Authors: Amy Basilious, Mary Feng, Rookaya Mather.

Disclosure Block: A. Basilious: None. M. Feng: None. R. Mather: None.

Abstract Title: Prevention of exposure keratitis in intensive care units: Case reports, survey, and proposed protocol

Abstract Body:

Purpose: Patients within intensive care units are at high risk of exposure keratopathy. Ocular protection can prevent significant ocular morbidity, yet little is known about the ocular care protocols followed in ICUs. This study reports three cases of globe-threatening neurotrophic ulcerative keratitis arising in the ICU setting following neurosurgical procedures. The current state of ocular care across ICUs in Ontario was investigated and an algorithmic ocular care protocol for ICU patients is discussed.

Study Design: Cross-sectional telephone survey and retrospective chart review

Methods: A telephone survey of intensive care units in major cities in Ontario was conducted. ICU charge nurses were asked about ocular care in their respective units. Questions focused on eye care protocols, assessment of eyelid closure, methods of ocular surface protection, recent complications, and ophthalmology consult practices. Data were reported in aggregate.

Results: Responses were collected from eighteen ICUs. Eyelid closure was formally assessed in 56% (10) of units. Less than one third (5; 28%) of ICUs reported having an eye care protocol which is routinely implemented for unconscious patients, while 72% (13) reported a protocol for patients who are unable to close their eyes. Lacrilube^TM (17; 94%) and artificial tears (16; 89%) were common components of protocols, with 15 ICUs (83%) using both. The majority of these protocols indicated tears or Lacrilube^TM on an as-needed basis, with one ICU reporting tears prn as the only component of their protocol. Infection and redness were the most commonly reported reasons an ophthalmology consult would be made. Additionally, five (28%) units indicated that a consult would be made for any eye problem encountered. Corneal abrasion, ulceration, unilateral blindness, infection, stretched optic nerve, and swollen orbit were reported as complications recently encountered.

Conclusions: Prevention of exposure keratitis is inconsistent between individual ICUs, with many operating in the absence of defined protocols for ocular protection. Variability and inconsistencies within ocular care practices may be causing undue harm to patients at risk. These findings highlight the importance of implementing a simple ocular care protocol across ICUs. We propose universal precautions be initiated on admission to ICU for all unconscious patients. This would require instillation of non-preserved ointment such as Lacrilube^TM every 4 hours. Prior to instillation of ointment, an ocular surface assessment should be conducted for purulent discharge, conjunctival hyperemia and opacification of any kind on the cornea. If any of these are identified, an ophthalmology consult should be initiated.