Most certainly, you need to have a thorough transfusion history on the patient, as my good friend Yanxia says above, but it also depends upon the condition of the patient. If the patient is exsanguinating, the old adage comes into play that it is not a medical triumph to give perfectly compatible blood to a corpse, when, in very many cases these days, a haemolytic transfusion reaction can be treated. HOWEVER, it is ALWAYS a decision to be made by a medically qualified person, rather than a laboratory qualified person to make, as to how urgent the transfusion may be.
IF there is time, it is always worthwhile doing a few more investigations. For example, is the patient DAT Positive, and, if so, is it IgG, IgM, IgA (rare), complement or a combination? Is the reaction seen in the auto-control due to a "cold" auto-antibody, or something else.
To repeat what I wrote above, it MUST always be a decision for a medically qualified person, rather than a "lab rat" (HATE that term, but I hope you know what I mean, without taking offence - being a retired "lab rat" myself), but, if it was a case with which I was dealing, apart from doing a few basic tests (see above) I would be happy to give the blood - and more importantly, receive the blood, if I were the patient.