Using 1.1 billion prescriptions to identify prescription drug trajectories since 1995

Disease and treatment 17. mar 2022 3 min Research director, professor Søren Brunak Written by Kristian Sjøgren

Researchers have made the first survey of an entire country’s drug use over 24 years. This can provide novel insight into how to optimise drug selection and identify suboptimal or futile prescriptions.

Doctors in Denmark issued 1.1 billion prescriptions to people in Denmark from 1995 to 2019.

They prescribed drugs to people with diabetes, hypertension, heart disease, chronic pain and other illnesses.

Statistics Denmark collects redeemed prescription data and has now analysed them in collaboration with researchers from the University of Copenhagen to find associations between the use of various drugs.

The huge quantity of data reveals that some diseases often precede others and that there are frequent prescription trajectories for alternative medicines for the same disease.

These trajectories provide unique insight into drug treatment in Denmark since 1995 and how doctors may be able to improve this in the future.

“The data are especially interesting for the major noncommunicable diseases, in which several hundred thousand people follow the same prescription trajectories, starting with one drug and then moving on to the next one. Linking data for drug prescription with data from the Danish National Patient Registry can show whether certain drugs lead to better health outcomes and whether they should be featured in future treatment guidelines,” explains the lead researcher behind the project, Søren Brunak, Professor, Novo Nordisk Foundation Center for Protein Research, University of Copenhagen.

The research has been published in NPJ Digital Medicine.

Found 9 million prescription trajectories

The researchers initially mapped each prescription trajectory in Denmark: how people are prescribed the drugs and in what order.

There can be a priority order of prescription for the same disease but can also relate to people being prescribed one drug for one disease and then another drug for another disease.

Mapping 1.1 billion prescriptions over 24 years requires more than a back-of-a-napkin calculation. The researchers therefore used a supercomputer to assemble the data and find meaningful associations.

The researchers set a threshold of at least 1,000 following the same prescription trajectory before attributing importance to it, but this still amounted to 9 million trajectories that comprise an atlas of frequent drug use in Denmark since 1995.

“This type of mapping has never been done before, and being able to do this in Denmark is unique because we have a complete dataset of the redeemed drugs prescribed to the entire population, which only a few countries in the world can emulate over such a long period,” says Søren Brunak.

Many thousands of people follow the same prescription trajectory

In the second part of the research, the researchers examined how they could use the various trajectories they identified.

They found a very interesting prescription trajectory for more than 750,000 people with hypertension (out of a population of 7.2 million), who were prescribed angiotensin-converting enzyme (ACE) inhibitors, or angiotensin receptor blockers either as first- or second-line treatment, shifting and not shifting.

However, the researchers also found differences in outcome. The people experiencing shifts away from ACE inhibitors generally survived longer than those who remained on ACE inhibitors.

Søren Brunak warns against thinking that the data can directly predict causal relationships, but the observation may be clinically significant anyway.

“There can be many explanations for why people change or do not change their prescription trajectories. This may include side-effects or certain drugs not being effective. But the data provide insight into some possible benefits of starting certain drugs. This may indicate that many people would benefit from skipping the cheap first-line drugs and starting on the second-line drugs that are associated with improved survival for those who switch more rapidly,” explains Søren Brunak.

The pharmaceutical industry is interested

The mapping of typical prescription trajectories can also provide interesting insight into the side-effects of various types of medicine.

The data can show whether some types of drugs are associated with an increased risk of being prescribed other types of drugs to reduce various side-effects.

This information is useful for pharmaceutical companies because it can give them a broad overview of the side-effects of their drugs in the real world outside clinical trials.

Søren Brunak also says that many pharmaceutical companies have shown great interest in the data the researchers have compiled and analysed.

“A post I put on LinkedIn has been viewed many thousands of times, and many people from the pharmaceutical industry have asked about the research and the availability of data,” he says.

Opportunity for improving treatment

The data can also show whether the prescribed drug actually has the anticipated effectiveness.

For example, doctors overprescribe drugs for ischaemic heart disease for many people to save the lives of a smaller group of people.

According to Søren Brunak, researchers can come closer to identifying this overtreatment through data, so that doctors can possibly stop prescribing this medicine for the people who may not benefit from it at all but only experience side-effects.

“Identifying the people who are being overtreated and therefore may benefit from stopping treatment is very difficult. But we think that we can do some of this by using this historical data,” explains Søren Brunak.

Stopping unnecessary prescribing that does not benefit people not only helps these individuals but also helps society by reducing expenditure on expensive drugs with no benefit.

“In the future, it will also be interesting when we can compare our data with those from other countries and find differences in prescription trajectories, which can teach us how to benefit people,” concludes Søren Brunak.

Optimizing drug selection from a prescription trajectory of one patient” has been published in NPJ Digital Medicine. Several authors are employed by the Novo Nordisk Foundation Center for Protein Research at the University of Copenhagen.

Søren Brunak is a leading pioneer in the biomedical sciences through invention and introduction of new computational strategies for analysis of biomed...

© All rights reserved, Sciencenews 2020