A Propensity Score Analysis of Clostridium difficile Infection among Adult Trauma Patients

Abstract

Background:

Clostridium difficile infection (CDI) is now the most common cause of health-care–associated infection and carries a mortality rate ranging from 5–30%. Previously, trauma patients in whom CDI developed were thought to represent a unique younger at-risk population. This study aimed to establish the incidence of CDI among adult trauma patients. We hypothesized that these patients would have increased risk of death, intensive care unit (ICU) length of stay (LOS), and hospital LOS compared with trauma patients without CDI.

Patients and Methods:

A retrospective study of all adult trauma patients admitted for greater than 48 hours to a single Level I trauma center between 2014 and 2016 was conducted. Analysis was performed using 1-to-5 propensity score matching with the aim to analyze the relationship between CDI, death, and other outcome variables.

Results:

Between 2014 and 2016, of 4893 trauma patients admitted for >48 hours, 27 (0.6%) patients received a diagnosis of CDI. These patients had a mean age of 55.6 years, mean injury severity score (ISS) of 22.4, and mortality rate of 9.1%. Of these patients, 22 were able to find appropriate propensity score matches. After adjusting for important covariables, there was no significant difference in death between CDI and non-CDI patients (odds ratio = 0.39, 95% confidence interval [CI]: 0.06–2.57, adjusted p = 0.66). In addition, there was no significant difference in ICU LOS between the two groups (relative mean [RM]: 1.55, 95% CI: 1.04–2.33, adjusted p = 0.0971). The CDI patients, however, did have a significantly longer hospital LOS, compared with non-CDI patients (RM = 1.39, 95% CI: 1.16–1.66, adjusted p = 0.0017).

Conclusions:

Among trauma patients admitted >48 hours CDI occurred at a rate of 0.6%, much lower than anticipated. Patients in whom CDI developed had a significantly longer hospital LOS however, had no significant difference in odds of mortality or ICU LOS compared to patients without CDI.

C lostridium difficile is a pathogen that has become a leading cause of nosocomial infections, including antibiotic-associated diarrhea and pseudomembranous colitis [1,2]. There has been a dramatic rise in incidence of C. difficile infections (CDI) worldwide, especially among surgical patients. CDI carries a mortality rate of 3–36% and is associated with toxic megacolon and septic shock [3]. CDI has been reported to cost an estimated $3.2 billion annually in the United States. Reveles et al. [4] demonstrated that the incidence of CDI nearly doubled from 2001 to 2010 with a startling increase in the rate of death. In addition, previous studies have demonstrated the rate of re-admissions with CDI to be as high as 22.5% [5 –7].

There has been a significant effort to identify relevant CDI risk factors. Currently, accepted risk factors for CDI include recent antibiotic exposure, age >65 years, proton pump inhibitor use, duration of hospitalization, and immunocompromised state [8]. According to the Centers for Disease Control and Infection, more than 2.8 million patients are hospitalized with traumatic injury annually [9]. Despite growing attention to CDI risk factors, epidemiology of CDI among hospitalized trauma patients still remains poorly understood.

A few small studies have identified a new population of CDI trauma patients, surprisingly without traditional risk factors for the disease. Lumpkins et al. [2] retrospectively reviewed 19 cases of CDI among 581 critically injured trauma patients admitted to the intensive care unit (ICU). They identified a unique subgroup of younger patients in whom DCI developed after minimal or no antibiotic exposure. Trauma patients with CDI were demonstrated to have increased intensive care unit (ICU) length of stay (LOS), hospital LOS, and ventilator days compared with non-CDI trauma patients. A larger and more recent study performed by Vanzant et al. [10] found CDI in 67 of 5042 total critically ill trauma patients. They also found the CDI patients have increased hospital LOS and ventilator days. Neither study found a significant difference in death between CDI and non-CDI patients.

Our study aimed to establish the incidence and unique risk factors for CDI among all adult trauma patients and to determine outcomes in this study population. We hypothesized that these patients would have increased risk of death, ICU LOS, and hospital LOS.

Patients and Methods

Patient selection

Before the initiation of this project, approval was obtained from the University of California, Irvine (UCI) Institutional Review Board to analyze the data collected in the UCI Trauma Database. We retrospectively queried the medical records of all adult trauma patients admitted for more than 48 hours to a single Level I trauma center from January 1, 2014 to December 31, 2016. Exclusion criteria were age <18 years, prisoners, and pregnant patients. The initial data set was cross-referenced with the CDI registry maintained by the UCI department of epidemiology and infection prevention to identify admitted trauma patients who received a diagnosis of CDI during their hospital stay. The diagnosis of CDI was established with a nucleic acid amplification testing (NAAT) positive for toxin produced by toxigenic strains of C. difficile (GeneXpert, Xpert C. Difficile/Epi, Cepheid, Inc.) with sensitivity of 97% and specificity of 95% for detection of CDI [11].

Patient information including age, gender, mechanism of injury, Glasgow Coma Scale (GCS), systolic blood pressure (SBP), heart rate (HR), Injury Severity Score (ISS), Abbreviated Injury Scale (AIS), Revised Trauma Score (RTS) on admission, as well as surgical data was collected. Data regarding antibiotic exposure, including antibiotic name, type, and duration, were also collected. Patients were defined as receiving “prophylactic” antibiotic agents if they received only pre-operative or intra-operative antibiotic prophylaxis. Patients were defined as having received “therapeutic” antibiotic agents if they received antibiotic agents besides those described as prophylactic antibiotic agents (i.e., electronic medical records were checked to see if any antibiotic agents not related to peri-operative prophylaxis were administered, or prophylactic antibiotic agents were continued post-operatively). Data relating to disease severity, including vital signs, laboratory values, radiologic studies, and any potential surgical intervention were collected.

To describe our CDI patient population, we used the current classification utilized by the Infectious Diseases Society of America and Society for Healthcare Epidemiology of America. This classification stratifies CDI into three categories: non-severe, severe, and fulminant (Supplementary Appendix A) [12]. In addition, we utilized the latest Sepsis 3 guidelines and quick sequential organ failure assessment score to diagnose sepsis in CDI patients [13].

We retrospectively analyzed outcomes of interest. The primary outcome measure was in-hospital death. Secondary outcome measures were hospital LOS, ICU LOS, and days on a ventilator.

Initial statistical analysis

Categoric variables were compared using the chi squared test or Fisher exact test; continuous variables were compared using the Student t test. To eliminate selection bias, we conducted propensity score matching analysis. The propensity scores were derived for each patient using a logistic regression to identify relevant risk factors for development of CDI. The estimated propensity score was the probability of CDI estimated from a fitted regression model. The propensity score was then used to obtain a 1-to-5 match of all CDI patients. Matching was done without replacement and using a greedy matching algorithm. The criteria for selecting matched pairs was based on nearest neighbor matching within a caliper distance equal to 0.2 of the pooled standard deviation of the logit of the propensity score. To assess the success of the matching procedure, standardized mean differences (SMD) were used to compare continuous and binary baseline covariates between groups.

To assess the association between the outcomes of interest and presence of CDI, logistic and poisson regression was performed. Relevant variables that were unable to be used in the matching procedure were used in the regression for the outcomes of interest. Odds ratios (OR) and relative means (RM) were be used to quantify the relationship between outcomes of interest and presence of CDI. These were reported along with 95% confidence intervals (CIs) for the true population parameters and the corresponding p values. The statistical analysis was performed using R statistical software (The R Project for Statistical Computing, R Foundation, Vienna, Austria).

Propensity score matching

Five patients with CDI were excluded from further analysis, because they did not produce a corresponding match. Thus, the propensity score matching algorithm produced 22 matches, each consisting of one CDI patient and five non-CDI controls. Certain data limitations prevented the inclusion of antibiotic regimens in the matching procedure. Antibiotic agents were hypothesized to be the main drivers of the development of CDI in trauma patients. Therefore, in addition to adjustments for HR and SBP, the final analysis of the outcomes of interest included adjustments for antibiotic data, including whether or not patients received therapeutic versus prophylactic antibiotic agents, difference in number, type, and duration of different antibiotic agents between the two groups.

Results

During the two-year study period of 12,705 trauma patients, 4893 patients were admitted for more than 48 hours. There were 27 patients who were identified to have CDI during their hospital stay, yielding an incidence of CDI of 0.6%. The average time to CDI diagnosis was 9.7 ±± 6.5 days. Baseline characteristics of the CDI patient population before matching are listed in Table 1. The mean age of the CDI patient group was 55.7 ± 21.9 years; 74.1% were males, and 48.1% were white. The majority of CDI (85.2%) patients sustained blunt trauma. Mean GCS score was 12.04 ± 4.01, and mean ISS was 22.4 ± 16.0. Before matching, the patients with CDI had a lower GCS score, lower RTS, higher ISS, more underwent laparotomy and thoracic procedures, and had a higher AIS Chest, AIS Head/Neck, AIS Abdomen, and AIS Extremity (Table 1). Post-matching, there were no significant differences between the CDI patients and non-CDI patients in any of the covariates (Table 2).

Table 1.

Patient Characteristics Pre-match^*

Characteristic	CDI (n = 27)	Non-CDI (n = 4866)	p	SMD
Age (y), mean (SD)	55.7 (21.9)	48.7 (22.9)	0.117	0.309
Gender = male, n (%)	20 (74.1)	3,191 (65.6)	0.471	0.185
GCS, mean (SD)	12.04 (4.01)	14.13 (2.42)	<0.001	0.630
RTS, mean (SD)	6.96 (1.28)	7.59 (0.87)	<0.001	0.576
ISS, mean (SD)	22.4 (16.0)	10.5 (10.0)	<0.001	0.923
Race, n (%)			<0.001	0.751
Am. Indian/Alaska Native	0 (0)	1 (0)
Asian-Hawaiian/Pac. Isl.	0 (0)	7 (0.1)
Black	2 (9.3)	114 (2.3)
White	13 (48.0)	3,804 (78.3)
Hispanic	6 (22.2)	184 (3.8)
Missing/Unknown	0 (4.7)	14 (0.3)
Other	1 (3.7)	128 (2.6)
BMI (kg/m²), mean (SD)	27.6 (4.8)	26.8 (6.0)	0.445	0.163
Laparotomy, mean (SD)	0.19 (0.4)	0.03 (0.2)	<0.001	0.605
Thoracic procedure, mean (SD)	0.11 (0.32)	0 (0.1)	<0.001	0.461
Max AIS		]
Head/Neck, mean, (SD)	2.04 (2.1)	0.94 (1.5)	<0.001
Face, mean, (SD)	0.33 (0.7)	0.22 (0.6)	0.370	0.160
Chest, mean, (SD)	1.67 (1.6)	0.95 (1.37)	0.007	0.484
Abdomen, mean, (SD)	1.04 (1.4)	0.47 (1.00)	0.004	0.456
Extremity mean, (SD)	1.48 (1.7)	0.83 (1.2)	0.005	0.453
External, mean, (SD)	0.85 (0.5)	0.71 (0.5)	0.153	0.267
Mechanism, % penetrating	4 (14.8)	345 (7.1)	0.236	0.249

CDI = Clostridium difficile infection; SMD = standardized mean difference; SD = standard deviation; GCS = Glasgow Coma Scale; RTS = Revised Trauma Score; ISS = injury Severity Score; BMI = body mass index; AIS = Abbreviated Injury Scale.

^*n = 4893.

Table 2.

Patient Characteristics Post-match^*

Characteristic	CDI (n = 22)	Non-CDI (n = 110)	p	SMD
Age (y), mean (SD)	55.2 (22.2)	56.8 (23.1)	0.802	0.059
Gender = male, n (%)	16 (72.7)	79 (71.8)	1.000	0.02
GCS, mean, (SD)	12.2 (4.2)	11.99 (4.4)	0.852	0.044
RTS, mean, (SD)	6.99 (1.4)	6.87 (1.6)	0.733	0.083
ISS, mean, (SD)	19.41 (15.2)	20.75 (16.4)	0.724	0.085
SBP on Arrival, mean (SD)	130.05 (39.4)	137.39 (39.1)	0.423	0.187
HR on Arrival, mean (SD)	94.05 (19.2)	90.64 (23.2)	0.519	0.160
Race, n (%)			0.682	0.301
Asian-Hawaiian/Pac. Isl.	4 (18.2)	22 (20.0)
Black	2 (9.1)	3 (2.7)
White	13 (59.1)	64 (58.2)
Hispanic	2 (9.1)	14 (12.7)
Other	1 (4.5)	7 (6.4)
BMI (kg/m²), mean (SD)	27.24 (5.0)	25.82 (5.0)	0.225	0.285
Laparotomy, mean (SD)	0.14 (0.4)	0.13 (0.3)	0.908	0.026
Thoracic procedure, mean (SD)	0.11 (0.32)	0.05 (0.2)	0.863	0.041
Max AIS
Head/Neck, mean, (SD)	1.86 (2.0)	2.21 (2.1)	0.478	0.169
Face, mean, (SD)	0.36 (0.8)	0.25 (0.7)	0.463	0.162
Chest, mean, (SD)	1.36 (1.6)	1.48 (1.7)	0.760	0.073
Abdomen, mean, (SD)	0.86 (0.6)	0.80 (1.4)	0.843	0.047
Extremity mean, (SD)	1.18 (1.6)	1.00 (1.4)	0.589	0.121
External, mean, (SD)	0.86 (0.6)	0.79 (0.5)	0.536	0.138
Mechanism, % penetrating	3 (13.6)	13 (11.8)	1.000	0.055

CDI = Clostridium difficile infection; SMD = standardized mean difference; SD = standard deviation; GCS = Glasgow Coma Scale; RTS = Revised Trauma Score; ISS = injury Severity Score; SBP = systolic blood pressure; HR = heart rate; BMI = body mass index; AIS = Abbreviated Injury Scale.

^*n = 132

Antibiotic agent exposure of CDI patients and non-CDI controls is listed in Table 3. Most patients in the CDI (54.6%) group received both prophylactic and therapeutic antibiotic agents. Mean number of days receiving therapeutic antibiotic agents was higher in the CDI (18.6 ± 16.9) group than in the non-CDI (5.6 ± 13.7) group. In addition, the mean number of prescribed therapeutic antibiotic agents was also higher in the CDI (2.6 ± 1.7) group than in the non-CDI (0.8 ± 1.3) group. Of the patients who received therapeutic antibiotic agents before CDI diagnosis, the three most common indications were ventilator-associated pneumonia (n = 7), empiric treatment of severe sepsis and septic shock (n = 3), and complicated urinary tract infection (n = 2). In addition, four patients had no documented indication for therapeutic antibiotic agents, and two patients had an inappropriately prolonged duration of antibiotic treatment.

Table 3.

Antibiotic Exposure of Patients with and without Clostridium difficile Infection^*

Characteristic	CDI (n = 22)	Non-CDI (n = 132)
Prophylactic antibiotics, n (%)	1 (4.6)	16 (14.6)
Therapeutic antibiotics, n (%)	7 (31.8)	15 (13.6)
Prophylactic and therapeutic antibiotics, n (%)	(14)12 (54.6)	30 (27.3)
No antibiotics, n (%)	2 (9.1)	71 (53.8)
Total days on therapeutic antibiotics, mean (SD)	18.6 (16.9)	5.6 (13.7)
Number of different therapeutic antibiotics, mean (SD)	2.6 (1.7)	0.83 (1.3)
Clindamycin	3 (13.6)	2 (1.8)
Fluoroquinolones	5 (22.7)	6 (5.5)
Cephalosporins	15 (68.2)	30 (37.3)
Penicillin (piperacillin/aminopenicillins)	12 (54.6)	23 (20.9)
Sulfonamides	0 (0.0)	0 (0.0)
Tetracyclines	0 (0.0)	0 (0.0)
Macrolides	0 (0.0)	3 (2.7)
Vancomycin	12 (54.5)	13 (11.8)
Carbapenems	0 (0.0)	4 (3.6)
Aminoglycosides	0 (0.0)	2 (1.8)
Rifamycin	0 (0.0)	1 (0.9)

CDI = Clostridium difficile infection; SD = standard deviation.

n = 132

The mortality rate among patients with CDI was 9.1%. The mortality rate among the non-CDI control group was 23.1%. There was no significant difference in odds of death, however, between the two groups (OR = 0.39, 95% CI = 0.06–2.57, p = 0.66). There was no significant difference in adjusted ICU LOS (RM = 1.55, 1.04–2.33, p = 0.10) and ventilator days (RM = 0.99, 0.57–1.73, p = 0.98). The adjusted hospital LOS among CDI patients was 1.39 times higher compared with non-CDI patients (RM = 1.39, 95% CI = 1.16–1.66, adjusted p value = 0.0017) (Table 4).

Table 4.

Short-term Outcomes Associated with Clostridium difficile Infection

CDI vs. Non-CDI	Value (95% CI)	p	Adjusted p
Odds ratio
Death	0.39 (0.06, 2.57)	0.33	0.66
Relative mean
ICU length of stay	1.55 (1.04, 2.33)	0.03	0.09
Hospital length of stay	1.39 (1.16, 1.66)	0.00044	0.0017
Ventilator days	0.99 (0.57, 1.73)	0.98	0.98

CDI = Clostridium difficile infection; CI = confidence interval; ICU = intensive care unit.

Associations based on relative means for quantitative outcomes and odds ratios for dichotomous outcomes. Associations were based on logistic regression for death. Poisson regression was used for ICU, hospital length of stay, and ventilator days.

Of the CDI patients with CDI, six (22%) met criteria for sepsis. Of the 27 patients, non-severe disease developed in 13 (48%), severe CDI developed in 11 (41%), and fulminant CDI developed in three (11%). No patients with CDI underwent surgical intervention including subtotal colectomy, loop ileostomy with colonic lavage, or fecal microbiota transplantation. The only patient who died with fulminant CDI was in multi-system organ dysfunction with a devastating brain injury. Therefore, surgical intervention was declined.

Discussion

Overall, CDI is a common nosocomial infection and presents a major cause of morbidity and death among hospitalized patients. Our study used a propensity matched design to estimate the incidence of CDI among trauma patients and compare the odds of death, hospital LOS, and ICU LOS between CDI and non-CDI hospitalized trauma patients. Interestingly, despite using a larger population than previous studies, we found a much lower incidence of CDI among trauma patients at only 0.6%. Further, and contrary to our hypothesis, we did not find a significant difference in odds of death between CDI patients and propensity matched non-CDI patients. We also did not see a significant difference in ICU LOS. The adjusted mean hospital LOS for CDI patients, however, was 1.39 times higher than patients without CDI.

The incidence of CDI among hospitalized patients has been quoted to range from 0.4% to 4%. The overall incidence of CDI in our inpatient trauma population of 4893 was 0.6%, which is more than five times lower than the 3.3% quoted by Lumpkins et al. [2] in their analysis of 581 critically injured trauma patients and two times lower than the 1%–2% quoted by Vanzant et al. [10] in their analysis of 5042 patients with blunt trauma.

There has been a considerable effort to develop infection control strategies to reduce incidence of CDI among hospitalized patients [14,15]. CDI transmission via spores resistant to conventional hand-hygiene products poses additional challenges to control this infection [14]. The following infection control strategies have been shown to be effective in preventing development of CDI and transmission: (1) Initiation of appropriate antibiotic stewardship program, (2) placement of patients with CDI in private rooms, (3) use of dedicated patient care items and equipment when caring for patients with CDI, (4) use of full barrier precautions (donning gown and gloves), (5) meticulous hand hygiene with soap and water when caring for patients with CDI, (6) environmental decontamination of rooms of patients with CDI with bleach [14,16 –20]. The low incidence of CDI in our trauma patient cohort is similar to our institution's CDI rate of 0.7%, demonstrating that institution-wide adoption of the aforementioned measures benefits all patients, including trauma patients.

The asymptomatic carriage rate for toxigenic C. difficile in hospitalized patients ranges from 4.4%–21% [21]. Because polymerase chain reaction-based CDI testing cannot distinguish between colonization and active colitis, it has been shown to lead to overdiagnosis of CDI when testing is performed on asymptomatic patients. In addition to the above-mentioned infection control strategies, our institution has implemented a real-time electronic clinical criteria verification for healthcare professionals that aimed to enforce appropriate CDI testing. At UCI, testing is performed by request only, and only if certain criteria are met. This resulted in reduction of inappropriate testing by 64% [22]. In addition, hospital-onset CDI rates decreased by 54%, from 17 per 10,000 patient days to seven cases per 10,000 patient days [22].

We were unable to corroborate the presence of a younger trauma population with no known risk factors at higher risk for CDI. The mean age of our CDI cohort was 55.7 years, greater than the mean age of non-CDI patients (48.7 years). It was also greater than the mean age of the CDI patients in the studies by both Lumpkins et al. [2] (48.7 years) and Vanzant et al. [10] (43.9 years). In addition, all but two patients in the CDI cohort received either prophylactic or therapeutic antibiotic agents.

Exposure to antibiotic agents is the most important risk factor for development of CDI. Patients receiving clindamycin, third or fourth generation cephalosporins, fluoroquinolones, and carbapenems are at an especially high risk. Both longer exposure to antibiotic agents and exposure to multiple antibiotic agents increase risk of CDI [12]. Most of the CDI patients in our cohort (54.6%) received both therapeutic and prophylactic antibiotics (Table 3). Only three CDI (11.1%) patients had no antibiotic exposure or prophylactic antibiotic agents only, compared with five (26.3%) patients in the study by Lumpkins et al [2]. Therefore, our study provides no basis for clinicians to consider the physiologic changes of trauma to increase the risk of CDI, although additional research studies may be necessary to confirm these findings.

Given the widespread use of antibiotic agents and multiple instances of unnecessary or prolonged use of antibiotic agents in our study, it is possible that implementation of antibiotic stewardship interventions, such as daily antibiotic audit and feedback, and restriction of high-risk antibiotic agents may help further decrease trauma CDI rates at our institution [23].

After adjusting for significant covariates, the odds for death were not significantly higher for CDI patients compared with those without CDI. The absolute mortality rate in our CDI patient cohort was 9.1%, however. This is comparable to mortality rates among CDI trauma patients reported by Glance et al. [24] (7.3%), Vanzant et al. [10] (6%), and Lumpkins et al. [2] (14.7%). Similar to our findings, the mortality rates among CDI trauma patients in these studies were lower than the mortality rate among non-CDI patients. This may be a result of a combination of a declining mortality rate among CDI patients and the unique characteristics of the trauma population being a relatively younger with fewer co-morbidities than the overall hospital population.

Luo et al. [25] reported that the rate of mortality among hospitalized CDI patients at acute care hospitals from the National Inpatient Sample (NIS) database during a 10-year study period ending in 2014 declined from 9.7% to 6.8%. Increased compliance with established CDI prevention and treatment guidelines [26], as well as increased awareness of healthcare personnel through in-hospital education programs may have contributed significantly to reduced mortality rates. Also, increased use of appropriate and accurate diagnostic testing, such as NAAT, leads to more timely diagnosis and appropriate treatment [27].

We found that hospital LOS for the CDI patients was significantly higher than hospital LOS for non-CDI patients. This is in agreement with previous reports of greater LOS in trauma patients in whom hospital-acquired infections (HAIs) developed, and CDI, in particular [28,29]. In their analysis of 155,891 trauma patients from the NIS, Glance et al. [24] found the median LOS was two-fold higher in patients with HAIs compared with patients without HAIs [29]. Both Lumpkins et al. [2] and Vanzant et al. [10] reported greater hospital LOS among CDI trauma patients as well. Hence, CDI in trauma patients does not appear to be associated with increased death; it is associated with increased morbidity and healthcare costs.

Our study had several important limitations including that we were unable to include HR and SBP as well as antibiotic data in the initial matching algorithm. These variables had to be accounted for when performing the logistic regression for the outcomes of interest. A major limitation in our study is the small population of patients with CDI; despite using a much larger overall trauma population, our observed incidence was surprisingly two to five times lower than previous reports. Finally, we were unable to examine whether the episode of CDI was an initial or recurrent episode, and we were not able to account for all risk factors associated previously with CDI, including serum albumin and creatinine, as well as significant comorbidities such as immunosuppression.

Conclusions

Patients in whom CDI developed had a significantly longer hospital LOS; however, they did not have increased ICU LOS or risk of death. The incidence of CDI is far lower in our study than previous reports. Future prospective studies comparing high performing to low performing centers in regard to CDI may help elucidate whether patient factors or preventative measures such as regulating the ability to test for CDI, hand hygiene, online education, or antibiotic stewardship programs are the source for this disparity.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Luciano

, Zuckerbraun

. Clostridium difficile infection: Prevention, treatment, and surgical management. Surg Clin North Am, 2014; 94:1335–1349.

Lumpkins

, Bochicchio

, Joshi

, et al. Clostridium difficile infection in critically injured trauma patients. Surg Infect, 2008; 9:497–501.

Desai

, Gupta

, Dubberke

, et al. Epidemiological and economic burden of Clostridium difficile in the United States: Estimates from a modeling approach. BMC Infect Dis, 2016; 16:303.

Reveles

, Lee

, Boyd

, Frei

. The rise in Clostridium difficile infection incidence among hospitalized adults in the United States: 2001-2010. Am J Infect Control, 2014; 42:1028–1032.

Deshpande

, Pant

, Olyaee

, Donskey

. Hospital readmissions related to Clostridium difficile infection in the United States. Am J Infect Control, 2018; 46:346–347.

Goyal

, Chatterjee

, Punjabi

, Rangaswami

. Thirty-day readmissions after hospitalization for Clostridium difficile enteritis measures and predictors: A nationwide analysis. Infect Control Hosp Epidemiol, 2017; 38:876–878.

Kelly

, LaMont

. Clostridium difficile—more difficult than ever. N Engl J Med, 2008; 359:1932–1940.

Elliott

, Androga

, Knight

, Riley

. Clostridium difficile infection: Evolution, phylogeny and molecular epidemiology. Infect Genet Evol, 2017; 49:1–11.

Hall

, DeFrances

, Williams

, et al. National Hospital Discharge Survey: 2007 summary. Natl Health Stat Report, 2010:28:1–32.

10.

Vanzant

, Ozrazgat-Baslanti

, Liu

, et al. Clostridium difficile infections after blunt trauma: A different patient population?. Surg Infect, 2015; 16:421–427.

11.

Bai

, Sun

, Jin

, et al. Accuracy of Xpert Clostridium difficile assay for the diagnosis of Clostridium difficile infection: A meta analysis. PloS One, 2017; 12:e0185891.

12.

McDonald

, Gerding

, Johnson

, et al. Clinical Practice Guidelines for Clostridium difficile Infection in Adults and Children: 2017 Update by the Infectious Diseases Society of America (IDSA) and Society for Healthcare Epidemiology of America (SHEA). Clin Infect Dis, 2018; 66:987–994.

13.

Singer

, Deutschman

, Seymour

, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA, 2016; 315:801–810.

14.

Dubberke

, Carling

, Carrico

, et al. Strategies to prevent Clostridium difficile infections in acute care hospitals: 2014 update. Infect Control Hosp Epidemiol, 2014; 35(Suppl 2):S48–S65.

15.

Mermel

, Jefferson

, Blanchard

, et al. Reducing Clostridium difficile incidence, colectomies, and mortality in the hospital setting: A successful multidisciplinary approach. Jt Commission J Qual Patient Saf, 2013; 39:298–305.

16.

Cohen

, Gerding

, Johnson

, et al. Clinical practice guidelines for Clostridium difficile infection in adults: 2010 update by the society for healthcare epidemiology of America (SHEA) and the infectious diseases society of America (IDSA). Infect Control Hosp Epidemiol, 2010; 31:431–455.

17.

Bauer

, Kuijper

, van Dissel

. European Society of Clinical Microbiology and Infectious Diseases (ESCMID): Treatment guidance document for Clostridium difficile infection (CDI). Clin Microbiol Infect, 2009; 15:1067–1079.

18.

Surawicz

, Brandt

, Binion

, et al. Guidelines for diagnosis, treatment, and prevention of Clostridium difficile infections. Am J Gastroenterol, 2013; 108:478–498.

19.

Boyce

, Pittet

. Guideline for Hand Hygiene in Health-Care Settings. Recommendations of the Healthcare Infection Control Practices Advisory Committee and the HICPAC/SHEA/APIC/IDSA Hand Hygiene Task Force. Society for Healthcare Epidemiology of America/Association for Professionals in Infection Control/Infectious Diseases Society of America. MMWR Recomm Rep, 2002; 51(RR-16):1–45.

20.

Davey

, Brown

, Fenelon

, et al. Interventions to improve antibiotic prescribing practices for hospital inpatients. Cochrane Database Syst Rev. 2005:Cd003543.

21.

Truong

, Schroeder

, Gaur

, et al. Clostridium difficile rates in asymptomatic and symptomatic hospitalized patients using nucleic acid testing. Diagn Microbiol Infect Dis, 2017; 87:365–370.

22.

Quan

, Yim

, Merrill

, et al. Reductions in Clostridium difficile infection (CDI) rates using real-time automated clinical criteria verification to enforce appropriate testing. Infect Control Hosp Epidemiol, 2018:39:625–627.

23.

Feazel

, Malhotra

, Perencevich

, et al. Effect of antibiotic stewardship programmes on Clostridium difficile incidence: A systematic review and meta-analysis. J Antimicrob Chemother, 2014; 69:1748–1754.

24.

Glance

, Stone

, Mukamel

, Dick

. Increases in mortality, length of stay, and cost associated with hospital-acquired infections in trauma patients. Arch Surg, 2011; 146:794–801.

25.

Luo

, Barlam

. Ten-year review of Clostridium difficile infection in acute care hospitals in the USA, 2005–2014. J Hosp Infect, 2018; 98:40–43.

26.

Patel

, Wungjiranirun

, Theethira

, et al. Lack of adherence to SHEA-IDSA treatment guidelines for Clostridium difficile infection is associated with increased mortality. J Antimicrob Chemother, 2017; 72:574–581.

27.

Bogaty

, Levesque

, Garenc

, et al. Trends in the use of laboratory tests for the diagnosis of Clostridium difficile infection and association with incidence rates in Quebec, Canada, 2010–2014. Am J Infect Control, 2017; 45:964–968.

28.

Watkins

, Mangira

, Muakkassa

, et al. Clostridium difficile Infection in trauma, surgery, and medical patients admitted to the intensive care unit. Surg Infect, 2018; 19:488–493.

29.

Bottiggi

, White

, Bernard

, Davenport

. Impact of device-associated infection on trauma patient outcomes at a major trauma center. Surg Infect, 2015; 16:276–280.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.02 MB

0.00 MB