NPS plans new drug interaction software for GPs
http://www.pharmainfocus.com.au/feature.asp?featureid=148
The National Prescribing Service believes current drug interaction
software used by GPs and pharmacists is less than useful and is
developing evidence-based software of its own that it hopes medical
professionals will embrace. Pharma in Focus looks at the reasons behind
the initiative and what it aims to achieve.
A couple of years ago the National Prescribing Service commissioned
research into the level of consistency of drug-drug interactions listed
in Australia’s four most commonly used clinical prescribing software
packages.
What it found was disturbing.
There was extraordinary variability between the packages, the NPS'
Stephen Phillips recalls. "There were different levels of information on
alerts between different packages, absence of alert in some packages but
not in others and so on. It made you ask the question 'what is the
usefulness and actual utility of that information with so much
variability?".
The NPS is soon to repeat the research exercise, but in the meantime has
begun work on an independent software system that will use an
evidence-based list of drug interactions. The NPS hopes to complete the
system within 12 months.
Phillips, a Sunshine Coast GP and chair of the NPS pharmaceutical
division support working group, says the lack of standardisation and the
sensitivity of alerts for drug-drug interaction and drug-disease
interaction is a recognised problem among doctors and pharmacists using
clinical prescribing and clinical dispensing software.
"In the context of drug interactions, for example, you can go to any
number of clinical prescribing systems in Australia and the drug
interaction, drug-drug interaction reference source that those
prescribing systems use may well be different for each system.
So doctors A, B and C can prescribe the same drug for the same patient
on the same other medications and depending on the system they’re using
may get a different warning - or no warning at all - for a potential
drug-drug interaction".
PI's the source of the problem
Phillips believes the problem stems in part from the fact that much of
the drug interaction and side effect information contained on current
software systems comes from PI’s (product information).
"There’s a debate starting to happen about the context of that
information itself. Most people would agree that the PI is largely a
medico-legal document that may not connect to the most relevant
information that you need to have on a clinical, day to day working
basis. If the interaction data bases that are available at the moment
are predominantly extracted or lifted from the PI, you may also have a
relatively insensitive and indiscriminate – in clinical terms – listed
information".
The end result of all of these issues, according to Phillips, is a
situation where doctors and pharmacists have become cynical about the
accuracy of the data presented to them.
"And if they are not confident of the evidence base or the place from
where that information is coming, then they tend to become desensitised
to it…[that results in] a flow-on loss of useful advice in a clinical
decision making process and potential loss of appropriate advice in a
prescribing decision".
However, Best Practice’s Director of Development, GP Dr Frank Pyefinch,
said he was not convinced that doctors have become cynical about the
accuracy of data. "I do believe that doctors ignore pop up prompts if
they are too frequent and if they often relate to relatively
insignificant interactions. [However] to get around this, some of the
software products in the market allow the user to set the level of
severity of warnings to display".
Pyefinch says finding the balance between not enough and too many
prompts is a challenge and it will be interesting to see how the NPS
handle this.
Meanwhile, the NPS has yet to decide what evidence based resource it
will use for its new database. It is currently scanning and analysing
resources in conjunction with an expert reference group. What is
certain, however, is that the resource they select will have multi-level
capacity in terms of testing clinical meaningfulness, clinical relevance
and so on.
No hard sell from NPS
Phillips says while the NPS has a working relationship with HCN, whose
product Medical Director boasts the lion’s share of the clinical
software prescribing market - it has not approached the company, or any
others, about swapping databases.
"That’s not the way we’re coming at it. We’ll develop the resource and
put it out there and then hopefully the market will see the benefit of
the resource [and can take up]".
"In the ideal world what we’d be saying is that the NPS has done the due
diligence on this work and we believe what we’ve put forward is
something you can rely on as being meaningful and useful and relevant
and therefore people would take notice of it and plug it into their
decision making.";
Chief executive officer of HCN, John Frost, said while he was aware that
some software products didn’t have comprehensive drug-drug and drug
disease interaction mechanisms, "we would argue that Medical
Director is by a long stretch the leader in that area in terms of the
various interactions that we do flag."
He said Medical Director’s interaction database was developed with input
from a number of sources including HCN, a pharmacist, a GP, pharma
companies and MIMS PI’s.
Open-minded vendors
As for whether HCN would ever consider replacing its database with the
NPS version, Frost says he’s open minded.
"Look, our goal is to provide the best clinical tool to our customer
bas... and we’d welcome any such improvement if indeed it was an
improvement".
Best Practice’s Frank Pyefinch is equally open-minded.
"I think that it is a positive step for the NPS to get involved in
setting some standards in this area and I am supportive of their effort
in principle".
"In practice though, we would need to assess any dataset of drug
interactions that they developed carefully before deciding whether we
would replace our existing database. We would need to be certain that
the new one was better than what we are currently using. If it were not
as extensive, we would need to consider the medico-legal ramifications
of not warning about interactions that it did not contain…. if it were
more extensive, we would need to look at whether the extra warnings
might contribute to the problem of pop up 'overload'".
Questions about maintenance and update frequency would also need to be
answered.
"Depending on the answers to the above questions, we would want to know
whether we could use the NPS dataset as a base, but then supplement it
if necessary with any interactions that it did not contain".
Despite tipping the apple cart, so to speak, Phillips says he hasn’t
received flak from any quarter over the NPS proposal – and doesn’t
expect to.
"We’re not trying to be controversial or confrontational about this,
we’re seeing it as an area of prescribing decision support that needs
some assistance. It’s probably acknowledged across the industry [that it
needs] work and we’re putting ourselves forward to do some of that work
without a proprietary interest".
"If we can refine information to make drug interaction warnings sensible
and relevant to daily practice, then I think it will be welcomed by
medical and pharmacy practitioners and consumers. As for pharma, I
really can’t see a downside at all".
---
Copied in the public interest
_______________________________________________
Gpcg_talk mailing list
[email protected]
http://ozdocit.org/cgi-bin/mailman/listinfo/gpcg_talk
