Evidence
View commentaries on this research

Two antibiotics (doxycycline and azithromycin) are not effective treatments for people with COVID-19, research found. The same study also showed that the anti-inflammatory drug (colchicine) makes little difference. These drugs should not be used to treat people in the community who have COVID-19.

However, an inhaled steroid (budesonide) improved recovery from COVID-19. The researchers say it is likely to reduce hospital admissions among people at high risk.

In normal times, it takes years to test drugs and to find out which groups of people might benefit. But during the pandemic, there was a desperate need for early, reliable data on treatments. Many treatments which are widely used in other conditions, and were thought likely to help, were given to people with COVID-19. Researchers set up a platform trial, which allowed them to test several of these drugs within a single study.

This trial, PRINCIPLE, included people in the community who had been unwell with confirmed or suspected COVID-19 for up to 14 days. They were all at high risk of hospitalisation or death from COVID-19 because of their age (65 years +) or because they had long-term health conditions.

The researchers hope their work is discouraging use of these antibiotics to treat people with COVID-19 in the UK and around the world. These drugs remain valuable for treating other illnesses. Using antibiotics only where they are likely to work reduces the chance that bacteria will become resistant. Avoiding ineffective treatments, including colchicine, spares people from the risk of side effects.

What’s the issue?

When the COVID-19 pandemic struck, safe and effective treatments were needed quickly, especially for those at high risk of needing admission to hospital. This includes people over 65, along with those who have multiple long-term conditions (for instance a weakened immune system, high blood pressure, diabetes or asthma).

Running several separate trials to see whether existing drugs work against COVID-19 would have taken too long. To speed up the process, a group of researchers in the UK set up a trial called PRINCIPLE (Platform Randomised Trial of Treatments in the Community for Epidemic and Pandemic Illnesses). Platform trials can test several drugs at the same time.

The drugs tested in PRINCIPLE were all commonly used to treat other illnesses; the question was whether they could help people with COVID-19. People in the trial all received the best available NHS care. Some were randomly chosen to receive, in addition, one of several drugs. Testing for each drug stopped when there were enough data to show whether or not it works.

Some previous research had suggested that 4 drugs in widespread use for COVID-19 might be effective: two safe and inexpensive antibiotics (doxycycline and azithromycin), an anti-inflammatory drug used to treat gout (colchicine) and an inhaled steroid used for asthma (budesonide).

The antibiotics especially were being used for people with COVID-19 before the study reported. Researchers used the platform trial to test whether these drugs are effective against the respiratory symptoms of COVID-19 in people at high risk of hospital admission or death.

What’s new?

Most participants were being treated at primary care centres across the UK (others joined the study online). They were either aged 65 years and older; or 50 years and older with multiple long-term conditions (average age was 61-64 years). Just over half were female.

Everyone taking part had been unwell for up to 14 days with suspected or confirmed COVID-19. They all received usual care, and some received one of the study drugs in addition: doxycycline, azithromycin, colchicine or budesonide.

People answered 8 questions about their symptoms every day for 28 days. The study looked at recovery times, hospitalisation or death within 28 days.

Doxycycline. An analysis of 1728 people included 780 (45%) who received doxycycline. Doxycycline had no meaningful benefit above usual care. Compared with usual care alone, the doxycycline group:

  • improved only half a day sooner (just before 10 days compared with just after 10 days)
  • had more hospitalisations or deaths (5.3% compared with 4.5%)
  • had more deaths (5 people [0.6%] compared with 2 people [0.2%]).

Azithromycin. An analysis of 1323 people included 500 (38%) who received azithromycin. Again, azithromycin had no meaningful benefit over usual care. Compared with usual care alone, the azithromycin group:

  • improved around 1 day sooner (7 compared with 8 days)
  • had around the same number of hospitalisations (3% in each group)
  • had the same number of deaths (no-one died in either group).

Colchicine. An analysis of 2755 people included 156 (6%) who received colchicine. Colchicine had no meaningful benefit over usual care. Compared with usual care alone, the colchicine group:

  • improved 1.4 days later (15 compared with 14 days)
  • had a similar number of hospital admissions (3% versus 4% )
  • had no deaths (there was one death in the usual care group).

Budesonide. An analysis of 1,586 people included 787 (50%) who received budesonide. People benefited from taking budesonide. Compared with usual care alone, the budesonide group:

  • improved almost 3 days sooner (12 compared with 15 days)
  • had fewer hospitalisations (9% compared with 12% people)
  • had a similar rate of deaths (1% in both groups).

In addition, people receiving budesonide had better wellbeing while recovering, felt more cheerful and active and were more likely to wake up feeling refreshed. They consulted their GP less often, and had a more sustained recovery (their symptoms were less likely to return).

Why is this important?

Doxycycline was recommended by the World Health Organization to treat people who were thought to have developed bacterial pneumonia alongside COVID-19. It was used for this purpose in the UK until March 2022, for people in the community at high risk of poor outcomes. Azithromycin was often used in the community but without evidence to back it. Emerging evidence suggests that it is rare for people to have bacterial infections alongside COVID-19 so, outside of hospitals, antibiotics are unlikely to benefit people with COVID-19.

Colchicine had been shown in a large trial to reduce hospital admissions from COVID-19. Inhaled budesonide had shown benefit in COVID-19 in a small UK trial.

This research shows that adding doxycycline, azithromycin or colchicine to usual care provides no meaningful benefit to people in the community with COVID-19. They should not be used. Antibiotics should always be used sparingly to prevent bacteria becoming resistant to them. Use of any drug which is not effective puts people at unnecessary risk of side effects, and wastes money.

Budesonide had not been explored for people at high risk of COVID-19 complications. This study showed that budesonide improved the time it takes for people to recover and their wellbeing. They had a more sustained recovery and made fewer demands on the healthcare system. These are important outcomes to the NHS.

Fewer than 1 in 5 published studies about COVID-19 are carried out in the community. Yet research can have its greatest impact outside of hospitals. In this study, the researchers couriered medicines to people, many of whom were too sick to collect their trial medications from a centre. This meant that people were able to join the trial regardless of where they live in the UK, making the findings highly generalisable.  

Before COVID-19, clinical research in epidemics of infectious diseases - such as swine flu - was only started once the outbreak had largely passed. It is therefore not known whether or not treatments used for swine flu are effective. PRINCIPLE’s novel platform trial design is an effective way of rapidly gathering reliable data that is improving care during the pandemic in the NHS and world-wide.

What’s next?

As a result of this study, the National Institute for Health and Care Excellence (NICE) updated guidance on doxycycline and azithromycin and these antibiotics are no longer recommended for people with COVID-19. NICE recommended further research into inhaled budesonide for COVID-19.

When budesonide was tested (March/April 2021), hospital admissions for COVID-19 were at low levels because of lockdown measures and the vaccination programme. The researchers say this could explain why budesonide did not reduce hospital admissions. It was not possible to show whether COVID-19 vaccination directly influenced the budesonide results. More research is needed to explore this in more detail.

The PRINCIPLE trial is ongoing; 7 drugs have been tested so far and trials of 2 more (the antiviral favipiravir and the antiparasite ivermectin) are still underway.

You may be interested to read

This NIHR Alert was based on 4 papers:

The antibiotic doxycycline: Butler C, and others. Doxycycline for community treatment of suspected COVID-19 in people at high risk of adverse outcomes in the UK (PRINCIPLE): a randomised, controlled, open-label, adaptive platform trial. Lancet Respiratory Medicine 2021;9:1010–20.

The antibiotic azithromycin: PRINCIPLE Trial Collaborative Group. Azithromycin for community treatment of suspected COVID-19 in people at increased risk of an adverse clinical course in the UK (PRINCIPLE): a randomised, controlled, open-label, adaptive platform trial. Lancet 2021;397:1063–1074.

The anti-inflammatory colchicine: Dorward J, and others. Colchicine for COVID-19 in the community (PRINCIPLE): a randomised, controlled, adaptive platform trial. British Journal of General Practice 2022; BJGP.2022.0083.

The inhaled steroid budesonide: Yu L-M, and others. Inhaled budesonide for COVID-19 in people at high risk of complications in the community in the UK (PRINCIPLE): a randomised, controlled, open-label, adaptive platform trial. Lancet 2021;398:843–855.


Funding: PRINCIPLE is funded by a grant to the University of Oxford from UK Research and Innovation and the Department of Health and Social Care through the NIHR as part of the UK Government's rapid research response fund.

Conflicts of Interest: Several authors have received grants and personal fees from various pharmaceutical companies. One was a former employee of GlaxoSmithKline.

Disclaimer: NIHR Alerts are not a substitute for professional medical advice. They provide information about research which is funded or supported by the NIHR. Please note that views expressed in NIHR Alerts are those of the author(s) and reviewer(s) and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care.


  • Share via:
  • Print article

Commentaries

Study authors

This platform trial enabled us to rapidly assess several drugs at the same time. We gathered rigorous large-scale, real-world evidence sooner than anyone else that doxycycline and azithromycin are not effective community treatments for COVID-19. Showing that doxycycline, in wide use to treat COVID-19 around the world, is not effective is a huge contribution. Preventing ineffective antibiotic use is critical since there is a parallel pandemic – that of drug resistance.

Showing that colchicine does not help people recover any quicker reduces unnecessary risk of side effects from this medicine that is ineffective against COVID-19. Inhaled budesonide can speed up recovery of people with COVID-19, and there is a high probability that it also reduces hospital admissions, but more research is needed on this.

The trial did not test against dummy drug since we wanted to see if adding drugs of interest to usual care would help people more than if they were not added. We also did not specify what ‘usual care’ was since that was likely to change during the course of the trial.

Christopher Butler and Richard Hobbs, Co-Chief Investigators, PRINCIPLE Study, Nuffield Department of Primary Care Health Services, University of Oxford

Lived experience

I contracted COVID-19 with bacterial pneumonia in April 2020 and was intubated in intensive care for 18 days. Any medical development which could help reduce the impact of this disease is critical; I am still debilitated 2 years after initial infection. This paper gives medical practitioners, especially GPs and doctors in intensive care, further information.

Nigel Simm, Public Contributor, Worcester

Clinician

These studies are informing guidance for the treatment of COVID-19. They should also be a source of information to the public, explaining why these drugs might not be suitable for them. It is helpful for clinicians to have evidence if they are challenged for omitting this treatment.

I hope the research will reduce inappropriate prescribing of antibiotics and support moves to tackle antibiotic resistance.

Hazel McPhillips, Advanced Nurse Practitioner, Lothians 

Pharmacist

Patients should not be offered antibiotics routinely. Alternative treatments are now available to people in the community. There needs to be more awareness of how to prescribe.

The trial has some limitations: recovery was self-reported and treatments were not blinded (everyone knew what they are taking). Also, new variants could make the findings difficult to interpret.

Ho Jan Senya, Clinical Trial Pharmacist, Sandwell and West Birmingham NHS Trust 

Back to top