Summary Care Records: the truth or nothing like it?

Richard Veryard has written a response to our report on inaccuracies and ommisions in the Summary Care Records database. He writes:”What’s wrong with the single version of truth.”

Researchers at University College, London, found in a confidential draft report that doctors were unable to trust the SCR database as a single source of truth.

NHS Connecting for Health has pointed out that errors and gaps in the SCR database are not its fault – they’re because of problems with GP-held data.

CfH’s response characterises its approach to the NPfIT: if problems are not of its own making, they don’t particularly matter.

But if large numbers of clinicians access the SCR and find the data is untrustworthy, they may decide not to take the time to use it again, which could be a disaster for the SCR, though not for CfH.

To doctors who access the SCR database it doesn’t matter how thedatabase came to be unreliable. Why would they care whether the faultlay with GPs or glitches on uploads of GP-held data to the NHS Spine?If the data isn’t sound, the database isn’t trustworthy.  

Fewer than 20 summary care records accessed every week

CfHhas already been bewildered by the lack of use of the SCR. Researchersat University College London found that CfH had commissioned aninternal inquiry into the poor uptake and use of the SCR, particularlyin secondary care.

Fewer than 20 summary care records were beingaccessed within secondary care across the country, UCL researchersfound; and their research is up to date – it continued until the secondweek of this month. It involved full cooperation with NHS Connectingfor Health and out-of-hours software supplier Adastra.

Should CfH care that the SCR database is inaccurate? Its job to roll out SCRs

CfHand the Department of Health give the impression that they care lessabout inaccuracies in the SCR database than rolling out the scheme asquickly as possible, which, perhaps, is not surprising.

UCL researchers found that CfH is a mainly political organisation,although it is also commercial and technical. Its responsibilities areclosely aligned with government policy. Its job is to deliver the SCR.

CfH, it seems, takes the approach that data quality can always betackled later.  But will inaccuracies and omissions ever be seriouslyconfronted?

Richard Veryard says in his article:

“…the pusillanimous way out is to build a database with imperfect data,and defer the quality problem until later. That’s what people havealways done, and will continue to do, and the poor quality data willnever ever get fixed.”

If the data doesn’t get fixed, will the SCR ever be of any real use, even if CfH and BT manage to deliver it?


Update – 23 March 2010 (pm)

In response to this article an NHS Connecting forHealth spokesperson kindly emailed the following comment:

“Having read your recent article –  ‘Summary Care Records: the truth or nothing like it’ , I wanted to make you aware of a couple of facts.

 –       A heading in the article says ‘Fewer than 20 Summary care records are being accessed every week’.  The truth is that over 400 SCRs are being viewed every week. The total number of SCRs accessed last week across out of hours and Secondary Care was 418.

 –       Also you say NHS CFH takes the approach that data quality can always be tackled later.

“We have published very clear guidelines as to the data quality requirements prior to the creation of records and have always been very clear about this-


My response:

NHS Connecting for Health is right to point out that the SCR database is accessed much more often when the numbers of out-of-hours accesses are added to the secondary care accesses. CfH says that the “total number of SCRs accessed last week across out of hours and Secondary Care was 418”.

When I asked for a breakdown, CfH was candid enough to say that last week there were 29 accesses in secondary care and 389 in out-of-hours.

This ties in, roughly, with research by University College London which found that fewer than 20 SCRs per week were being accessed in secondary care across England. It also ties in with the fact that CfH commissioned an internal inquiry into why the SCR is accessed so little in secondary care.  

My understanding is that UCL researchers found that about 15,000 SCRs have been accessed in total, and that the access rate varies from 2 to 20% in out-of-hours and walk-in centres, but is only about 0.1% in secondary care.   

That said, productive and widespread access to accurate SCRs by out-of-hours services would be an undoubted success for the NPfIT, provided of course that the hit rate is high. A hit is when a clinician accessing the SCR finds the correct patient’s records on the database.

A high SCR access rate by out-of-hours services doesn’t mean much if the clinicians can’t find the record they are looking for. Statistics, statistics …


What’s wrong with the single source of truth – Richard Veryard

Confidential report on Summary Care Records finds database is inaccurate – IT Projects Blog

Clinical champions and the NPfIT – Open Source Medicine

Why I’m opting out of Summary Care Records – Ianvisits

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

I'm not sure that the business case for SCR - and for potential uses - has been thought through or presented well.

Taking the question of where a SCR would be useful.

The most benefit will be derived for patients on a lot of medication and with complicated medical problems: it is unlikely that any emergency service would routinely check an SCR for a healthy 18year old knocked down by a bicycle!

(If I'm right, this would reduce the number of accesses significantly).

The first uploads will (not sure about the early adopter sites) only contain medication, allergies and adverse reactions: vital for medicine reconciliation after admission and in an OOH situation:less vital in an emergency.

After this, other data will be uploaded - but there does not appear to be a consistent plan for what should be uploaded: it may depend on individual GPs or practices - which means that the record will be incomplete - but the end-user just won't know where: not a good basis for making decisions.

The quality of the medication information is likely to be good - but does not include medication prescribed outside general practice such as chemotherapy: allergies and adverse reactions are useful in the practice, and may be reasonably complete: other information is gathered for enabling the practice to take care of the patient, and not to feed the SCR: there can be no guarantee that, even if it was all uploaded (which it won't be) it would be usable or complete from the end-user in emergency care's point of view.

As for uploads, my understanding is that in East of England the plan is to start as soon as the opt-out period is over - so in most areas, mid-June.

I just wanted to say that this blog helped me enormously when writing my own blog and getting to grips with the complexity of Summary Care Records.

Thank you

Thank you for the kind comments - and for the reference on your blog. I am glad the posts have been useful.