Editing Draize test (section)

==Reliability==
In 1971, before the implementation in 1981 of the modern Draize protocol, [[Toxicology|toxicologists]] Carrol Weil and Robert Scala of [[Carnegie Mellon University]] distributed three test substances for comparative analysis to 24 different university and state laboratories. The laboratories returned significantly different evaluations, from non-irritating to severely irritating, for the same substances.<ref>{{cite journal |vauthors=Weil CS, Scala RA |title=Study of intra- and interlaboratory variability in the result of rabbit eye and skin irritation tests |journal=Toxicol. Appl. Pharmacol. |volume=19 |issue=2 |pages=276–360 |date=June 1971 |pmid=5570968 |doi= 10.1016/0041-008X(71)90112-8|bibcode=1971ToxAP..19..276W }}</ref>  A 2004 study by the U.S. Scientific Advisory Committee on Alternative Toxicological Methods analyzed the modern Draize skin test. They found that tests would:<ref>William S. Stokes, ''[https://wayback.archive-it.org/all/20090925042928/http://ntp.niehs.nih.gov/index.cfm?objectid=BD09E3A8-F1F6-975E-7C5A023D74150157 Preliminary Evaluation of the Underprediction Rate of the In Vivo Dermal Irritation Test Method]'', U.S. Scientific Advisory Committee on Alternative Toxicological Methods, retrieved 29 June 2009.</ref>
* Misidentify a serious irritant as safe: 0–0.01%
* Misidentify a mild irritant as safe: 3.7–5.5%
* Misidentify a serious irritant as a mild irritant: 10.3–38.7%