A proposed classification scheme for chemical injury to the human eye

H. F. Edelhauser, James P McCulley, R. B. Nussenblatt

Research output: Contribution to journalArticlepeer-review


Current eye injury classifications for products are based on the results of a Draize eye test using three rabbits. This test has been shown to be extremely variable and validation of new assays against Draize test results has been unsuccessful. In order to facilitate development and acceptance of mechanistically based in vitro tests that more accurately predict the human response, the ILSI Technical Committee requested that Drs. Edelhauser, McCulley, and Nussenblatt propose a classification scheme for human eye injuries that could result from acute chemical or product exposure. The scheme they proposed was based primarily on initial injury since these main endpoints considered were: 1) degree of injury to the outer ocular coat vascular plexus, and 2) the depth of injury into the cornea. A scheme is proposed that classifies injuries into five categories based on predicted outcomes: 0) full recovery without treatment, 2) full recovery with treatment, 2) possible recovery with treatment, 3) measurable loss of vision, and 4) total loss of vision. The initial scheme has been modified based on input from the broader ophthalmology and vision research community. After critical review by ophthalmological toxicologists, the scheme will provide support for development of more accurate and predictive in vitro alternatives, help standardize reporting of human eye injury, and provide a scientifically sound basis for cautionary eye labeling.

Original languageEnglish (US)
Pages (from-to)278-279
Number of pages2
JournalJournal of Toxicology - Cutaneous and Ocular Toxicology
Issue number3
StatePublished - Jan 1 1999

ASJC Scopus subject areas

  • Ophthalmology
  • Toxicology
  • Health, Toxicology and Mutagenesis


Dive into the research topics of 'A proposed classification scheme for chemical injury to the human eye'. Together they form a unique fingerprint.

Cite this