- As you know, mistakes in IVF, as in many other medical areas, can have serious consequences. In 2002, after confirmation of a few incidents in the Leeds General Infirmary unit, the HFEA enforced witnessing, and 4 years after the electronic witness was authorised as an alternative to manual double-witnessing. - So, witnessing can be defined as the “ Double checking performed on all clinical and laboratory procedures ” in the expectation that if a person makes an error, it will be caught by the other.
- But is double manual witnessing as effective as should be? Despite its apparently self-evident value, m anual witnessing is a burden: it increases the workload and distractions within the lab and adds more witness-related paperwork in the laboratory. Furthermore, the effects of the involuntary automaticity described by Toft in 2005 can reduce the effectiveness of double witnessing. That is to say: when the same action is performed repeatedly by the same person, their attention levels decrease, and that is when mistakes can be made. - As a result, several alternative options have been developed and trialled in order to replace the majority of manual witnessing steps: (i) systems based on barcodes such as Matcher, (ii) systems based on silicon barcodes that are injected directly into eggs or embryos and (iii) systems based on Radio Frequency Identification tags. In our case, we decided to use the Research Instruments electronic witness based on radio frequency tags.
The electronic witness of Research Instruments uses self-adhesive RFID tags that are placed on all dishes and tubes containing gametes or embryos. All workstations are equipped with readers and touch screens that register all tagged cultureware and the actions performed according to the cultureware present. Each lab can design its own flowchart incorporating every procedure that they perform. These 3 elements ensure that when gametes or embryos from 2 different patients are brought into the same workstation, the alarm is activated.
In this retrospective analysis, we had 3 objectives: First: calculate the incidence of additional interventions. In other words, how many times a second embryologist is needed to manually double witness a step that could been have witnessed electronically. - Second: analyse the distribution of the mismatches. A mismatch it´s defined as.... - Finally we assessed the true mismatch rate, that is the proportion of mismatches actually corresponding to situations where there is a risk of mixing gametes or embryos from different patients.
- Over 21000 witness steps have been carried out since Dec ’ 10. Of those, only 54 corresponded to mismatches. 25 of those 54 were caused by donation procedures when the flowchart was still not optimised or during training thereby ‘ forcing ’ a mismatch. In fact, only 23 were true mismatches demonstrating a true mismatch rate of 0.11%. If we have a look at the evolution of the true mismatch rate, we can see that after the first two periods, it remains constant around 0.10%. So we are inadvertently bringing dishes or tubes to the same workstation from 2 different patients once every 1000 witness steps. (T) True: when dishes from two different patients in treatment are in the same workstation (even if they are going to be discarded or one of the dishes is empty, as still there is a “ theoretical ” risk of mixing up embryos) (F) Forced: training or when donation at the beginning when the flowchart wasn ’ t optimised at the beginning (U) Use: tags pre-assigned because they were to close to the workstation
Almost 25000 RFID tags were used, but only 17 broke or were unusable – a rejection rate of less than 1 in a 1000. Furthermore, admin assigns and mismatches can be considered as additional “ unwanted ” interventions that require additional efforts (a second person to manual witness). Finally, excluding steps that absolutely require a second manual witness as per SOP, only 1% of all witness steps required the intervention of a second person.
We can conclude that the RI electronic witness system accurately records all witness steps and reduces the workload and pressure that manual witness created when it was introduced. It identified in our lab a low true mismatch rate that compares favourably with other studies already published, like the one that studied wristband mix-ups in an American Hospital setting, that reported an error rate of 0.1%. In our series, all true mismatches were rectified in under 10 seconds In the future we aim to expand the study with data from other centres in order to corroborate our data. With these data we aim to estimate what the likely incidence of mix-ups in IVF laboratories may be worldwide. We all make mistakes...
Mistakes in the IVF lab can lead to catastrophic events (including loss of business, lawsuits, wasted time, increased regulatory burden and in this case a devastating outcome for a couple). For this reason alone, irrespective of the system you use (electronic or manual) some form of witnessing is essential to reduce the chance of this kind of error.
RI Witness in the IVF lab - Alan Thornhill
Increasing efficiency and
effectiveness in the IVF
Laboratory using RI-Witness
Dr Alan Thornhill
Assisted Conception Unit
How often do you make mistakes?
very rare in
mistake was in
We think it is
Witnessing in UKWitnessing in UK
Origin of witnessing
“Double-checking performed on all clinical and laboratory
Electronic WitnessingElectronic Witnessing
Is double manual witnessing as effective as it should be?
Involuntary automaticity (Toft et al., 2005)
RI Witness records all laboratory procedures
Type of procedure + Patient + Operator + Time
Identification of the true mismatch rate (0.11%)
Error Rate Type of study Author
6% Mismatches using barcode based system (MatcherTM
) Schnauffer et al.
0.2% 1000 drugs administration (by 2 nurses) Krause et al.
7 - 10 % Pathology studies Elson D.
0.3 % Mistakes in 4 hospital departments Carraro et al.
0.8 % Data input: keyboard vs. Barcode system Shaw et al.
0.1% Wristband mix-ups involving 2 patients Valenstein P.
•Monitored (constant monitoring)
– Duration and timing
•Operator recorded (who is responsible?)
• Numbers of procedures per person
• Workload planning (who is needed?)
• Time saved per procedure
– Duration, no double witness
• Printed labels
– Efficient, legible, accurate