Bacterial keratitis in a Swedish county hospital: management and clinical outcome.

Abstract

Bacterial keratitis is a serious ocular infection which may lead to severe visual impairment. At present, there are no up to date national guidelines in Sweden for handling and treatment of bacterial keratitis. A previous review of clinical keratitis care in Sweden was performed in 1992 (Neumann & Sjöstrand 1992). The aims of this study were to investigate current clinical management and treatment outcome of suspected bacterial keratitis in a Swedish county hospital. A retrospective review of electronic patients' charts at the Department of Ophthalmology, County Hospital Council of Västmanland during the period 1 January 2011 until 31 December 2012 was performed. The study followed the Tenets of the Declaration of Helsinki and was approved by the Regional Ethical Review Board at Uppsala University. The ICD-10 (International Statistical Classification of Diseases and Related Health Problems) code keratitis (H16) was used in the search of patients' charts. Cases with corneal infiltrates (n = 159), cases with corneal epithelial defects combined with intra-ocular inflammation (n = 7) and clinical bacterial keratitis (n = 6) were included in the review. All cases included had received antibiotic treatment. In all, the study comprised 172 cases. Follow-up data were available in 120 cases, whereas final visual acuity (VA) was available and obtained in 99 cases. Seven patients (4%) had bilateral keratitis. Gender distribution was close to 1:1 (88 women and 84 men). The age of the patients ranged from 0 to 96 years. Most patients were aged 20–39 years (40%) and 40–59 years (28%). Mean duration of subjective symptoms before diagnosis was 2.7 days (SD 2.4). A predisposing factor was identified in 71% of the cases. Contact lens wear was the most frequent risk factor (53%) followed by corneal trauma (11%) and corneal pathology (5%). In 58% of cases, a culture was taken. Out of 106 cultures taken, 97 were conjunctival smears, seven were corneal scrapings, and two were from contact lenses. In 20 cultures (71% of corneal scrapings and 13% of conjunctival smears), a causative agent/bacteria could be identified (Table 1). In total, 93% of the cases were treated empirically with topical levofloxacin. No difference in choice of antibiotic treatment was seen between age groups. The antibiotic was given as monotherapy (66%) or in combination with another antibiotic (34%). Cycloplegics were given in 42 cases, and topical corticosteroid was added during follow-up in 24 cases. Surgical adjunctive therapy was provided in seven cases: cross-linking (4), amniotic membrane transplantation (3), conjunctival flap (1) and evisceration (1). Two cases had both an amniotic membrane transplantation and underwent cross-linking. Four patients were hospitalized due to lack of treatment response, insufficient compliance or threatening corneal perforation. A causative agent was identified in three cases that required surgical therapy. In the eviscerated eye, staphylococcus aureus was identified in a conjunctival smear. Initial VA ranged from amaurosis to Snellen 30/20. In 90% of cases, preserved or improved VA was observed at the last follow-up visit. An improvement in VA was seen in three of the four patients that underwent cross-linking. In conclusion, the epidemiologic data in this study are in line with previous reports (Schaefer et al. 2001; Bourcier et al. 2003). The majority of cases were successfully treated empirically. Overall, observed clinical care adhered to the AAO recommendations (AAO 2013) including choice of first-in-line antibiotic. However, corneal scrapings for culture could be taken more frequently, especially in cases of severe keratitis.