Download Risk factors for a Clostridium difficile colonization in humans.

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Common cold wikipedia , lookup

Gastroenteritis wikipedia , lookup

Urinary tract infection wikipedia , lookup

Hygiene hypothesis wikipedia , lookup

Neonatal infection wikipedia , lookup

Traveler's diarrhea wikipedia , lookup

Henipavirus wikipedia , lookup

Transmission (medicine) wikipedia , lookup

Hospital-acquired infection wikipedia , lookup

Cysticercosis wikipedia , lookup

Sociality and disease transmission wikipedia , lookup

Infection control wikipedia , lookup

Clostridium difficile infection wikipedia , lookup

Transcript
Risk factors for a Clostridium difficile
colonization in humans.
Date:
Student:
Supervisor:
Education:
Course:
13-5-2013
M. van Schaik (3185338)
E.C. Keessen (IRAS-VPH)
Veterinary medicine, University of Utrecht
research internship
Abstract
Clostridium difficile infection (CDI) can occur in a large number of animal species and in humans. An
increase in incidence and severity of CDI in humans is reported since the beginning of this century.
Symptomless colonization with the bacterium has been described in many animal species, including
production animals. This could lead to an increased risk of infection of humans through direct
contact with these animals or through indirect contact via the environment. Consumption of
contaminated food products could also increase the risk of infection. The finding of overlapping C.
difficile ribotypes in humans, animals and food products has led to the hypothesis that transmission
of C. difficile via animals to humans contributes to the increasing number of patients with CDI. This
study tries to determine the risk factors that lead to a colonization of C. difficile in humans working
with pigs or live at a pig farm. To analyze the data a multivariate logistic regression model is used.
The software program for the multivariate logistic regression model is SAS 9.2 (SAS institute Inc.,
Cary, North Carolina, USA). No significant correlation is found between the risk factors and
colonization of C. difficile in human using the multivariate logistic regression model with backward
elimination. An univariate model is used to select variables for the multivariate logistic regression
model. In the univariate model two significant risk factors are determined. One risk factor is a high
contact frequency between persons and pigs (p=0.03). The other risk factor is administration of
antibiotics to the pigs (p=0.06).
Introduction
C. difficile is a gram positive anaerobe bacterium. C. difficile produces endospores and toxins.
C. difficile infection (CDI) is a growing problem in hospitals in Europe and northern America. The
incidence of outbreaks of C. difficile ribotype 027 is rising In Dutch hospitals (3,4). The incidence of
infection with C. difficile ribotype 078 rises in hospitals in the Netherlands (15) and in Europe(16). On
the basis of geno- and phenotypical analyses it is concluded that strains C. difficile ribotype 078 from
piglets were indistinguishable from human strains in the Dutch population. A common origin of
human and animal strains of C. difficile should be considered (3). The results of multiple studies
suggest that animal reservoirs are possible sources, via food, direct contact, or via the environment
of human C. difficile infection (2,8). However there are still many uncertainties about the connection
of the specific C. difficile ribotypes in humans and animals.
CDI has emerged as a significant economic concern for swine producers, but little is known about
transmission patterns in hog operations (5, 6). In neonatal piglets, C. difficile is currently one of the
major causes of enteritis (11). In this study the risk factors at pig farms that lead to colonization with
C. difficille in humans are determined. Some case control studies about risk factors for CDI are done
but the risk factors in these studies concern the use of medicine in human patients and the intervals
of visiting healthcare institutions. The questions in the questionnaire used for this study are related
to animal contact, use of antibiotics, the management at the pig farm and the working procedures,
with a focus on the hygienic measures and contact frequencies between pigs and humans. The risk
factors that are selected in this study are derived from other studies with different pathogens, like
MRSA and based on epidemiological knowledge. Multiple studies suggest that antibiotic treatment is
a main risk factor for developing CDI in humans (9) and in horses (10), that is the reason why
antibiotic treatment is a risk factor that is investigated in this study.
The protocol of the study was approved by the Medical Ethical Committee of Utrecht University.
Methods and Material
The colonization with C. difficile was investigated by taking human stool samples of 138 participants.
The participants all work or live at a pig farm in the Netherlands. 40 farms joined the Bactopath
project. The human stool samples are collected by the person itself, the stool sample is put in a
plastic air tight container and in an airtight bag. The bag is send by normal mail to the laboratory. At
arrival the samples were frozen to -20°C. All humans who work or live on pig farms of which samples
are collected completed a personal questionnaire about possible risk factors. This includes the work
method on the pig farm, the method of cleaning, (personal) hygienic methods and procedures, and
the way the farm is managed. Other questions were related to whether the person administered
antibiotics to the pigs and concerned the contact frequency between the person and the pigs.
C. difficile detection is performed using standard protocols for samples from human and animal
origin as described by Hopman et al. (2011). One sample of each person is used. The samples are
defrosted once.
In short, 1 gram of faeces is mixed with 9 ml of the CDMN broth and incubated anaerobe at 37°C for
7 days. The broth is alcohol-shocked for spore selection by mixing 2 ml homogenized broth with 96%
ethanol in a sterile tube and leaving it for 1 hour at room temperature. After centrifuging this
mixture (4000 x g for 10 minutes) the supernatant is discarded, and the pellet is streaked onto a CLO
agar plate and incubated anaerobe for 48 hours at 37°C. Suspicious colonies based on gram stain
appearance, colony morphology (swarming, non-haemolytic, greyish, rough) and characteristic odour
are sent to the Leiden University Medical Center (LUMC) for PCR ribotyping. In some cases Molecular
characterization of selected type 078 and 045 isolates is performed to determine genetic relatedness
of the isolates.
Data analysis
A multivariate logistic regression model with backward elimination is performed to investigate the
risk factors. Criteria for variables to be included in multivariate analysis are that no more than 10% of
the data is missing in relation with the outcome. The next step is to perform an univariate statistical
test using the chi² method. For this test SPSS 20 (IBM Corporation Software Group, Somers, NY,
USA)software program is used. All results with a p-value < 0.3 are considered for the multivariate
model. In the multivariate model P-values <0.05 are considered statistically significant. To perform
the logistic regression model the software program SAS 9.2 (SAS institute Inc., Cary, North Carolina,
USA) is used.
Results
138 persons participated in this study. C. difficile was found in fecal samples of 13 farmers and
employees, 5 partners, and in none of the children. Only participants who reported to have direct
contact with pigs were colonized with the bacterium, however this was not a significant risk factor
(p=0.072). The following variables meet the criteria (p-value < 0.3 ) to be used in the multivariate
logistic regression model. Weekly working hours; residence on the pig farm; relocate sows; relocate
piglets; relocate weaning pigs; feeding sows; feeding piglets; feeding weaning pigs; cleaning sows;
performing healthcare to piglets; contact with fattening pigs; cleaning departments of piglets; use of
a facemask; contact with pigs on other farms; contact with own pigs; presence in the departments of
own pigs; presence in the departments of pigs on other farms; presence in stables with sheep;
administration of antibiotics; cleaning hands before entering pig farm; cleaning hands after entering
pig farm; and frequency of contact with own pigs.
Univariate results
Two risk factors are statistical significant in the univariate analysis. Contact frequency between
persons and pigs (p=0.03) and the administration of antibiotics to the pigs (p=0.006).
The contact frequency is divided into two variables, contact on a daily/weekly base versus contact on
a monthly/yearly base. The administration of antibiotics is divided into two groups. A group who
administrated antibiotics to pigs and a group who didn’t.
Multivariate results
Due to redundancy all effects were removed from the multivariable logistic regression model and no
variables in the model are available for display on the X-axis. This means there is no significant risk
factor determined in this model.
Discussion
Due to the fact that there are no previous risk factors documented a number of possible risk factors
is selected in order to make this study possible. The selection was made on the base of risk factors
described in previous studies about CDI and the knowledge gained from documented risk factors
from other bacteria then C. difficile. It is likely that risk factors concerning C. difficile are similar as to
other bacteria with the same characteristics.
Two results suggest that the contact frequencies between humans and pigs are an important risk
factor of C.difficile colonization in humans. The first is that only participants who reported to have
direct contact with the pigs were colonized with the bacterium, however this was not a significant
risk factor (p=0.072). The second is the contact frequency between persons and pigs (p=0.03).
Both results are expected because the transmission of C. difficile between animals and humans can
occur due to direct contact.
The data used to measure the influence of antibiotic treatment in this model is derived from the
questionnaire. The question referred to whether participants personally administrated antibiotics to
the animals or not. A significant result was found in the univariate model (p=0.006). Antibiotics are a
known risk factor for developing CDI in humans (9). A possible explanation is the direct contact
between pigs and the administrator. Another explanation is the direct contact between the
antibiotics and the skin of the administrator during administration. Because of this contact there is a
disturbance in the micro flora of the administrator that gives C. difficile an opportunity to colonize. In
a parallel study in the Bactopath project, quantitative data is available about how much antibiotics a
farm uses and what sort of antibiotics they use. In a follow up study it would be interesting to use
this data in combination with the colonization of C. difficile in humans.
Questionnaires always carry the risk of bias. The questionnaire was intensive, that is why it is possible
that people did not fill out the questions as precise as they could due to lack of time or loss of
interest.
In further studies it would be interesting to focus on the contact structures between humans and
pigs, how intense is the contact and how do specific methods of handling pigs differ between
humans who work with pigs. In further studies a different study design would be recommended. For
example follow participants that are not yet colonized with C. difficile for a period of time. Divide the
participants in two groups. One group has frequent contact with pigs or administer antibiotics to the
pigs. The other groep does not. One downside to our methodology is that the authors had no
previous experience with the softwareprogram SAS.
Conclusion
None of the risk factors are significant in a logistic regression statistical analysis with backward
elimination. However the risk factors contact frequency and antibiotic administration give a
significant outcome in a univariate analysis. Only participants who reported to have direct contact
with the pigs were colonized with the bacterium, however this was not a significant risk factor. The
conclusion in this study is that none of the risk factors have had a significant role in the prevalence of
C. difficile in humans. Further research is necessary to explain the role of the univariate results that
have a significant outcome. This research should focus on the contact structures between humans
and pigs, and on quantitative data concerning antibiotic administration to pigs.
Literature
1. Keel, K., Brazier, J. S., Post, K. W., Weese, S. & Songer, J. G. Prevalence of PCR ribotypes among
Clostridium difficile isolates from pigs, calves, and other species. J. Clin. Microbiol. 45, 1963-1964
(2007).
2. Indra A., Lassnig H., Baliko N., et al. Clostridium difficile: a new zoonotic agent? Wiener klinischer
Wochenschrift( 2009) 121:91-95
3. Kuijper, E.J., Van den Berg, R.J., Debast, S., Visser, C.E.,
Veenendaal, D., Troelstra, A., et al. (2006b) Clostridium
difficile PCR ribotype 027, toxinotype III in The Netherlands. Emerg Infect Dis 12: 827–830.
4. Goorhuis, A. et al. Spread and epidemiology of Clostridium difficile polymerase chain reaction
ribotype 027/toxinotype III in The Netherlands. Clin. Infect. Dis. 45, 695-703 (2007).
5. Songer, J., K. Post, D. Larson, B. Jost, and R. Glock. 2000. Infection of
neonatal swine with Clostridium difficile. Swine Health Prod. 8:185–189.
6.Waters, E. H., J. P. Orr, E. G. Clark, and C. M. Schaufele. 1998. Typhlocolitis
caused by Clostridium difficile in suckling piglets. J. Vet. Diagn. Investig.
10:104–108.
7. Keessen ,E.C., W. Gaastra,, and L.J.A. Lipman. 2010. Clostridium difficile infection in humans and
animals, differences and similarities. Veterinary Microbiology VETMIC-5243
8. Marwah M. Bakri, Derek J. Brown, John P. Butcher, and Alistair D. Sutherland, 2009, Clostridium
difficile in Ready-to-Eat salads, Scotland Emerging Infectious Diseases, Vol. 15, No. 5, May 2009
9. Kuijper, E.J., Van Disselt, J.T. 2008.Spectrum of Clostridium difficile infections outside health care
facilities. Canadian Medical Association journal, journal de l'Association medicale canadienne 179,
747-748.
10.Madewell, B.R., Tang, Y.J., 1995.Apparent outbreaks of Clostridium difficile-associated diarrhea in
horses in a veterinary medical teaching hospital. Journal of veterinary diagnostic investigation :
official publication of the American Association of Veterinary Laboratory Diagnosticians, Inc 7, 343346.
11. Songer, J.G., Anderson, M.A. (2006). Clostridium difficile: an important pathogen of food animals.
Anaerobe 12, 1-4.
12. Baverud, V., Gustafson, A. 1997. Clostridium difficile associated with acute colitis in mature
horses treated with antibiotics. Equine veterinary journal 29, 279-284.
13.Ruby, R., Magdesian, K.G., Kass, P.H. 2009. Comparison of clinical, microbiologic, and
clinicopathologic findings in horses positive and negative for Clostridium difficile infection. Journal of
the American Veterinary Medical Association 234, 777-784
14. Lawley, T.D., Clare, S., 2009. Antibiotic treatment of clostridium difficile carrier mice triggers a
supershedder state, spore-mediated transmission, and severe disease in immunocompromised hosts.
Infection and immunity 77, 3661-3669.
15. Kuijper, E.J., Hensgens, M.P.M. Clostridium difficile infection in an endemic setting in the
Netherlands. European journal of clinical microbiology and infectious diseases, Volume: 30 Issue: 4
(April 1, 2011), pp: 587-593
16. Dubberke, E.R., Hoppe-Bauer, J. Impact of Clinical Symptoms on Interpretation of Diagnostic
Assays for Clostridium difficile Infections. Journal of clinical microbiology, Volume: 49 Issue: 8 (August
1, 2011), pp: 2887-2893