Antibiotic associated diarrhoea: My new primary care strategy

A woman in her mid 50s had been prescribed doxycycline 5 days ago for persistent upper respiratory symptoms.  She returned to us because of watery diarrhoea since yesterday. There was no fever, pain or vomiting. She had eaten nothing out of the ordinary and no contacts were affected.

She was clinically well hydrated and had no other medical problems.

Old me: OMG! C diff! Admit!

New me: This is a possibility, but only 10-20% of antibiotic associated diarrhoea is caused by C diff. Most of the time the problem is an alteration in the bowel flora. The third potential mechanism is a direct prokinetic effect of the antibiotic, typically erythromycin.

Old me: Still, it’s C. diff till proven otherwise. The antibiotic must be stopped and a stool sample for C diff toxin sent.

New me: The antibiotic should ideally be stopped. Whether or not stool samples need to be sent routinely is a matter of risk. The two questions I need to ask myself are:


a)   How likely is it that in this patient there is C. diff?

b)   If it were C. diff and I missed it, how bad would that be?


a) What makes it likely to be C. diff?

This can be split into factors in the history of the presenting complaint and patient factors.

HPC: Fever, blood in stools, crampy pain and a tender abdomen go more with C diff; non C.diff antibiotic induced diarrhoea tends to be more of a ‘nuisance diarrhoea’ (not too severe).

Patient factors (according to NICE CKS):

1. Advanced age — the rate of positive C. difficile assay results in people older than 65 years of age is 20–100 times greater than the rate in people 10–20 years of age.

2. Antibiotic treatment — although none can be excluded, the most commonly implicated antibiotics are:


Cephalosporins — in particular second- and third-generation cephalosporins (such as cefuroxime axetil, cefixime, ceftriaxone, and cefotaxime)

Fluoroquinolones (such as ciprofloxacin, norfloxacin)


Ampicillin and amoxicillin (which may relate mainly to the volume of their use rather than being high risk)

3. Underlying morbidity such as abdominal surgery, cancer, chronic renal disease, and tube feeding.

4. Current use of a proton pump inhibitor (such as omeprazole and lansoprazole) or other acid-suppressive drugs (such as H2-receptor antagonists).

Two further important risk factor are hospitalisation (C diff colonisation rates of about 20-30 percent, compared to 3 percent in the general population) and previous C. diff infection. In fact, the recurrence rate is around 20% for the first infection and 45-60% after the second episode in hospitalised patients.


b) If it were C. diff, how bad would it be to miss it?

C. diff infection can cause anything from a self resolving bout of diarrhoea to pseudomembranous colitis, toxic megacolon, perforation of the colon, sepsis and death. As with any diarrhoea, there is always a risk of dehydration.


Old me: OK, so what is the mechanism of C. diff infection then?

New me: It’s not as simple as ‘take a culprit antibiotic –> kill off the good bacteria –> let C. diff proliferate’. This mechanism does not explain why C diff occurs in outbreaks on wards, as antibiotic exposure is pretty constant from week to week.

It’s like Quasar.

When I was a kid, we used to go to Quasar. If you got shot, a voice came from your gun saying ‘Defence shield active, active. Warning, warning’. The idea was that for two seconds, the defence shield was active (you could not shoot, but could not be shot) and then you were in trouble for two more seconds (you could be shot, but could not shoot). This was terrible game design. The defence shield bit was OK, but the warning bit encouraged picking on one victim over and over.

Clearly I got over this and was definitely not one of the people who got picked on over and over on at his own Quasar birthday party by his own friends.

Antibiotics put you in the ‘warning, warning’ phase. The disruption to the bowel flora is the first hit. If you then meet C. diff whilst you are in the warning, warning phase, that is the second hit. You then go one of two ways. Most people get colonised asymptomatically. A minority get C. diff infection. It’s not clear what determines your fate after the second hit, but it is probably a mixture of virulence factors and host factors.



Old me: Oh, I see. Most of the time you can tell it’s unlikely to be a C. diff infection then.

New me: Yeah. NICE CKS say that if C. diff is suspected, then get a stool sample sent. If not, don’t.

The usual rules for deciding who to admit with gastroenteritis apply. If you suspect C. diff, assess severity.

If you are going to send a stool sample for C.diff toxins, bear in mind there is a 10-20% false negative rate. Because of this, you may have to repeat the C. diff toxin test 24 hours after a negative result if clinical suspicion remains. This is because quite a bit of toxin needs to be detected to give a positive test.

Old me: What about loperamide? The BNF says absolutely no way in antibiotic-associated colitis.

New me: This caused me a problem. My patient was using loperamide, and was going on a long distance journey the next day. She wanted to know whether loperamide was safe.

Loperamide is safe in non colitis / non C. diff antibiotic associated diarrhoea However, it leads to retention of the toxin as well as potentially increasing intraluminal pressures in C. diff infection.

My patient seemed very well and clinically had ‘nuisance diarrhoea’ rather than C. diff. Not taking loperamide would clearly affect her life the next day. I advised her it should be OK as long as she is pain free, but to restrict use to during the coach journey and to keep hydrated.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s