It’s finger lickin’ blood! An artefactual cause of raised amylase

Pre-analytical factors affecting test results are well recognised. We observed a number of cases of significantly increased serum amylase results in children that decreased to normal faster than would have been expected given its known circulating half life. In the majority of cases the results were considered to be random errors and patients were sent home without follow-up investigation. One patient, a two-year old with complex medical problems, presented with unexplained variable increases in serum amylase activity ranging from 206 to 3055 U/L over a six-week period. On questioning it was established that all samples had been obtained from finger prick (capillary) collection and that the child was known to habitually suck his fingers and chew his hands. It was therefore suspected that the high amylase results may have been due to contamination from saliva on the skin. A repeat venous sample, taken two days after the peak amylase result of 3055 U/L had been reported, was normal (78 U/L). Subsequent weekly venous samples have also all been within normal limits. Electrophoresis on a sample with raised amylase from the same child showed it to be predominantly salivary amylase and no macro-amylase was detected.

To further investigate the potential for salivary amylase contamination, paired (clean versus sucked) finger prick samples were taken from seven volunteers. An increase in serum amylase activity due to contamination from saliva was observed in all cases. The magnitude of the increase varied. However, increases of >500 U/L were seen in three out of the seven cases, with a mean increase of 318 U/L (median 199 U/L; range 44–724 U/L). Cleaning the finger with an alcohol swab reduced the measured amylase activity although results remained higher than those obtained from clean (washed) fingers. Salivary amylase activity of one volunteer was measured as 92,458 U/L, indicating that as little as 1 µL of saliva in a 0.5 mL blood sample could increase the amylase activity by 200 U/L.

Contaminated samples may result in further unnecessary investigations including repeat blood sampling and abdominal ultrasound, in addition to more invasive investigations. It is therefore important that both clinicians and laboratory staff are aware of this potential for misleading results. We have changed our policy to request that where amylase measurement is required the sample should ideally be a venous sample. Where this is not possible, thorough hand washing is advised. All high amylase results from finger prick samples should be confirmed on a venous sample.