This study concerns the safety of blood transfusion. Prior to transfusion, staff should perform a number of essential safety checks to ensure that the correct patient is receiving the correct blood product. Evidence suggests that these safety checks are not always done. This study has been designed to assess the effect of a simple intervention on the performance of the bedside safety check.

The hypothesis is that a simple intervention will improve the performance of the bedside check.

The local collaborator at each participating hospital will identify two clinical areas comparable for the following elements: similar hours of service, similar use of red cells, similar level of intensity, and similar nurse-to-patient staffing ratios. For the purposes of this study, a clinical area is defined as a discrete area where one type of clinical care takes place. Comparability will be established by completion of a checklist.

The local collaborator will transmit comparability data to the study co-ordinator who will pass it on to a nominated medical consultant (Professor Mike Murphy) and the study statistician. If both parties are satisfied that the two clinical areas are comparable, confirmation will be transmitted to the local collaborator together with authority to commence the two-week baseline audit. Once the baseline is complete the local collaborator will contact the statistician, via the co-ordinator, to be informed of the area for intervention. The local collaborator will then inform the blood bank of the intervention area.

The intervention will be applied to one clinical area or ward in each participating hospital. Both clinical areas will be audited during a two-week baseline observation period. The two areas will then be randomly allocated to either the intervention of labeled blood bags or no intervention. Auditing will carry on in both areas until 2 weeks after the intervention has been introduced. To assess the longer-term impact of the intervention, there will be a further two-week period of audit 8 weeks from the start of the intervention. The intervention will be carried out for the full 10-week period.

The randomized comparison will be based on the data collected in the 2-week period after one area has received the labeled bags, and both areas have been observed by an observer for the same length of time. Thus the effect of the intervention will be separated from any initial effect that the observer may have on bedside checking practice. This comparison will also be made on the long-term audit data collected at 8-weeks.

The intervention under investigation is a simple label applied to the blood bag in such a way to cover (obstruct entry) the "ports" used to administer blood. The label will be applied to units of crossmatched red cells only. The position of the label will require that the user remove the label to administer the red cells. The label will state: "STOP: Check the patient's wristband". The wording will be translated to the appropriate language as necessary.

Observers (auditors)- not blinded to the test or control areas - will be asked to audit the bedside check at the time of transfusion. This will be done by direct observation. The auditor will take up a position to observe practice and will therefore be obvious to the staff. However the auditor will not participate in the transfusion check, not answer questions, and not reveal observed errors in the check (unless they observe that the wrong blood will be transfused, in which case they should intervene). Observations will be recorded on the data form.

Participants are invited from Europe, North America, South America, Asia and Australia. The target is for at least 15 hospitals.

Inclusion Criteria:

• Crossmatched red cells

Exclusion Criteria:

• Uncrossmatched red cells, platelets, fresh frozen plasma