Reducing antibiotic prescribing

What was the aim of the project?

Antibiotic stewardship is vital to reduce antimicrobial resistance but, in clinical practice it can be difficult to achieve a balance between using antimicrobials when they are really needed and reducing use when perhaps they are not.

C reactive protein (CRP) testing is a blood test used to determine if a person has acute infection or inflammation. CRP testing carried out as part of the GP consultation (near-patient or point-of-care) provides a useful marker, along with clinical judgement, to indicate necessity of treatment with antibiotics.

Within Swindon Clinical Commissioning Group (CCG), while there is a trend of year-on-year reducing antibiotic prescribing by GPs, prescribing has increased within the out-of-hours GP setting. Dr Caroline Ward, GP Clinical Evidence Fellow with Swindon CCG set up a pilot study of near patient CRP testing within the Urgent Care Centre in Swindon. This is a central unit where patients are seen when their own GP practice is closed.

The aim of the pilot was to:

  • Reduce antibiotic prescribing in the local urgent care and out-of-hours setting to match the trend ‘in-hours’
  • Enable the CCG to meet its antibiotic prescribing targets and receive the funding that meeting the target will attract
  • Fulfil responsibilities around antibiotics stewardship.

Who was involved?

Swindon CCG worked with Seqol (superceded by Great Western Hospitals) who provided the urgent care, Alere who provided near patient testing equipment, and urgent care centre clinicians.

What was the aim of the evaluation?

The aim of the evaluation was to evaluate the effectiveness of CRP near-patient testing in reducing antibiotic prescriptions.

What resources and people were involved?

Dr Caroline Ward, GP Clinical Evidence Fellow with Swindon CCG, was assisted by Dr Chris Turner, Associate Medical Director at the Urgent Care Centre and Paul Clarke, Pharmaceutical Advisor, Swindon CCG.

The CRP testing machine was provided by Alere for free for a three to six month trial period. Local out-of-hours and Rapid Assessment Unit clinicians in the Urgent Care Centre were also involved.

What did they do?

Presentations of respiratory tract infections are common in primary care and it can sometimes be difficult to distinguish patients with viral respiratory tract infections from those with bacterial infections.

The team developed a proposal and secured funding from the CCG for a six-month pilot study of near-patient C-reactive protein testing (CRP) with the Urgent Care Unit in Swindon, which started in August 2016. The test analyser machine was supplied by Alere for free for the duration of the pilot, and the CCG agreed to allocate £3,600 based on a projected approximate use of 10 disposable testing cartridges a day for three months.

CRP is a marker of infection. A raised CRP indicates that the patient is more likely to have a bacterial infection and require antibiotics, while a lower result indicates a likely viral infection, which does not require antibiotics. The test involves a simple finger prick blood test in the clinic (point-of-care testing) with a result available in four minutes.

Clinicians used this test in line with recommendations from NICE guidance: Pneumonia in adults: diagnosis and management (December 2014). Staff were provided with a CRP testing guidance sheet which was based on NICE guidance for use of testing in cases of diagnostic uncertainty in lower respiratory tract infection (a flowchart), and audit form to complete each time so that accurate data could be collected.

What did they find?

They greatly overestimated the use of the machine and found only 208 tests were done in a six month period, costing approximately £800 – less than a third of the predicted cost.

Results show that for patients who have had a CRP test, immediate antibiotic prescribing reduced by more than 50 percent.

Of 55 patients who would have been given immediate antibiotics, only 21 still received an immediate prescription after the test, and 18 received a delayed prescription instead and the rest received no prescription.

Who was the evidence shared with and why?

Data is being shared with Swindon CCG.

What next?

Data analysis is still in progress. However, early results indicate that CRP testing is likely to have changed prescribing behaviour, and patients who underwent CRP testing were less likely to receive an unnecessary antibiotic.

There appears to be a shift towards delayed rather than immediate prescribing. Previous studies show that only around 30% of patients prescribed a delayed antibiotic will use them.

Now the pilot has finished, an evaluation report will be written and submitted to Swindon CCG. This will inform a decision whether to fund continuing point-of-care (near-patient) testing.

What has changed as a result?

An early cultural shift has seen Urgent Care clinicians increasing their use of delayed antibiotic prescribing. Immediate prescribing has reduced.

Successes and challenges


The staff implementing on the ground have done a great job in stressing the importance of completing the audit forms and following the guidelines – compliance with this type of change in practice is notoriously difficult to achieve.


The pilot took time to get set up. Also, it has not been possible to collate data from all CRP cartridges used as there have been machine errors at times, and sometimes clinicians have not completed the audit correctly, making a minority of the data uninterpretable. In addition, what we couldn’t track was whether performing a CRP test, and confirming that antibiotics are not indicated, reduces the rate of patient presenting to another health care professional during the same period of illness. Evidence from other trials shows that it does. This was out of scope of this project.


Near patient CRP testing is recommended by NICE CG191: “Pneumonia in Adults: Diagnosis and Management” . The excerpt from the NICE guidance to which this pilot study relates to reads as follows:

“For people presenting with symptoms of lower respiratory tract infection in primary care, consider a point-of-care [near-patient] C-reactive protein [CRP] test if after clinical assessment a diagnosis of pneumonia has not been made and it is not clear whether antibiotics should be prescribed.”