Acanthamoeba keratitis: multicentre survey in England 1992-6

Aim - To investigate the frequency, outcomes, and risk factors for acanthamoeba keratitis (AK) in England during the past 4 years. Methods - An ophthalmologist in 12 of the 14 regional health authorities (RHAs) coordinated identification of patients in their region presenting with AK between 1 October 1992 and 30 September 1996. Clinical and postal patient questionnaire data were analysed. Results - 243 patients (259 eyes) with an AK diagnosis were identified, equating to an annualised incidence of 0.14 per 100000 individuals. UK resident patients for each year numbered 50, 71, 73, and 32 respectively. Among patients with sufficient data 170/237 (72%) were diagnosed early (within 30 days of presentation), 197/218 (90%) were treated with polyhexamethyl biguanide and/or chlorhexidine, and 40/243 (16%) underwent surgery. Visual acuities of 6/12 or better were achieved by 222/259 (86%) eyes, including 84 eyes of patients under review or lost to follow up. Non-contact lens (CL) wearers were associated with delayed diagnosis, increased need for surgery and a poorer visual outcome (only 10/18 eyes, 56%, achieved 6/12 acuity). 225/243 (93%) patients were CL wearers, and 205/243 (84%) were soft CL (SCL) users. Among SCL user respondents, previously identified risk factors - swimming with CL (47/138, 34%), non-sterile CL rinsing (11/138, 8%), omitted disinfection (85/138, 62%), and chlorine release disinfection (65/138, 47%) - were identified for 125/138 (91%) patients. Conclusions - Earlier diagnosis and more effective medical therapy have improved the prognosis for most AK patients. The study demonstrates the highly preventable nature of the disease: 91% of the SCL wearers could have avoided the disease by refraining from inadvisable practices, and a marked fall in frequency was seen after intensive media attention to AK, possibly in conjunction with increasing penetrance of new CL products. Since the frequency of AK appears to be largely determined by the ever changing trends in CL use, continued monitoring is indicated.