This site is intended for health professionals only

At the heart of general practice since 1960

pulse june2020 80x101px
Read the latest issue online

GPs go forth

Don't use steroids for chest infections, GPs advised

GPs should not use steroids for treating acute lower respiratory tract infection in adults who do not have asthma or other chronic lung disease, UK researchers have said.

The study, published in the Journal of the American Medical Association (JAMA), concluded that oral corticosteroids are not effective in the treatment of chest infections in non-asthmatic adult patients in primary care.

Researchers from the Universities of Bristol, Nottingham, Oxford and Southampton randomly tested 398 non-asthmatic adults presenting with acute chest infections but with no evidence of pneumonia and not requiring immediate antibiotic treatment.

One group of 198 participants received 40mg of the oral steroid ‘prednisolone’ for five days, and the other ‘control’ group of 200 participants received an identical placebo over the same time period.

The study, funded by the National Institute for Health Research School for Primary Care Research, found that there was no reduction in the duration of cough or the severity of the accompanying symptoms between two and four days after treatment in the prednisolone group compared with the placebo group.

Participants were selected from 54 practices in England and had a mean age of 47 years, while 63% were female and 17% smokers.

Alastair Hay, lead author of the study and a GP and professor of primary care at the University of Bristol Medical School, said: 'Chest infections are one of the most common problems in primary care and often treated inappropriately with antibiotics.

'Corticosteroids, like prednisolone, are increasingly being used to try to reduce the symptoms of chest infections, but without sufficient evidence.

'Our study does not support the continued use of steroids as they do not have a clinically useful effect on symptom duration or severity. We would not recommend their use for this group of patients.'

Professor Mike Moore, study co-author and a GP and professor of primary health care research at the University of Southampton, added: 'Oral and inhaled steroids are known to be highly effective in treating acute asthma as well as infective flares of other long-term lung conditions but need to be used carefully because of the risk of unwanted side effects.

'We chose to test the effect of steroids for chest infections as some of the symptoms of chest infections, such as shortness of breath, wheeze and cough with phlegm, overlap with acute asthma. However, we have conclusively demonstrated they are not effective in this group of patients.'

Dr Andrew Green, clinical and policy prescribing lead at the GPC, said: 'Oral steroids have significant and potentially serious side effects, and systemic infection is considered a contra-indication without simultaneous anti-infective therapy. I am therefore relieved that there has been no symptomatic benefit shown, as this will remove any potential pressure on GPs to prescribe contrary to established practice.

'It should be noted that although no adverse consequences were found in this study, the sample size is too small to exclude serious but uncommon events.'





Readers' comments (3)

  • Young patients, small sample size. Not much help to me really on Planet Earth. Get back to work.

    Unsuitable or offensive? Report this comment

  • How were these patients diagnosed? It appears from the abstract that diagnosis of those patients included, was made exclusively on symptoms rather than examination or investigation. The sample included patients with cough and at least one LRT symptom. According to the paper 77% had phlegm and 46% had chest pain. Mucus production is common in viral URTIs and MSK chest pain is also common in people with coughs of all causes. So if these were some of the defining criteria used for diagnosing LRTI, I doubt the validity of the diagnoses made in this cohort which may well have been a mixture of people with upper and lower respiratory, bacterial and viral infections. Furthermore, the cohort included patients with "no evidence of pneumonia." I can only assume this means that examination was used to exclude patients with objective chest signs, making it even more likely that those included did not actually have LRTI.

    Unsuitable or offensive? Report this comment

  • No signs of pneumonia? Assuming that CXRs were not done then this suggests no chest signs on examination. Which would further suggest none of these patients had LRTIs. Few assumptions there but does make one wonder.

    Unsuitable or offensive? Report this comment

Have your say