed among organisms. However, delay in reepithelialization was observed by Srinivasan et al.6 The difference in healing time in our study could be due to the lack of standardization of steroid treatment, although overall the delay in healing did not translate to worse clinical outcomes.

In the subgroup analysis of patients with P. aeruginosa infection, a higher proportion of patients in the topical steroid group presented with more severe keratitis, at the baseline visit, with >50% of the steroid group presenting with a hypopyon. A greater number of patients presence of a corneal scar and the duration of hospital admission and healing times were longer in the steroid group compared with no steroid group. Hypopyons occur in more severe keratitis,25 which likely led to the prescribing of steroids. A subgroup of the SCUT study found topical steroids to improve visual outcomes after 3 months.9 Our study found no significant benefit with visual outcomes in the topical steroid group, possibly due to the smaller sample size in our study.

Contact lens wearers prescribed topical steroids with topical antibiotics for P. aeruginosa infection had more severe keratitis at presentation, longer hospital admission, and healing times. However, at the final visit, the final VA, change in VA, and reepithelization rates were the same with and without steroids use. Furthermore, the steroid group had a greater improvement in VA change, but this was not statistically significant; however, this was possibly due to the small sample size. Topical steroids may aid visual outcome in CLW with P. aeruginosa infection, but a prospective RCT is needed to confirm this finding.

Topical steroid treatment in our study population did not seem to be associated with any trends in terms of adverse effects. There was no evidence of an increase in any adverse effects with topical steroid treatment, except for a greater proportion of nonhealed defects at day 21. Topical steroids may increase intracular pressure (IOP), and although there were more patients with increased IOP in the steroid group, this was not statistically significant.

The study is limited by its retrospective design and the lack of standardization of clinical measurements which compromised the accuracy of the outcome measures. Furthermore, the Sydney Eye Hospital is a quaternary hospital and may have an overrepresentation of more severe keratitis. However, it should be noted that although the SCUT study was a large RCT, the patient population did not represent that of patients in Sydney, Australia. Retrospective studies have limitations, but they are a cost-effective way to evaluate treatment outcomes in bacterial keratitis in the real world. Although there was a lack of standardization in treatment protocols, this allowed for customization of treatment, which is commonly seen in clinical practice. For instance, at our institution, the introduction of topical steroids in cases of microbial keratitis is at the discretion of the treating clinician. They are typically administered once the causal organism has been identified and dependent on the severity of the inflammation.

The study reports the current use of adjunctive steroid use in bacterial keratitis in a quaternary hospital in Australia. This study showed topical steroid use produced no overall difference in final VA nor adverse events but resulted in significantly longer healing times. Subgroup analyses suggested that topical steroids may have a differential effect depending on the specific causative organisms. However, a well-designed prospective RCT is needed to provide conclusive findings to further guide the management of bacterial keratitis to optimize outcomes.


The authors acknowledge the Corneal Unit at Sydney Eye Hospital who managed the patients with microbial keratitis. The Medical Records Department, Sydney Eye Hospital, and Prof Monica Lahra from New South Wales Health Pathology.


1. Kaye S, Tuft S, Neal T, et al. Bacterial susceptibility to topical antimicrobials and clinical outcome in bacterial keratitis. Invest Ophthalmol Vis Sci 2010; 51:362–368.
2. Morlet N, Daniell M. Microbial keratitis: what's the preferred initial therapy? View 2: empirical fluoroquinolone therapy is sufficient initial treatment. Br J Ophthalmol 2003; 87:1169–1172.
3. Herretes S, Wang X, Reyes JM. Topical corticosteroids as adjunctive therapy for bacterial keratitis. Cochrane Database Syst Rev 2014; 10:CD005430.
4. Carmichael TR, Gelfand Y, Welsh NH. Topical steroids in the treatment of central and paracentral corneal ulcers. Br J Ophthalmol 1990; 74:528–531.
5. Blair J, Hodge W, Al-Ghamdi S, et al. Comparison of antibiotic-only and antibiotic-steroid combination treatment in corneal ulcer patients: double-blinded randomized clinical trial. Can J Ophthalmol 2011; 46:40–45.
6. Srinivasan M, Lalitha P, Mahalakshmi R, et al. Corticosteroids for bacterial corneal ulcers. Br J Ophthalmol 2009; 93:198–202.
7. Srinivasan M. Corticosteroids for bacterial keratitis. Arch Ophthalmol 2012; 130:143–150.
8. Lalitha P, Srinivasan M, Rajaraman R, et al. Nocardia keratitis: clinical course and effect of corticosteroids. Am J Ophthalmol 2012; 154:934–939.
9. Sy A, Srinivasan M, Mascarenhas J, et al. Pseudomonas aeruginosa keratitis: outcomes and response to corticosteroid treatment. Invest Ophthal Vis Sci 2012; 53:267–272.
10. Khoo P, Cabrera-Aguas MP, Nguyen V, et al. Microbial keratitis in Sydney, Australia: risk factors, patient outcomes, and seasonal variation. Graefes Arch Clin Exp Ophthalmol 2020; 258:1745–1755.
11. Shah A, Sachdev A, Coggon D, et al. Geographic variations in microbial keratitis: an analysis of the peer-reviewed literature. Br J Ophthalmol 2011; 95:762–767.
12. Srinivasan M, Mascarenhas J, Rajaraman R, et al. The Steroids for Corneal Ulcers Trial (SCUT): Secondary 12-Month Clinical Outcomes of a Randomized Controlled Trial. Am J Ophthalmol 2014; 157:327–333.
13. Green M, Hughes I, Hogden J, et al. High-dose steroid treatment of bacterial keratitis. Cornea 2019; 38:135–140.
14. Khoo P, Cabrera-Aguas M, Robaei D, et al. Microbial keratitis and ocular surface disease: a 5-year study of the microbiology, risk factors and clinical outcomes in Sydney, Australia. Curr Eye Res 2019; 44:1195–1202.
15. Ngo J, Khoo P, Watson SL. Improving the efficiency and the technique of the corneal scrape procedure via an evidence based instructional video at a quaternary referral eye hospital. Curr Eye Res 2020; 45:529–534.
16. Khoo P, Cabrera-Aguas M, Nguyen V, et al. Micorbial keratitis in Sydney, Australia: Risk factors, patient outcomes and seasonal variation. Graefes Arch Clin Exp Ophthalmol 2020; 258:1745–1755.
17. Watson S, Cabrera-Aguas M, Khoo P, et al. Keratitis antimicrobial resistance surveillance program, Sydney, Australia: 2016 Annual Report. Clin Exp Ophthalmol 2019; 47:20–25.
18. Bharathi M, Ramakrishnan R, Meenakshi R, et al. Microbial keratitis in South India: influence of risk factors, climate, and geographical variation. Ophthalmic Epidemiol 2007; 14:61–69.
19. Cabrera-Aguas M, Khoo P, George CRR, et al. Antimicrobial resistance trends in bacterial keratitis over 5 years in Sydney, Australia. Clin Exp Ophthalmol 2020; 48:183–191.
20. Wilhelmus KR. Indecision about corticosteroids for bacterial keratitis: an evidence-based update. Ophthalmology 2002; 109:835–843.
21. Gritz DC, Kwitko S, Trousdale MD, et al. Recurrence of microbial keratitis concomitant with antiinflammatory treatment in an animal model. Cornea 1992; 11:404–408.
22. Tomas-Barberan S, Fagerholm P. Influence of topical treatment on epithelial wound healing and pain in the early postoperative period following photorefractive keratectomy. Acta Ophthalmol Scand 1999; 77:135–138.
23. Chung JH, Kang YG, Kim HJ. Effect of 0.1% dexamethasone on epithelial healing in experimental corneal alkali wounds: morphological changes during the repair process. Graefes Arch Clin Exp Ophthalmol 1998; 236:537–545.
24. Nien CJ, Flynn KJ, Chang M, et al. Reducing peak corneal haze after photorefractive keratectomy in rabbits: prednisolone acetate 1.00% versus cyclosporine A 0.05%. J Cataract Refract Surg 2011; 37:937–944.
25. Keay L, Edwards K, Stapleton F. Signs, symptoms, and comorbidities in contact lens-related microbial keratitis. Optom Vis Sci 2009; 86:803–809.

bacterial keratitis; corneal infection; topical steroids; treatment

Copyright © 2020 Asia-Pacific Academy of Ophthalmology. Published by Wolters Kluwer Health, Inc. on behalf of the Asia-Pacific Academy of Ophthalmology.