Users have their say on drug interaction alerts
A survey of users of common prescribing and dispensing software has found they want information on drug interactions and its presentation to be improved.
Software users indicated they want drug interaction alerts to be more relevant to practice and the format and presentation of alerts to be more easy to read.
They also want the alerts to include information about the severity of the interaction and to be visually differentiated.
The NPS survey of over 300 GPs and community pharmacists, published yesterday in the Medical Journal of Australia, found that there appears to be a gap between the current offerings of some software systems and what users are looking for.
Michelle Sweidan, team leader of e-Health and Decision Support at NPS and an author of the article, said the survey was devised following a paper written by NPS in 2009 which evaluated six prescribing systems – Best Practice, Genie, Medical Director, MedTech, Plexus and Profile, along with three dispensing systems.
That paper evaluated the quality of drug interaction alerts and made a number of recommendations from an expert panel on potential improvements to the software packages.
Ms Sweidan said based on the results of that paper, NPS thought it would be useful to talk to the end users as well.
“The respondents wanted information about the severity of the drug interaction, the clinical effects of an interaction and they wanted management advice,” she said. “Those things came through very strongly.”
She said the three most important things users would like addressed in drug interaction alerts were increasing the relevance of the alerts to practice; improving the format and presentation of drug interaction alerts; and the differentiation of alerts by severity.
“They wanted information to be better presented on the screen,” she said. “Currently in some software the information is presented as a paragraph, which can be hard to read when you are in a rush.
“We believe that presenting that information as dot points under headings is much more easily read, but that relies on the underlying knowledge base being structured in a way that information can be presented like that.
“The software vendors need to work together with the publishers of the knowledge bases to look at how the information is structured and how it can be incorporated into the alerts.”
Margaret Gehrig, national business manager for MIMS Australia, welcomed the study and said MIMS had had discussions with NPS in the past about the need to find a way to deal with the varying ways the information is displayed.
“We have indicated to the NPS during our discussions that we would welcome the opportunity to work with them to get some guidelines/standards around this issue,” she said.
MIMS provides data to more than 70 vendors, including some of those in the NPS study such as Genie, Best Practice, Medtech32 and Profile.
Medical Director, which has approximately 50 per cent of the GP clinical software market, uses a dataset that vendor Health Communications Network (HCN) developed through its subsidiary Phoenix Medical Publishing called AusDI Advanced.
“Most dispensing vendors use their own dataset as they need a wide range of data e.g. general consumer pharmacy lines,” Ms Gehrig said. “A couple of them have recently taken our data/drug interaction module but most have their own drug interaction info.”
Ms Gehrig said MIMS provides its information in the same format to vendors and gives them advice on integrating and displaying the data appropriately.
“However, we have no control over their development of functionality other than to review their development and provide feedback from a clinical perspective (which we do as part of their agreement in taking our data),” she said.
“We also give them a copy of our reference product, eMIMS, so that they can see how to display the data correctly and provide the right functionality.”
Respondents to the survey also indicated there were several basic user interface issues that need to be addressed, Ms Sweidan said.
“For example, the size of the window and the amount of scrolling they have to do,” she said. “In some cases the information is squashed up into a small box and you have to scroll a lot. That could be relatively easy to fix.
“Another example is that they get the same warnings appearing more than once when a drug is supplied in different strengths. The classic one is warfarin – it comes in a number of strengths and it comes with a lot of drug interactions.
“If you prescribe a new antibiotic that interacts for example, that alert pops up three times if there are three different strengths of warfarin already prescribed for that patient, so there is a usability issue that the vendors could address at their end.”
Severity is a more complex issue to address, she said. “Users want severity ratings so that they can concentrate on the more important information, which is understandable. Some systems already have severity ratings but with different types of classifications. At their simplest it might be low, medium or high. Users want the severe interactions to grab their attention.”
Ms Sweidan said there were a number of international studies comparing different knowledge bases which have found there are variations in severity ratings, and much of the reason for this is the subjective nature in assigning these ratings.
“We would like to see more consistency in severity ratings but it won't be easy to do. Having an international centralised knowledge base would be a possible solution to this and other inconsistencies.”
Posted in Australian eHealth