20 May 2019

Out with the antibiotics, in with the antivirals

Communicable Disease Respiratory

GPs appear to be ditching antibiotics in favour of antivirals for suspected flu cases, which is good news for the fight against antibiotic resistance.

But chronically ill patients, who are more likely to benefit from antivirals, are being prescribed the drug at the same rate as other Australians, a study based on NPS MedicineWise data shows.

The database included 4.2 million patients of all ages with a range of socio-economic backgrounds and health issues.

In the study, antibiotics prescriptions for suspected flu dropped from 31% to 27% over three years, while antiviral prescriptions climbed from 21% to 30%.

The study showed that public health campaigns to clamp down on antibiotic prescribing appeared to be working, at least in primary care settings.

“On the other hand, antivirals are increasing, but that’s not a bad thing,” Dr Carla Bernardo (PhD), the lead author on the study and a postdoctoral research fellow at the University of Adelaide, said.

There is some evidence that antivirals can reduce mortality in people who are hospitalised with the flu, and reduce complication rates in people with comorbidities.

For this reason, you would expect that GPs would be more likely to prescribe antivirals for more vulnerable patients, Dr Bernardo said.

However, the NPS MedicineWise data didn’t show that.

“Our results were very surprising because we didn’t find a difference in antiviral prescribing between healthy people and people with chronic conditions,” Dr Bernardo said. The rate of antiviral prescribing was 24.8% in patients without comorbidities and 25.9% for patients with comorbidities, which was not a clinically meaningful difference, she said.

This was concerning because antivirals such as Tamiflu, Relenza and Rapivab only shortened flu symptoms by about one day in otherwise healthy adults, Dr Bernardo said.

The medication has potential side-effects such as nausea, vomiting, headache, renal events and psychiatric events.

“If you weigh the benefits against the risks of side-effects and the risk of viral resistance, antivirals are not worth it in these otherwise healthy patients,” Dr Bernardo said.

As with many big data projects, the data used in the study wasn’t perfect.

Around 7% of the time, the GP notes were incomplete, and symptoms and diagnoses were not recorded. These consultations were counted as influenza-like illnesses if an antiviral was prescribed.

Another issue was the rate of influenza-like illness presentations. There were only 98,086 influenza-like illnesses recorded for a total of 32 million GP consultations (a rate of three per 1000 consultations).

“That just seems a bit low to me,” Allen Cheng, a professor in infectious disease epidemiology at Monash University, said.

Dr Bernardo said the numbers would have been higher if the research team hadn’t used the strict definition of flu-like illness as the presence of all three symptoms (fatigue, fever and cough).

“What they are trying to do is mine data that GPs enter and, obviously, that’s really difficult,” Professor Cheng said. “Influenza might be spelled incorrectly, or it might be called ‘the flu’ or there could be a lot of variations on fever, cough and fatigue.”

“That said, a lot of that matches with a lot of other things we know about flu. For example, 2017 was a really big year and that seems to be reflected in the figures. And we do think that we are using more antivirals so that matches with what we think is going on.”

Professor Cheng said there was a perception among GPs that patients sometimes expected them to prescribe something for flu symptoms. “And antivirals are certainly more appropriate than antibiotics,” he said.

“The evidence suggests that antivirals shorten the illness by a modest amount, maybe a day. And particularly if you give it early. And so, the decision for a GP to give it really depends on is the patient’s time worth more than the $50 to $60 that it costs the patient to get the prescription.”

While the link between antibiotics use and antibiotic resistance was strong, there was little evidence that antiviral use led to antiviral resistance, particularly when use rates were still quite low, he said.

BMJ Open 2019, 11 May