Re: Warning of major NHS IT overspend

[Brian Beesley <[EMAIL PROTECTED]>]

On Friday 29 October 2004 20:35, Dave Howe wrote:
>
> I think it will be an "at need" system - new records will be added to
> the database, but old records will only be transcribed when there is a
> good reason to (for instance, it is possible the "few sheets of paper in
> a brown envelope" will be transcribed when you move GP, by the simple
> process of sending the transferred files not to the new GP, but to the
> central office) 

The last couple of phrases are one of the bones of contention.

Why does everything have to be stored centrally?

Why cannot the files be retained where they are and links made to them in the 
new GP's system?

The point here is that the security of the records remains with the 
_original_ keeper, so the data subject does not lose out by having records 
transferred to a different system which may not protect the data subject's 
privacy - at the worst case, moving the records to an offshore contractor, 
who may not apply customs and laws relating to privacy in this country.

> and that locally relevant documents that require a lot
> of storage (such as a decently readable scan of an x-ray) will be
> retained in analogue film form at the point of generation, and only a
> reference made in the database as to their existance.

Yes. The point here is that "decently readable" scans of x-rays require a 
huge amount of storage. A hi-res scan of a 35mm frame can easily run to 100 
MBytes; X-ray plates are over 100 times the area. Of course data compression 
techniques (or reduced resolution) can be employed but sometimes vital data 
will be obscured, making the value of retaining the scan moot - at best - it 
stands to reason that if a scan failed to show a critical detail which was 
present on the original film, a misdiagnosis could result.

ISTR the requirement to electronically transfer x-ray images between 
university departments and teaching hospitals was the critical factor in the 
design of two or three generations of the UK academic network, so far as 
bandwidth was concerned.

If you're prepared to store potentially terabytes of information per patient 
(hi-res scans of a few dozen x-ray plates), there doesn't seem any reason to 
transcribe handwritten notes at all - just scan them & record the scans. This 
reduces the cost and skill needed in the area of transcribing notes, and 
probably the error rate in interpreting the notes when they are referred to.
>
>    Data Entry clerks are cheap and entering the results of a few
> tickboxes on a form doesn't take long per each.

But not as cheap or reliable as scanner operators.

> Human entry is more likely for
> transcription of handwritten records in other than a
> scanned-highres-image form.

... with errors resulting from judgements made in interpreting other peoples' 
handwriting....

>    Direct digital entry would be a pain - as you point out, everyone
> would have to have little electronic gadgets to view and update records;
> the patient's life could easily be in the balance due to a battery
> failure or broken ethernet wire.

Yes, so at the very least existing paper records should not be thrown out for 
the sake of "modernization", and systems for storing, transmitting and 
displaying electronic records should be designed to resist breakdown. (Up 
goes the cost again, reducing the cost effectiveness of whatever value the 
switch to an electronic system may have.)

Brian Beesley