Wessex Intensive Care Society (WICS)

Microbiology Regional ICM Teaching


1. Microbiology Quiz (and answers!) - download here

2. Clinical Implications of antibiotic pharmacokinetic principles in the critically ill (NEJM 2013) - critique by Dr Emma Fitzgerald - download here

3. Summary of lectures

Author: Dr Ben Harris
Follow @Harben81

Fungal infections in the ICU – Dr Chesterfield

  • Very few places have a specific fungal infection (ITU) policy – national survey, Chalmers et al, 2011.
  • A few units screen for candida on admission
  • 10% colonised on admission to hospital
  • 80% are colonised by the end of their ITU stay
  • Of that 80% in ITU, 5-30% develop invasive candida – blood/peritoneum etc.
  • Colonisation is a risk factor for invasive disease so in PHT they try and identify & type & find out sensitivities for all candida positive cultures so if it becomes a problem later on they are ahead of the game.
  • Diagnosing invasive candida is not easy
  • Blood cultures only have a 50% chance of being positive in a known case of candidaemia.
  • Slow growing 2-3 days
  • There are some other PCR and beta D glutan tests but they have to go off to a reference lab and therefore this limits their clinical use

Clinical predictors of invasive candida infection are
Increasing numbers of positive cultures over time in an increasing number of sites (e.g. NBL, then wound swab, then urine)
Candida present in more than 2 sites
Colonisation index: Number of distant sites colonised:total number of body sites
cultured. If >0.5 suggests high risk of invasive candida

Other risk factors for invasive candida:
  • Surgery
  • Multifocal colonisation o TPN
  • Severe sepsis
  • ITU stay >7 days

Empirical antifungal therapy
  • Very variable the approach to this. Depends a lot on the individual microbiologist.
  • Virtually all agree candida on an NBL or BAL or ET aspirate is not significant. Candida pneumonia is very rare.
  • Threshold for empirical antifungals falls with multiple sites of colonisation, still spiking temperatures, or if they have liver cirrhosis (higher risk).
  • Who benefits most from empirical antifungal treatment?
  • Those who have risk factors (liver cirrhosis, GI perforation), have multiple colonisation sites, who are immunosuppressed, who continue to spike fevers despite antibiotics etc.

What is the empirical antifungal of choice in the ICU?
  1. Caspofungin
  2. Fluconazole
Used to go straight for fluconazole in PHT but they have recently started to see candida glabrata that can be resistant to fluconazole. The approach now is to start
caspofungin and then change to fluconazole when sensitivities come back. Remember that lots of people are on prophylactic fluconazole in the community so no point giving them the same agent as treatment.
Up to 40% mortality with invasive candida, worse if a non-albicans species.

Infectious disease society of America is the ‘gold standard’ guidelines for invasive candida.
Use caspofungin for invasive candidia and switch to fluconazole if sensitivities allow.
Need a two week course (clock starts from time of last negative culture). Things to look out for in invasive candidia infections
Need to do retinal examinations to rule out endophthalmitis (15% risk of this, treated with surgery or intrathecal amphotericin)
Echo to exclude seeding to the heart.
Lines are often the focus of candida, remove for 48 hours after a positive culture if possible.
Ideally try to be ‘plastic free’ or ‘plastic minimal’ for a period of time after diagnosis of invasive candida.
Remember many antifungals can be given orally so might not need a line (less likely in ITU)
Common ITU scenarios
Candida on NBL/BAL/ET aspirate has a VERY poor predictive value for candida pneumonia.
This is different if you have candida in NBL, urine and elsewhere. This represents an increasing risk of invasive disease as described above.

Candida in the urine
Don’t treat unless high risk of dissemination such as neutropenia, sepsis, undergoing
urological intervention.
If treating you should get urological imaging to exclude fungal balls in ureters. o Antifungals in those groups not at risk of disseminated infection make little
difference, 75% clear candida on their own with no treatment. Don’t forget to remove the catheter!

Single positive culture + cirrhosis + not getting better
  • Start treatment with antifungals.
  • Stop if subsequent cultures negative? No as only 50% ever are positive.
  • Better to continue for 14 days and then review.
  • Remember swabs are hard to interpret due to colonisation so if feasible try to send tissue samples

Resistance patterns: Dr Wyllie
  • Main ITU infections are pneumonia, urine and surgical wound infections
  • Catheter related bloodstream infections are getting really rare!
  • Ecoli and pseudomonas most common in ITU
  • It is increasingly common for community acquired pathogens to be resistant to common antibiotics, a relatively new phenomenon in the last few years.
  • ‘Replacement pathogens’ common in the ITU such as candida.
  • It is more critical than ever to get specimens because of all the multi-resistant bugs in the
  • 100% of NBL’s in 2012 from ITU grew something.
  • Co-amoxiclav resistance rates in coliforms, most common NBL isolate, now 30-40%
  • The big worry is the EXPLOSION of Tazocin use for HAP.
  • Use of Tazocin has gone up 200% in the last 4 years in PHT
  • Use of meropenem has gone up 100% in the last 4 years at PHT
  • Ciprofloxacin resistance is very low but very good to give you CDIFF and MRSA!
  • Tazocin usually has good pseudomonal cover
  • Rotation of antibiotics over time throughout the region is a good idea to reduce resistance
    but there are not enough antibiotics out there to do this!
  • Gentamicin resistance is low in PHT so far but a big problem in some London hospitals.

Most wound swabs will grow something but is it important?
A new beta lactaminase has been found that can inactivate Tazocin.

MRSA is now very rare in PHT, not in USA though.
  • Meropenem resistant pseudomonas is increasing in prevalence, especially in southern Europe.
  • In Europe, southern Europe is really bad for multi-drug resistant bugs. Italy seems to be especially bad.
  • Need to remember that any exposure to healthcare systems in Southern Europe is a really big risk factor for the patient having a multi-drug resistant bug. Remember it’s not just holidays but also medical tourism – for cosmetic and other surgery.

The problems are not just due to antibiotic overuse. It is the whole package
  • Lines
  • Hand washing
  • Isolation
  • Use of antibiotics in animals
  • Seeing as MRSA has plummeted after a focus on hand washing etc. practice before must have been terrible!
  • With these nasty bugs the aim is to Identify & Isolate ASAP
  • Faecal transplantation: Five done in Portsmouth due to CDiff diarrhoea. Close family members screened or there is a pool of screened lab staff who can donate. NJ tube. Well tolerated. 80% success rate for difficult to eradicate CDIFF in PHT. Some patients are so fed up of diarrhoea they have asked for faecal transplantation.

Q&A session: Dr Wyllie & Dr Adeniji
What can ITU do better? Nothing really, no inappropriate antibiotic prescriptions in PHT ITU.
Next pandemic of flu, are we ready? Yes, after SARS the communication and surveillance mechanisms are much better now. There is the technology and communication now for a quick response and quick vaccine production so yes, we probably are ready.
CSF specimens: Ideally get CSF in meningitis BUT not always straight away if worried about ICP. PCR can be done 48 hours after antibiotics. CT scan ok but ?raised ICP? This is a difficult one because without fundoscopy (which unless you are doing it regularly) is not that helpful, and papilloedema is a late sign) then you only really have the CT and clinical examination to go on. Beware the time lag between the CT report and the LP – raised ICP might have developed. You have to balance the risk of finding out what the bug is vs doing the LP and coning. Balance of risk. Now that we have molecular tests you don’t have to rush in to get a CSF sample necessarily.
Do aprons make a difference? As part of the whole package (hand washing etc.) yes they do.
Line sepsis: PICC lines have less incidence of infection for several reasons. They are further from the mouth and all its associated bugs, they are often put into ‘less sick’ patients. They are often less exposed to bacteria as they commonly are used in patients at home rather than in hospital. Always wait for the chlorhex in alcohol to dry when putting in any line.
Suspected line sepsis – issues.
No evidence to remove lines at 7 days BUT there are severe financial penalties attached for infections.
Arterial lines – have a much higher threshold to remove as they are difficult to put in, are painful to put in and are associated with blood coming out rather than stuff going in i.e.cvc’s.
Although MRSA bacteraemias are rare, MSSA, bacteraemias are still occurring. This is a really significant infection with significant morbidity attached.
  • Remove the line if worried and try to get paired line and peripheral cultures.
  • Selective gut de-contamination in the ITU
  • We already do this a bit, chlorhexidine mouthwash!
  • The trend of all the evidence is that it probably helps but the worries about resistance will only really be answered with trials going on for 10 years or when its use becomes
  • The antibiotic to use will depend on the local flora and patterns of resistance.
  • Abdominal sepsis
  • Remember gentamicin and metronidazole is for SURGICAL PROPHYLAXIS only
  • If they have abdominal sepsis you should use;
  • Amoxicillin + gentamicin (single dose) + metronidazole o Or tazocin + gentamicin
  • If using Tazocin you don’t need metronidazole.
Sepsis doses of Gentamicin 3 or 5mg/kg
Should be 5mg/kg, maybe even 7mg/kg as that’s what the original nomogram was based on.