Group Name: Duke Center for Antimicrobial Stewardship and Infection Prevention

Background: Wastewater drains in hospital patient rooms have been identified as environmental reservoirs for multidrug-resistant organisms, and they have been linked to outbreaks of carbapenem-resistant Enterobacteriaceae (CRE). We studied the colonization of wastewater drains in a new hospital bed tower. Methods: A patient care unit in a new bed tower opened on July 18, 2020. In-room sinks were located in each hospital room opposite the patient head wall. Patients admitted to this unit underwent weekly rectal cultures to survey for carbapenemase-producing CRE. Additionally, infection preventionists performed routine surveillance of all clinical cultures for CRE. Cultures were performed from all patient room sinks in this unit monthly beginning September 14, 2020. Samples were obtained from the drain cover, handles, and top of bowl using sponges soaked in neutralizing buffer and processed using the stomacher technique. The tail-pipe was sampled using a flocked mini-tip swab soaked in neutralizing buffer; the P-trap water was sampled with sterile tubing attached to a 50-mL syringe. All samples were plated on HARDYCHROM-ESBL and KPC Colorex media and were incubated at 37°C for 24 hours. Results: The first identified CRE-positive patient was admitted to the new unit on December 4, 2020; urine culture obtained at the time of admission grew KPC–producing Klebsiella pneumoniae (KPC-KP). The sink in this patient’s room had been sampled 3 prior times (most recently on November 9, 2020) and was negative for CRE. On December 7, 2020, KPC-KP was found on the drain cover (6,750 colony-forming units, CFU) and in the sink’s P-trap (1,840 CFU) of the index patient’s room during routine sink surveillance. Additional samples from other room surfaces were taken on December 9, 2020, and KPC-KP was recovered from the computer keyboard (452 CFU) and patient bedrails (880 CFU). The patient was discharged from this room December 13, 2020, and the room underwent enhanced terminal room cleaning including UV-C light. On the next routine sink sampling on January 4, 2021, KPC-KP was recovered again in the index room sink P-trap (9,800 CFU) but at no additional sites. MLST was performed, and all isolates were ST-258. Conclusions: In a new bed tower with no prior evidence of CRE-positive patients, the first identified case of a CRE (KPC-KP) in a patient resulted in widespread environmental contamination of the room after only 3 days of hospitalization and contamination of the in-room sink drain that persisted after 1 month. Given the ease with which CRE colonizes wastewater drains, new strategies are needed to mitigate drain colonization and to prevent CRE transmission to subsequent patients.

Funding: No

Disclosures: None

Background: Care bundles comprise evidence-based practices and interventions that are easily and consistently implemented while improving patient outcomes. As patient acuity and task overload continue to increase, infection prevention bundle and process measure compliance and data collection may become a lower priority for registered nurses (RNs). In early 2019, a certified nursing assistant (CNA) began full-time quality liaison work on a 53-bed inpatient adult oncology unit at UNC Medical Center to provide targeted compliance data collection and to correct deficits in real time when possible and within the appropriate scope of practice. Methods: The quality liaison CNA is highly motivated, with a relevant clinical background and effective communication skills. After conducting a gap analysis, the unit developed specific responsibilities for several areas of quality improvement, including infection prevention. In addition to rounding on all patients daily, the quality liaison (1) performs direct patient care tasks like Foley catheter care, (2) conducts patient education on topics such as chlorhexidine gluconate treatments, (3) performs all relevant process measure audits, and (4) easily relays missed or needed care to RNs with a door sign created as part of this initiative. High-risk findings, such as a loose central-line dressing, prompt immediate communication to the RN, with follow-up and escalation when necessary. Results: Patients and staff received the quality liaison well, and the increased attention to care bundle components and auditing ensured consistent, evidence-based care along with accurate and reliable data collection. Compared to the previous calendar year, the number of central-line audits on the unit increased by >1,400 by the end of 2019. Patient outcomes improved, and during 1 fiscal year, the unit achieved rate reductions between 40% and 55% for central-line–associated bloodstream infections, catheter-associated urinary tract infections, and healthcare-associated C. difficile infections. Staffing and logistical challenges imposed by the COVID-19 global pandemic have hampered this work because the quality liaison was redeployed to direct patient care intermittently. Correspondingly, from July to October 2020, the same infection rates increased between 30% and 353%. Conclusions: Having a designated quality liaison is an effective means to achieving quality improvements while remaining an integral member of the patient care team. As staffing has improved on this unit, the quality liaison has refocused efforts, and infection rates are beginning to improve. Given the success of the quality liaison role in improving quality outcomes on this unit, the hospital is exploring expansion of this model to additional units.

Funding: No

Disclosures: None

Group Name: Watussi Study Group

Background: A positive safety climate is an important precursor of safe care outcomes. However, only limited evidence supports the association of low surgical-site infection (SSI) rates and positive safety climate. We investigated the role that perceptions of SSI prevention measures play for both safety climate level and strength as a subjective norm, that is, the social pressure perceived to perform the prevention measures, commitment to observe SSI prevention measures despite other situational pressures, and the level of knowledge about the prevention measures. Methods: The safety climate scale of the Safety Attitudes Questionnaire and 3 scales assessing subjective norm, commitment, and knowledge were used. All items were translated and retranslated from German to French and to Italian. All translated scales were pretested for understandability. Operating room (OR) personnel in 54 Swiss acute-care hospitals were surveyed, resulting in 2,769 analyzed responses with data aggregated on the hospital level. Two regression analyses were conducted: one using the percentage of positive responses per hospital as a safety climate level indicator, and another using the standard deviation of the safety climate ratings per hospital as a safety climate strength indicator. As independent variables, the hospital means of subjective norm, commitment, and knowledge were investigated and appropriately adjusted for number of respondents and sample composition. Results: The sample consisted of 1,495 nurses (54%) and 1,101 physicians (40%). Commitment and subjective norm were significant predictors (p < 0.001 and p < .05, respectively) of safety climate level, in the expected positive direction, but KNOW was not (R2, adjusted: 0.48); for safety climate strength, only COM was significant p < 0.001 (R2, adjusted: 0.27). Conclusions: The extent to which OR personnel were committed to perform the measures, such as timely administration of antibiotics, was associated with their safety climate rating level and strength. Thus, the rather general safety climate assessments are related to more specific safety behaviors necessary to achieve good outcomes such as low infection rates. Subjective norm was related to safety climate level only, indicating that in work environments with a good safety climate, the perceived social pressure to adhere to infection prevention measures may be higher. Knowledge about SSI prevention had no significant impact on safety climate, pointing to future research regarding the role of education in implementing prevention measures. Investigating how attitudes and knowledge about measures to prevent specific patient safety outcomes furthers our understanding of the role of safety climate in patient safety improvement.

Funding: No

Disclosures: None

Background: Infection following ambulance transport, or medical-transport–associated infection (MTAI), is understudied. Although medical-transport vehicles are routinely contaminated with methicillin-resistant Staphylococcus aureus (MRSA) and/or vancomycin-resistant Enterococcus (VRE), an association between vehicle exposure and disease development has not been identified. We estimated the relative risk (RR) of developing MRSA or VRE colonization or infection within 30 days of ambulance exposure. Methods: We performed a retrospective cohort study of patients with a principal diagnosis of chest pain presenting to our emergency department (ED) from January 1, 2016, through December 31, 2019. To control for confounding by healthcare exposure, patients were included if they presented from and were discharged to nonhealthcare locations without being admitted to the hospital. Encounters were stratified by whether the patient arrived at the ED via ambulance or private vehicle. Propensity scores were calculated using multivariable logistic regression with ambulance exposure as the dependent variable. Age, smoking status, history of myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, diabetes mellitus, and chronic kidney disease were included as covariates because their standard differences were >0.10. Propensity score matching was performed in a 2:1 ratio, but not all exposed patients received 2 matching unexposed patients due to a low sample size. A multivariable logistic regression was performed on the matched cohort to estimate the RR of newly diagnosed MRSA or VRE infection or colonization within 30 days following ambulance exposure. Results: In total, 321,229 patients had ED encounters during the study period. After applying inclusion criteria and propensity score-matching there were 11,324 patients: 3,903 in the ambulance group and 7,421 in the unexposed group. Moreover, 12 patients (0.11%) had the outcome of interest, including 9 (0.08%) with MRSA and 3 (0.03%) with VRE. The 30-day prevalences of MRSA and VRE were larger in the ambulance group than in the unexposed group: 8 (0.20%) and 4 (0.05%), respectively (P = .02). Patients who presented to the ED via ambulance were almost 4 times more likely to have MRSA or VRE within 30 days of their encounter (RR, 3.72; 95% CI, 1.09–12.71; P = .04). The RRs for MRSA and VRE alone were 3.33 (95% CI, 0.79–13.94; P = .10) and 4.14 (95% CI, 0.37–46; P = .25), respectively. Conclusions: To our knowledge, our cohort study is the first to demonstrate an association between ambulance exposure and the development of disease. These results represent the first step in evaluating MTAI burden to eventually develop targeted interventions with the purpose of reducing it.

Funding: No

Disclosures: None

Background:Pseudomonas aeruginosa uncommonly causes illness in neonatal intensive care units (NICU). A cluster of 4 infections was appreciated over 6 weeks in our inborn–delivery NICU, prompting an investigation. Methods: Upon recognition of a cluster of infections, we retrospectively audited all cultures positive for P. aeruginosa from all sites (sterile and nonsterile) over the prior year in the index NICU (NICU 1, inborn) and for comparison in the adjacent NICU (NICU 2, larger, outborn–surgical). We performed multilocus sequence testing (MLST) of available clinical isolates to identify clonality. We initiated quarterly prospective surveillance of P. aeruginosa colonization of infants through nares, perirectal swabs, and tracheal aspirates of intubated infants or oropharyngeal swabs of nonintubated infants. We also swabbed incubators, ventilatory equipment, and sinks for selective P. aeruginosa culturing. Results: We identified 7 invasive P. aeruginosa infections in the inborn NICU (5 bloodstream and 2 pneumonia) over an 11-month period (Figure 1). Over the same period, there were no P. aeruginosa bloodstream infections in the adjacent NICU. Affected neonates were high risk: gestational age ranged from 22 weeks and 4 days to 26 weeks and 3 days; day of life at infection ranged from 6 to 37; 6 infants were on a jet ventilator; and all infants were receiving enteral nutrition (6 of 7 with donor human milk and 7 of 7 with expressed mother’s milk). Two infants died from their infection, and 5 infants survived to hospital discharge. All 7 isolates were pansusceptible to routine antimicrobials. MLST of the first 4 available isolates demonstrated 4 different sequence types; however, the first 2 were from the same clonal complex, indicating relatedness (Figure 1). For environmental samples, 8 obtained 8 cultures (swabs) of incubators and ventilatory equipment, and 24 cultures of faucets and drains of all sinks. Only sink cultures were positive, yielding 3 isolates identified as P. aeruginosa and 4 isolates identified as P. aeruginosa–like. Whole-genome sequencing (WGS) is underway to identify relatedness to the clinical isolates. We initiated quarterly infant surveillance by swab culture for P. aeruginosa nasal colonization then escalated to perirectal and oropharyngeal swab or tracheal aspirate cultures (intubated infants) in subsequent quarters. We did not detect any infants colonized with P. aeruginosa. Conclusions: We identified a cluster of P. aeruginosa in high-risk neonates with no point source. Molecular typing indicated a multiclonal cluster. This finding poses a management dilemma. A colonized water system is suspected and WGS of environmental samples is underway.

Funding: No

Disclosures: None

Figure 1.

Background: Healthcare-associated infections (HAIs) affect patient health and are tracked closely by infection prevention. Patients in a pediatric intensive care unit (PICU) acquired viral respiratory infections had longer use of respiratory support. We sought to determine the types of viral respiratory HAIs (VR-HAIs) acquired in the PICU and the characteristics of those affected. Methods: CHOC Children’s Hospital is a 334-bed tertiary-care center. Charts were reviewed on patients with VR-HAIs from fiscal years (FY) 2005–2020. High-risk VR-HAI (HR-VR-HAI) were influenza A and B, respiratory syncytial virus (RSV), adenovirus, parainfluenza, and human metapneumovirus (hMPV, added in FY 2014). Patients in the PICU, cardiovascular ICU (CVICU), and oncology ICU (OICU) with HR-VR-HAIs were reviewed. Patients were categorized according to underlying pathology, immunosuppression, and isolation prior to HR-VR-HAI. Increased respiratory support was defined as any increase from a patient’s baseline support ±24 hours of viral diagnosis: increase in oxygen flow or transition from nasal cannula to high-flow nasal cannula or ventilator support. Antibiotic escalation, defined as initiation of antibiotic therapy for ≥2 days ±24 hours of viral diagnosis or broadening the spectrum of antimicrobials for ≥2 days ±24 hours of viral diagnosis. Results: During FY 2005–2020, there were 204 VR-HAIs: 143 HR-VR-HAIs (70%), of which 39 (27.2%) occurred in ICUs (Figure 1). Most of the HR-VR-HAIs were RSV, parainfluenza, and hMPV (Figure 2). Of 39 patients, 10 (25.6%) had underlying oncologic conditions, 9 of whom were immunosuppressed. Of 39 patients, 16 (41%) had structural cardiac disease, 4 (10.3%) had pulmonary disease, 5 (12.8%) had neurologic disease, and the remaining 4 (10.3%) had other comorbidities. Of 39 patients, 12 (31%) required an increase in respiratory support and 13 (33%) had escalation of antibiotics. Of 39 HR-VR-HAI patients, 2 died within 2 weeks of acquisition. Conclusions: HR-VR-HAIs are uncommon in ICUs. RSV, parainfluenza, and hMPV are the most common, and 1 of 3 of patients required escalation in respiratory support and/or escalation in antibiotics. All patients had underlying comorbidities. In our series, there were 2 deaths within 2 weeks of infection.

Funding: No

Disclosures: None

Figure 1.

Figure 2.

Background: Healthcare personnel (HCP) in crowded and resource-poor countries (eg Bangladesh), might be at risk of exposure to and transmission of respiratory illnesses to coworkers, patients, and caregivers. The infection control practices in public hospitals are inadequate in Bangladesh. We estimated the incidence of respiratory illness episodes among HCP, and proportion of HCP who worked during respiratory illnesses, including influenza virus infection, at 2 tertiary-care public hospitals in Bangladesh. Methods: From May 2008 to February 2016, HCP (defined as physicians, nurses, interns, patient care assistant, cleaners, and administrative staff working in adult and pediatric medicine wards) were asked to self-report to study physicians when they experienced new onset of cough, rhinorrhea, difficulty breathing, or fever during the April–September influenza epidemic period each year. Study physicians followed HCP throughout their respiratory illness episodes and recorded respiratory symptoms, onset dates, duration of illness, and days of presenteeism and absenteeism during illness. Nasopharyngeal and oropharyngeal swabs were collected after informed written consent and were tested for influenza by rRT-PCR. We used hospital records to enumerate total HCP working in the study wards during influenza season and multiplied by 6-months follow-up per year to calculate person-time contribution for estimating respiratory illness incidence. Results: HCP self-reported 107 episodes of respiratory illness during 656 person years of follow-up, for an estimated incidence of 16.3 per 100 person years (95% CI, 13–20). Of 107 episodes, 33 (31%) included fever and cough. The mean illness length was 3.9 days (SD, ±1.8). HCP worked an average of 3.4 days (SD, ±1.4) while ill. HCP missed work for a median of 1 day (IQR, 1–2) during 29 (27%) of 107 illness episodes. HCP consented to collect swabs during 56 (52%) episodes, and among them 8 (14%) of 56 tested positive for influenza (flu-A, n = 5; flu-B, n = 3). Also, 63% of HCP with influenza reported fever and cough. HCP experiencing either respiratory illness or influenza worked for similar periods of days while ill: mean, 4 (SD, ±2.2) versus mean, 3.3 (SD, ±1.4) (P = .257). HCP worked during 105 (98%) of 107 respiratory illness and 7 (88%) of 8 influenza episodes. Conclusions: Most HCP in Bangladesh, including those with influenza, worked during respiratory illnesses. The potential value of stay-at-home policies, compensation for sick days, and influenza vaccination in reducing HCP-associated respiratory pathogen transmission could be assessed in Bangladesh and similar settings.

Funding: No

Disclosures: None

Background: Despite significant morbidity and mortality, estimates of the burden of healthcare-associated viral respiratory infections (HA-VRI) for noninfluenza infections are limited. Of the studies assessing the burden of respiratory syncytial virus (RSV), cases are typically classified as healthcare associated if a positive test result occurred after the first 3 days following admission, which may miss healthcare exposures prior to admission. Utilizing an expanded definition of healthcare-associated RSV, we assessed the estimates of disease prevalence. Methods: This study included laboratory-confirmed cases of RSV in adult and pediatric patients admitted to acute-care hospitals in a catchment area of 8 counties in Tennessee identified between October 1, 2016, and April 30, 2019. Surveillance information was abstracted from hospital and state laboratory databases, hospital infection control databases, reportable condition databases, and electronic health records as a part of the Influenza Hospitalization Surveillance Network by the Emerging Infections Program. Cases were defined as healthcare-associated RSV if laboratory confirmation of infection occurred (1) on or after hospital day 4 (ie, “traditional definition”) or (2) between hospital day 0 and 3 in patients transferred from a chronic care facility or with a recent discharge from another acute-care facility in the 7 days preceding the current index admission (ie, “enhanced definition”). The proportion of laboratory-confirmed RSV designated as HA-VRI using both the traditional definition as well as with the added enhanced definition were compared. Results: We identified 900 cases of RSV in hospitalized patients over the study period. Using the traditional definition for HA-VRI, only 41 (4.6%) were deemed healthcare associated. Adding the cases identified using the enhanced definition, an additional 12 cases (1.3%) were noted in patients transferred from a chronic care facility for the current acute-care admission and 17 cases (1.9%) were noted in patients with a prior acute-care admission in the preceding 7 days. Using our expanded definition, the total proportion of healthcare-associated RSV in this cohort was 69 (7.7%) of 900 compared to 13.1% of cases for influenza (Figure 1). Although the burden of HA-VRI due to RSV was less than that of influenza, when stratified by age, the rate increased to 11.7% for those aged 50–64 years and to 10.1% for those aged ≥65 years (Figure 2). Conclusions: RSV infections are often not included in estimates of HA-VRI, but the proportion of cases that are healthcare associated are substantial. Typical surveillance methods likely underestimate the burden of disease related to RSV, especially for those aged ≥50 years.

Funding: No

Disclosures: None

Figure 1.

Figure 2.

Background: Research testing human study participants regarding the effectiveness of face masks in preventing influenza transfer or transmission is limited. In this pilot study, we investigated the following question: In influenza-positive veterans, what is the effect of face-mask wearing in comparison to not wearing a face mask on influenza transfer to bedside tables measured for 2 hours per condition over a 10-week period during the 2019–2020 influenza season Methods: Influenza-positive veterans with influenza symptom onset ≤ 120 hours admitted to the Salem Veterans Affairs Medical Center were recruited to participate in this study. Exclusion criteria included critical illness requiring an oxygen mask or intubation. The Precept® FluidGard® 160 Procedure Mask 15300, Precept Medical Products, Inc., Arden, NC was worn by all participants during the two-hour intervention period. Surface swabs were used to measure the presence of influenza on bedside tables. CDC/NIOSH tested for influenza A and B from surface samples and facemasks using real-time polymerase chain reaction (PCR) assay (TaqMan ThermoFisher Scientific). Demographic information was collected (Table 1). A study questionnaire collected qualitative data on tolerability and feasibility of wearing a facemask when hospitalized with influenza. Institutional Review Board approval was granted. Results: From January 2, 2020, to March 11, 2020, 8 participants completed the study. Mean age was 67 years, all were male. Of these 8 participants, 6 had influenza A and 2 had influenza B. Half were diabetic; all received oseltamivir. Relative room humidity ranged from 15.6% to 39.8%. Neither influenza A nor B was detected by qPCR on bedside tables for any of the 8 participants under either face-mask–wearing condition. All participants reported that wearing the face mask was easy or very easy; of these, 5 reported experiencing warmth from the mask. Also, 50% of participants selected 2 hours as the time they could tolerate wearing a mask; the other 25% specified they could wear the face mask for 3 hours or 5 hours or more, respectively. Conclusions: In this pilot study, we demonstrated that wearing face masks is a tolerable infection control practice for providing source control for inpatients with influenza and will guide future research. Because a major limitation was the small size of the study, associated with lack of viral capture, a larger study is planned. Using face masks for source control among inpatients with influenza and other respiratory virus infections should be considered a standard infection control practice.

Funding: No

Disclosures: None

Table 1.

Table 2.

Table 3.

Background: Deep and organ-space surgical site infections (SSIs) are serious complications of coronary artery bypass graft (CABG) procedures. It is unclear whether the use of bilateral versus single internal mammary artery (BIMA vs SIMA) and surgical approach to internal mammary artery (IMA) harvest (pedicled vs skeletonized) are independent risk factors for SSI. The use of BIMA grafting redirects blood flow away from the sternum to the heart and may increase SSI risk due to lower tissue perfusion. A skeletonized approach to graft harvest, wherein the IMA is dissected free of surrounding tissue to preserve collateral sternal blood flow, may decrease SSI risk as compared to a pedicled approach in which the IMA is mobilized within a tissue pedicle. We describe the incidence and potential risk factors for post-CABG SSI in an academic tertiary-care center performing ~500 IMA procedures annually. Methods: Data were abstracted on adult patients who underwent a CABG procedure using at least 1 IMA graft between July 2017 and June 2020. Additional data on potential risk factors for SSI were obtained electronically from hospital data marts and the Division of Cardiac Surgery database, including demographics, comorbidities, number of arterial grafts, surgical approach, surgeon, and discharge location. Using standard NHSN definitions, infection control practitioners identified post-CABG deep and organ-space SSIs. Patient and procedure characteristics were evaluated as potential risk factors for deep and organ-space SSI using the Fisher exact test. Results: Of 1,591 CABG procedures performed during the study period, 1,244 (78.2%) were performed using a SIMA technique and 347 (21.8%) were performed using a BIMA technique. The overall post-CABG SSI incidence was 1.2 per 100 procedures, with 1.0 SSIs per 100 SIMA procedures and 1.7 SSIs per 100 BIMA procedures. Table 1 demonstrates an increase over time in proportion of CABG procedures performed using SIMA and skeletonized IMA grafts. We also observed a decrease in overall SSI incidence over this period. See Table 2 for univariate predictors of post-CABG SSI. Conclusions: Female sex, BMI ≥40, age ≥75 years, diabetes, and discharge to a rehabilitation setting were associated with development of post-CABG SSI. Although the overall incidence of deep and organ-space SSI in our cohort was very low, making it difficult to draw conclusions about potentially modifiable risk factors, an increase in the use of SIMA and skeletonized grafts appears to be accompanied by a decrease in SSI incidence. More data from our institution and others are needed to determine the significance of this trend.

Funding: No

Disclosures: None

Table 1.

Table 2.

Background: Left-ventricular-assist device (LVAD)–related infections occur in 20%–40% of LVAD recipients and may result in up to 10% of LVAD-related deaths. Optimal surgical infection prophylaxis for LVAD implantation is not well defined. Our institution historically used a 4-drug surgical infection prophylaxis regimen of fluconazole, ciprofloxacin, rifampin, and vancomycin as recommended by the device manufacturer. In January 2020, a 2-drug surgical infection prophylaxis regimen of vancomycin and cefazolin was implemented to reduce broad-spectrum antibiotic use while preserving gram-positive coverage. The primary objective of this study was to compare LVAD-associated infection rates before and after changing surgical infection prophylaxis. Methods: A retrospective review of patients who underwent LVAD implantation between January 2018 and January 1, 2021, was performed. Definitions of LVAD-associated infections and non-LVAD infections were based on the International Society for Heart and Lung Transplantation guidelines. Infection rates at 2 weeks and 30 days after implantation and 30-day mortality were compared between the 4-drug surgical infection prophylaxis regimen (January 2018–December 2019) and the 2-drug regimen (January 2020–January 2021). Additional data collected included demographics, cause of cardiomyopathy, type of infection, and causative organism. Results: In total, 51 patients were in the 4-drug surgical infection prophylaxis group and 23 patients were in the 2-drug surgical infection prophylaxis group. Baseline characteristics between the groups were similar. The cause of cardiomyopathy in both groups was predominantly nonischemic (67% vs 70%, = .81), and most patients received a Heartmate III device (84% vs 100%, P = .06). There was no statistical difference between infection rates in the 4-drug and 2-drug prophylaxis groups at 2 weeks or 30 days (Table 1). The 30-day mortality rate was 4% in the 4-drug group versus 13% in the 2-drug group (P = .17). No deaths were due to infections. Gram-negative and fungal LVAD–associated infections were rare: 4% versus 4% (P = .99) for gram-negative infections and 2% versus 0% (P = .99) for fungal infections. The most commonly isolated organisms were Staphylococcus aureus and coagulase-negative Staphylococcus in both groups. Pneumonia was the most common infection in both groups (Figure 1). Conclusions: We did not observe a significant increase in infection or mortality with narrowing of perioperative antibiotics. However, these results should be interpreted cautiously given the small sample size. Larger studies are needed to confirm these findings.

Funding: No

Disclosures: None

Table 1.

Figure 1.

Background: Healthcare-associated infections are important because they constitute a public health problem due to the increase in morbidity and mortality that they produce in hospitalized patients, increased hospitalization costs due to prolonged stay, expensive antibiotic treatments and surgical reinterventions, not counting the social costs due to loss of wages and production, among others. Methods: We report the specific prevalence of healthcare-associated infections (HCAIs) in Edgardo Rebagliati Martin National Hospital, Perú, in 2020. We performed a descriptive cross-sectional study from July 27 to July 31, 2020. The medical records of hospitalized patients were reviewed according to the inclusion criteria. STATA software was used for descriptive statistical analyses. Results: In total, 1,217 hospitalized patients were included in the study. The prevalence of HCAI was 12.2% (149 patients). The prevalence of HCAI in areas where patients with the diagnosis of COVID-19 were hospitalized was higher (8.1%) than in common hospitalization areas (4.1%). Men represented 92% of the total number of patients with HCAIs. The most frequent infections were clinically defined pneumonia (30.9%) and bloodstream infections (20.1%). The most frequently isolated microorganism was Pseudomonas aeruginosa. Conclusions: The prevalence of HCAI was 12.2%. The most frequent HCAIs were pneumonia and bloodstream infection.

Funding: No

Disclosures: None

Background: NDM/OXA-23 carbapenemase-producing Acinetobacter baumannii isolates have been reported worldwide, but rarely in the United States. A California acute-care hospital (ACH) A identified 3 patients with pan-nonsusceptible A. baumannii during May–June 2020, prompting a public health investigation to prevent further transmission among the regional healthcare network. Methods: A clinical isolate was defined as NDM/OXA-23–producing A. baumannii from a patient at ACH A or B, or an epidemiologically linked patient identified through colonization screening during May 2020–January 2021. ACHs A and B are sentinel sites for carbapenem-resistant A. baumannii surveillance through the Antibiotic Resistance Laboratory Network (AR Lab Network), where isolates are tested for carbapenemase genes. The California Department of Public Health with 3 local health departments conducted an epidemiological investigation, contact tracing, colonization screening, and whole-genome sequencing (WGS). Results: In total, 11 cases were identified during May 2020–January 2021, including 3 cases at ACH A during May–June 2020, and 8 additional cases during November 2020–January 2021: 5 at ACH A, 1 at ACH B, and 2 at skilled nursing facility (SNF) A. Isolates from ACHs A and B were identified through testing at the AR Lab Network. Of the 11 patients (including the index patient), 4 had exposure at SNF A, where 2 cases were identified through colonization screening. Screening conducted at ACH A and 5 other long-term care facilities (LTCFs) identified no additional cases. WGS results for the first 8 cases identified showed 2–13 single-nucleotide polymorphism differences. Antibiotic resistance genes for all isolates sequenced included NDM-1 and OXA-23. On-site assessments related to a COVID-19 outbreak conducted at ACH A identified infection control gaps. Conclusions: Hospital participation in public health laboratory surveillance allows early detection of novel multidrug-resistant organisms (MDROs), which enabled outbreak identification and public health response. A high COVID-19 burden and related changes in infection control practices have been associated with MDRO transmission elsewhere in California. This factor might have contributed to spread at ACH A and hampered earlier screening efforts at SNF A, likely leading to undetected transmission. Extensive movement of positive patients among a regional healthcare network including at least 6 ACHs and 7 LTCFs likely contributed to the prolonged duration of this outbreak. This investigation highlights the importance of enhanced novel MDRO surveillance strategies coupled with strong infection prevention and control practices as important factors in identifying outbreaks and preventing further transmission in regional networks.

Funding: No

Disclosures: None

Background: Candidemia is the fourth most common bloodstream infection in hospitalized patients in the United States, with an attributable mortality rate between 30% and 50%. Understanding the epidemiology of candidemia is critical due to its prevalence and association with extended hospital stays, high treatment cost, and significant morbidity and mortality. In 2019 the Connecticut Department of Public Health deemed candidemia a mandatory reportable condition and began state-wide surveillance in conjunction with the CDC’s Emerging Infections Program (EIP). Previously, the EIP had conducted population-based statewide surveillance of candidemia in Connecticut from 1998 to 2000, allowing an opportunity to assess how the epidemiology of candidemia has evolved. The goal of this study is to compare state-wide Connecticut EIP candidemia data from 2 periods (1998–2000 and 2019) to identify trends in infections and incidence, providing insight for potential improvements to current prevention measures and treatments. Methods: The sample population included all Connecticut residents aged ≥20 who tested positive for a candidemia infection during 1998–2000 and 2019. Patients who had positive blood cultures for Candida spp but were < 20 years old or were not Connecticut residents were excluded. Connecticut EIP candidemia case report forms from each time period were compared and matching fields were chosen as variables for univariate analysis to search for statistically significant differences. Selected variables include: Candida species present in blood culture, patient demographics, previous exposures to healthcare settings, length of stay, presence of central venous catheter (CVC), and location of the patient at diagnosis (community vs. hospital onset). De-identified patient-level information was provided by the EIP. Results: In total, 381 candidemia episodes from 1998–2000 were compared to 247 episodes in 2019. The proportion of C. albicans species in 1998–2000 was 49.9% and declined to 40.5% of cases in 2019 (P = .02). Outcomes improved as well, with 65.2% of patients in 2019 having survived compared to 51.4% in 1998–2000 (P = .001). Other findings indicate that patients with candidemia in 2019 were less likely to have a central venous catheter, less likely to have undergone a recent surgery, and were more likely to have community-onset infection (all p < 0.05). Conclusions: The epidemiology of candidemia has changed over the past 20 years, with significant improvements in patient survival and a shift toward community-onset infections and non-Albicans spp. These findings have important implications in designing prevention strategies and optimizing candidemia management, particularly in the community setting where increased intravenous drug use and the availability of home healthcare may be important factors.

Funding: No

Disclosures: None

Table 1.

Background: Mechanical ventilation is commonly seen in critical ill patients. The vulnerability of these patients is high, and a wide range of associated conditions can stem from this intervention. To objectively identify nosocomial respiratory conditions and provide conformed surveillance definitions of these events, the Centers for Disease Control and Prevention (CDC) established the ventilator-associated event (VAE) criteria. They denote 3 categories of increasing progression in mechanically ventilated patients from a ventilator-associated condition (VAC) to an infection-related ventilator-associated complication (IVAC) and finally to a possible ventilator-associated pneumonia (PVAP). Manipulation of ventilator settings, such as starting on higher values to not trigger VAC criteria, has been criticized by some experts as not only ‘gaming the system,’ but potentially harming patients. In October 2018, our institution began a baseline of 8 cm H2O as the starting positive end-expiratory pressure (PEEP) protocol for mechanical ventilation but exempting neurosurgical patients. We sought to determine whether an 8 PEEP protocol is an effective strategy for reducing VAEs in our institution. Methods: We retrospectively examined patient data at our institution from January 2014 through February 2020. VAEs were separated by VAC only and IVAC positive (+), which are a combination of IVACs and PVAPs. Using the days between VAEs, a daily event probability can be calculated based on the geometric distribution. Furthermore, as VAEs occur, the likelihood of the event can be assessed as expected or unexpected using a strict probability limit of 0.99865 to reduce type 1 errors. Results: In total, 307 patients were identified in our hospital’s VAE surveillance. Of those, 180 met CDC-defined VAC-only criteria, and 127 patients met IVAC+ definitions. After implementation of an 8-PEEP protocol, the daily event probability for VACs decreased from 0.083 to 0.047. The last event occurred 162 days after the previous VAC, which was unexpected, because the probability of occurrence extended beyond the probability limit. With regard to IVAC+ events, the daily event probability decreased from 0.057 to 0.039 without significant reduction in the IVAC+ rate. Conclusions: Although a change in the VAC-only rate occurred, signified by a longer time between events, it took more than a year to achieve in our institution. Additionally, we did not see a reduction in the IVAC+ rate. These findings suggest that an 8-PEEP protocol may be able to reduce VAEs due to noninfectious etiologies, such as congestive heart failure and atelectasis.

Funding: No

Disclosure: None

Figure 1.

Figure 2.

Background: Mechanical ventilation is a lifesaving therapy for critically ill patients. Hospitals perform surveillance for the NHSN for ventilator-associated events (VAE) by monitoring mechanically ventilated patients for metrics that are generally thought to be objective and preventable and that lead to poor patient outcomes. The VAE definition is met in a stepwise manner; initially, a ventilator-associated condition (VAC) is triggered with an increase in positive end-expiratory pressure (PEEP, >3 cm H2O) or fraction of inspired oxygen (FIO2, 0.20 or 20 points) after a period of stability or improvement on the ventilator. We believe that many reported VAEs could be avoided by provider and respiratory therapy attention to “knobmanship.” We define knobmanship as knowledge of the VAE definition and trigger points combined with appropriate clinical care for mechanically ventilated patients while avoiding unnecessary triggering of the VAE definition by avoiding small unneeded changes in PEEP or FIO2. Methods: We performed a chart review of 283 patients who had a reported VAE to the NHSN between January 1, 2019, and December 31, 2020. We collected data including type of VAE, VAE triggering criteria, and clinical course. Results: Of the 283 VAEs, 59 were triggered by a PEEP increase from 5 to 8 with stable or decreasing FIO2. Of the 59 VAEs, 33 were VACs, 18 were infection-related ventilator- associated complications (IVACs), and 8 were possible ventilator-associated pneumonia (PVAP). Most of these transient changes in PEEP were deemed clinically unnecessary. A 21% reduction of VAEs reported to the NSHN over the 2-year review period could have been avoided by knobmanship. Conclusions: The VAE definition may often be triggered by provider bias to the ventilator settings rather than what the patient’s clinical-condition requires. Attention to knobmanship may result in substantial decrease in reported VAE.

Funding: No

Disclosures: None

Figure 1.

Background: We evaluated the impact of a comprehensive SARS-CoV-2 (COVID-19) infection prevention (IP) bundle on rates of non–COVID-19 healthcare-acquired respiratory viral infection (HA-RVI). Methods: We performed a retrospective analysis of prospectively collected respiratory viral data using an infection prevention database from April 2017 to January 2021. We defined HA-RVI as identification of a respiratory virus via nasal or nasopharyngeal swabs collected on or after hospital day 7 for COVID-19 and non–COVID-19 RVI. We compared incident rate ratios (IRRs) of HA-RVI for each of the 3 years (April 2017 to March 2020) prior to and 10 months (April 2020 to January 2021) following full implementation of a comprehensive COVID-19 IP bundle at Duke University Health System. The COVID-19 IP bundle consists of the following elements: universal masking; eye protection; employee, patient, and visitor symptom screening; contact tracing; admission and preprocedure testing; visitor restrictions; discouraging presenteeism; population density control and/or physical distancing; and ongoing attention to basic horizontal IP strategies including hand hygiene, PPE compliance, and environmental cleaning. Results: During the study period, we identified 715 HA-RVIs over 1,899,700 inpatient days, for an overall incidence rate of 0.38 HA-RVI per 1,000 inpatient days. The HA-RVI IRR was significantly higher during each of the 3 years prior to implementing the COVID-19 IP bundle (Table 1). The incidence rate of HA-RVI decreased by 60% after bundle implementation. COVID-19 became the dominant HA-RVI, and no cases of HA-influenza occurred in the postimplementation period (Figure 1). Conclusions: Implementation of a comprehensive COVID-19 IP bundle likely contributed to a reduction in HA-RVI for hospitalized patients in our healthcare system. Augmenting traditional IP interventions in place during the annual respiratory virus season may be a future strategy to reduce rates of HA-RVI for inpatients.

Table 1.

Figure 1.