In this post we look at a data quality failure that comes very close to home as my newborn son falls foul of a faulty NHS patient ID allocation process.

We look at how easy it was for the situation to arise and the potentially serious issues that this kind of data quality failure raises for the quality of patient care.

NHS unique ID failure leads to duplicate newborn patient records

My son was born several weeks ago. As for all newborn babies, the first duty of the midwife (after bringing him screaming into the world) was to put an identity tag onto his wrist and ankle.

That tag made his identification unique as it provides what appeared to be a unique local hospital number, time of birth and name.

All went well so we returned to the ward.

However, the following day there was confusion when one of the staff asked whether we lived in Peterborough. We actually live in Stratford-upon-Avon which as the map below illustrates, is some considerable distance away.

We therefore enquired why my son was incorrectly identified as living at a different location. It then became clear that there had been an error in the NHS unique patient ID allocation process.

We were told that because another baby with a surname Jones had been born on the same day then the system (or staff) had mistakenly allocated his number to our son ie. my son's records showed the same NHS ID as the child with the same surname in Peterborough. The staff informed us that the issue would be resolved (he received a new NHS ID) and we left hospital.

Several weeks later we expected to receive notification regarding one of the scans that is mandatory for newborns but the details never arrived.

In light of the earlier confusion, we proceeded to chase up the appointment.

What transpires is that our appointment had gone to the parents of the other baby Jones in Peterborough. (We hope they also realised the error and didn't make the 90 mile trip to attend our son's appointment).

So now we're trying to unravel exactly what happened.

This is the chain of events so far:

1. Our son, surname Jones, born 27/04/2009 and recorded on local medical records with unique hospital number
2. Other child born 27/04/2009  in hospital 90 miles away with surname Jones
3. Medical staff use local hospital number to identify our son and then link this number to the NHS unique ID of the other child (we have a computer printout of the first entered record showing local hospital record mapping to a Peterborough address, we didn't spot this at the time)
4. Our son's local hospital record has the name and address corrected, no longer points to Peterborough address and a new NHS ID issued for our son but this is not updated on the local records
5. NHS ID of our baby still appears to be incorrectly mapped to Peterborough baby across local medical records in hospital
6. Hospital staff finally update records to show new (correct) NHS ID
7. Several weeks elapse and medical staff issue scan instructions against allocated NHS number, appointment sent to Peterborough
8. Deadline for scan approaches so we chase appointment, discover that the records are still incorrect
9. Health service promises to resolve issue

What are the implications for this?

Right now, we don't know how many systems and departments have the wrong details. Clearly, changes made in his local records and our paper records failed to adequately filter through the system and perhaps we'll never know if there are departments or systems which still reference our son incorrectly.

When we registered him with the local GP the secretary had to make an educated guess as to which NHS ID was accurate. I'm sure the parents in Peterborough of their baby Jones are equally concerned.

What is even more alarming is this snippet I found from this post:

"...one group member pointed out that the NHS Number did not prevent patients impersonating others.

She said there had been an incident where an imposter attempting to obtain free NHS treatment had slipped through the system to the point when a blood transfusion was about to be given. A discrepancy between the imposter’s blood group and that of the genuine NHS Number holder alerted staff to the problem."

So, the NHS number is clearly used to hold vital medical information. I'm not a medical specialist but this caused me real alarm knowing that in the event of an emergency, a medical team working off the medical records associated with my son's NHS number could have incorrectly administered medicines based on the wrong blood group or clinical background.

Clearly there is a major flaw here.

Yes, the staff member should have picked up the fact that the postal region was different to that of the hospital my son was born in but is it really their responsibility? Surely there should be additional measures here to prevent this occurring.

For example, what if another child with the same surname in the same town had been born in the same hospital?

If the NHS number allocation system is so flawed that it cannot detect a difference in births between two completely different hospitals, with different names, in different towns, 90 miles apart - what is the likelihood of the system allocating the correct NHS ID to two children with the same name, on the same day, in the same hospital?

Clearly, in this case there was no immediate danger to either child but as there is a continued effort to standardise on the unique NHS ID it certainly points to major flaws in the current process.