NPC Archive Item: Routine use of antibiotics to prevent serious complications of URTIs is not justified

NOTE – This is an archive post from the NPC and has not been updated since first publication. Therefore, some hyperlinks may no longer be working.
MeReC Rapid Review

NPC Logo

Petersen I et al.  Protective effect of antibiotics against serious complications of common respiratory tract infections: retrospective cohort study with the UK General Practice Research Database.  BMJ published online 18th October 2007 doi:10.1136/bmj.39345.405243.BE

What is the background to this? – Most antibiotic prescribing is in primary care, and most of it is for respiratory tract infections.  Evidence-based clinical guidelines such as CKS (formerly Prodigy) advise against routine use of antibiotics in upper respiratory tract infections (URTIs) including coughs, colds, sore throat and otitis media.  Although rates of prescribing for acute respiratory tract infections in UK general practice fell by 45% between 1994 and 2000, this seems to be more because patients are less likely to present with URTIs than because GPs (and now nurse prescribers) are less likely prescribe when patients present.  More information about this and other aspects of antibiotic prescribing can be found on the common infections section of NPC.  The authors of this study looked at the risk of serious complications among people with respiratory tract infections given antibiotics and those not given antibiotics, in a large UK general practice sample.

What does this study claim? – Primary care prescribers should not base their prescribing for sore throats, otitis media or other URTIs on a fear of serious complications:  more than 4000 people would have to be treated to prevent one case of quinsy, mastoiditis or pneumonia respectively.  However, in patients with chest infections, especially older people, antibiotics may be useful in preventing pneumonia: the number needed to treat (NNT) to prevent one case of pneumonia in people 65 years and older = 39 (95%Confidence interval [CI] 36 to 42).  NNT (95%CI) in people aged 16-64 years = 119 (105 to 136)

How does this relate to other studies? – Clinical guidelines (such as CKS (formerly Prodigy) advise against routine use of antibiotics in upper respiratory tract infections (URTIs) including sore throat and otitis media.  However, the advice for lower respiratory tract infections (chest infections, LRTIs) is more complex.  Antibiotics are essential in pneumonia.  CKS (formerly Prodigy) advises that in other types of chest infection in adults, antibiotics are not indicated in people who are otherwise well.  However, health professionals should carefully assess the symptoms and signs of people presenting with acute cough in people with characteristics and / or pre-existing conditions that increases the probability of pneumonia.  More information and advice on determining who is likely to benefit from antibiotics is given in the CKS (Prodigy) guidance and on the common infections – respiratory tract section of NPC.

So what? – This study reassures us that, although antibiotics have a statistically significant effect in reducing the rate of complications of URTIs, these are now so rare that many thousands of patients would need to be treated – and exposed to side effects – to prevent one case.  Widespread prescribing of antibiotics is a major contribution to the rate of development of  antibiotic resistance.  For people presenting with LRTIs it is important to assess their baseline risk of pneumonia, assess the signs and symptoms and their severity and then consider, on the basis of all this, whether an immediate prescription, a  delayed prescription, or no prescription (with watchful waiting) is most appropriate.

Action – Primary care prescribers should not base their prescribing for sore throats, otitis media or other URTIs on a fear of serious complications.  Prescribers should think carefully about the best strategies for individual patients who have LRTIs.

Study details

  • Design – retrospective cohort study using the records of the 162 UK general practices contributing to the UK GPRD database, 1st July 1991 to 30th June 2001.
  • Patients –  patients presenting with common respiratory tract infections: upper respiratory tract infection (1,081,000 episodes), sore throat (1,065,088 episodes), otitis media (459,876 episodes) and chest infection (749,389 episodes).
  • Outcomes – risk of developing serious complications (pneumonia, quinsy, mastoiditis and pneumonia, respectively) in the month after consultation and the odds ratio for the development of complications (antibiotics prescribed on the day of consultation  vs. no antibiotics
  • Results :

Infection/ adverse outcome

Adjusted odds ratio (95%CI)

NNT (95%CI)

P value

Otitis media/ mastoiditis (all ages)

0.56 (0.37 to 0.86)

4,064 (2,393 to 13,456)


Sore throat/ quinsy (all ages)

0.84 (0.73 to 0.97)

4,300 (2,522 to 14,586)


Other URTI/ pneumonia (all ages)

0.68 (0.58 to 0.79)

4,407 (2,905 to 9,126)


Chest infection/ pneumonia

0-4 years

0.22 (0.17 to 0.27)

101 (85 to 125)


5 -15 years

0.18 (0.13 to 0.24)

96 (73 to 137)


16 -64 years

0.27 (0.23 to 0.32)

119 (105 to 136)


65 years and older

0.35 (0.33 to 0.38)

39 (36 to 42)


  • Sponsorship – UK Department of Health