Tom Beale Wrote:
"John S. Gage" wrote:
> Brian Bray wrote:
> >
> > The more I think about EMR, the more I think of a document for each
> > patient. No database or middleware needed ;-)
Well, except for:
- physician access (as noted by John)
- administrative access
- data filtering, querying and searching
- security, encryption
- visual filtering
- multimedia
- automated processing of populations of records
If I may add, the role of database in:
- decision-support particularly self-learning tools - ANN, CBR etc. and data
mining.
- Different views for same data
- Monitoring kind of situations where same data fields are required to be
filled repeatedly for a single patient (in contrast to use of database to
access records for a group)
- Relational access from one process to another e.g. from clinical to
administrative
> You just have to ask yourself: how much do we really need to know about
> patients? Do we need images, or just reports on images? Can we get
> away with 90% historical data. Mike Roizen's editorial in the NEJM
> seems to suggest that we can.
It seems to me (as a non-physician) that there are two questions here.
1. Are all the images (tests, scans etc) we make of patients necessary? (If
not, there needs to be a process to reduce clinical practice to its
necessary level for medical care to occur properly.)
2. If (once such a process has taken place) we have images, scans etc -
i.e. ones we deem necessary, then surely they want to be available in the
record. Why take them in the first place?
How much data is right? In medicine, we constantly struggle with this basic
question. Given the ambiguity in this area, an audit committee imbued with
Spanish inquisition zeal can always point out deficiencies in most
meticulous of records. Does that mean that we should record all data for all
situations for all the patients? Clearly, not a very practical answer! It is
only the clinical context that decides what is an appropriate dataset for a
situation. By prescribing certain amount of data (or, for that matter,
clinical actions) for a clinical problem, one is making an arbitrary
decision. This implies that this much amount of data is appropriate for a
subset of patients, but certainly not all. The cut off point is decided by
the one who prescribes the data to be collected. If advances including those
in information technology make it feasible to collect and use more data then
datasets for each situation would enlarge. This does not mean that some data
which has been recorded for another situation may not have a bearing on
another clinical problem or situation. Such data may influence seemingly
unconnected problems in serendipitous ways. Therefore, it stands to reason
that one opportunistically exploits such data in a situation even if it is
not within the prescribed dataset for it. All data recorded for a patient
has to be available for possible future use.
Hemant Shah
