Determination of blood groups for identification tags (ie, dog tags) is not done in medical laboratories, and here may be a 3% to 6% error rate in either the ABO or the Rh blood groups, although a recent study indicates that the risk may be lower. During World War II, however, when blood had to be given in emergencies in forward hospitals and tests for compatibility were impractical, the blood group of the recipient had to be accepted on the basis of his identification tag, in which the known error was from 5 to 25 percent.
Previously, these percentages of error had been acknowledged in the context that dog tags were designed only to identify potential donors and were not designed to be used for transfusions on individual patients.
However, even contemporary medical officers who are unfamiliar with this philosophy may assume that identity tags are provided to guide transfusions without confirmatory testing, as World War II–era medical officers did:
Circular Letter No. 170, Office of The Surgeon General, War Department, Services of Supply. In this letter [dated 2 December 1942], it was pointed out that the policy of blood grouping of military personnel and its purpose seems to have been misunderstood by some medical officers.... [T]he assumption that the first [dog tag or identification tag] typing test would be the only one performed was based on a false premise. To correct the misconception it was pointed out again that the purpose of the program was to simplify assembling donors whose blood would probably crossmatch with that of intended recipients.