http://www.fbi.gov/hq/lab/fsc/backissu/july1999/ponce.htm
It appears that this problem has been recognised since 1999.
There is probably a lot of subsequent work either confirming or denying this finding.
It appears that this problem has been recognised since 1999.
Our study questions the dependability usually attributed to presumptive tests. Throughout the process, we found that a stain that was clearly a bloodstain gave a negative result in the presumptive test. This result was obviously a false negative, a nonlegitimate negative that might have detained a perfectly viable and necessary investigation.
<snip>
The second experiment we performed demonstrates the possibility of obtaining false negatives with reagents frequently used in presumptive tests. Just as false positives are sometimes encountered, our tests affirm that false-negative results may be obtained as a result of sample contamination.
The presence of ascorbic acid may not be the only cause of false results. It is possible for a bloodstain to go unnoticed if it has had sufficient contact with a product with a high reduction strength (detergents and foods). Evidence examiners should also consider that a variety of circumstances, such as washing, rain, heat, and time, may reduce the concentration of blood in the sample, with small amounts of contaminant being sufficient to give rise to a false negative in the presumptive test. Examiners may, therefore, encounter stains that look like blood, but these stains would not be recognized as such in a presumptive test. This would give rise to a false negative in an otherwise reliable presumptive test, with the result that important evidence may be lost.
There is probably a lot of subsequent work either confirming or denying this finding.