Viva

Question 1

Why have you stated that depressed skull fractures, craniotomy, IVH and neurosurgical intervention are risk factors and not analysed them later?

Answer 1

Because these factors were not well represented in the data and are well established already

Question 2

What is the pfausler cell index?

Why did you not use it?

Answer 2

Pfausler et al. 2004 devised a measure to aid in csf-infection diagnosis. It is ratio of leukocytes:erythrocytes in csf divided by same measure in peripheral blood. Rise in this is characteristic of infection.

It is not used because calculating the cell index requires routine sampling and this is not advocated in the QMC as it is a risk factor.

Beer and lackner both utilised this though

Question 3

What do Mayhall et al., Holloway et al. And Lo et al. 07 all say about elective revision?

Answer 3

Mayhall et al. After 5 days review as inf rate soars
Holloway et al. Rises up to ten days but then drops off
Lo et al. Argued against elective revision saying infection risk increases with each new EVD, retrograde col

Question 4

What is the issue with antibiotic prophylaxis?

Answer 4

Resistant pathogens.
Poon et al. 1998 compared one course peri vs high dose dual therapy prophylaxis and in spite of lower inf rates selection for mrsa and candida spp. Occurred

Question 5

What did Wyler and Kelly find?

Limitations?

Answer 5

If EVD was to be in situ longer than 3 days then prophylactic abx should be used.

Limited study though, only 11 infections and average duration of EVD in situ was very short (2.5 and 5 days)

Question 6

What was the issue with studies by rebuck et al. And alleyne et al. (2000)?

Answer 6

Antibiotic prophylaxis,
Rebuck- no significant effect of prophylaxis on rate
Alleyne - no sig effect
Both weak because of poor sample size.
Rebuck et al. Chose narrow spec cephalosporins as management and this negatively impacted result.
Poon et al chose broad spec

Question 7

Why is inadequate antibiotic use a problem?

Answer 7

Can result in resistant/ different microbiological profiles. Ibrahim et al. Bloodstream infections 2000 - translates to csf infections, adequate and early abx is important to prevent neurological impairment

Question 8

What is the view on systemic infection as a risk factor for csf infection?

Answer 8

Kim et al. 2012 no sig result
Clark et al. 1989 found relationship so did Holloway et al. 1996
Schultz et al found no relationship

Kim et al. Found similarities in organism between systemic and EVD culture isolates. This conflicts Clark et al. Where they stated it is different. Requires further research as it may be important in terms of managing it

Question 9

Why would flushing a catheter be a risk factor? Any evidence?

Answer 9

Breaching a closed system and providing potential site of entry for pathogens.
Aucoin et al. 1986 flushed EVDs with antibiotic solution but actually found 6% increase of rr ventriculitis in those that received the flush

Question 10

Which papers commented on leaks as a risk factor?

Answer 10

Mayhall et all. 1984 and Schultz et al. 1993 found no link between leaks and EVD-infection.

Lyke et al. Found that leaks were a risk factor and irrigation was not! However small sample size of flushed patients is reason and they think leaks provide longer opportunity for pathogens

Question 11

What is the bactiseal EVD made of?

Answer 11

Silicone matrix flexible catheter, imoregnated with 0.15% clindamycin and 0.054% rifampicin.

Question 12

How do you know Bactiseal EVD works?

Answer 12

In vitro studies by prof Bayston have shown protection against bacterial colonisation from staphylococci.
In vivo studies by zabramski 03 lackner 08 and Harrop 10 show efficacy of AICs fighting infection

Pople et al. 2012 dismissed because of poor recruitment

Question 13

Why is the Wong 2010 study important if it shows no benefit from using an AIC compared to prophylaxis?

Answer 13

It shows non-inferiority.
3 x cases of C.Diff colitis in prophylaxis group shows that the risk of resistant pathogens is not evident in AICs as it is with systemic long term prophylaxis.

Also investigated if impregnation made the catheter stiffer which it did not, which did happen in catheter mentioned by Stevens et al. 09

Question 14

What is the problem with the study by Gutierrez-Gonzalez et al. 2010?

Answer 14

Assessed shunts and EVDs together in order to adjust for low numbers of EVD infections. In terms of EVDs infection rates were very high in both control and bactiseal. Bactiseal EVDs showed a trend towards lower inf rates

Question 15

What were the outcomes from an in vitro analysis of the minocycline and rifampin AIC?

Answer 15

Stevens et al. 2009, showed antibiotic effects on colonisation as expected but unexpectedly AICs were more likely to give false negatives.

Question 16

Stevens et al. 2009s findings were overturned in 2015, why?

Answer 16

Bayston 2015 evaluated effect of AICs in bacterial viability in vitro. Ran study over a period of 21 days as opposed to just one day like in Stevens. Evaluated 3 AICs, and showed that after day 1 bacterial viability was no longer an issue and argued that samples are not taken in the first day post insertion anyway

Question 17

Tell me about the findings of Thomas et al. 2012

Answer 17

Meta-analysis to review antibiotic impregnated csf devices. Found significant results in favour of AICs but concluded that this is largely on basis of observational studies and RCTs are needed

Question 18

Some of these papers you have quoted on silver catheters say they are significant in reducing infection, is that so?

Answer 18

Lackner et al. 2008 pilot study - Didn’t have one single infection in the silver arm of study - selection bias and small sample size.

Lacjak et al. 2013 only obtained significant result when data were manipulated - added different brand plain catheter with much higher infection rate in order to make silver one significantly better.

Keong et al. Double blind prospective RCT recorded a very high infection rate 21.4% and plain and silver only became significantly different after 10 days in situ.

All of these have doubts in integrity and accuracy

Question 19

What did Hagel et al. 2014 find?

Answer 19

Trend towards silver catheters increasing infection rates. However v small number of infections in both groups

Question 20

What studies compared the efficacy of silver and AICs?

Answer 20

Winkler et al. 2013 RCT found no sig difference but had a small sample size and unfamiliar diagnostic criteria

Wang et al 2013 meta analysis found no stat sig difference between silver and plain but did between AIC and plain.

Wang et al. Is a good study of high importance.

Question 21

What is the basics trial?

Answer 21

British antibiotic and silver impregnated catheters for ventriculoperitoneal shunts.
Multi-centre RCT
Ongoing trial recruits from QMC and others across the country. Will hopefully provide solid evidence to support reduced infection rates in both cases and also inspire a similar trial in EVDs as is needed.

Question 22

Why was your sample chosen and not done at random?

Answer 22

Of 874 EVDs placed only 111 from 59 pts were included. They were chosen because they were previously identified as having suspected EVD-infections but there were doubts on the legitimacy of this. All patients not selected were definitely not EVD-infections and hence the rate could still be calculated.

Question 23

Why was your criteria used?

Answer 23

It was in accordance with the literature and professional opinion (supervisor).

Question 24

Why were the recognised pathogens chosen?

Answer 24

S aureus E. coli klebsiella enterococcus pseudomonas and candida are common csf pathogens

Question 25

Why did you use the overall rate and not the definite one?

Answer 25

It is unclear as to whether or not a definite rate was calculated in the previous audit. Therefore using the overall is more appropriate.

Question 26

What does the breakdown of organisms mean?

Answer 26

In accordance with microbiological profiles of other studies CONS and gram negative bacilli are predominant. No evident selection is apparent which means no selective pressure is exerted by the AIC however a breakdown from the previous internal audit would be more useful.

Question 27

What are the tests used for goodness of fit?

Answer 27

Hosted-lemeshow test and omnibus tests of model coefficients also gave chi square value

Question 28

Why is the probable criterion so complex?

Answer 28

Clinical signs added as these are used in other studies (smith and Alksne, Clark et al and sundbarg et al) as well as csf pleocytosis/ glucose along with cultures from tips, one culture only and a gram stain result. 2 of the 3 categories eliminates possibility of contaminants (reduced false positive). Note clinical signs should not be depended upon when population is mostly unconscious

Question 29

You’ve used EVD tip cultures but said they’re not useful?

Answer 29

Wong et al. 2010 stated that 7 EVD tips gave positive results without evidence of ventriculitis. Hence tip results may produce false positives but combining with other elements should help prevent this

Question 30

What vital pieces of info are you missing from the previous internal audit?

Answer 30

Breakdown of organisms
Criteria used
Proportion definite
Antibiotic protocols

Question 31

Wang et al. 2013 produced what result that adds substance to mine?

Answer 31

Systematic review on AICs that yielded a 3.6% infection rate (close to 3.09%). Suggests they are effective

Question 32

What trend was seen with each new catheter per patient?

Answer 32

Protection against infection. Contrary to other studies (Lo et al 07 and Clark et al 89) infection gets less likely - does bactiseal have a protective effect? Larger sample sizes of patients with several EVDs required to test this. Only 12 of 59 patients analysed had 3+ EVDs.

Question 33

Which study does Wong et al 2002 directly contradict?

Answer 33

Mayhall et al 84 as it says there is no benefit to predetermined exchange after 5 days post insertion when compared with clinical presentation demanding it

Question 34

Why are peaks of infection seen so early in studies?

Why is our study so much later?

Answer 34

Most studies have short duration EVDs approx 5 or 6 days. (They aren’t like urinary catheters). E.g aucoin Mayhall and Clark

This lack of sample size could explain peaks at these times. A large sample of long duration EVDs (our average duration was 14.2 days) could explain the later peak in incidence as well as the protective effect of the bactiseal carheter.

Question 35

Why are so many of the patients on abx and what does this mean?

Answer 35

Due to nature of the population, 58 of 111 were receiving abx at insertion and 75 of 111 experienced a concomitant infection. A random sample would not have experienced this. Could contribute to low rate of infection recorded

Question 36

How did the irrigation and sampling protocols affect the results?

Answer 36

Flushes as a last resort only. Nearly 75% of EVDs not flushed. Wider sample required to assess this as a risk factor but could contribute to low rate

Sampling when suspicion of infection only. Hoefnagel et al 08 linked samples and breaches to EVD-infection. Lots of samples were often taken to check if infection had cleared in a new EVD negatively impacting analysis

These two risk factors should be analysed in a randomised sample but they both may contribute to low rate.

Question 37

What did Lyke et al. 2001 find?

Answer 37

9 of 11 infections were g neg bacilli they attributed this to use of broad spectrum gram pos prophylaxis for every single patient. This is problematic as gram neg infections are harder to treat and are associated with high rates of mortality and morbidity.

Question 38

Why is a rise in gram positive infections observed?

Answer 38

Data is warped by just one or 2 cases. Monthly reputation may be required. Perhaps promotes rise of resistant skin flora.

Question 39

How do you know bactiseal doesn’t promote resistance?

Answer 39

Gram negatives are still in minority, organisms are concordat with normal skin flora and literature and there is no obvious selection for organisms. Also the drugs are released at non toxic concentrations very slowly - not in mutant selection window

Question 40

Is the Bactiseal EVD effective against all organisms?

Answer 40

No it is proven to be effective against staphylococci. Wang et al. 2013 showed a non sig trend do wards decreasing infection rates regarding gram negative bacilli but this requires further research

Question 41

What are the strengths and weaknesses of the study?

Answer 41

Strengths: same researcher, use of predetermined trusted criteria, similarities with wang et al. 2013.

Weaknesses: retrospective, selection bias, mismatch of info with previous internal audit

Question 42

Why is this study incomplete?

Answer 42

Missing data and sample means risk factor analysis is incomplete. It also requires multi centre randomised controlled trials like the BASICS trial

Question 43

Why is the range of infection rates so wide?

Answer 43

0-22% Lozier et al. 2002

Varies depending on the patient population, antibiotic protocols and criteria for defining infection