The two sample, or “group check”, rule has been widely implemented in the UK. It's use increases the chances of identifying a Wrong Blood in Tube (WBIT) and thus preventing ABO incompatible transfusion.

Following a series of WBITs, we became aware of a practice in our hospital whereby junior doctors were taking two transfusion samples at the same time and labelling them with different times.

We undertook a survey to ascertain understanding of the two sample rule and the implications of taking two samples at the same time but labelling them with different times.

All had received transfusion training and were aware of the existence of the two sample rule. 17/22 identified the reason for the rule was “safety” or “to avoid error” (freetext responses). The rest gave either no response, or incorrect responses such as “antibodies changing”.

When posed a scenario of a patient with a historic sample 5 years ago, 13/22 (59.1%) incorrectly stated two new samples were required.

In response to being asked what the correct procedure was if the doctor was the only person available to perform phlebotomy, only 7/22 selected the correct answer: take two samples as long as you identify the patient twice and perform phlebotomy twice...

...Most erroneously stated the two samples have to be taken 30 minutes apart.

Respondents were asked to state the rate of WBIT rates for full blood counts and transfusion samples; 81.8% and 90.9% overestimated the true rate (approx 0.5% and 0.05% respectively).

**82%** doctors confessed to having taken two samples at the same time and labelled them with different times. 41% did this frequently and a further 27% whenever the patient needed a second sample.

Five doctors (22.7%) reported taking two samples themselves and asking someone else to sign the second tube.

We demonstrated a lack of understanding among junior doctors of the reason for the two sample rule as well as the requirements of the two sample rule. Almost all overestimated the WBIT rate.

Despite this a significant majority admitted to undertaking unsafe practice by circumventing processes designed to reduce harm due to error.

How can we help people taking transfusion (or any) samples to understand the consequences of their workarounds? We see how much harm is done not just to patients but to the healthcare workers involved when things go wrong.

Teaching and training don't seem to help. IT may provide a long term solution but is expensive, takes time to implement and workarounds still exist. How can we convince people to do the right thing?