Study did not start due to COVID and long term staff sickness

Microbial keratitis is a common and serious eye disease in Edinburgh. Prompt treatment with antibiotics can prevent sight loss, and identification of the micro-organism and its antibiotic sensitivities are key to appropriate management. Standard practice of collecting infected material from the cornea using a blade can be distressing and time-consuming. Corneal impression membranes (CIM) have recently been introduced to another National Health Service (NHS) eye unit (St Paul's Eye Unit, Liverpool) as they detect more micro-organisms and are more patient-friendly than corneal scrape. The aim of this study is to compare CIM with reference to current standard practice of corneal scrape. If CIM have greater sensitivity and fewer adverse events than scrape then the investigators will consider using CIM instead of scrape in routine clinical care. The investigators will also collect additional CIM to help develop new microbiological tests being developed at the University of Edinburgh, which if successful could be applied to CIM at the bedside to further improve the speed of diagnosis in the future. Development of the new microbiological tests is facilitated by having samples of germs from eye infections.

Study design: cross-sectional study comparing diagnostic techniques Participants: recruited from the Acute referral clinic at the Princess Alexandra Eye Pavilion, or emergency on-call ophthalmology service What is involved: subjects with microbial keratitis will have standard investigations to identify the germ causing the infection. In addition the investigators will capture germs using CIM, and will compare CIM with the standard test to see which is better.

Funding: departmental funding

Microbial keratitis is a common serious eye disease in Edinburgh. Prompt treatment with antibiotics can prevent sight loss, and identification of the micro-organism and its sensitivities are key to appropriate management. It is standard practice to collect multiple samples from the cornea using a needle or blade and plate material from an ulcer onto culture media. However this can be distressing to the patient, and the procedure can be especially difficult in those who are not co-operative or where there is a risk of corneal perforation. Corneal scrapes are also time consuming to prepare and perform. Corneal impression membranes (CIM) have recently been introduced to clinical practice in St Paul's Eye Unit in Liverpool, after a study found significantly higher isolation of micro-organisms from CIM compared to corneal scrape (40.8% vs 26.9% in n=130 eyes). However, CIM are not currently used in clinical practice in Scotland, and the results from Liverpool have not yet been replicated in a second study.

What questions does the study aim to answer?

1. Is CIM better at capturing the germ causing corneal infection than corneal scrape in Edinburgh?
2. Is CIM acceptable in terms of side effects, compared to corneal scrape?
3. Can new diagnostic techniques be used on CIM to detect micro-organisms?

What new information will the research provide? Whether or not CIM are clinically useful in Edinburgh. It is possible that this study will find a higher frequency of contaminants with CIM, and the investigators will record the specific organism as well as the frequency for analysis. The project will also determine if new diagnostic techniques being developed in the University of Edinburgh can be used on CIM from patients.

This is a cross-sectional study aiming to compare the diagnostic accuracy of corneal impression membrane (CIM) against a clinical reference standard of corneal scrape.

The hypothesis is that the CIM is better at picking up germs from the eye than the traditional corneal scrape. A second hypothesis is that CIM is more acceptable for the patient.

The study follows the methods of a previous study in Liverpool, and will recruit a consecutive series of patients with bacterial, fungal, or acanthamoeba microbial keratitis. Patients seen in the Princess Alexandra Eye Pavilion Acute Referral Clinic, or in evenings or weekends on-call, will be invited to take part. The study aims to recruit 120 eyes (to detect a difference of 0.15 between the discordant proportions (80% power, 2-sided significance 5%; McNemar's test; based on the results of the Liverpool study).

Following informed consent a doctor will anaesthetise the cornea and collect corneal material from the ulcer using surgical blades. Corneal scrape will be done once for each plate of culture media, once for a Gram stain smear, and once for acanthamoeba (if indicated), per eye. This is standard clinical care.

After this, a series of four CIM will be applied to the ulcer, while the eye is still anaesthetised. The CIM is a circle of filter paper on a plastic applicator. It is touched to the ulcer for 3-5 seconds, and then placed in a sterile container with broth.

Samples from the cornea will be sent to microbiology for processing, according to standard clinical practice, along with one CIM for comparison. Three further CIM will be sent to the University of Edinburgh for analysis using new microbiological tests. Patients will be treated empirically according to current local standard practice while awaiting microbiology results.

How is keratitis diagnosed? A doctor can tell if there is an infection of the cornea (i.e. diagnose microbial keratitis) simply by examining the eye. However, in order to find out which germ is causing the infection samples of the ulcer must be sent to the microbiology lab. Traditionally this is done by scraping the cornea with a blade or needle, but this is unpleasant and doesn't always pick up germs from the eye. The investigators think that the CIM will pick up more germs and be less unpleasant for the patient.

What will the new microbiological tests at the University involve? CIM sent to the University will be tested in a lab using new methods currently under development. These use special probes to detect bacteria. The investigators will determine whether optical Smart Probes, which are able to detect Gram positive or Gram negative bacteria or fungi in real-time (developed within the Proteus group (proteus.ac.uk) at the University of Edinburgh) are able to identify the infective pathogen on the CIM.

How and where will patients be recruited? Patients presenting to the Princess Alexandra Eye Pavilion will be invited to take part if they have symptoms and signs that are typical of microbial keratitis. These patients will be recruited from the Acute Referral Clinic, or the on-call service out of hours.

How long will visits take? A normal visit to the Acute Referral Clinic or on-call eye doctor for microbial keratitis can take up to an hour. The extra work of consenting the patient and applying CIM is unlikely to add a significant amount of time to the visit. The investigators estimate an addition of less than 30min in most cases. Although patients can take longer to decide to take part, the process of applying the CIM is itself very quick (approximately 5min overall, the CIM are only on the eye for 3-5 seconds each).

Are there any follow up visits? Patients will be followed up as part of standard care, but this project does not involve any follow up for research purposes.

How are participants treated? Standard care for microbial keratitis is broad spectrum antibiotic eye drops. These are given straight away after the corneal scrape. Antibiotic treatment is changed later if the lab test results show this is necessary. The investigators will treat according to standard practice.

Do ophthalmologists require special training to apply the CIM? The investigators will give the doctors one training session on correct procedures for scrape and CIM. In addition there will be instructions for how to apply the CIM in a bag with the CIM to be used for each participant.

If there is an ulcer in both eyes will you sample both? Yes

Will the participant have had any treatment before recruitment? Possibly. Sometimes patients with microbial keratitis have been treated by their optician or General Practitioner (GP) before being sent to the Eye Pavilion. This will not affect eligibility for the study, but the investigators will record which antibiotics have been used before recruitment.

How are CIM prepared? CIM are supplied from Merck in sterile packaging. They are ready to use. There is no further processing.

Inclusion Criteria:

• Adult patients 18 years old or older
• Appearances typical of a new bacterial, fungal or acanthamoeba infection of the cornea, in one or both eyes
• Able to give informed consent

Exclusion Criteria:

• Patients who don't give consent
• Patients who would not normally have a corneal scrape as part of routine care. This includes patients with viral rather than bacterial corneal infection, such as herpetic keratitis.
• Patients with corneal perforation or descemetocele

Patients attending the Princess Alexandra Eye Pavilion in Edinburgh, who are given a clinical diagnosis of microbial keratitis in one or both eyes