EEPD. Volume 8. Signposts.
3. Specific Electronic “Trigger Data” and “Action Suggestions”
Newborn baby dies after doctors’ error (Daily Telegraph, 21 Oct 2010)
According to this report the (maternity) staff were said to have been had been warned (presumably at the initial pregnancy assessment) that “there was a strong family history of MCAD deficiency” but the baby “was discharged with his mother the day he was born” and staff “failed to carry out any tests to determine if he had this condition” Two days later he died from a disease of which it was said “it is a treatable condition and sufferers can live a normal life if they follow dietary advice”
A statement by the family said “we did all we could to bring this to the attention of the hospital, so it is difficult to understand why this still happened”
As is usual in such cases the hospital “hoped it could learn from it’s mistakes” - presumably by extra staff training. Yet such aspirations do nothing to prevent an identical or similar human error happening at another time in another place.
Partly this is because of the widespread public delusion that with enough training, with tough examinations, with continuing medical education and with regular re-validation clinicians can be trained to be perfect. All that is needed is to “punish the guilty” despite the acceptance by most clinicians that “it could almost as easily have been me”.
Doctors and midwives are being expected year by year to memorise more and more, despite the fact that most human errors occur not because clinicians don’t know the correct action but rather that they happened not to have thought of the possible need for a particular form of management in a particular situation. After all they are only human!
When computers were first introduced there had been hopes that the risk of such human errors would be reduced. And software could technically be written to allow flow patterned questions about such a risk to be entered electronically at the start of pregnancy. It would then have been technically possible for this to have been drawn to the attention of the paediatricians both on a VDU and on both the birth summary and later in the discharge printout that there was a risk and that tests should have been carried out before the baby’s discharge. Indeed alarm bells could have rung if such tests had not been confirmed to have been performed prior to the baby going home.
Such “human errors” are fortunately uncommon they still occur even in major teaching hospitals as illustrated by the following recent notice on the labour ward wall in a major British teaching hospital.
In a 1990 paper by Lilford et all entitled Risk Factors identified at booking with generated Action Suggestions in a book entitled “Computing and Decision Support in Obstetrics and Gynaecology” that 99 Risk Factors were identified together with appropriate Action Suggestions.
Yet 20 years later very few such Action Suggestions are to be found in the average British maternity computer system.
As with so much medical computing it has turned out to be much more difficult that we ever imagined.
For such Action Suggestions to become more common, three problems need to be addressed
1. The detailed documentation required for such Software in c
2. The cost of writing the software
The main obstacles in Britain has been a) a focus on using
rival commercial confidentiality leading to every maternity computer system being incompatible with any other b) the amount of detail analysis required for each item and c) the market for maternity software being too small to afford the cost.
By using an Open Source internet approach it is intended that the eepd should
i) encourage the standardisation of perinatal data as suggested in the EEPD Volume 5. Logical Priority
ii) encourage the detailed documentation required for each entity as set out below.
Human errors cannot be totally prevented,
but there is no technical reason why computers cannot significantly reduce the risk.
Until this is done we will continue to have to spend more money on lawyers and to get more “Doctors’ Mistakes” headlines than there need to be.