Since September 2015, UW Health has been focusing on the use of a prevention bundle aimed at reducing Clostridium difficile Infection (CDI) risk in our inpatient population.
Successes so far include:
- A 26 percent decrease in hospital-onset CDI rate through May 2016
- Nine out of the last 10 weeks have seen two or fewer hospital-onset CDI cases
- Only four hospital-onset CDI cases in May 2016, the lowest since beginning to track CDI
Please visit the following links to learn more about CDI:
- CDI Webpage – View unit level data and measures on key bundle measures
- Improvement Stories – Read stories of how units have adopted the CDI bundle
You may contact Marc Oliver-Wright, Infection Control Specialist, or Anne Leclaire, Nursing Quality and Safety CNS, with any questions about the CDI initiative.
This is a truly remarkable achievement and would not possible without the hard work of all of you. Thank you for making our facility a safer place to be for our patients!
Please view the A3 for more detail
What’s wrong with testing patients preemptively to identify colonized patients? (rather than infected) Don’t we want to identify colonized patients and get them in isolation?
Neither the CDC guidelines nor the more recent SHEA guidelines propose an active surveillance program (identifying colonizers) in fact the latter says "Routine screening for asymptomatic carriage of C. difficile is not recommended.." and it goes on to say "The benefits of placing these patients under contact precautions (decreased transmission of C. difficile to other patients) must be balanced with the risks (in the absence of universal screening, it will be unlikely to impact CDI incidence) and potential for adverse events due to contact precautions. Do not attempt to decolonize asymptomatic patients since this has not been effective and may increase the patient’s risk of developing CDI in the future."
How infectious are patients who are colonized with C. diff?
"How contagious are colonizers?" The short answer to that is we just don't know for sure. We know that colonized patients without diarrhea are much lower from a transmission risk point of view than their symptomatic counterparts, But it is worth considering the following:
- If you test and they are positive, you will feel you have no choice but to treat.
- If you treat and they are not infected, you increase their risk of developing CDI and failing treatment when they actually need it.
- Such failure increases the potential for sequelae, recurrence, and generally worse outcomes.
- We know that environmental contamination from C diff is very heavy even weeks after effective treatment when symptoms have resolved
Don’t we want to treat colonized patients?
Actually, no, we don’t. It doesn’t actually work very well and the attempt to do so puts patients at greater risk for recurrence and treatment failure later on.
What are the positive predictive values of different clinical features (fever, abdominal pain etc) in helping to diagnose C. diff.
The table below shows the various risk factors. As for the “insignificant” p-values, they should be weighed against common sense. For example, in the below, the p-value for abdominal pain as a significant variable in CDI cases was 0.10-it stands to reason that despite the value not achieving statistical significance that patients with CDI-or any diarrheal illness for that matter are more apt to experience abdominal pain than their healthier counterparts.
I am confused about testing; isn’t PCR supposed to be better?
When properly utilized, on patients whose diarrhea is caused by an infection, PCR is a great test. PCR (100% sensitivity, 99.6% specificity) detects only the toxin producing strains and when properly ordered has higher sensitivity and specificity than EIA or culture. However, it fails to differentiate colonization from infection.
At any point in time, more than 12% of hospitalized patients are experiencing diarrhea (The Annals of Pharmacotherapy. 2006;40(6):1030-1034.) and this is more common in neutropenic patients (Ann Oncol. 2007; 18(1):183-9). C. difficile isn’t the only cause of diarrhea-even in hospitalized patients. More than a quarter of the time, no cause (including C. diff) can be identified. In fact, C. diff only accounts for 15-20% of antibiotic associated diarrhea (Am Fam Physician. 2005 Mar 1;71(5):921-928), yet colonization can range from 6-19% of all hospitalized patients (unpublished data). In other words, diarrhea is common in hospitalized patients and C. difficile accounts for a small (but clinically significant) proportion of these cases. Thus, it’s critical to test the right patient at the right time.
I want to test the patient, but the lab is refusing because it’s been less than 7 days since the last negative test. What gives?
Multiple studies have confirmed that repeat testing following an initially negative result is useless. Between 97.1 and 99.5% of the time you will get the same negative result. Our own data has shown that there has never been a subsequent positive within the 7 days after a negative test
I have a patient who I’ve treated for CDI. I want to test them to get them out of isolation.
- Most patients who initially test positive will continue to test positive by PCR for several weeks after completing therapy and clinically improving. A repeat positive test result should not be used as an indicator of treatment failure.
- Negative subsequent test results do not have any impact on the decision to keep a patient in enhanced contact precautions. All patients who test positive remain in isolation for a minimum of 90 days following their positive test result. Isolation may be extended beyond 90 days if the patient is still symptomatic.
- Recurrent infections happens about 20% of the time.
Does this algorithm work for pediatric patients?
First, the algorithm was developed based on recommendations available in the literature-a literature dominated by studies of the adult population. There is a minimal amount of literature addressing Clostridium difficile in pediatric patients-not enough to develop an algorithm.
What is known is that neonates and young children are much more likely to be colonized with C. difficile (>30%) and that the incidence of Clostridium difficile infection in children has been increasing over the past few years.
Although parts of this algorithm may be applicable/logical for all patient populations (e.g. do not test patients actively on new laxatives), it is not intended for use with pediatric patients.
How does this algorithm work for the immunosuppressed patient population?
While immunosuppressed patients are at greater risk for infection and more commonly experience diarrhea from a non-infectious etiology, most of this algorithm is applicable.
Patients who recently complete or are actively on chemotherapy often experience diarrhea proximal to their treatment window-and at variable frequencies. It can be challenging to differentiate anticipated diarrhea/loose stools such as this from diarrhea of a potentially infectious origin.
I reviewed the algorithm. I still want to test the patient.
First and foremost, this is an algorithm, not an edict; it is written on paper, not in stone. As the disclaimers embedded in it articulate, clinical judgment can and should be utilized throughout the process. If that judgement indicates testing is still warranted, test the patient. All cases are reviewed and as such, it would be helpful if your rationale for testing was spelled out in the patient’s record.
My patient has a history of CDI. Should I retest?
General considerations: There is no evidence pinpointing a specific time (days, weeks or months) after a patient initially tests positive when retesting the patient can best distinguish reinfection vs relapse vs ongoing colonization. Treatment failure one week following antibiotic administration rages from 14-22% and recurrence after successful treatment occurs between 24-27% of the time (http://www.ncbi.nlm.nih.gov/pubmed/22398198). Colonization as detected by PCR following successful treatment has been reported as high as 50% (J Clin Gastroenterol 2012;46:846-849). However, general recommendations can be made based on the patient’s clinical course following their initial diagnosis.
If the patient’s symptoms did resolve during/after treatment, especially if the patient’s most recent positive was >90 days ago, testing the patient should be considered. Evaluate the patient’s current clinical situation with respect to the CDI testing algorithm and decide accordingly.
If the patient’s symptoms have never resolved - that is, if they have had ongoing diarrhea despite being on treatment - repeat testing may not be warranted.
In limited circumstances, patients with a history of prolonged CDI, including treatment failure or relapse, may benefit from a prophylactic oral vancomycin dosing strategy (125mg PO once or twice daily); especially if broad-spectrum antibiotic exposure is inevitable. You may want to request an Infectious Diseases consult for evaluation.
We have a patient who has developed diarrhea, but this is a known and expected complication of the patient’s therapy. We feel confident that the diarrhea is not due to Clostridium difficile infection and would like to give the patient loperamide (a.k.a. Imodium). Do we need to test the patient first for Clostridium difficile infection?
If a patient develops diarrhea, particularly when it is a known potential complication of a therapy or treatment regimen, loperamide or similar anti-diarrheal agents may be used in the absence of a negative test, presuming suspicion for CDI is low.
For example, patients undergoing chemotherapy are anticipated to develop diarrhea during or shortly after their treatment regimen. If the treatment team appropriately evaluates the patient, and determines that the diarrhea is pharmacologic, rather than infectious in nature, it is appropriate for them to treat the patient’s symptoms without the need to test (or retest) the patient for CDI.
The patient has an elevated white blood cell (WBC) count and I would like to rule out CDI.
Leukocytosis can be predictive of more severe CDI (a) or more predictive for recurrence of CDI (b) in previously positive and treated patients. But it is not a good sole predictor for CDI in the absence of other signs or symptoms such as radiologic evidence of obstruction, diarrhea or abdominal pain.
Is CDI increasing in pediatric patients?
It appears to be. As in adults, CDI has increased in pediatric patients over the past 10-20 years, reportedly 53-75% higher between the late 90s and 2006 in two large multi-institutional studies (1). Also, community onset CDI appears to be increasing in pediatric patients; Clostridium difficile was the single most commonly recovered bacterial pathogen in pediatric patients presenting to the Emergency Department (2).
Should I use the adult CDI testing algorithm for pediatric patients?
The algorithm was developed based on recommendations available in the literature - literature dominated by studies of the adult population. There are some distinct differences with pediatric patients (addressed below).
The pediatric testing algorithm is intended to account for these population differences by utilizing a combination of evidence from the pediatric literature and consensus from internal subject matter experts at UW Health.
How are pediatric patients different with regards to testing for CDI?
There is a high prevalence of asymptomatic carriage of toxigenic C. difficile in infants and young children up to 2 years of age. As such, testing for CDI is generally not reccomended and if done should be conducted with testing for alternative causes of diarrhea such as norovirus and rotavirus. Even if positive, CDI should not be assumed to be causative of diarrhea unless there is no other plausible explanation (6).
How are pediatric patients the same as adults with regards to testing?
The laboratory uses the same test for both populations - the PCR - which fails to differentiate colonization from infection. Estimates of colonization range widely for both adults and pediatric patients, but have been reported to be as high as 20% in hospitalized children and adults (8, 9). With high rates of colonization, the tendency to treat patients with positive test results and the potential for unintended consequences with inappropriate therapy, it’s critical to test the right patient at the right time, for the right reason.
Common themes regarding CDI testing between adults and pediatrics include:
- Don’t test asymptomatic patients (less than 3 liquid/watery/loose stools per day). The lab rejects specimens that do not take the shape of the container.
- Don’t test patients on laxatives (see below)
- Don’t repeat testing
- With negatives: The lab rejects tests/specimens submitted within 7 days of a prior positive test. (7)
- With positives: Do not test for cure. The test used by the lab is a PCR test and often remains positive for many weeks after completion of therapy and successful resolution of symptoms (6).
My pediatric patient is on a home bowel regimen that includes a laxative or stool softener. How should this impact my decision to test the patient?
Home bowel regimens are often adjusted by patients or families themselves in response to the patient’s bowel movements. Deescalation of dose or holding the purgative altogether, albeit temporarily, is preferred to testing and treating the colonized patient. In short, do not test a patient within 24 hours of receiving a laxative. Observe for bowel pattern changes. If the diarrhea subsides, they do not have C. difficile infection.
Does CDI occur in pediatric patients at UW as often as it does in adults?
Not even close. In 2015, the rate of hospital onset CDI (detected >2 days after admission) in American Family Children’s Hospital was 2.7 per 10,000 patient days. Compare that to the adult population of University Hospital and the American Center where the rate was 13.1. The rate among adults is not only 4.8 times higher, but statistically significant.
How are pediatric patients different with regards to risk factors for CDI?
Some pediatric patients do not have any of the common risk factors for CDI. In a recent study of community onset CDI, nearly 25% of pediatric patients had no risk factors prior to being positive for CDI (5). Among outpatients with risk factors, recent cephalosporin use and the use of gastrointestinal feeding devices were associated with increased risk.
Additional risk factors unique to pediatric patients (inpatients) include: cancer (15 times higher risk)(3), solid organ transplant recipients and those with either a gastrostomy or jejunostomy tube have an eightfold and threefold higher risk, respectively (4). Renal insufficiency, impaired humoral immunity and diaper use have been recognized as risk factors in older children as well (8).
Common risk factors for both adult and pediatric cases include inflammatory bowel disease, proton pump inhibitors and antimicrobial therapy (8).
CDI Prevention and Testing Bundle
Fluoroquinolone Reduction Pilot
Guidelines, Policies, Protocols and Tools
Prevention, Diagnosis and Treatment of Clostridium Difficile Infection - Clinical Practice Guideline
Adult Inpatient Clostridium difficile Infection Testing Algorithm
Clostridium Difficile - Adult Protocol
Enhanced Precautions Audit Tool Training Video
Pediatric Inpatient Clostridium difficile Infection Testing Algorithm
Isolation Precautions for Pediatric Patients (HFFY #6415)
Isolation Precautions for Pediatric Patients Spanish (HFFY #7187)
Isolation Precautions for Adult Patients (HFFY #6980)
Isolation Precautions for Adult Patients (Spanish HFFY #7190)
What You Need to Know about Clostridium difficile Infection (CDI) (HFFY #7219)
What You Need to Know about Clostridium Difficile Infection (CDI) (Spanish HFFY #7929)
You’re at Risk for C. diff (HFFY #7853)
Enhanced Contact Precautions Video
Environmental Services Quality Assurance Program
Isolation Precautions: Transmission Stops with You