Tenth of antibiotics prescribed by GPs over 22 years fail to cure
26 September 2014
A 22-year study by Cardiff University has found that more than
one in 10 of all antibiotic treatments in a primary care setting
failed. This rate has increased and continues to rise, according to
the study, which analysed almost 11 million antibiotic prescriptions
in the UK.
Much data has been gathered about antibiotic resistance in
hospitals, but virtually nothing is known about the frequency and
pattern of antibiotic failure in primary care.
Researchers from Cardiff University therefore set out to assess
antibiotic treatment failure rates in UK primary care with
particular focus on four of the most common kinds of infection:
upper respiratory tract infections (URTIs), lower respiratory tract
infections (LRTIs), skin and soft tissue infections (SSTIs), and
acute otitis media (AOM – middle-ear infection).
Findings revealed that between 1991 and 2012 overall antibiotic
treatment failures had risen from 13.9% (1991) to 15.4% (2012),
indicating an increase of 12%. Throughout this period, antibiotic
treatment for LRTIs such as bronchitis and pneumonia was the least
successfully treated of all infection classes, showing an increased
failure rate of 35%.
Treatment failure rates for commonly prescribed antibiotics such
as amoxicillin, penicillin and flucloxacillin remained below 20%
throughout the studied period, while antibiotics not normally
recommended for first line therapies showed concerning rates of
effectiveness. Most notably, Trimethoprim, an antibiotic normally
used in the treatment of URTIs and listed on the WHO’s register of
‘essential medicines’, showed a failure rate rise of 40%.
“There is a strong link between the rise in antibiotic treatment
failure and an increase in prescriptions,” said Professor Craig
Currie from Cardiff University’s School of Medicine. “Between 2000
and 2012, the proportion of infections being treated with
antibiotics rose from 60% to 65% which is the period in which we see
the biggest increase in antibiotic failure rates. These episodes of
failure were most striking when the antibiotic selected was not
considered first choice for the condition treated.
“Given the lack of new antibiotics being developed, the growing
ineffectiveness of antibiotics delivered through primary care is
very worrying indeed. There is a mistaken perception that antibiotic
resistance is only a danger hospitalised patients, but recent
antibiotic use in primary care is the single most important risk
factor for an infection with a resistant organism. Furthermore, what
happens in primary care impacts on hospital care and vice versa.
“Antibiotic resistance in primary care needs to be more closely
monitored, which is actually quite difficult given that primary care
clinicians seldom report treatment failures. The association between
antibiotic resistance and antibiotic treatment failure also needs to
be further explored. From the general level of feverish debate, it’s
not quite the "cliff" we would have imagined, but clearly this is
In Europe, hospital infections caused by antibiotic resistant
bacteria result in 25,000 deaths every year, increased morbidity and
EUR 1.5 billion in healthcare and societal costs. Microbial
resistance to antibiotics has increased at an alarmingly rate in
recent decades, to an extent that WHO has declared the issue a
‘public health crisis’.
Professor Currie added: “We need to ensure that patients receive
the appropriate medication for their condition and minimise any
unnecessary or inappropriate treatment which could be fuelling
microbial resistance to antibiotics, prolonging illness and in some
cases killing people.”
Criteria used in the study to define antibiotic treatment failure
included the prescription of a different drug within 30 days of
first-line antibiotic; record of patient hospitalisation for an
infection related diagnosis within first 30 days of the first-line
prescription; GP referral to infection-related specialist within 30
days of treatment initiation; and patient death with an infection
related diagnosis within 30 days of commencing antibiotic treatment.
Researchers used data drawn from the Clinical Practice Research
Datalink (CPRD) to assess the failure of first-line antibiotics in
the UK from 1991-2012. The data source stores the records of over
14M individuals obtained from 700 primary care practices in the UK.
The study emphasises why researchers need access to anonymous
healthcare data from the UK population.
The study, published in The BMJ, was funded by Abbott
Healthcare Products and is available here: