- Our Research
- Get Involved
While it’s impossible to imagine the conduct of modern medicine without the use of multiple medications, errors in medication are a major source of preventable patient morbidity and health system costs.
The Australian Commission on Quality and Safety in Healthcare estimates that medication errors which cause negative drug interactions result in at least 190,000 hospital admissions and cost the health system $660 million each year. Such errors can also prolong hospital admissions, result in new co-morbidities during admission, or increase the risk of readmission because of adverse interactions between drugs.
A project funded by Sydney Health Partners is aiming to reduce errors by improving the consistency and quality of the medication alerts for possible drug interactions provided to prescribing clinicians when working in electronic medical records.
Chief investigator, Dr Angus Ritchie, says a lot of the automatically-generated medication alerts seen by physicians are of low quality because they are triggered by drug interaction risks which are clinically insignificant.
“Even when we turn the system settings down so that only the highest priority alerts get through to the prescribers, there are still a lot of low value alerts being issued by the systems,” Dr Ritchie said.
“We want the prescriber to get fewer alerts of higher quality but it is incredibly difficult to achieve that at the moment.”
The project aims to incorporate additional information about the patient to support better targeting of alerts and make decision support more specific and more effective.
“The decision support tools in the EMR systems we use at the moment are not designed to incorporate specific patient information. And because you get no additional information about that patient to help you make a decision, the alerts are unhelpful.”
As the Chief Medical Information Officer at Sydney Local Health District, Dr Ritchie is also concerned about the inconsistencies in the medication alerts provided by different systems.
“If you use more than one electronic medical record system in the course of your work, you’ll find they each give you different drug interaction information.”
Dr Ritchie and his colleagues began getting to grips with the thorny problem by first examining the different drug interaction alerts for the prolongation of QT intervals. A measure of electrical conductivity of the heart, QT intervals can be prolonged by some drug interactions to such a degree that there is a risk of severe or even fatal cardiac arrhythmias. Risks related to QT intervals are responsible for about half of all the drug interaction alerts seen in Australian hospitals.
“The drugs that cause QT prolongation are generally prescribed in post-operative care, or are psychotropic drugs used in the treatment of mental illnesses,” said Dr Ritchie.
The Sydney Health Partners project is comparing alert information shown for the same drug in different systems and cross referencing those with reported incidents of medication error. By doing so the research team is able to distinguish “good catches” and “missed catches” from alerts which were unnecessary.
“It’s mainly about having the subject matter experts decide whether that particular alert for that particular reason is clinically meaningful,” said Dr Ritchie.
“First, we need to understand that interaction, then make it clinically relevant and thirdly translate it back into clinical practice. If we can whittle the drug alerts down to just the ones that are both statistically relevant and clinically relevant, we can then reincorporate those back into a decision support system.”