Development, Evaluation, and Multisite Deployment of a Machine Learning Decision Tree Algorithm To Optimize Urinalysis Parameters for Predicting Urine Culture Positivity.

PittUDT, a recursive partitioning decision tree algorithm for predicting urine culture (UC) positivity based on macroscopic and microscopic urinalysis (UA) parameters, was developed in support of a broader system-wide diagnostic stewardship initiative to increase appropriateness of UC testing. Reflex algorithm training utilized results from 19,511 paired UA and UC cases (26.8% UC positive); the average patient age was 57.4 years, and 70% of samples were from female patients. Receiver operating characteristic (ROC) analysis identified urine white blood cells (WBCs), leukocyte esterase, and bacteria as the best predictors of UC positivity, with areas under the ROC curve of 0.79, 0.78, and 0.77, respectively. Using the held-out test data set (9,773 cases; 26.3% UC positive), the PittUDT algorithm met the prespecified target of a negative predictive value above 90% and resulted in a 30 to 60% total negative proportion (true-negative plus false-negative predictions). These data show that a supervised rule-based machine learning algorithm trained on paired UA and UC data has adequate predictive ability for triaging urine specimens by identifying low-risk urine specimens, which are unlikely to grow pathogenic organisms, with a false-negative proportion under 5%. The decision tree approach also generates human-readable rules that can be easily implemented across multiple hospital sites and settings. Our work demonstrates how a data-driven approach can be used to optimize UA parameters for predicting UC positivity in a reflex protocol, with the intent of improving antimicrobial stewardship and UC utilization, a potential avenue for cost savings.

Remediation of Mucorales-contaminated Healthcare Linens at a Laundry Facility Following an Investigation of a Case Cluster of Hospital-acquired Mucormycosis.

BACKGROUND: In an investigation of hospital-acquired mucormycosis cases among transplant recipients, healthcare linens (HCLs) delivered to our center were found to be contaminated with Mucorales. We describe an investigation and remediation of Mucorales contamination at the laundry supplying our center. METHODS: We performed monthly RODAC cultures of HCLs upon hospital arrival, and conducted site inspections and surveillance cultures at the laundry facility. Remediation was designed and implemented by infection prevention and facility leadership teams. RESULTS: Prior to remediation, 20% of HCLs were culture-positive for Mucorales upon hospital arrival. Laundry facility layout and processes were consistent with industry standards. Significant step-ups in Mucorales and mold culture-positivity of HCLs were detected at the post-dryer step (0% to 12% [P = .04] and 5% to 29% [P = .01], respectively). Further increases to 17% and 40% culture-positivity, respectively, were noted during pre-transport holding. Site inspection revealed heavy Mucorales-positive lint accumulation in rooftop air intake and exhaust vents that cooled driers; intake and exhaust vents that were facing each other; rooftop and plant-wide lint accumulation, including in the pre-transport clean room; uncovered carts with freshly-laundered HCLs. Following environmental remediation, quality assurance measures and education directed toward these sources, Mucorales culture-positivity of newly-delivered HCLs was reduced to 0.3% (P = .0001); area of lint-contaminated rooftop decreased from 918 m2 to 0 m2 on satellite images. CONCLUSIONS: Targeted laundry facility interventions guided by site inspections and step-wise culturing significantly reduced Mucorales-contaminated HCLs delivered to our hospital. Collaboration between infection prevention and laundry facility teams was crucial to successful remediation.

Whole-Genome Sequencing Surveillance and Machine Learning of the Electronic Health Record for Enhanced Healthcare Outbreak Detection.

BACKGROUND: Most hospitals use traditional infection prevention (IP) methods for outbreak detection. We developed the Enhanced Detection System for Healthcare-Associated Transmission (EDS-HAT), which combines whole-genome sequencing (WGS) surveillance and machine learning (ML) of the electronic health record (EHR) to identify undetected outbreaks and the responsible transmission routes, respectively. METHODS: We performed WGS surveillance of healthcare-associated bacterial pathogens from November 2016 to November 2018. EHR ML was used to identify the transmission routes for WGS-detected outbreaks, which were investigated by an IP expert. Potential infections prevented were estimated and compared with traditional IP practice during the same period. RESULTS: Of 3165 isolates, there were 2752 unique patient isolates in 99 clusters involving 297 (10.8%) patient isolates identified by WGS; clusters ranged from 2-14 patients. At least 1 transmission route was detected for 65.7% of clusters. During the same time, traditional IP investigation prompted WGS for 15 suspected outbreaks involving 133 patients, for which transmission events were identified for 5 (3.8%). If EDS-HAT had been running in real time, 25-63 transmissions could have been prevented. EDS-HAT was found to be cost-saving and more effective than traditional IP practice, with overall savings of $192 408-$692 532. CONCLUSIONS: EDS-HAT detected multiple outbreaks not identified using traditional IP methods, correctly identified the transmission routes for most outbreaks, and would save the hospital substantial costs. Traditional IP practice misidentified outbreaks for which transmission did not occur. WGS surveillance combined with EHR ML has the potential to save costs and enhance patient safety.

Outbreak of Pseudomonas aeruginosa Infections from a Contaminated Gastroscope Detected by Whole Genome Sequencing Surveillance.

BACKGROUND: Traditional methods of outbreak investigations utilize reactive whole genome sequencing (WGS) to confirm or refute the outbreak. We have implemented WGS surveillance and a machine learning (ML) algorithm for the electronic health record (EHR) to retrospectively detect previously unidentified outbreaks and to determine the responsible transmission routes. METHODS: We performed WGS surveillance to identify and characterize clusters of genetically-related Pseudomonas aeruginosa infections during a 24-month period. ML of the EHR was used to identify potential transmission routes. A manual review of the EHR was performed by an infection preventionist to determine the most likely route and results were compared to the ML algorithm. RESULTS: We identified a cluster of 6 genetically related P. aeruginosa cases that occurred during a 7-month period. The ML algorithm identified gastroscopy as a potential transmission route for 4 of the 6 patients. Manual EHR review confirmed gastroscopy as the most likely route for 5 patients. This transmission route was confirmed by identification of a genetically-related P. aeruginosa incidentally cultured from a gastroscope used on 4of the 5 patients. Three infections, 2 of which were blood stream infections, could have been prevented if the ML algorithm had been running in real-time. CONCLUSIONS: WGS surveillance combined with a ML algorithm of the EHR identified a previously undetected outbreak of gastroscope-associated P. aeruginosa infections. These results underscore the value of WGS surveillance and ML of the EHR for enhancing outbreak detection in hospitals and preventing serious infections.

Two Cases of Newly Characterized Neisseria Species, Brazil.

We describe 2 human cases of infection with a new Neisseria species (putatively N. brasiliensis), 1 of which involved bacteremia. Genomic analyses found that both isolates were distinct strains of the same species, were closely related to N. iguanae, and contained a capsule synthesis operon similar to N. meningitidis.

Performance of the T2Bacteria Panel for Diagnosing Bloodstream Infections: A Diagnostic Accuracy Study.

Background: Blood cultures, the gold standard for diagnosing bloodstream infections (BSIs), are insensitive and limited by prolonged time to results. The T2Bacteria Panel (T2 Biosystems) is a direct-from-blood, nonculture test that identifies the most common ESKAPE bacteria (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Escherichia coli). Objective: To assess performance of the T2Bacteria Panel in diagnosing suspected BSIs in adults. Design: Prospective patient enrollment (8 December 2015 through 4 August 2017). Setting: Eleven U.S. hospitals. Patients: 1427 patients for whom blood cultures were ordered as standard of care. Intervention: Paired blood culture and T2Bacteria testing. Measurements: Performance of T2Bacteria compared with a single set of blood cultures in diagnosing proven, probable, and possible BSIs caused by T2Bacteria-targeted organisms. Results: Blood culture and T2Bacteria results were positive for targeted bacteria in 3% (39 of 1427) and 13% (181 of 1427) of patients, respectively. Mean times from start of blood culture incubation to positivity and species identification were 38.5 (SD, 32.8) and 71.7 (SD, 39.3) hours, respectively. Mean times to species identification with T2Bacteria were 3.61 (SD, 0.2) to 7.70 (SD, 1.38) hours, depending on the number of samples tested. Per-patient sensitivity and specificity of T2Bacteria for proven BSIs were 90% (95% CI, 76% to 96%) and 90% (CI, 88% to 91%), respectively; the negative predictive value was 99.7% (1242 of 1246). The rate of negative blood cultures with a positive T2Bacteria result was 10% (146 of 1427); 60% (88 of 146) of such results were associated with probable (n = 62) or possible (n = 26) BSIs. If probable BSIs and both probable and possible BSIs were assumed to be true positives missed by blood culture, per-patient specificity of T2Bacteria was 94% and 96%, respectively. Limitation: Low prevalence of positive blood cultures, collection of a single set of culture specimens, and inability of T2Bacteria to detect nontargeted pathogens. Conclusion: The T2Bacteria Panel rapidly and accurately diagnoses BSIs caused by 5 common bacteria. Primary Funding Source: T2 Biosystems.

How Clean Is the Linen at My Hospital? The Mucorales on Unclean Linen Discovery Study of Large United States Transplant and Cancer Centers.

Mucormycosis outbreaks have been linked to contaminated linen. We performed fungal cultures on freshly-laundered linens at 15 transplant and cancer hospitals. At 33% of hospitals, the linens were visibly unclean. At 20%, Mucorales were recovered from >10% of linens. Studies are needed to understand the clinical significance of our findings.

High-Level Carbapenem Resistance in OXA-232-Producing Raoultella ornithinolytica Triggered by Ertapenem Therapy.

OXA-232 is an OXA-48-group class D ß-lactamase that hydrolyzes expanded-spectrum cephalosporins and carbapenems at low levels. Clinical strains producing OXA-232 are sometimes susceptible to carbapenems, making it difficult to identify them in the clinical microbiology laboratory. We describe the development of carbapenem resistance in sequential clinical isolates of Raoultella ornithinolytica carrying blaOXA-232 in a hospitalized patient, where the ertapenem MIC increased from 0.5 µg/ml to 512 µg/ml and the meropenem MIC increased from 0.125 µg/ml to 32 µg/ml during the course of ertapenem therapy. Whole-genome sequencing (WGS) analysis identified loss-of-function mutations in ompC and ompF in carbapenem-resistant isolates that were not present in the initial carbapenem-susceptible isolate. Complementation of a carbapenem-resistant isolate with an intact ompF gene resulted in 16- to 32-fold reductions in carbapenem MICs, whereas complementation with intact ompC resulted in a 2-fold reduction in carbapenem MICs. Additionally, blaOXA-232 expression increased 2.9-fold in a carbapenem-resistant isolate. Rapid development of high-level carbapenem resistance in initially carbapenem-susceptible OXA-232-producing R. ornithinolytica under selective pressure from carbapenem therapy highlights the diagnostic challenges in detecting Enterobacteriaceae strains producing this inefficient carbapenemase.

Reduced ceftazidime and ertapenem susceptibility due to production of OXA-2 in Klebsiella pneumoniae ST258.

BACKGROUND: OXA-2 is a class D ß-lactamase that confers resistance to penicillins, as well as narrow-spectrum cephalosporins. OXA-2 was recently reported to also possess carbapenem-hydrolysing activity. Here, we describe a KPC-2-encoding Klebsiella pneumoniae isolate that demonstrated reduced susceptibility to ceftazidime and ertapenem due to production of OXA-2. OBJECTIVES: To elucidate the role of OXA-2 production in reduced ceftazidime and ertapenem susceptibility in a K. pneumoniae ST258 clinical isolate. METHODS: MICs were determined by the agar dilution method. WGS was conducted to identify and compare resistance genes between isolates. Expression of KPC-2 was quantified by quantitative RT-PCR and immunoblotting. OXA-2 was expressed in Escherichia coli TOP10, as well as in K. pneumoniae ATCC 13883, to define the relative contribution of OXA-2 in ß-lactam resistance. Kinetic studies were conducted using purified OXA-2 enzyme. RESULTS: K. pneumoniae 1761 belonged to ST258 and carried both blaKPC-2 and blaOXA-2. However, expression of blaKPC-2 was substantially reduced due to an IS1294 insertion in the promoter region. K. pneumoniae 1761, K. pneumoniae ATCC 13883 and E. coli TOP10 carrying blaOXA-2-harbouring plasmids showed reduced susceptibility to ertapenem and ceftazidime, but meropenem, imipenem and cefepime were unaffected. blaOXA-2 was carried on a 2910 bp partial class 1 integron containing aacA4-blaOXA-2-qacEΔ1-sul1 on an IncA/C2 plasmid, which was not present in the earlier ST258 isolates possessing blaKPC-2 with intact promoters. Hydrolysis of ertapenem by OXA-2 was confirmed using purified enzyme. CONCLUSIONS: Production of OXA-2 was associated with reduced ceftazidime and ertapenem susceptibility in a K. pneumoniae ST258 isolate.

Verification of Ceftazidime-Avibactam and Ceftolozane-Tazobactam Susceptibility Testing Methods against Carbapenem-Resistant Enterobacteriaceae and Pseudomonas aeruginosa.

Ceftazidime-avibactam and ceftolozane-tazobactam are newly approved agents for the treatment of infections caused by multidrug-resistant Gram-negative bacteria. Resistance to both agents has been described clinically. Susceptibility testing on automated systems is unavailable for either agent. Our objective was to compare the disk diffusion and Etest methods to standard broth microdilution (BMD) methods for testing ceftazidime-avibactam and ceftolozane-tazobactam against a diverse collection of carbapenem-resistant Enterobacteriaceae (CRE) and carbapenem-resistant Pseudomonas aeruginosa (CRP) isolates, respectively. Among 74 ceftazidime-avibactam-susceptible and -resistant CRE isolates, BMD categorical agreement was higher with Etest (96%) than with disk diffusion (72%; P = 0.0003). Twenty-eight percent of ceftazidime-avibactam-susceptible CRE isolates were classified as resistant by disk diffusion. Results were comparable to those obtained with resistance defined genotypically. Among 72 ceftolozane-tazobactam-susceptible and -resistant CRP isolates, the levels of BMD categorical agreement with disk diffusion and Etest were 94% and 96%, respectively; the only errors identified were minor. Our findings demonstrate that Etest measurements of ceftazidime-avibactam and ceftolozane-tazobactam susceptibility correlate closely with standard BMD methods, suggesting a useful role clinically. On the other hand, disk diffusion measurements overcalled CRE resistance to ceftazidime-avibactam. A better understanding of ceftazidime-avibactam interpretive breakpoints is needed before disk diffusion is used routinely in the clinic. Until clinicians and microbiologists understand Etest and disk diffusion performance at their centers, test results should be interpreted cautiously.

Frequency and Mechanisms of Spontaneous Fosfomycin Nonsusceptibility Observed upon Disk Diffusion Testing of Escherichia coli.

Fosfomycin maintains activity against most Escherichia coli clinical isolates, but the growth of E. coli colonies within the zone of inhibition around the fosfomycin disk is occasionally observed upon susceptibility testing. We aimed to estimate the frequency of such nonsusceptible inner colony mutants and identify the underlying resistance mechanisms. Disk diffusion testing of fosfomycin was performed on 649 multidrug-resistant E. coli clinical isolates collected between 2011 and 2015. For those producing inner colonies inside the susceptible range, the parental strains and their representative inner colony mutants were subjected to MIC testing, whole-genome sequencing, reverse transcription-quantitative PCR (qRT-PCR), and carbohydrate utilization studies. Of the 649 E. coli clinical isolates, 5 (0.8%) consistently produced nonsusceptible inner colonies. Whole-genome sequencing revealed the deletion of uhpT encoding hexose-6-phosphate antiporter in 4 of the E. coli inner colony mutants, while the remaining mutant contained a nonsense mutation in uhpA The expression of uhpT was absent in the mutant strains with uhpT deletion and was not inducible in the strain with the uhpA mutation, unlike in its parental strain. All 5 inner colony mutants had reduced growth on minimal medium supplemented with glucose-6-phosphate. In conclusion, fosfomycin-nonsusceptible inner colony mutants can occur due to the loss of function or induction of UhpT but are rare among multidrug-resistant E. coli clinical strains. Considering that these mutants carry high biological costs, we suggest that fosfomycin susceptibility of strains that generate inner colony mutants can be interpreted on the basis of the zone of inhibition without accounting for the inner colonies.

Rate of FKS Mutations among Consecutive Candida Isolates Causing Bloodstream Infection.

Precise FKS mutation rates among Candida species are undefined because studies have not systematically screened consecutive, disease-causing isolates. The Sensititre YeastOne (SYO) assay measures echinocandin MICs against Candida with less variability than reference broth microdilution methods. However, clinical breakpoint MICs may overstate caspofungin nonsusceptibility compared to other agents. Our objectives were to determine Candida FKS mutation rates by studying consecutive bloodstream isolates and to determine if discrepant susceptibility results were associated with FKS mutations. FKS hot spots were sequenced in echinocandin-intermediate and -resistant isolates and those from patients with breakthrough candidemia or ≥ 3 days of prior echinocandin exposure. Overall, 453 isolates from 384 patients underwent susceptibility testing; 16% were echinocandin intermediate or resistant. Intermediate susceptibility rates were higher for Candida glabrata than for other species (P < 0.0001) and higher for caspofungin than for other agents (P < 0.0001). Resistance rates were similar between agents. FKS mutations were detected in 5% of sequenced isolates and 2% of isolates overall. Corresponding rates among C. glabrata isolates were 8% and 4%, respectively. Among Candida albicans isolates, rates were 5% and <1%, respectively. Mutations occurred exclusively with prior echinocandin exposure and were not detected in other species. Isolates with discrepant susceptibility results did not harbor FKS mutations. Mutation rates among isolates resistant to ≥ 2, 1, and 0 agents were 75%, 13%, and 0%, respectively. In conclusion, FKS mutations were uncommon among non-C. glabrata species, even with prior echinocandin exposure. Discrepancies in echinocandin susceptibility by SYO testing were not driven by mutations and likely reflect imprecise caspofungin clinical breakpoints.

Prevalence and duration of asymptomatic Clostridium difficile carriage among healthy subjects in Pittsburgh, Pennsylvania.

Previous studies suggested that 7 to 15% of healthy adults are colonized with toxigenic Clostridium difficile. To investigate the epidemiology, genetic diversity, and duration of C. difficile colonization in asymptomatic persons, we recruited healthy adults from the general population in Allegheny County, Pennsylvania. Participants provided epidemiological and dietary intake data and submitted stool specimens. The presence of C. difficile in stool specimens was determined by anaerobic culture. Stool specimens yielding C. difficile underwent nucleic acid testing of the tcdA gene segment with a commercial assay; tcdC genotyping was performed on C. difficile isolates. Subjects positive for C. difficile by toxigenic anaerobic culture were asked to submit additional specimens. One hundred six (81%) of 130 subjects submitted specimens, and 7 (6.6%) of those subjects were colonized with C. difficile. Seven distinct tcdC genotypes were observed among the 7 C. difficile-colonized individuals, including tcdC genotype 20, which has been found in uncooked ground pork in this region. Two (33%) out of 6 C. difficile-colonized subjects who submitted additional specimens tested positive for identical C. difficile strains on successive occasions, 1 month apart. The prevalence of C. difficile carriage in this healthy cohort is concordant with prior estimates. C. difficile-colonized individuals may be important reservoirs for C. difficile and may falsely test positive for infections due to C. difficile when evaluated for community-acquired diarrhea caused by other enteric pathogens.

Use of multilocus variable number of tandem repeats analysis genotyping to determine the role of asymptomatic carriers in Clostridium difficile transmission.

BACKGROUND: Previous studies have suggested that asymptomatic carriers of toxigenic Clostridium difficile are a source of hospital-associated (HA) infections. Multilocus variable number of tandem repeats analysis (MLVA) is a highly discriminatory molecular subtyping tool that helps to determine possible transmission sources. METHODS: Clostridium difficile isolates were recovered from perirectal swabs collected for vancomycin-resistant Enterococcus (VRE) surveillance as well as from clinical C. difficile toxin-positive stool samples from July to November 2009 at the University of Pittsburgh Medical Center Presbyterian (UPMC). MLVA was performed to determine the genetic relationships between isolates from asymptomatic carriers and patients with HA C. difficile infection (HA-CDI). Asymptomatic carriage and HA-CDI isolates were considered to be associated if the carriage isolate was collected before the HA-CDI isolate and if the MLVA genotypes had a summed tandem-repeat difference of ≤ 2. RESULTS: Of 3006 patients screened, 314 (10.4%) were positive for toxigenic C. difficile, of whom 226 (7.5%) were detected only by VRE surveillance cultures. Of 56 incident cases of CDI classified as HA at UPMC during the study with available isolates, 17 (30%) cases were associated with CDI patients, whereas 16 (29%) cases were associated with carriers. Transmission events from prior bed occupants with CDI (n = 2) or carriers (n = 2) were identified in 4 of 56 cases. CONCLUSIONS: In our hospital with an established infection control program designed to contain transmission from symptomatic CDI patients, asymptomatic carriers appear to have played an important role in transmission. Identification and isolation of carriers may be necessary to further reduce transmission of C. difficile in such settings.

Endoscopic retrograde cholangiopancreatography and endoscopic ultrasound endoscope reprocessing: Variables impacting contamination risk.

OBJECTIVE: To evaluate variables that affect risk of contamination for endoscopic retrograde cholangiopancreatography and endoscopic ultrasound endoscopes. DESIGN: Observational, quality improvement study. SETTING: University medical center with a gastrointestinal endoscopy service performing ∼1,000 endoscopic retrograde cholangiopancreatography and ∼1,000 endoscopic ultrasound endoscope procedures annually. METHODS: Duodenoscope and linear echoendoscope sampling (from the elevator mechanism and instrument channel) was performed from June 2020 through September 2021. Operational changes during this period included standard reprocessing with high-level disinfection with ethylene oxide gas sterilization (HLD-ETO) was switched to double high-level disinfection (dHLD) (June 16, 2020-July 15, 2020), and duodenoscopes changed to disposable tip model (March 2021). The frequency of contamination for the co-primary outcomes were characterized by calculated risk ratios. RESULTS: The overall pathogenic contamination rate was 4.72% (6 of 127). Compared to duodenoscopes, linear echoendoscopes had a contamination risk ratio of 3.64 (95% confidence interval [CI], 0.69-19.1). Reprocessing using HLD-ETO was associated with a contamination risk ratio of 0.29 (95% CI, 0.06-1.54). Linear echoendoscopes undergoing dHLD had the highest risk of contamination (2 of 18, 11.1%), and duodenoscopes undergoing HLD-ETO and the lowest risk of contamination (0 of 53, 0%). Duodenoscopes with a disposable tip had a 0% contamination rate (0 of 27). CONCLUSIONS: We did not detect a significant reduction in endoscope contamination using HLD-ETO versus dHLD reprocessing. Linear echoendoscopes have a risk of contamination similar to that of duodenoscopes. Disposable tips may reduce the risk of duodenoscope contamination.

Stewardship-Guided T2Candida Testing Shortens Time to Antifungal Treatment and Reduces Antifungal Usage Among Medical Intensive Care Unit Patients With Septic Shock.

Background: Diagnosis of invasive candidiasis (IC) is limited by insensitivity and slow turnaround of cultures. Our objectives were to define the performance of T2Candida, a nonculture test, under guidance of a diagnostic stewardship program, and evaluate impact on time to antifungal initiation and antifungal utilization. Methods: This was a retrospective study of adult medical intensive care unit (MICU) patients with septic shock for whom T2Candida testing was performed from March 2017 to March 2020. Patients with positive T2Candida results during this period were compared to MICU patients who did not undergo T2Candida testing but had septic shock and blood cultures positive for Candida from January 2016 through March 2020. Results: Overall, 155 T2Candida tests from 143 patients were included. Nine percent of T2Candida tests were positive compared to 4.5% of blood cultures. Sensitivity, specificity, positive predictive value, and negative predictive value of T2Candida for proven and probable IC were 78%, 95%, 50%, and 99%, respectively. Patients who tested positive for T2Candida (n = 14) were diagnosed earlier and initiated on antifungal therapy sooner than patients with IC (n = 14) diagnosed by blood culture alone (median, 5.6 vs 60 hours; P < .0001). Median antifungal days of therapy/1000 patient-days were 23.3/month preimplementation and 15/month postimplementation (P  = .007). Following a negative T2Candida result, empiric antifungals were either not administered in 58% or discontinued within 72 hours in 96% of patients. Conclusions: Diagnostic stewardship guided T2Candida testing resulted in reduced time to IC diagnosis, faster initiation of antifungal therapy, and lower antifungal usage among MICU patients with septic shock.

Perirectal swab surveillance for Clostridium difficile by use of selective broth preamplification and real-time PCR detection of tcdB.

Active surveillance testing to identify and isolate asymptomatic carriers of toxigenic Clostridium difficile has been limited by the lack of a test that is sensitive, specific, and timely enough to serve as an infection control tool. We tested DNA preamplified from perirectal surveillance specimens in a liquid medium selective for C. difficile by using a modified commercial real-time PCR assay. All fermenting specimens were subcultured, and isolates were tested for toxigenicity. Culture-positive toxigenic isolates served as the gold standard for comparison with the broth preamplification/PCR assay. The limit of detection for the assay was 1 CFU. Relative to toxigenic anaerobic culture, the sensitivity, specificity, and positive and negative predictive values of this assay were 70/70 (100.0%), 422/426 (99.1%), 70/74 (94.6%), and 422/422 (100.0%), respectively. These data demonstrate that selective broth preamplification and real-time PCR of perirectal swab specimens constitute a practical approach to the detection of asymptomatic C. difficile carriage.