The main goal of ARK is that when doctors review antibiotic prescriptions, they will feel confident it is safe to stop them more often than they do now.

The examples that follow on the next pages illustrate the importance and feasibility of stopping antibiotics more often when they are not needed.

The risk of antibiotics causing harm to patients can feel less serious than the risk of stopping antibiotics, even when that risk is very small

At our trust, a 72 year old woman recently presented with confusion. Ceftriaxone and aciclovir were started to cover the possibility of meningo-encephalitis.

Imaging showed a subdural haematoma and although bloods showed no inflammatory response and the patient had no fever, the medical team decided to stay with antibiotics until a lumbar puncture was done. The team talked about the safety of doing a lumbar puncture and then two attempts were made but failed. Five days later, it was decided to finish two weeks of antibiotic treatment because a lumbar puncture could not be done.

After 10 days of treatment, the patient developed a severe C. difficile infection needing fluid resuscitation, insertion of a rectal tube and lengthy hospital treatment. A root-cause analysis review found that there was no need for antibiotic treatment after a non-infective diagnosis was made. C. difficile risk is higher when antibiotic treatment goes on longer than 3 days (Wiström 2001) so harm could have been avoided if the antibiotics been stopped earlier.

The risk of antibiotics causing harm to patients can feel less serious than the risk of stopping antibiotics, even when that risk is very small (2)

Another case was a 91 year old woman who presented to acute medicine with confusion and fever. UTI was suspected and she was started on a 5 day course of ciprofloxacin, as her GP had recently given a course of trimethoprim.

On day 2 she had developed a widespread rash. The rash slowly got worse over the next 72 hours and became toxic epidermal necrolysis (TEN).

Antibiotics were stopped and she was transferred to a local burns unit, but her condition worsened, made worse still by poor left ventricular function, and she died a few days later.

Click the reference to see an article about over-treatment of Nursing Home residents (Mody & Crnich, 2015).

One issue with using case stories about over-using antibiotics is that these are single cases. Clinicians may think “Well, I wouldn’t have done that.”

In our hospital we wanted to find out how often patients are started on antibiotics which then continue even though there is little evidence they are still needed.

We looked at new patients who were started on antibiotics for respiratory symptoms when they were admitted. That’s about six patients every day at our hospital.

With these patients, after 48 hours, 1 in 5 have no radiological or clinical evidence of infection, a low CRP and a low NEWS score. Only 1 in 5 of these (so, 1 in 25 patients started on antibiotics) had their treatment stopped as test results came in. The rest went on to complete a ‘full course’ even though, often, they didn’t need to.

One issue with using case stories about over-using antibiotics is that these are single cases. Clinicians may think “Well, I wouldn’t have done that.” (2)

The new antibiotic CQUIN means trusts have to check how often antibiotic prescriptions are reviewed 48-72 hours after they are written.

In our trust, the audits found that more than 85% of prescriptions were reviewed — but antibiotic treatment carried on unchanged in 80% of the patients. Even when changes were made, most patients were just switched from IV to oral treatment and antibiotics were only stopped at 3% of reviews.

The results:
Standard 4.1 – Outcomes of 48 – 72 hour review of antibiotic prescription

People might not think that changing antibiotic prescribing
a) can be done in hospitals and
b) will actually benefit patients

At our hospital, rates of C. difficile increased for several years up to 2007, even though there were efforts to improve infection control.

In 2007, we diagnosed an average of 30 cases of CDI every month. We changed our antibiotic policy to replace cephalosporins and quinolones with penicillin antibiotics.

Within a year, our use of cephalosporins and quinolones had been cut by more than half. Rates of C. difficile started to fall and now we diagnose only 3-4 cases each month.

What happened with the Clostridium difficile epidemic shows that changes can be made and really do change outcomes.