Background: Gastroenteritis causes significant morbidity and mortality in long-term care facility (LTCF) residents, a growing population within the United States. Methods: We conducted a retrospective cross-sectional study in LTCFs in Philadelphia County from 2009 to 2018. Outbreak characteristics and interventions were extracted from Philadelphia Department of Public Health’s (PDPH) database, and quality data on all LTCFs was extracted from the CMS Nursing Home Compare database. Results: We identified 121 gastroenteritis outbreaks in 49 facilities. Numbers of affected patients ranged from 2 to 211 patients (median patient attack rate, 17%). Staff were reported ill in 94 outbreaks (median staff attack rate, 5%). Outbreak facilities were associated with higher occupancy rates (91% vs 88%; P = .033) and total bed numbers (176 vs 122; P = .071) when compared to nonoutbreak facilities. Higher rates of staff illness were associated with prolonged outbreaks (13% vs 4%; P < .001) and higher patient illness rates (9% vs 4%; P = .012). Prolonged outbreaks were associated with lower frequency of cohorting for outbreak management (13% vs 41%; P = .046). Conclusions: This study is the largest published analysis of gastroenteritis outbreaks in LTCFs. Facility characteristics and staff disease 20 activity were associated with more severe outbreaks. Heightened surveillance for gastrointestinal symptoms among staff and increased 21 use of cohorting might reduce the risk of prolonged gastroenteritis outbreaks in LTCF.

Funding: None

Disclosures: None

Background: Isolation precautions are recommended when caring for patients identified with highly resistant microorganisms (HRMOs). However, the direct costs of isolating patients are largely unknown. Therefore, we aimed to obtain detailed information on the daily direct costs associated with isolating patients identified with HRMO. Methods: This study was performed from November until December 2017 on a 12-bed surgical ward. This ward contained solely isolation rooms with an anteroom. The daily direct costs of isolation were based on three cost items: (1) additional personal protective equipment (PPE); measured by counting the consumption of empty packaging materials, (2) cleaning and disinfection of the isolation room; based on the costs of an outsourced cleaning company, and (3) additional workload for healthcare workers; based on literature and multiplied by the average gross hourly salary of nurses. A distinction was made between the costs for strict isolation, contact-plus isolation, and contact isolation. Results: During the study period, 26 patients were nursed in isolation because of HRMO carriage, resulting in a total of 304 isolation days (median 7 isolation days; range 1-44). Gloves were consumed the most and hair caps the least. The average daily direct costs of isolation were the least expensive for contact isolation, €28/$31, and the most expensive for strict isolation, €41/$47. Conclusions: By using a novel, easy method to estimate consumption of PPE, we conclude that the daily direct costs of isolating a patient, differs per type of isolation. Insight into the direct costs of isolation is of utmost importance when developing or revising policies.

Funding: None

Disclosures: None

Background: Contaminated surfaces within patient rooms and on shared equipment is a major driver of healthcare-acquired infections (HAIs). The emergence of Candida auris in the New York City metropolitan area, a multidrug-resistant fungus with extended environmental viability, has made a standardized assessment of cleaning protocols even more urgent for our multihospital academic health system. We therefore sought to create an environmental surveillance protocol to detect C. auris and to assess patient room contamination after discharge cleaning by different chemicals and methods, including touch-free application using an electrostatic sprayer. Surfaces disinfected using touch-free methods may not appear disinfected when assessed by fluorescent tracer dye or ATP bioluminescent assay. Methods: We focused on surfaces within the patient zone which are touched by the patient or healthcare personnel prior to contact with the patient. Our protocol sampled the over-bed table, call button, oxygen meter, privacy curtain, and bed frame using nylon-flocked swabs dipped in nonbacteriostatic sterile saline. We swabbed a 36-cm2 surface area on each sample location shortly after the room was disinfected, immediately inoculated the swab on a blood agar 5% TSA plate, and then incubated the plate for 24 hours at 36°C. The contamination with common environmental bacteria was calculated as CFU per plate over swabbed surface area and a cutoff of 2.5 CFU/cm2 was used to determine whether a surface passed inspection. Limited data exist on acceptable microbial limits for healthcare settings, but the aforementioned cutoff has been used in food preparation. Results: Over a year-long period, terminal cleaning had an overall fail rate of 6.5% for 413 surfaces swabbed. We used the protocol to compare the normal application of either peracetic acid/hydrogen peroxide or bleach using microfiber cloths to a new method using sodium dichloroisocyanurate (NaDCC) applied with microfiber cloths and electrostatic sprayers. The normal protocol had a fail rate of 9%, and NaDCC had a failure rate of 2.5%. The oxygen meter had the highest normal method failure rate (18.2%), whereas the curtain had the highest NaDCC method failure rate (11%). In addition, we swabbed 7 rooms previously occupied by C. auris–colonized patients for C. auris contamination of environmental surfaces, including the mobile medical equipment of the 4 patient care units that contained these rooms. We did not find any C. auris, and we continue data collection. Conclusions: A systematic environmental surveillance system is critical for healthcare systems to assess touch-free disinfection and identify MDRO contamination of surfaces.

Funding: None

Disclosures: None

Background: Despite clear guidance for appropriate testing of symptomatic patients for Clostridioides difficile testing (McDonald et al), the ideal testing methodology remains unresolved. Laboratories currently use different algorithms that incorporate enzyme immunoassay (EIA) testing for toxin, glutamate dehydrogenase (GDH) antigen, and polymerase chain reaction (PCR) testing in combination or as a single test. At UNC Hospitals, a large academic hospital with nearly 1,000 beds in the ninth most populous state in the United States, patients are currently tested by an EIA test for toxin and GDH antigen first, and discordant toxin/GDH results are referred for PCR testing. Previous studies have demonstrated that detection of toxin by EIA is a better predictor of C. difficile infection (CDI) complications (Polage et al). Methods: We investigated all patients who were tested for C. difficile from July 2018 to June 2019. Within each testing methodology and result, we assessed the percentage of patients with at least 3 loose stools documented within a 24-hour period, percentage with a severe episode based on white blood cell (WBC) counts >15,000 cells/mL, or percentage with a serum creatinine level >1.5 mg/dL. Fisher-type confidence intervals were calculated for each proportion. Results: Patients positive for C. difficile by the EIA method had 66.9% appropriate loose stool documentation (95% CI, 57.4%–75.5%), whereas patients with EIA-indeterminate (toxin negative, GDH positive) and positive by only PCR had 49.7% appropriate loose stool documentation (95% CI, 42.7%–56.8%). C. difficile patients that tested negative had 48.1% appropriate loose stool documentation (95% CI, 46.0–50.2%). In addition, patients positive by the EIA method had nearly double the proportion of severe disease by WBC or creatinine criteria compared to patients who were either positive by PCR or who tested negative (Table 1). Conclusions: Patients positive for C. difficile by the EIA method were statistically more likely to meet criteria for loose stool documentation. There was no statistically significant difference between patients that tested positive only by PCR or who tested negative. The percentage of patients with severe episode criteria based on WBC or creatinine was nearly doubled between those who tested positive by EIA and PCR (20% vs 10%), although this finding was not statistically significant. The percentage with severe disease (WBC or creatinine) was nearly identical among patients who were positive by PCR and who tested negative. These findings demonstrate that documentation of loose stool is a more sensitive indicator of toxin detection than either clinical parameter, reinforcing the importance of stool documentation in evaluating patients for C. difficile testing.

Funding: None

Disclosures: None

Backround: In American hospitals alone, the CDC estimates that hospital acquired infections (HAIs) account for an estimated 1.7 million infections and 99,000 associated deaths each year.1 Although the United states and most industrialized nations have made strides in lowering the overall HAI rate by taking critical steps to reduce HAIs, an overall formula that combines a global risk assessment per patient for HAI acquisition has yet to be established. To address this issue, we developed the ICEL equation. This equation uses a probabilistic argument to estimate the likelihood of HAI acquisition and to promote infection control dialogue among healthcare practitioners from diverse healthcare disciplines. Methods: We defined HAI risk using the ICEL acronym as follows: HAI risk = (I + C + E + L), where I is invasive devices present; C is patient-specific characteristics; E is the average number of pathogenic organisms in the patient environment; and L is the length of stay. A simple scale of 1–10 points is subjectively assigned for each of the following categories:

I = (number of invasive devices / surgeries / % body surface areas open)

C = Patient specific characteristics (immune system integrity / immunomodulators / radiation exposure / chemotherapy, etc)

E = Environmental conditions / cleaning (average number of pathogenic bacteria in room, 100% hand hygiene compliance, patient / staff colonization, etc)

L = Length of stay days risk, where 0–3 days is low risk, 4–7 is moderate risk, and 8–10+ is high risk

Summing the points for each of the 4 categories, the greatest possible total is 40. A total score of 0–10 indicates low risk of HAI; 11–20 indicates low-to-medium risk of HAI; 20–30 indicates a high risk of HAI; and 30–40 indicates a very high risk of HAI. Results:This equation was designed to stimulate thought and encourage multidisciplinary cooperation among providers, nursing, environmental services, and facilities departments rather than provide an exact number for HAI risk. All of these categories are key players in the determining patient risk of acquiring an HAI. If any of the 4 hospital departments mentioned fails in their duties, the patient is at higher risk of HAI. Conclusions: This categorical HAI risk assessment relies on the subjective medical and environmental knowledge of the assessor to assign risk across the continuum of the healthcare environment. Although it is nearly impossible to provide exact numbers regarding total risk in these risk categories, the goal of the scoring system is to encourage clinical dialogue among hospital staff so that they communicate and collaborate within their specialties and with their peers to assure that each category poses as low a risk as possible, thus driving the total risk for HAI lower.

1. https://www.cdc.gov/hai/data/portal/progress-report.html

Funding: None

Disclosures: None

Background: Antibiotic “time outs” have been identified as a way to decrease inappropriate use of antibiotics in hospitals.1 The University of Vermont Medical Center created a best-practice advisory (BPA) to alert clinicians to review piperacillin-tazobactam prescriptions after 72 hours (Fig. 1). Data examining the use of a BPA as a method to prompt clinicians to perform an antibiotic “time out” are limited. Objective: The purpose of our retrospective study was to evaluate the effectiveness of the BPA on the rate of piperacillin-tazobactam prescribing as measured by defined daily dose per 1,000 patient days (DDD). Methods: The BPA was integrated into the electronic health record and designed to activate once piperacillin-tazobactam has been prescribed for ≥72 hours. Under approval of the University of Vermont’s Institutional Review Board, administered data for piperacillin-tazobactam and 3 control antibiotics (cefazolin, ceftriaxone, and meropenem) were collected for 1 year prior to and 1 year following the launch of the BPA. Administered data were converted to DDD, and an interrupted time-series analysis was performed to evaluate for changes in antibiotic use. Results: The data included 7,094 patients in the preintervention group and 6,661 patients in the postintervention group. The BPA fired 1,478 times. The prescribing rate of piperacillin-tazobactam 1 year prior to the BPA was 32.34 DDD and decreased every month both before (−1.22 DDD) and after (−0.27 DDD) the BPA initiation, with no significant difference in prescribing trends (P = .10). Meropenem prescribing in the BPA era increased each month compared to the pre-BPA period (1.16 DDD; P = 0.02), whereas cefazolin use (P = .93) and ceftriaxone (P = .09) use did not significantly change. Conclusions: The data show that piperacillin-tazobactam utilization at our institution is decreasing. Considering that this trend started prior to the launch of the BPA and that rate of decline remained unchanged post-BPA, we conclude that the BPA did not further impact our piperacillin-tazobactam consumption. It is possible that other factors influencing prescribing account for the observed decline, including an institution-wide educational campaign regarding the appropriate use of broad-spectrum antibiotics that was initiated in the months prior to the BPA. The reason for the significant rise in meropenem post-BPA is unclear. This may be unrelated to the BPA; however, it requires further investigation.

1. Core elements of hospital antibiotic stewardship programs. Centers for Disease Control and Prevention website. https://www.cdc.gov/antibioticuse/healthcare/implementation/core-elements.html. Updated July 19, 2019. Accessed October 6, 2019.

Funding: None

Disclosures: None

Background: Polymerase chain reaction (PCR) testing for the diagnosis of Clostridioides difficile infection (CDI) detects the presence of the organism; a positive result therefore cannot differentiate between colonization and the pathogenic presence of the bacterium. This may result in overdiagnosis, overtreatment, and risking disruption of microbial flora, which may perpetuate the CDI cycle. Algorithm-based testing offers an advantage over PCR testing as it detects toxin, which allows differentiation between colonization and infection. Although previous studies have demonstrated the clinical utility of this testing algorithm in differentiating infection from colonization, it is unknown whether the test changes CDI treatment decisions. Our facility switched from PCR to an algorithm-based testing method for CDI in June 2018. Objective: In this study, we evaluated whether clinicians’ decisions to treat patients are impacted by a test result that implies colonization (GDH+/Tox−/PCR+ test), and we examined the impact of this decision on patient outcomes. Methods: This is a retrospective cohort study of inpatients with a positive C. diff test between June 2017 and June 2019. The primary outcome was the proportion of patients treated for CDI. We compared this outcome in 3 groups of patients: those with a positive PCR test (June 2017–June 2018), those who had a GDH+/Tox−/PCR+ or a GDH+/Tox+ test result (June 2018–June 2019). Secondary outcomes included toxic megacolon, critical care admission, and mortality in patients with GDH+/Tox−/PCR+ who were treated versus those who were untreated. Results: Of patients with a positive PCR test, 86% were treated with CDI-specific antibiotics, whereas 70.4% with GDH+/Tox+ and 29.25% with GDH+/Tox−/PCR+ result were treated (P < .0001). Mortality was not different between patients with GDH+/Tox−/PCR+ who were treated versus those who were untreated (2.7% vs 3.4%; P = .12), neither was critical care admission within 2 or 7 days of test result (2% vs 1.4%; P = .15) and (4.1% vs 5.4%, P = .39), respectively. There were no cases of toxic megacolon during the study period. Conclusions: The change to an algorithm-based C. difficile testing method had a significant impact on the clinicians’ decisions to treat patients with a positive test, as most patients with a GDH+/Tox−/PCR+ result did not receive treatment. These patients did not suffer more adverse outcomes compared to those who were treated, which has implications for testing practices. It remains to be explored whether clinicians are using clinical criteria to decide whether or not to treat patients with a positive algorithm-based test, as opposed to the more reflexive treatment of patients with a positive PCR test.

Funding: None

Disclosures: None

Background: Ventilator-associated pneumonia (VAP) represents the highest burden among all healthcare-associated infections (HAIs), with a particularly high rate in patients in neurosurgical ICUs. Numerous VAP risk factors have been identified to provide a basis for preventive measures. However, the impact of individual factors on the risk of VAP is unclear. The goal of this study was to evaluate the dynamics of various VAP risk factors given the continuously declining prevalence of VAP in our neurosurgical ICU. Methods: This prospective cohort unit-based study included neurosurgical patients who stayed in the ICU >48 consecutive hours in 2011 through 2018. The infection prevention and control (IPC) program was implemented in 2010 and underwent changes to adopt best practices over time. We used a 2008 CDC definition for VAP. The dynamics of VAP risk factors was considered a time series and was checked for stationarity using theAugmented Dickey-Fuller test (ADF) test. The data were censored when a risk factor was present during and after VAP episodes. Results: In total, 2,957 ICU patients were included in the study, 476 of whom had VAP. Average annual prevalence of VAP decreased from 15.8 per 100 ICU patients in 2011 to 9.5 per 100 ICU patients in 2018 (Welch t test P value = 7.7e-16). The fitted linear model showed negative slope (Fig. 1). During a study period we observed substantial changes in some risk factors and no changes in others. Namely, we detected a decrease in the use of anxiolytics and antibiotics, decreased days on mechanical ventilation, and a lower rate of intestinal dysfunction, all of which were nonstationary processes with a declining trend (ADF testP > .05) (Fig. 2). However, there were no changes over time in such factors as average age, comorbidity index, level of consciousness, gender, and proportion of patients with brain trauma (Fig. 2). Conclusions: Our evidence-based IPC program was effective in lowering the prevalence of VAP and demonstrated which individual measures contributed to this improvement. By following the dynamics of known VAP risk factors over time, we found that their association with declining VAP prevalence varies significantly. Intervention-related factors (ie, use of antibiotics, anxiolytics and mechanical ventilation, and a rate of intestinal dysfunction) demonstrated significant reduction, and patient-related factors (ie, age, sex, comorbidity, etc) remained unchanged. Thus, according to the discriminative model, the intervention-related factors contributed more to the overall risk of VAP than did patient-related factors, and their reduction was associated with a decrease in VAP prevalence in our neurosurgical ICU.

Funding: None

Disclosures: None

Emily Feyes, The Ohio State University College of Veterinary Medicine; Dixie Mollenkopf, The Ohio State University College of Veterinary Medicine; Thomas Wittum, The Ohio State University College of Veterinary Medicine; Dubraska Diaz-Campos, The Ohio State University College of Veterinary Medicine; Rikki Horne, The Ohio State University College of Veterinary Medicine

Background: The Ohio State University College of Veterinary Medicine (OSU-CVM) Antimicrobial Stewardship Working Group (ASWG) uses monthly environmental surveillance to understand the effectiveness of our veterinary medical center (VMC) infection control and biosecurity protocols in reducing environmental contamination with multidrug resistant organisms. Monthly surveillance allows us to monitor trends in the recovery of these resistant organisms and address issues of concern that could impact our patients, clients, staff, and students. Methods: The OSU-CVM ASWG collects samples from >100 surfaces within the companion animal, farm animal, and equine sections of our hospital each month. Sampling has been continuous since January 2018. Samples are collected from both human–animal contact and human-only contact surfaces using Swiffer electrostatic cloths. These samples are cultured for recovery of Salmonella spp, extended-spectrum cephalosporin-resistant Enterobacteriaceae, carbapenemase-producing Enterobacteriaceae (CPE), and methicillin-resistant Staphylococcus spp. Results: The recovery of these antibiotic resistant target organisms is low in the environment of our hospital. Recovery from human-only contact surfaces (19.8%) is very similar to recovery from human–animal contact surfaces (25.5%). We commonly recover Enterobacteriaceae (E.coli, Klebsiella spp, and Enterobacter spp) that are resistant to extended-spectrum cephalosporins (496 of 2,016; 24.6%) from the VMC environment. These antibiotic-resistant indicator bacteria are expected in a veterinary hospital setting where use the of β-lactam drugs is common. Recovery of both Salmonella spp and CPE has remained very low in our hospital environment over the past 19 months: 16 of 2,016 (0.7%) for Salmonella and 15 of 2,016 (0.8%) for CPE. Discussion: The active environmental surveillance component of our antimicrobial stewardship program has allowed us to reduce the threat of nosocomial infections within our hospital and address environmental contamination issues before they become a problem. Our consistently low recovery of resistant organisms indicates the effectiveness of our existing cleaning and disinfection protocols and biosecurity measures. Due to the nature of our patient population, we do expect to find resistant organisms in the patient-contact areas of the hospital environment. However, our similar rates of resistant organisms from human-only surfaces (eg, computer keyboards, door handles, telephones, and Cubex machines) indicates a need to improve our hand hygiene practices. These data are now supporting the implementation of a new hand hygiene campaign in our veterinary hospital.

Funding: None

Disclosures: None

Background: The undisputed versatility of smart devices makes them integral to modern-day society, especially within our National Health System. Despite the benefits, there are increasing concerns regarding their contamination and the associated infection risk. Bacteria under antimicrobial selective pressure can rapidly acquire resistant mechanisms leading to the assumption; mobile phones used within clinical environments may harbor bacteria associated with a higher infection mortality rate. Using next-generation sequencing technology, we characterized the true extent of bacterial contamination on mobile devices of hospital staff to determine the level of antimicrobial resistance within Staphylococcus aureus and Enterococcus faecalis. Smart phones of 250 hospital staff and 191 control group participants were swabbed over a 6-month period. The microbiome of devices were characterized using Illinima MiSeq metabarcoding pipeline. Cultured isolates were quantified and underwent Kirby-Bauer disc diffusion. Primer version 6 and SPSS version 23 software were used to analyze the statistics. Nearly all mobile devices were contaminated with potential pathogens regardless of environment. Metabarcoding revealed far greater bacterial diversity and abundance of gram-negative bacterial contamination than did culture-based methods. In total, 198 bacterial genera of were discovered across both groups, of which 34 were unique to the hospital. Bacillus was significantly higher within the hospital group (P = .036). Differences were also detected within the hospital (high-risk vs low-risk areas, P = .048). Methicillin-resistant Staphylococcus aureus and vancomycin-resistant Enterococcusfaecalis were only isolated from hospital mobile phones (P < .001 and P = .038, respectively). Our results indicate that traditional culture-dependent swabbing methods do not provide an accurate picture of contamination. Metabarcoding reinforces the need for mobile phone infection control practices to mitigate the risks associated with the increase use of smart device technology in clinical environments. These devices are currently exposing immunocompromised patients to unknown levels of pathogenic and multidrug-resistant bacteria. Departmental differences may suggest that the mobile phone is not just an extension to its owners but to their environment and that routine decontamination should be required to prevent the undermining of hand hygiene and the transmission of pathogenic bacteria.

Funding: None

Disclosures: None

Background: Most studies on improving hand hygiene compliance (HHC) focus on clinical wards. The 5 Moments of Hand Hygiene, as stated by the WHO, are less easy to identify in an outpatient setting or procedure rooms. Therefore, observing compliance of these moments in an outpatient clinic or among healthcare workers (HCWs) in the operating room (OR), is far more difficult. Nonetheless, proper hand hygiene in the OR is of utmost importance to prevent postoperative wound infection. Objective: We developed and implement a scoring instrument with simplified moments of hand hygiene for nonsterile HCWs in the OR. Methods: All 13 hospitals of the Antibiotic Resistance Network Southwest Netherlands were asked to submit their guidelines on hand hygiene in the OR. These guidelines were, after discussion, combined into 1 guideline, describing different hand hygiene areas for different groups of nonsterile HCWs in the OR. After asking for feedback and incorporating these adjustments, the guideline was converted into a policy document. Based on this document, a paper scoring instrument was developed to observe HHC in the OR in a uniform way across all hospitals. The Erasmus University Medical Center Rotterdam, the Netherlands (Erasmus MC) acted as a pilot hospital where the implementation of the scoring instrument was studied. Results: The policy document has been approved by the infection committees of all 13 hospitals. The preliminary data of the pilot in the Erasmus MC, although still ongoing, are as follows. Hand hygiene in the OR was observed at 4 different time points. The anesthetist was observed once during 4 procedures in 3 different ORs. At the other 3 time points, the OR assistants (ie, OR nurses and circulating nurses) were observed during 4 procedures in 4 different ORs. Hand hygiene moments were easy to identify; the paper scoring instrument could be used to record observations of HHC in the OR. Conclusions: The guideline with the simplified moments of hand hygiene for nonsterile HCWs in the OR has been successfully implemented. The pilot test in the Erasmus MC already showed that, after defining the hand hygiene moments in the OR, the HHC in the OR is easier to observe and record using the scoring instrument. Moreover, the instrument has provided clarity for HCWs regarding the moments ate which they should disinfect their hands.

Funding: None

Disclosures: None

Background: Appropriate doffing of personal protective equipment (PPE) prevents healthcare worker (HCW) self-contamination and spread of pathogens. HCWs may encounter an array of PPE types (eg, gloves, gowns, masks) and designs (eg, masks with elastic ear loops vs ties) during their duties, some of which may be unfamiliar. We know little about how HCWs strategize when doffing unfamiliar PPE. As part of a larger study examining the doffing process and the risk of self-contamination, we used qualitative methods to explore factors influencing HCW PPE doffing strategies. Methods: In total, 70 HCW participants from 2 Midwestern academic hospitals were assigned to 1 of 4 doffing simulation scenarios. In the first 3 scenarios, participants were asked to doff 3 mask designs (n = 10), 2 gown designs (n = 10), or 2 glove designs (n = 10). In the fourth scenario, HCWs with different levels of training (n = 40) participated in 2 doffing simulations randomized per participant: a distraction simulation and a nondistraction simulation (using identical PPE types and designs). In all scenarios, participants were instructed to doff in their usual manner. Doffing performances were video-recorded. Participants then reviewed the recordings and took part in short semistructured interviews about their performance. Interviews were audio-recorded, transcribed, and analyzed using thematic analysis. Results: When faced with unfamiliar PPE during the simulations, participants were required to problem solve. In so doing, participants reported drawing on their day-to-day routine practices with familiar PPE to inform their doffing strategies. Aspects of routine practice identified as influential included PPE types typically worn, PPE donning and doffing order, doffing frequency, familiar PPE design cues, and experience tailoring strategies to specific patient care contexts. Participants frequently reported the desire to avoid self-contamination as driving doffing strategies and problem solving, but they also noted unique patient care demands related to their specific roles when they explained their doffing decisions. At the same time, HCWs identified lack of familiarity, lack of training, and nonintuitive design as barriers to doffing appropriately when encountering unfamiliar PPE. Conclusions: Different PPE designs may not be interchangeable, and proper doffing techniques may not be intuitive. The previous experiences of HCWs informed their strategies when doffing unfamiliar PPE. However, this practice sometimes caused them to use inappropriate doffing techniques and resulted in self-contamination. This finding has important implications for hospital policies and procedures regarding the introduction of new PPE and indicates that HCWs need training when new items are introduced.

Funding: None

Disclosures: None

Background: Improving adherence to hand hygiene (HH) of healthcare workers (HCWs) is a challenge for health institutions, and the use of technologies has been considered an important strategy within this process. Methods: To evaluate the impact of the use of alcohol-based hand rub gel (ABHR) dispensers with light sensors on the adherence to HH by HCWs. This is a prospective quasi-experimental study with comparative analysis between two 4-bed adult intensive care units at a private, tertiary-care hospital conducted over 22 weeks. An approach detection module with an LED lamp was attached to the ABHR dispenser. As a healthcare personnel approached it, the sensor was activated, and a red light turned on as a visual stimulus for HH. The color of the light changed to blue when HH was performed. All ABHR dispensers had electronic counters, but light sensors were installed only in the 4-bed dispensers of the intervention unit. Throughout the period, direct observation of adherence to HH was performed by 4 nurses who had previously been rated with an excellent coefficient of agreement (κ test = 0.951 and 0.902). At the end of the study, a perception survey was performed with the HCWs. Results: The median activation of ABHR dispensers per week was higher in the intervention unit with 1,004 (IQR, 706–1,455) versus 432 (IQR, 350–587) in the control group (P < .001). The same occurred when compared to the median activation per 1,000 patient days, with 53,069 (IQR, 47,575–67,275) versus 19,602 (IQR, 15,909–24,500) in the control group (P < .001). However, there was no evidence of difference in adherence to HH during direct observation between the 2 groups: 51.0% HH compliance (359 of 704) in the control group and 53.8% HH compliance (292 of 543) in the intervention group (P = .330). The same result emerged when we evaluated the “My Five Moments for HH” and by professional category. HCWs (N=66) answered the perception survey: 66.6% stated that lighting devices caught their attention regularly or most of the time and 59% agreed that the devices motivated HH. Conclusions: Using light sensors in ABHR dispensers can be an effective technology for improving HH. This finding was evident in the evaluation of the number of uses of the ABHR dispensers and in the HCW perceptions. Although direct observation did not show statistical evidence of difference between the groups, adherence was higher in the intervention group.

Funding: None

Disclosures: None

Johanna Blaak, W21C, University of Calgary; Rachel DiMaio, University of Calgary; Julia Kupis, University of Calgary; Ross Sweetzir, Cisco Systems; Conny Betuzzi, Alberta Children’s Hospital, Alberta Health Services; Corey Dowler, Alberta Children’s Hospital, Alberta Health Services; Krista McIntytre, Alberta Children’s Hospital, Alberta Health Services; Jaime Kaufman, University of Calgary; Greg Hallihan, University of Calgary; John Conly, Foothills Medical Centre; Joseph Vayalumkal, Alberta Childrens Hospital

Background: Interaction design offers a novel interventional strategy to enhance hand-hygiene compliance (HHC) and reduce hospital-acquired infections (HAIs) in the pediatric setting. A quality improvement initiative in collaboration with the University of Calgary and Alberta Health Services led to the implementation of a pilot project with sensor-embedded alcohol -based hand rub (ABHR) dispensers at a hematology-oncology and hematopoietic stem cell transplant unit at Alberta Children’s Hospital (ACH). Methods: Internet of things (IoT) sensors were installed in ABHR dispensers (n = 3) on the unit. Usage data were transmitted to a local server using an MQTT messaging protocol for 16 weeks. Real-time data visualization was presented on a central display next to the nursing station with 11 unique pediatric themes including dinosaurs, transportation, and Canadian animals. Data were collected with and without visualization, and frequency of use (FoU) was determined for both periods. Qualitative interviews with unit stakeholders (n = 13) were held to determine perceptions of the intervention. Results: During the first 8 weeks of the study period, the mean daily use without visualization was 47 times (SD, 14.5) versus 99 times (SD, 23.9) with visualization. When accounting for novelty, by removing the first week of data, the mean daily use was 92 (SD 19.6). The percentage increase from period 1 to period 2 was 96.6%, accounting for novelty. Qualitative interviews with stakeholders (n = 13) on the unit indicated that the intervention increased their personal awareness of hand hygiene (75%) and acted as a constant reminder to perform hand hygiene for everyone on the unit including nonclinical staff, patients, and family members (92%). Conclusions: These limited data suggest that interaction design may improve HH frequency and show promise as a tool for increased HH awareness and education. Interaction design provides a unique, innovative, and acceptable hand hygiene improvement strategy for staff, patients, and families in the pediatric inpatient setting.

Funding: None

Disclosures: None

Background: Transmission of healthcare-associated pathogens such as Clostridioides difficile and methicillin-resistant Staphylococcus aureus (MRSA) is a persistent problem in healthcare facilities despite current control measures. A better understanding of the routes of pathogen transmission is needed to develop effective control measures. Methods: We conducted an observational cohort study in an acute-care hospital to identify the timing and route of transfer of pathogens to rooms of newly admitted patients with negative MRSA nares results and no known carriage of other healthcare-associated pathogens. Rooms were thoroughly cleaned and disinfected prior to patient admission. Interactions of patients with personnel and portable equipment were observed, and serial cultures for pathogens were collected from the skin of patients and from surfaces, including those observed to come in contact with personnel and equipment. For MRSA, spa typing was used to determine relatedness of patient and environmental isolates. Results: For the 17 patients enrolled, 1 or more environmental cultures became positive for MRSA in rooms of 10 patients (59%), for C. difficile in rooms of 2 patients (12%) and for vancomycin-resistant enterococci (VRE) in rooms of 2 patients (12%). The patients interacted with an average of 2.4 personnel and 0.6 portable devices per hour of observation. As shown in Figure 1, MRSA contamination of the floor occurred rapidly as personnel entered the room. In a subset of patients, MRSA was subsequently recovered from patients’ socks and bedding and ultimately from the high-touch surfaces in the room (tray table, call button, bedrail). For several patients, MRSA isolates recovered from the floor had the same spa type as isolates subsequently recovered from other sites (eg, socks, bedding, and/or high touch surfaces). The direct transfer of healthcare-associated pathogens from personnel or equipment to high-touch surfaces was not detected. Conclusions: Healthcare-associated pathogens rapidly accumulate on the floor of patient rooms and can be transferred to the socks and bedding of patients and to high-touch surfaces. Healthcare facility floors may be an underappreciated source of pathogen dissemination not addressed by current infection control measures.

Funding: None

Disclosures: None

Background: In the United States, ∼5 million endoscopies are performed annually. Contaminated endoscopes account for more nosocomial infections than any other medical device, but the vast majority of such events go unreported. We found no reports in the literature or the FDA Manufacturer and User Facility Device Experience (MAUDE) database of the incident described below. Methods: During a colonoscopy, the operator noticed resistance while advancing a clipping wire through the channel. A balloon-tipped catheter sheath was then extruded into the colonic lumen. The sheath and endoscope were withdrawn without incident, and the procedure was completed with a different endoscope. According to equipment logs, the last time that type of balloon-tipped catheter was used occurred 20 days prior, resulting in 20 patients having potentially been exposed to an incompletely disinfected device. Interrogation of the endoscope with various inserts revealed that the presence of a retained sheath would allow passage of all types of guide wires, (snips, snares, etc), including the cleaning brush. The only device whose passage would have been prevented by a retained sheath was a vascular clipping device. A review of procedure notes and interviews of involved physicians revealed that such clippings were performed as recently as 2 days prior to the incident, thus reducing the number of potentially exposed to 2, plus the index. The county and state health departments were notified, a MAUDE report was filed, and patients were notified and offered free testing for bloodborne and enteric pathogens. Discussion: The root causes of the exposure included the absence of a closed-loop feedback for removable components (similar to an operating room sponge count), an inexperienced endoscopy technologist, and, in our opinion, a design flaw of the sheath that allows the sheath to enter the channel. Specifically, unlike other sheath brands, this brand lacks a large, irremovable warning flag that precludes channel entry (Fig. 1). Had we not been able to trace the use of each individual endoscope (n = 45) in the clinic and link each to specific patients, procedures, and reprocessing logs, we would have had no way to determine the extent of exposures. This incident, which we present as a cautionary tail to others, highlights (1) a possible equipment design flaw, (2) the importance of closed-loop feedbacks for removable components, (3) the criticality of detailed procedure notes along with granular cleaning and reprocessing logs traceable to every endoscope, and (4) the challenge of communicating risk of disease transmission to patients.

Funding: None

Disclosures: None

Background: Catheter-related bloodstream infections (CRBSIs) are one of the most frequent causes of heathcare-associated infections and an important cause of preventable death. Central venous catheters (CVC) have been considered the most worrisome origin of CRBSI. Implemented preventive measures at most hospitals and published guidelines for the prevention of vascular catheter infections have focused mostly on CVCs. However, peripherally inserted venous catheters (PIVC)–related bloodstream infections have increased in recent years and are currently among the top 10 patient safety concerns for 2019. Objective: We describe the changes in the epidemiology of catheter-related bloodstream infections among acute-care hospitals reporting at the VINCat program (Infection Control and Antimicrobial Stewardship Catalan Program) from 2008 to 2018. Methods: Data on 55 hospitals in Catalonia reporting all the episodes of CRBSI diagnosed according to standardized definitions during 2008–2018 were used for the analysis. Participating hospitals were classified into 3 groups according to size: group 1 (>500 beds), 9 hospitals; group 2 (500–200 beds), 17 hospitals; and group 3 (<200 beds), 29 hospitals. Catheters were classified in 3 categories: CVCs, PICVCs, and short peripheral venous catheters (PVCs). Rates of catheter-related bloodstream infection (CR-BSI) were obtained by adjusting the total number of episodes by 1,000 hospital stays. Simple linear regressions were performed. Values of P ≤ .05 were considered statistically significant. Results: During the study period, 8,221 nosocomial episodes of CRBSI were diagnosed among the 55 participating hospitals. In total, 37,587,967 hospital stays were counted. The CRBSI rate was 0.22 episodes per 1,000 hospital stays (group 1, 0.28; group 2, 0.15; and group 3, 0.16), following a downward trend from 2008 to 2018 from 0.28 to 0.21 per 1,000 hospital stays (P < .005). Among them, CVC-BSI showed a downward trend (from 610 annual episodes in 2008 to 312 in 2018), and PICVC and PVC showed an upward trend (from 51 and 120 annual episodes in 2008 to 130 and 312 in 2018, respectively). Annual rates of PICVCs and PVCs showed an upward trend, but CVCs showed a downward trend in 2018 (P < .05): 0.09 per 1,000 hospital stays for PICVCs; 0.07 per 1,000 hospital stays for PVCs, and 0.04 episodes per 1,000 hospital stays for CVCs (Fig. 1). Conclusions: PIVC-related bloodstream infections have increased in recent years, whereas bloodstream infections related to CVC have followed a downward trend. Our hospitals should implement preventive measures to specially address the prevention of PICVC infections.

Funding: None

Disclosures: None

Background: Antimicrobial resistance (AMR) is an increasingly critical global public health challenge. An initial step in prevention is the understanding of resistance patterns with accurate surveillance. To improve accurate surveillance and good clinical care, we developed training materials to improve the appropriate collection of clinical culture samples in Ethiopia. Methods: Specimen-collection training materials were initially developed by a team of infectious diseases physicians, a clinical microbiologist, and a monitoring and evaluation specialist using a training of trainers (ToT) platform. Revisions after each training session were provided by Ethiopian attendees including the addition of regional and culturally relevant material. The training format involved didactic presentations, interactive practice sessions with participants providing feedback and training to each other and the entire group as well as assessments of all training activities. Results: Overall, 4 rounds of training were conducted from August 2017 to September 2019. The first 2 rounds of training were conducted by The Ohio State University (OSU) staff, and Ethiopian trainers conducted the last 2 rounds. Initial training was primarily in lecture format outlining use of microbiology laboratory findings in clinical practice and steps for collecting specimens correctly. Appropriate specimen collection was demonstrated and practiced. Essential feedback from this early audience provided input for the final development of the training manual and visual aids. The ToT for master trainers took place in July 2018 and was conducted by OSU staff. In sessions held in February and August 2019, these master trainers provided training to facility trainers, who provide training to personnel directly responsible for specimen collection. In total, 144 healthcare personnel (including physicians, nurses, and laboratory staff), from 12 representative Ethiopian public and academic hospitals participated in the trainings. Participants were satisfied with the quality of the training (typically ranked >4.5 of 5.0) and strongly agreed that the objectives were clearly defined and that the information was relevant to their work. Posttraining scores increased by 23%. Conclusions: Training materials for clinical specimen collection have been developed for use in low- and middle-resource settings and with initial pilot testing and adoption in Ethiopia. The trainings were well accepted, and Ethiopian personnel were able to successfully lead the trainings and improve their knowledge and skills regarding specimen collection. The materials are being finalized in an online format for easier open access dissemination. Further studies are planned to determine the effectiveness of the trainings in improving the quality of clinical specimen submissions to the microbiology laboratory.

Funding: None

Disclosures: None

Background: In April 2019, the Georgia Department of Public Health (DPH) initiated whole-genome sequencing (WGS) on NDM-producing Enterobacteriaceae identified since January 2018. The WGS data analyzed at CDC identified related Klebsiella pneumoniae isolates with hypervirulence markers from 2 patients. Carbapenemase-producing hypervirulent K. pneumoniae (CP-hvKP) are rarely reported in the United States, but they can to cause serious, highly resistant, invasive infections. We conducted an investigation to identify cases and prevent spread. Methods: We defined a case as NDM-producing K. pneumoniae with ≥4 hypervirulence markers identified by WGS, isolated from any specimen source from a Georgia patient. We reviewed the case patient’s medical history to identify potentially affected facilities. We also performed PCR-based colonization screening and retrospective and prospective laboratory-based surveillance. Finally, we assessed facility infection control practices. Results: Overall, 7 cases from 3 case patients (A, B, and C) were identified (Fig. 1). The index case specimen was collected from case-patient A at ventilator-capable skilled nursing facility 1 (vSNF1) in May 2018. Case-patient A had been hospitalized for 1 month in India before transfer to the United States. Case-patient B’s initial isolate was collected in January 2019 on admission to vSNF2 from a critical access hospital (CAH). The CAH laboratory retrospectively identified case-patient C, who overlapped with case-patient B at the CAH in October 2018. The CAH and the vSNF2 are geographically distant from vSNF1. Case-patients B and C had no known epidemiologic links to case-patient A. Colonization screening occurred at vSNF1 in May 2018, following detection of NDM-producing K. pneumoniae from case-patient A ∼1 year before determining that the isolate carried hypervirulence markers. Among 30 residents screened, 1 had NDM and several had other carbapenemases. Subsequent screening did not identify additional NDM. Colonization screening of 112 vSNF2 residents and 13 CAH patients in 2019 did not reveal additional case patients; case-patient B resided at vSNF2 at the time of screening and remained colonized. At all 3 facilities, the DPH assessed infection control practices, issued recommendations to resolve lapses, and monitored implementation. The DPH sequenced all 27 Georgia NDM–K. pneumoniae isolates identified since January 2018; all were different multilocus sequence types from the CP-hvKP isolates, and none possessed hypervirulence markers. Conclusions: We hypothesize that CP-hvKP was imported by a patient hospitalized in India and spread to 3 Georgia facilities in 2 distinct geographic regions through indirect patient transfers. Although a response to contain NDM at vSNF1 in 2018 likely limited CP-hvKP transmission, WGS identified hvKP and established the relatedness of isolates from distinct regions, thereby directing the DPH’s additional containment activities to halt transmission.

Funding: None

Disclosures: None

Background: Addiction medicine consultation and medication-assisted treatment (MAT) have been promoted as a way to improve outcomes for patients hospitalized with injection drug use–associated endocarditis (IDU-IE). However, IDU-IE outcomes have not been evaluated in settings where these services are commonplace. Objective: In this study, we evaluated IDU-IE outcomes in a setting where involvement of addiction medicine consultants and use of MAT is well integrated into patient care. Methods: Medical records of patients hospitalized with a diagnosis of bacteremia or infective endocarditis (IE) between October 1, 2015, and December 31, 2017, at a safety-net hospital in Boston were screened for evidence of active injection drug use (IDU) within 6 months of hospitalization (as documented by providers or as supported by urine toxicology assays) for suspected or definite IE using modified Duke criteria. Patients without active IDU or IE were excluded, as were those with a diagnosis of IDU-IE over the 6 months prior to the index hospitalization. Demographic parameters, receipt of antibiotics and MAT, other clinical information, and details of rehospitalizations were recorded. Analyses of descriptive statistics were performed. Results: Of 567 subjects screened for inclusion, 47 patients met inclusion criteria. All had opiate use disorder (OUD); 41 patients (87.2%) had polysubstance abuse. Addiction medicine consultation was completed for 41 patients (87.2%). Of the 47 subjects, 23 patients (54.8%) received MAT (methadone or buprenorphine/naloxone) over their entire hospitalization, and 31 patients (73.8%) received MAT for >75% of the index admission. Moreover, 43 patients (91.5%) survived to discharge, of whom 28 (59.6%) completed antibiotic therapy. Relapsed IDU was observed in 33 patients (76.7%). Relapsed IDU trended toward significance among undomiciled patients (OR, 4.07,; 95% CI, 0.93–17.85; P = .06). Also, 24 patients (55.8%) were rehospitalized within 1 year due to infectious complications of IDU; undomiciled patients were readmitted more frequently (OR, 20.45; 95% CI, 1.09–383.99; P = .04). Completion of IDU-IE antibiotic therapy, relapse of IDU, and rehospitalization were not associated with prior AMA discharges, duration or variety of IDU, receipt of MAT during the index admission, or addiction medicine consultation. The rate of readmission due to an infectious complication of IDU within 1 year was unrelated to the proportion of hospital days where MAT was prescribed. Conclusions: In settings with high rates of addiction medicine consultation and in-hospital MAT administration, inpatient interventions targeting OUD may not necessarily be protective against morbidity and rehospitalization. Focusing on housing instability and outpatient continuation of MAT may be beneficial.

Funding: None

Disclosures: None