Written by Noirin O’Herlihy 1 , Robert Gaffney 2 , Patrick Henn 2 , Anthony G. Gallagher 3 , Mary R. Cahill 2,4
- Red House Family Practice, Mallow Primary Healthcare Centre, Mallow, Co. Cork
- School of Medicine, University College Cork
- Faculty of Life and Health Sciences, Ulster University, Magee Campus, Londonderry, UK
- Department of Haematology, Cork University Hospital
Have you ever looked at a blood result on a patient of yours that is very surprising? Maybe they came to clinic, looked well, didn’t feel tired but unexpectedly had a haemoglobin of 8g/dL? Or you are doing your ward rounds and someone you are about to discharge has a Potassium is 2.2 – you do an ECG and this is normal? If this or scenarios like this ring a bell, you should consider WRONG BLOOD IN TUBE – WBIT pronounced ‘wi-bit’. Proportionately speaking WBITs are rare events in laboratory medicine occurring in <0.01% of samples 1 However, when a laboratory, such as ours churns out over 4 million FBC’s a year, this translates into quite a few WBITs.
In CUH and University Hospital Kerry we have done much ongoing work in documentation and prevention. 2, 3 We describe some of this below.
Since 2008 blood banks in CUH and UHK have been meticulously followed up in documenting samples that could be ‘wrong blood in tube’ (WBIT). Data is recorded in our quality system along with corrective and preventive action. The haematology laboratories in CUH have undertaken a major project on documentation and prevention in recent years and all the laboratories – biochemistry, microbiology and histopathology are concerned about the vigilance for these wrong blood in tube instances.
How do we detect them?
Let me give you some clinical scenarios based on actual cases with names and places altered to preserve anonymity.
Stories from the hospital front line
A patient is admitted into the busy Emergency Department on Friday afternoon unwell and febrile. Blood cultures are taken and sent to the laboratory. On Saturday morning the blood cultures grow a gram negative bacillus. The microbiology doctor makes a call to ED to say that “Mr. Sean O’Sullivan has gram negative Septicaemia” and a puzzled ED doctor replies “Mr. Sean O’Sullivan had an ankle fracture, never had blood cultures and is gone home”.
What happened here? The patient whose blood cultures were taken has been accidently labelled with sticky labels from the patient with the ankle fracture- Mr Sean O’Sullivan. As the patient with the ankle fracture had no blood tests taken, has gone home, was never systemically unwell, has a normal temperature, there are now no implications for that patient. However, the challenge of finding the correct patient and transmitting this urgent report on Saturday morning remains.
A young man was admitted non-specifically unwell. The patient is non verbal and is unable to confirm his identity. His Haemoglobin is 8g/dl and it is decided to transfuse him. Two units are cross matched and a post transfusion Haemoglobin is 15g/dl.
What happened? In this case two patients side by side both had a full blood count taken and sent to the laboratory. Bloods were taken to the main desk in the wards in separate dishes for labelling. Bloods were not labelled at the bed side. At the desk, labelling of tubes resulted in transposed tubes. As a result of this WBIT incident, a patient was unnecessarily transfused while another patient with a Hb of 8g/dl was not detected. In this case, the cross match could have been affected and incorrect blood crossmatched and transfused. We were fortunate that this didn’t happen.
How are WBITs suspected or detected?
In the early years of our campaigns to reduce wrong blood in tubes (WBIT), these samples were mostly detected in the laboratory when a routine ‘DELTA’ check is carried out on samples. This means for all samples falling outside a prespecified range (- including pre-set change within ‘normal’ from any previous result), the previous sample is checked. For example, if the patient in hospital that has a haemoglobin of 8 one day and 12 the next, this would alert the laboratory staff to check if that patient has been transfused. If the patient has not been transfused it is likely that at least one other sample was incorrect and the ward is was notified. Investigation of the cause for the WBIT, usually unearths a circumstance around the bed side which led to sample transposition and/or mislabelling. In the case of the transfusion laboratory, if a patient has had a previous sample (this is why we encourage double sampling in the non-emergency transfusion situation) perhaps blood group is B, the sample comes in and this time it is blood group 0, then we know that one of these samples is incorrect. In this circumstance of possible transfusion, this discovery can be life saving.
There are of course times when WBITs arrive in the laboratory and the results of the sample analysis are unremarkable and so we do not pick them up. Sometimes these samples are followed up by the sender. One scenario where WBIT can be detected is if the requester rings the laboratory and says that they were “expecting a result on Sean O’Mahony” and the lab looks up and finds that there is no sample from Sean O’Mahony. If the results emanate from a small area such as a GP setting, we can track through all bloods received that day and we may find that there are blood samples received labelled with another patient – let’s say ‘John Sheehan’ who the practice can confirm did not attend that day. So we can infer that there might have been a mislabelling of the tube on a patient who did attend with the details of a patient who did not attend. This scenario usually occurs when unlabelled tubes are moved away from the patient.
In another scenario a ward or requesting department will sometimes ring the laboratory and say we have just realised that there was an error with these 2 patients and we have noticed that there are labels on the desk that should have been on a sample bottle. That some other circumstance alerts the medical or nursing staff to the fact that there may have been a labelling error. In this circumstance we are delighted with vigilance of our laboratory user and will happily retrieve the affected samples before any potentially erroneous results are issues and analyse fresh sample pairs from the patients involved.
During the last 5 years, we have documented the progress and origin of many if these errors. Many of these WBIT and we have attempted to implement systems and training for prevention.
In 2017, we engaged with stakeholders in the process of performing phlebotomy in the hospital including nurses, interns, NCHDs, medical scientists from the laboratory and IT staff to discuss the barriers and facilitators to performing phlebotomy safely. Videos were performed of experts and interns performing phlebotomy and a consensus was reached using the methodology of Proficiency Based Progression (PBP) Training developed by Gallagher et al. 4 We produced a metric to describe the correct process of performing phlebotomy in CUH from the instruction to take blood to the dispatch of the sample to the laboratory. A particular focus on avoiding WBIT s was ensured with critical steps highlighted including positive patient identification, labelling at the bedside and if labels were printed ensuring the label matched the name on the blood bottle. The metric was validated. 2
In 2018, we attempted to replicate this training with a smaller team and less support and it was less impactful.
A PBP training programme was developed and delivered to interns and a non -randomised controlled trial was performed to examine the effectiveness of the intervention. 3 In the haematology laboratory, the study demonstrated that interns who received PBP training in 2017 had significantly fewer samples rejected than in the control group (there was a 25% reduction in errors in 2017). A reduction in errors was recorded in 2018 but was not statistically significant. 3
A qualitative study, comprised of observation of interns performing phlebotomy according to the metric (correct process) in their clinical environment followed by short interviews, identified factors which could potentially increase the risk of WBIT, including patients not wearing identification wristbands, difficulty accessing essential equipment, insufficient hardware (e.g computers and printers) and stress. This provided an insight into the environment and context within the interns are expected to perform. 3 From this a subsequent less formal observational studies we can infer that provision of barcoded wristbands and printers to doctors, nurses and phlebotomists at the bed side is an optimal solution. Pre-labelling of tubes and non labelling at the bed side are major risk factors for WBITS
Wrong blood in tube is a feature of sampling and labelling in all systems in all countries in which it has been sought. 1 It is clear from studies elsewhere (US study on blood transfusion from Harvard 5 ) that electronic bar coding, scanning of the patients wrist band and generation of the label at the bed side provides the maximum measure of protection against this error. These systems are being introduced across the country for transfusion samples in the form of ‘Bloodtrack © ’. We would support the introduction of such a system for all sampling and labelling. In the interim labelling should always take place at the bed side -with a pen-to-tube approach where a bed side printer is unavailable.
Meanwhile it is unlikely that we will ever completely eliminate the propensity for the blood in the tube to belong to a different patient. As clinicians, if we receive a result that is not in keeping with clinical details or clinical judgement, always be suspicious and consider repeating the test if needed.
- Bolton-Maggs PH, Wood EM, Wiersum-Osselton JC. Wrong blood in tube–potential for serious outcomes: can it be prevented? British journal of haematology. 2015;168(1):3-13.
- O’ Herlihy N GS, Henn P, Gaffney R, Cahill MR, Gallagher AG. . Validation of phlebotomy performance metrics developed as part of a proficiency based progression intiative to mitigate wrong blood in tube. Postgraduate medical journal. 2020.
- O’Herlihy N GS, Gaffney R, Henn P, Khashan AS, Ring M, Gallagher A, Cahill MR. . Proficiency-based progression intern training to reduce critical blood sampling errors including ‘wrong blood in tube’. HRB Open Research 2021;Dec 20;4:77.
- Gallagher AG OSG. Fundamentals of Surgical Simulation; Principles & Practices: London: Springer; 2011.
- Kaufman RM DA, Cohn CS, Fung MK, Gorlin J, Melanson S, Murphy MF, Ziman A, Elahie AL, Chasse D, Degree L Electronic patient identification for sample labeling reduces wrong blood in tube errors. Transfusion. 2019;59(3):972-80.
Sign up to receive a FREE digital magazine sent directly to your inbox: Subscribe