Background: Medications, medical devices, biological products, and other medical products can cause healthcare-associated infections related to contamination in production or transportation (intrinsic contamination) or contamination at the point of use (extrinsic contamination). Rapid identification of contaminated medical products can lead to actions to decrease further patient harm. We sought to describe events that prompted public health investigations of contaminated medical products in healthcare facilities. Methods: We reviewed records of CDC consultations with health departments and healthcare facilities from January 2015 through August 2019 to identify public health investigations in which medical products were identified as a likely source of patient infection or pathogen transmission to at least 1 patient. We collected data on products, contamination type, pathogens, route of patient exposure, healthcare setting where exposure occurred, and resulting actions. Results: There were 34 investigations involving medications (n = 15, 44%), medical devices (n = 12, 35%), biological products (n = 3, 9%), and other medical products (n = 4, 12%). Intrinsic contamination was suspected in 15 investigations (44%), with 13 (87%) based on isolation of a pathogen from unopened products and 2 (13%) based on isolation of similar pathogens from patients in contact with a medical product at multiple facilities. Extrinsic contamination was suspected in 19 investigations (56%) based on evidence of pathogen transmission at a single healthcare facility and concurrent infection control gaps at that facility supporting a mechanism of contamination. The most common pathogens prompting investigation were nontuberculous mycobacteria (n = 9, 26%), Burkholderia spp (n = 7, 21%), Klebsiella spp (n = 3, 9%), Serratia spp (n = 2, 6%), and other environmental and commensal organisms. Patients were most commonly exposed in hospitals (n = 19, 56%) and outpatient settings (n = 9, 26%). The most common patient exposures that resulted in transmission of the pathogen were infusions and injections (n = 15, 44%), diagnostic and therapeutic procedures (n = 9, 26%), and surgery (n = 5, 15%). Patient were notified and offered testing in at least 6 investigations (18%) . Interventions included product removal, healthcare provider alerts, patient notification and testing, modification of injection safety practices and other general infection control practices, correction of improper storage and handling, and changes in product design, manufacturing processes, or instructions for use. Conclusions: Public health investigations identified intrinsic and extrinsic contamination of medications, devices, and other products as a cause of healthcare-associated infections. Healthcare facilities should consider contaminated products in investigations of healthcare-associated infections, take steps to identify local infection control concerns, and alert public health authorities to events that could suggest widespread contamination.

Funding: None

Disclosures: None

Background: Hospital-onset bacteremia and fungemia (HOB) may be a preventable hospital-acquired condition and a potential healthcare quality measure. We developed and evaluated a tool to assess the preventability of HOB and compared it to a more traditional consensus panel approach. Methods: A 10-member healthcare epidemiology expert panel independently rated the preventability of 82 hypothetical HOB case scenarios using a 6-point Likert scale (range, 1= “Definitively or Almost Certainly Preventable” to 6= “Definitely or Almost Certainly Not Preventable”). Ratings on the 6-point scale were collapsed into 3 categories: Preventable (1–2), Uncertain (3–4), or Not preventable (5–6). Consensus was defined as concurrence on the same category among ≥70% expert raters. Cases without consensus were deliberated via teleconference, web-based discussion, and a second round of rating. The proportion meeting consensus, overall and by predefined HOB source attribution, was calculated. A structured HOB preventability rating tool was developed to explicitly account for patient intrinsic and extrinsic healthcare-related risks (Fig. 1). Two additional physician reviewers independently applied this tool to adjudicate the same 82 case scenarios. The tool was iteratively revised based on reviewer feedback followed by repeat independent tool-based adjudication. Interrater reliability was evaluated using the Kappa statistic. Proportion of cases where tool-based preventability category matched expert consensus was calculated. Results: After expert panel round 1, consensus criteria were met for 29 cases (35%), which increased to 52 (63%) after round 2. Expert consensus was achieved more frequently for respiratory or surgical site infections than urinary tract and central-line–associated bloodstream infections (Fig. 2a). Most likely to be rated preventable were vascular catheter infections (64%) and contaminants (100%). For tool-based adjudication, following 2 rounds of rating with interim tool revisions, agreement between the 2 reviewers was 84% for cases overall (κ, 0.76; 95% CI, 0.64–0.88]), and 87% for the 52 cases with expert consensus (κ, 0.79; 95% CI, 0.65–0.94). Among cases with expert consensus, tool-based rating matched expert consensus in 40 of 52 (77%) and 39 of 52 (75%) cases for reviewer 1 and reviewer 2, respectively. The proportion of cases rated “uncertain“ was lower among tool-based adjudicated cases with reviewer agreement (15 of 69) than among cases with expert consensus (23 of 52) (Fig. 2b). Conclusions: Healthcare epidemiology experts hold varying perspectives on HOB preventability. Structured tool-based preventability rating had high interreviewer reliability, matched expert consensus in most cases, and rated fewer cases with uncertain preventability compared to expert consensus. This tool is a step toward standardized assessment of preventability in future HOB evaluations.

Funding: None

Disclosures: None

Background: Antibiotic resistance has increased at alarming rates, driven predominantly by antibiotic overuse. Although most antibiotic use occurs in outpatients, antimicrobial stewardship programs have primarily focused on inpatient settings. A major challenge for outpatient stewardship is the lack of accurate and accessible electronic data to target interventions. We sought to develop and validate an electronic algorithm to identify inappropriate antibiotic use for outpatients with acute bronchitis. Methods: This study was conducted within the University of Pennsylvania Health System (UPHS). We used ICD-10 diagnostic codes to identify encounters for acute bronchitis at any outpatient UPHS practice between March 15, 2017, and March 14, 2018. Exclusion criteria included underlying immunocompromising condition, other comorbidity influencing the need for antibiotics (eg, emphysema), or ICD-10 code at the same visit for a concurrent infection (eg, sinusitis). We randomly selected 300 (150 from academic practices and 150 from nonacademic practices) eligible subjects for detailed chart abstraction that assessed patient demographics and practice and prescriber characteristics. Appropriateness of antibiotic use based on chart review served as the gold standard for assessment of the electronic algorithm. Because antibiotic use is not indicated for this study population, appropriateness was assessed based upon whether an antibiotic was prescribed or not. Results: Of 300 subjects, median age was 61 years (interquartile range, 50–68), 62% were women, 74% were seen in internal medicine (vs family medicine) practices, and 75% were seen by a physician (vs an advanced practice provider). On chart review, 167 (56%) subjects received an antibiotic. Of these subjects, 1 had documented concern for pertussis and 4 had excluding conditions for which there were no ICD-10 codes. One received an antibiotic prescription for a planned dental procedure. Thus, based on chart review, 161 (54%) subjects received antibiotics inappropriately. Using the electronic algorithm based on diagnostic codes, underlying and concurrent conditions, and prescribing data, the number of subjects with inappropriate prescribing was 170 (56%) because 3 subjects had antibiotic prescribing not noted based on chart review. The test characteristics of the electronic algorithm (compared to gold standard chart review) for identification of inappropriate antibiotic prescribing were the following: sensitivity, 100% (161 of 161); specificity, 94% (130 of 139); positive predictive value, 95% (161 of 170); and negative predictive value, 100% (130 of 130). Conclusions: For outpatients with acute bronchitis, an electronic algorithm for identification of inappropriate antibiotic prescribing is highly accurate. This algorithm could be used to efficiently assess prescribing among practices and individual clinicians. The impact of interventions based on this algorithm should be tested in future studies.

Funding: None

Disclosures: None

Background: Bundles have been proven to reduce the risk of healthcare-associated infections and to provide for rapid recognition and response for the best outcome in patients with sepsis. Each element alone does not provide the statistical significance that all elements together allow. Providing near real-time compliance with bundle measures to clinical staff can drive performance improvement with the bundle during the patient’s hospital stay, resulting in improved clinical care and prevention of infection. Methods: In 2019, 3 clinical initiatives were chartered that applied evidence-based bundles for early identification and treatment of sepsis, prevention of healthcare-associated pneumonia (HAP), and prevention of surgical site infection. The bundle included the following elements: assessment of sepsis, measurement of lactic acid, collection of blood culture, timely administration of antibiotics. The HAP bundle included the following elements: assessment of aspiration risk, elevation of the head of the bed, oral care twice daily and preoperatively, and incentive spirometry postoperatively. And the SSI bundle included the following elements: preoperative CHG bath, appropriate preoperative antibiotic, perioperative glucose control, and perioperative temperature control. A multidisciplinary team developed and implemented dashboards that extracted bundle elements from the electronic medical record (EMR) nightly. Bundle compliance was calculated at the individual element level as well as the aggregate. Bundle failure data were available at the patient level as well as in aggregate by care location and provider, allowing for real-time feedback to staff and creation of improvement plans. An unanticipated benefit was the identification and correction of charting inconsistencies. Results: Collection, aggregation, and analysis of bundle compliance data were displayed in a system dashboard, and data were refreshed nightly. This approach allowed us to display overall bundle compliance at the facility and system level, including a heat map showing each facility’s compliance with the bundle and each associated element. Utilization of an EMR dashboard allowed for performance review on 100% of eligible patients rather than a sample, as occurs with manual review and abstraction processes. Routine review of performance via the dashboards with frontline staff, clinical leaders, medical staff, and executives has resulted in month-by- month improvement in bundle compliance. Conclusions: Direct data mining, data aggregation and analysis, followed by direct feedback to frontline staff, has resulted in steady improvement in overall bundle compliance, compliance with individual bundle components, and standardization of charting in the EMR. This approach has ultimately resulted in better outcomes for sepsis patients, reduction in healthcare-associated pneumonia, and reduction in surgical site infections.

Funding: None

Disclosures: None

Background: Enhanced terminal room cleaning with ultraviolet C (UVC) disinfection has become more commonly used as a strategy to reduce the transmission of important nosocomial pathogens, including Clostridioides difficile, but the real-world effectiveness remains unclear. Objectives: We aimed to assess the association of UVC disinfection during terminal cleaning with the incidence of healthcare-associated C. difficile infection and positive test results for C. difficile within the nationwide Veterans Health Administration (VHA) System. Methods: Using a nationwide survey of VHA system acute-care hospitals, information on UV-C system utilization and date of implementation was obtained. Hospital-level incidence rates of clinically confirmed hospital-onset C. difficile infection (HO-CDI) and positive test results with recent healthcare exposures (both hospital-onset [HO-LabID] and community-onset healthcare-associated [CO-HA-LabID]) at acute-care units between January 2010 and December 2018 were obtained through routine surveillance with bed days of care (BDOC) as the denominator. We analyzed the association of UVC disinfection with incidence rates of HO-CDI, HO-Lab-ID, and CO-HA-LabID using a nonrandomized, stepped-wedge design, using negative binomial regression model with hospital-specific random intercept, the presence or absence of UVC disinfection use for each month, with baseline trend and seasonality as explanatory variables. Results: Among 143 VHA acute-care hospitals, 129 hospitals (90.2%) responded to the survey and were included in the analysis. UVC use was reported from 42 hospitals with various implementation start dates (range, June 2010 through June 2017). We identified 23,021 positive C. difficile test results (HO-Lab ID: 5,014) with 16,213 HO-CDI and 24,083,252 BDOC from the 129 hospitals during the study period. There were declining baseline trends nationwide (mean, −0.6% per month) for HO-CDI. The use of UV-C had no statistically significant association with incidence rates of HO-CDI (incidence rate ratio [IRR], 1.032; 95% CI, 0.963–1.106; P = .65) or incidence rates of healthcare-associated positive C. difficile test results (HO-Lab). Conclusions: In this large quasi-experimental analysis within the VHA System, the enhanced terminal room cleaning with UVC disinfection was not associated with the change in incidence rates of clinically confirmed hospital-onset CDI or positive test results with recent healthcare exposure. Further research is needed to understand reasons for lack of effectiveness, such as understanding barriers to utilization.

Funding: None

Disclosures: None

Background: The National Healthcare Safety Network’s (NHSN) Antibiotic Resistance (AR) Option offers hospitals a way to report antibiotic resistance data from their facility’s laboratory information system and create facility-specific antibiograms. Suppression of select antibiotic susceptibility results may be used by antibiotic stewardship teams to prevent unnecessary use of broad-spectrum therapies by not making those susceptibilities available to providers. To be of use, antibiograms should offer a complete picture of antibiotic resistance. We wanted to understand the impact of data suppression. Methods: A retrospective cross-sectional study was conducted including data from 2017 and 2018. The clinical susceptibility data for cefotaxime, ceftriaxone, ceftazidime, ertapenem, imipenem, and meropenem against carbapenem-resistant Enterobacteriaceae (CRE), Pseudomonas aeruginosa (CRPA), Acinetobacter baumannii (CRAB), and extended-spectrum β-lactamase–producing Enterobacteriaceae (ESBL) were collected from commercial antimicrobial susceptibility testing instruments (cASTI) in 3 Tennessee healthcare networks that also report to the NHSN AR Option. These data were linked to the NHSN data using 4 keys: date of birth, isolate collection date, pathogen, and specimen source. An isolate was defined as suppressed when susceptibility results were observed from the cASTI but not in NHSN. The proportions of suppressed results were calculated and stratified by genus, facility, and antibiotic. Results: Overall, 1,009 isolates were matched between the NHSN AR data and the laboratory test results. Of these, 4.1% were CRAB, 23.3% were CRPA, and 72.6% were Enterobacteriaceae. In total, 4,948 susceptibility results were available from cASTIs. Suppressed results in NHSN data were observed in 918 isolates (91.0%) and accounted for 2,797 results (56.6%). Of the 817 isolates tested against imipenem, 18.7% were found to be suppressed. Moreover, 100%, 57.9%, and 8.6% of imipenem tests against CRAB, CRPA, and Enterobacteriaceae, respectively, were suppressed. Of the suppressed results, 38.3%, 3.6%, and 58.1% were susceptible, intermediate, and resistant respectively. Of the 363 isolates tested against meropenem, 48.2% were found to be suppressed. In addition, 12.2%, 53.0%, and 52.2% of meropenem tests against CRAB, CRPA, and Enterobacteriaceae, respectively, were suppressed. Of the suppressed results, 47.4%, 11.4%, and 41.1% were susceptible, intermediate, and resistant, respectively. Conclusions: A large proportion of isolates had at least 1 analyzed antibiotic suppressed within the NHSN AR Option. It will be important to develop and implement strategies to ensure that nonsuppressed data are available to be reported to the NHSN AR module.

Funding: None

Disclosures: None