The investigators will conduct a two-period, two-intervention, cluster randomized crossover (CRXO) designed study at Foothills Hospital in Calgary, Alberta. Each cluster (hospital ward) will receive each of the two interventions in a separate approximately 6-month period of time leading to two "cluster-periods" with associated "wash-in" and "wash-out" periods attached. The two interventions are: 1) a rapid diagnostic test (RDT) using near-patient testing (NPT) for C. difficile infections (CDI); 2) testing for CDI using centralized testing facilities (standard operating procedure, 2 step algorithm).

The investigators hypothesize that NPT for CDI will result in reduced patient isolation days in the hospital, decreased morbidity and mortality, reduced unnecessary antibiotic use and overall reduced costs to the health care system, including both hospital and laboratory.

Our primary endpoint is to examine the differential effect of NPT on the duration of contact precautions (i.e. patient isolation days). Secondary outcomes will compare turn-around times from specimen submission to result, days of hospitalization, the number of days of antibiotic therapy, the incidence of CDI, severe CDI, and in-hospital mortality. A comprehensive economic evaluation will be performed to determine the cost of testing, patient isolation, and hospitalization for all patients in the control and treatment arms.

Antimicrobial resistance (AMR) has been designated as a major priority by the World Health Organization (WHO), Centers for Disease Control and Prevention (CDC) and the Public Health Agency of Canada (PHAC). PHAC has identified Clostridium difficile infection (CDI) as a Tier 1 AMR priority pathogen. CDI occurs directly as a consequence of antibiotic use and falls under the clinical syndrome of antibiotic-associated diarrhea (AAD). It is the most common cause of infectious diarrhea among hospitalized patients in developed countries like Canada (Skidmore, 1993). The economic cost of CDI in the United States is estimated at USD 5.4 billion, close to 87% of which occurs in healthcare settings (Desai et al., 2016).

Clinical microbiologists have developed a two-step algorithm in which an EIA or an RDT is first used to screen stools for the presence of glutamate dehydrogenase (GDH) antigen derived from C. difficile. The GDH antigen test is very specific (> 95%) and negatives can be reported right away, only requiring GDH-positive or toxin-positive specimens (by EIA or RDT) to be confirmed by NAAT. In Calgary, the two-step algorithm is performed for a population of close to 1.5 million people at an offsite Diagnostic and Scientific Centre (DSC). The use of a testing centre undermines some of the advantages of the rapid testing methodologies developed in recent years, because of increased transportation and processing time.

A key issue in many settings is that testing falls into a separate budget from the treatment of patients with CDI infections. This has the potential to distort the decision of whether to run a test and, if so, what sort of test. The distortions are of two types. First, the hospital may decide to order more tests than is optimal, since the tests are perceived as "free" from the perspective of decision-makers. At the same time, the diagnostic facility may choose low-cost, slow diagnostics, because the costs associated with delays in test results are borne by the hospital. This is a classic case of silos in health care costs inefficiently distorting policy decisions, with potentially negative results for both patient health and costs.

Many institutions follow the clinical practice guidelines provided by the Infectious Diseases Society of America for the management of C. difficile (Cohen et al., 2010). As part of routine Infection Prevention and Control (IPC) management, patients with acute diarrhea of unknown cause are pre-emptively placed in private rooms with additional contact precautions, which mandate the use of gowns and gloves for healthcare workers. This is also known as patient isolation. Investigation for gastrointestinal pathogens is typically initiated and the duration of patient isolation depends largely on the result of microbiological testing. When severe or complicated CDI is suspected, patients are initiated on empiric treatment as soon as the diagnosis is suspected. Generally, oral metronidazole is used for mild-to-moderate CDI and oral vancomycin for severe CDI. Among hospitalized patients with acute diarrhea, initiation of pre-emptive management for CDI is common.

The intervention of interest is implementing a new RDT to Near Patient Testing (NPT); initial screening for CDI to be performed at the on site hospital rapid response laboratory (RRL) instead of the DSC. Techlab Inc. offers a lateral flow assay (C. DIFF QUIK CHEK COMPLETE® test) that detects both GDH and Toxin simultaneously and further simplifies the testing algorithm. As prescribed in the two-step algorithm, GDH and Toxin negatives will be reported out immediately as negative for CDI, while positive or discrepant results will be confirmed at the DSC using NAAT. The vast majority of requests for CDI testing are negative thus allowing a majority of results to be reported immediately without confirmation at the DSC.

The Investigators will conduct a two-period, two-intervention, cluster randomized crossover (CRXO) design study at Foothills Hospital. Each cluster (ward) will receive each of the two interventions in a separate 6-month period of time leading to two "cluster-periods" with associated "wash-in" and "wash-out" periods attached. The interventions are RDT using NPT for CDI, and testing for CDI using our centralized testing facility.

This study is cross-sectional CRXO in that, each cluster-period consists of different individuals (patients). The order in which the interventions are delivered to each cluster (ward) is randomized to control for potential period effects. Within Foothills Hospital there are 40 wards which will be observed over a 12-month period. Each ward is randomly assigned to administer one of the two interventions for the first 6-month period (Period 1). In the second 6-month period (Period 2), the ward will administer the other intervention. Within each ward (cluster) there are two cluster-periods. Study outcomes during Period 1 and Period 2 intervention will be compared in both arms (NPT and centralized laboratory testing).

NPT testing can deliver results within several hours, while typically off-site testing takes about one day. If NPT and the off-site testing deliver identical test results as our pilot data suggest, the investigators can follow the clinical outcomes for each arm without the confounder of differences in test performance. As a contingency, if variation in test performance characteristics is found, analysis of patient outcomes will take this into consideration.

For patients without CDI, the antibiotics, and in many cases, the isolation, is unnecessary. NPT testing has the potential to avoid this unnecessary use of antibiotics and to avoid costly isolation procedures for the day+ that standard testing requires. It also enables a better-directed diagnostic process for the cause of diarrhea. For patients who are positive for CDI, the earlier test result will influence downstream events such as initiation of antibiotic therapy, resolution of disease, hospitalization days, and other clinical outcomes.

The investigators hypothesize that NPT for CDI will result in reduced patient isolation days in the hospital, decreased morbidity and mortality, reduced unnecessary antibiotic use and overall reduced costs to the health care system, including both hospital and laboratory.

Previous data at this facility has indicated that the mean number of patients per ward is 39/month (SD=43) for a total of 1578 patients in a 6-month period and the mean number of negative patients is 33/month (SD=34) for a total of 1300 negative pts in a 6-month period. The inclusion criteria will be diarrhea in adult patients admitted to Foothills Medical Center in Calgary, Alberta being tested for CDI. Outpatients and patients in the emergency room will be excluded.

Our primary endpoint is to examine the differential effect of NPT on the duration of contact precautions (i.e. patient isolation days). Secondary outcomes will compare turn-around times from specimen submission to result, days of hospitalization, the number of days of antibiotic therapy, the incidence of CDI, severe CDI, and in-hospital mortality. A comprehensive economic evaluation will be performed to determine the cost of testing, patient isolation, and hospitalization for all patients in the control and treatment arms.

Contact precautions (i.e. patient isolation) due to CDI will be defined as an order by healthcare workers for contact precautions within the 7-day infection window of a CDI diagnosis. The length of time on contact precautions will be defined as the calendar date contact precautions are completed or patient discharge date minus the calendar date contact precautions are ordered. Data on contact precautions will be extracted from the clinical patient management system in Calgary, known as Sunrise Clinical Manager (SCM). It is used extensively in all Calgary acute care facilities and emergency departments. These data include patient identifiers (e.g. patient health number, names, date of birth), start and stop dates of isolation, facility, and patient care unit where isolation order was placed, and the reason for isolation (e.g. CDI, diarrhea). These data will be linked to the specimen results tested for CDI by the patient's Personal Health Number (PHN). Turn-around times from specimen submission to result will be provided by Calgary Laboratory Services. Days of hospitalization will be defined as the difference between the admission and discharge date. The specimen results will be linked to the Discharge Abstract Database using the patient's PHN to determine during which acute care encounter the specimen was collected and the admission/discharge dates of that encounter. Antibiotic therapy will be defined as the administration of antibiotics initiated on the day of and up to 48 hours after specimen collection for CDI testing. Total days of antibiotic therapy during the acute care encounter will be defined as the calendar date antibiotic therapy is completed minus the calendar date antibiotic therapy is ordered. The data will be extracted from SCM.

The investigators will also account for the severity of CDI. The epidemiological definition for severe CDI is that primarily proposed by Cohen et al. This definition has been used by infectious disease specialists in Alberta to guide the management and treatment of patients with CDI. Patients with severe CDI will have leukocytosis with a white blood cell (WBC) count >15,000 cells/µL OR an elevated serum creatinine (i.e. 200 µmol/L or greater). The inpatient laboratory results that will be considered to meet the definition for severe CDI will be those collected within a 7-day infection window around the date the laboratory tested positive for C. difficile. The presence of either measure meeting the criteria within the defined infection window will qualify as meeting the definition of severe CDI. If either criterion is measured more than once during the infection window, the highest value will be used to represent the severity of CDI. The laboratory values will be provided by Calgary Laboratory Services and linked to the positive C. difficile test results. In-hospital, all-cause mortality will be defined as death occurring during the acute care encounter where specimens were collected for C. difficile testing.

The investigators predict that, given the large number of patients and randomized study design, matching of patients by age, gender, and co-morbidities is not required. Potential hurdles in the study model are logistical challenges associated with building a testing and reporting system for a hospital on 24hrs/7 days per week basis. The strength of the proposal lies in the scale and nature of the study which seeks to establish the medical and economic benefit of NPT versus centralized testing. The results could dramatically influence public health policy in this country and seek to bridge traditional silos, namely the laboratory and hospital budgets. Data on costs will be furnished by our partners at Calgary Laboratory Services/Alberta Health Services (Infection Prevention and Control).

Beyond evaluating the relative merits of NPT versus centralized testing, studies will also look at the microbiology of CDI in this study. Specifically, the investigators will conduct several laboratory methods to examine antibiotic resistance and genetic variation in CDI. For example, whole genome sequencing of cultured CDI isolates may inform us about the transmission of specific strains in our setting. Indeed, certain strains and states of gut dysbiosis may result in increased morbidity and mortality (Shahinas et al., 2012). The investigators will look at the microbiome of patients in this study to determine if dysbiosis contributes to adverse clinical outcomes. The investigators will also correlate the strain types with the results of testing to determine if genetic variation affects test performance. Cultured toxin-producing CDI isolates will serve as a gold standard to evaluate test performance.

Inclusion Criteria:

• Patients located on inpatient wards for whom C. difficile test is ordered at Foothills Medical Center. Usually defined by having 3 or more loose bowel movements in a 24 hour period.

Exclusion Criteria:

• No diagnostics for C. difficile requested. Inpatient at a different facility. All outpatients.

Not possible as integrating diagnostics into the hospital system.