Search

Background Previous estimates of methicillin-resistant and -sensitive Staphylococcus aureus (MRSA, MSSA) bacteremia rates in hospitalized patients showed decreases in hospital-onset (HO) MRSA, no changes in community-onset (CO) MRSA and HO MSSA, and slight increases in CO MSSA rates from 2012–2017; more recent trends have not been well characterized. Methods We calculated monthly S. aureus bacteremia rates using microbiology data from hospitals reporting antibiotic susceptibility results in the Premier Healthcare Database from January 2017- June 2021. S. aureus blood isolates resistant to methicillin, oxacillin, or cefoxitin were categorized as MRSA. Positive blood cultures collected on or before hospital day 3 were categorized as CO; those collected on day 4 or later were categorized as HO. Annual rate differences were assessed using generalized estimating equation models with a negative binomial distribution adjusting for hospital bed size, teaching status, urban/rural designation, discharge month, census division, distributions of patient age, sex and race and hospital-level clustering. Results Among 11 million discharges from 356 hospitals, we identified 5,627 HO S. aureus bacteremia events: 2,558 (45%) MRSA and 3,069 (55%) MSSA. In 2017, we observed HO MRSA bacteremia rates of 0.41 and HO MSSA rates of 0.49 per 10,000 patient-days. Compared with 2017, adjusted rates of HO MRSA and MSSA bacteremia were significantly higher in 2019, 2020, and 2021 (Figure 1). Of 42,587 CO S. aureus bacteremia events, 19,107 (45%) were MRSA and 23,480 (55%) MSSA. CO MRSA bacteremia rates were lowest in 2017 (1.52 per 1,000 discharges), and adjusted rates were significantly higher in all later years (Figure 2); we observed significant regional variability in annual differences (p=0.018). The observed CO MSSA bacteremia rate was 2.03 per 1,000 discharges in 2017, and adjusted rates were significantly higher in 2020 and 2021. Figure 1.Hospital-Onset MRSA and MSSA Bacteremia Incidence Rate Ratios* with 95% confidence intervals, 2017 – 2021**.*Reference year =2017.**2021 includes data during January – June.***Statistically significant at p Conclusion In contrast with previous trends, recent S. aureus bacteremia rates in this cohort have increased in comparison to 2017. Potential explanations warranting exploration include differences in regional trends and pandemic-associated changes in inpatient risk, severity of illness, length of stay, and hospital utilization Disclosures All Authors: No reported disclosures.

Background Studies describing risk factors for Clostridioides difficile infection (CDI) are often limited in their ability to identify potentially important exposures occurring long before diagnosis. We describe hospitalizations and antibiotic use (AU) occurring up to one year prior to CDI diagnosis among Medicare beneficiaries. Methods We studied incident CDI cases (positive C. difficile test in a person ≥65 years without a positive test in the prior 8 weeks) identified during 2016–2018 through population-based CDI surveillance from four states participating in the Centers for Disease Control and Prevention’s Emerging Infections Program. The analysis included specimens collected in all settings and was limited to case patients who were identified as having fee-for-service Medicare and Part D drug coverage for the year preceding specimen collection. Inpatient hospitalization data was extracted from Medicare Provider Analysis and Review (MEDPAR) files and outpatient AU (prescriptions filled) was determined using Part D drug event files. Timing of hospitalizations and antibiotic prescriptions were described as recent (0–3 months prior to specimen collection) or remote (4–12 months prior). Results Of 1,953 CDI cases, 1,594 (82%) filled ≥1 course of outpatient antibiotics in the prior year; 805 (41%) filled an antibiotic both recently and remotely, 497 (25%) only remotely, and 292 (15%) only recently. Cases with outpatient AU received a median of 23.5 (IQR 12–46) total days supplied, and a median of 2 different antibiotic classes (IQR 1 – 3). The most frequent antibiotic classes filled include fluoroquinolones (17% of all antibiotics filled), 1st generation cephalosporins (10%), and folate pathway inhibitors (10%). Overall, 1,314 (67%) cases were hospitalized in the prior year; 569 (29%) were hospitalized both recently and remotely, 446 (23%) only recently, and 299 (15%) only remotely. Median length of stay was 13 days (IQR 6–28). A total of 142 cases (7%) did not have hospitalization or outpatient AU in the prior year, and 1,097 (56%) had both. Conclusion Incident CDI cases have substantial exposure to recent and remote hospitalization and outpatient AU. Understanding cumulative effects of multiple risk factors can guide prevention strategies, including antibiotic stewardship efforts. Disclosures Scott Fridkin, MD, Pfizer: Grant/Research Support Ghinwa Dumyati, MD, Pfizer: Grant/Research Support.

Background The 2014 United States National Strategy aimed to reduce inappropriate inpatient antibiotic use by 20% for monitored conditions by 2020. The Hospital Core Elements of Antibiotic Stewardship highlight opportunities to improve treatment of common infections, including optimizing length of therapy (LOT) for community-acquired pneumonia (CAP). A minimum of 5 days of antibiotic therapy for patients with uncomplicated CAP is recommended, with > 7 days, or > 3 days after clinical improvement, rarely necessary. In this study, we evaluated annual trends in LOT from 2013-2020. Methods We conducted a retrospective cohort study using IBM MarketScan® database to evaluate LOT annual trends among adults 18-64 years in the United States hospitalized with uncomplicated CAP from 2013-2020. We included patients with a primary diagnosis of bacterial or unspecified pneumonia using International Classification of Diseases 9th and 10th revision codes, length of stay (LOS) of 2-10 days, discharged home with self-care, and not re-hospitalized in the 3 days following discharge. Discharge home was used as a surrogate for clinical improvement. We obtained inpatient LOS and post-discharge LOT data from MarketScan. We estimated annual inpatient LOT based on LOS from the Premier Healthcare Database (PHD). Total LOT was calculated by summing predicted inpatient LOT (from PHD) and actual post-discharge LOT (from MarketScan). Proportion of total LOT > 7 days and post-discharge LOT > 3 days were considered indicators of likely excessive LOT. Results There were 44,976 uncomplicated CAP hospitalizations among patients 18–64 years in MarketScan, 2013-2020. Patients had a median age of 54 years, median LOS of 3 days, were more likely to be female (56%) and in the South region (49%). The median LOT decreased from 9.6 days in 2013 to 8.6 days in 2020. The proportion of patients with total LOT > 7 days decreased from 68% in 2013 to 51% in 2020 (% change: -25%); the proportion with post-discharge LOT > 3 days decreased from 73% in 2013 to 63% in 2020 (% change: -14%; Figure 1). Conclusion The proportion of CAP patients with likely excessive LOT decreased by 25% from 2013-2020, surpassing the 2020 goal. Antibiotic stewardship programs should continue to pursue interventions to reduce excessive length of therapy for common infections. Disclosures All Authors: No reported disclosures.

Replication-competent virus has not been detected in individuals with mild to moderate coronavirus disease 2019 (COVID-19) more than 10 days after symptom onset. It is unknown whether these findings apply to nursing home residents. Of 273 specimens collected from nursing home residents >10 days from the initial positive test, none were culture positive.

Undetected infection with SARS-CoV-2, the virus that causes coronavirus disease 2019 (COVID-19) contributes to transmission in nursing homes, settings where large outbreaks with high resident mortality have occurred (1,2). Facility-wide testing of residents and health care personnel (HCP) can identify asymptomatic and presymptomatic infections and facilitate infection prevention and control interventions (3-5). Seven state or local health departments conducted initial facility-wide testing of residents and staff members in 288 nursing homes during March 24-June 14, 2020. Two of the seven health departments conducted testing in 195 nursing homes as part of facility-wide testing all nursing homes in their state, which were in low-incidence areas (i.e., the median preceding 14-day cumulative incidence in the surrounding county for each jurisdiction was 19 and 38 cases per 100,000 persons); 125 of the 195 nursing homes had not reported any COVID-19 cases before the testing. Ninety-five of 22,977 (0.4%) persons tested in 29 (23%) of these 125 facilities had positive SARS-CoV-2 test results. The other five health departments targeted facility-wide testing to 93 nursing homes, where 13,443 persons were tested, and 1,619 (12%) had positive SARS-CoV-2 test results. In regression analyses among 88 of these nursing homes with a documented case before facility-wide testing occurred, each additional day between identification of the first case and completion of facility-wide testing was associated with identification of 1.3 additional cases. Among 62 facilities that could differentiate results by resident and HCP status, an estimated 1.3 HCP cases were identified for every three resident cases. Performing facility-wide testing immediately after identification of a case commonly identifies additional unrecognized cases and, therefore, might maximize the benefits of infection prevention and control interventions. In contrast, facility-wide testing in low-incidence areas without a case has a lower proportion of test positivity; strategies are needed to further optimize testing in these settings.

Background Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection can spread rapidly within skilled nursing facilities. After identification of a case of Covid-19 in a skilled nursing facility, we assessed transmission and evaluated the adequacy of symptom-based screening to identify infections in residents. Methods We conducted two serial point-prevalence surveys, 1 week apart, in which assenting residents of the facility underwent nasopharyngeal and oropharyngeal testing for SARS-CoV-2, including real-time reverse-transcriptase polymerase chain reaction (rRT-PCR), viral culture, and sequencing. Symptoms that had been present during the preceding 14 days were recorded. Asymptomatic residents who tested positive were reassessed 7 days later. Residents with SARS-CoV-2 infection were categorized as symptomatic with typical symptoms (fever, cough, or shortness of breath), symptomatic with only atypical symptoms, presymptomatic, or asymptomatic. Results Twenty-three days after the first positive test result in a resident at this skilled nursing facility, 57 of 89 residents (64%) tested positive for SARS-CoV-2. Among 76 residents who participated in point-prevalence surveys, 48 (63%) tested positive. Of these 48 residents, 27 (56%) were asymptomatic at the time of testing; 24 subsequently developed symptoms (median time to onset, 4 days). Samples from these 24 presymptomatic residents had a median rRT-PCR cycle threshold value of 23.1, and viable virus was recovered from 17 residents. As of April 3, of the 57 residents with SARS-CoV-2 infection, 11 had been hospitalized (3 in the intensive care unit) and 15 had died (mortality, 26%). Of the 34 residents whose specimens were sequenced, 27 (79%) had sequences that fit into two clusters with a difference of one nucleotide. Conclusions Rapid and widespread transmission of SARS-CoV-2 was demonstrated in this skilled nursing facility. More than half of residents with positive test results were asymptomatic at the time of testing and most likely contributed to transmission. Infection-control strategies focused solely on symptomatic residents were not sufficient to prevent transmission after SARS-CoV-2 introduction into this facility.

Background Multidrug-resistant (MDR) bacteria that are commonly associated with health care cause a substantial health burden. Updated national estimates for this group of pathogens are ne...

Previously reported associations between hospital-level antibiotic use and hospital-onset Clostridioides difficile infection (HO-CDI) were reexamined using 2012–2018 data from a new cohort of US acute-care hospitals. This analysis revealed significant positive associations between total, third-generation, and fourth-generation cephalosporin, fluoroquinolone, carbapenem, and piperacillin-tazobactam use and HO-CDI rates, confirming previous findings.

Background: Epidemiological studies have utilized administrative discharge diagnosis codes to identify methicillin-resistant and methicillin-sensitive Staphylococcus aureus (MRSA and MSSA) infections and trends, despite debate regarding the accuracy of utilizing codes for this purpose. We assessed the sensitivity and positive predictive value (PPV) of MRSA- and MSSA-specific diagnosis codes, trends, characteristics, and outcomes of S. aureus hospitalizations by method of identification. Methods: Clinical micro biology results and discharge data from geographically diverse US hospitals participating in the Premier Healthcare Database from 2012–2017 were used to identify monthly rates of MRSA and MSSA. Positive MRSA or MSSA clinical cultures and/or a MRSA- or MSSA-specific International Classification of Diseases, Ninth/Tenth Revision, Clinical Modification (ICD-9/10 CM) diagnosis codes from adult inpatients (aged ≥18 years) were included as S. aureus hospitalizations. Septicemia was defined as a positive blood culture or a MRSA or MSSA septicemia code. Sensitivity and PPV for codes were calculated for hospitalizations where admission status was not listed as transfer; true infection was considered a positive clinical culture. Negative binominal regression models measured trends in rates of MRSA and MSSA per 1,000 hospital discharges. Results: We identified 168,634 MRSA and 148,776 MSSA hospitalizations in 256 hospitals; 17% of MRSA and 21% of MSSA were septicemia. Less than half of all S. aureus hospitalizations (49% MRSA, 46% MSSA) and S. aureus septicemia hospitalizations (37% MRSA, 38% MSSA) had both a positive culture and diagnosis code (Fig. 1). Sensitivity of MRSA codes in identifying positive cultures was 61% overall and 56% for septicemia, PPV was 62% overall and 53% for septicemia. MSSA codes had a sensitivity of 49% in identifying MSSA cultures and 52% for MSSA septicemia; PPV was 69% overall and 62% for septicemia. Despite low sensitivity, MRSA trends are similar for cultures and codes, and MSSA trends are divergent (Fig. 2). For hospitalizations with septicemia, mortality was highest among those with a blood culture only (31.3%) compared to hospitalizations with both a septicemia code and blood culture (16.6%), and septicemia code only (14.7%). Conclusions: ICD diagnosis code sensitivity and PPV for identifying infections were consistently poor in recent years. Less than half of hospitalizations have concordant microbiology laboratory results and diagnosis codes. Rates and trend estimates for MSSA differ by method of identification. Using diagnosis codes to identify S. aureus infections may not be appropriate for descriptive epidemiology or assessing trends due to significant misclassification.Funding: NoneDisclosures: Scott Fridkin reports that his spouse receives consulting fees from the vaccine industry.

We compared methicillin-resistant Staphylococcus aureus (MRSA) bloodstream infections (BSIs) captured by culture-based surveillance and MRSA septicemia hospitalizations captured by administrative coding using statewide hospital discharge data in Connecticut from 2010 to 2018. Observed discrepancies between identification methods suggest administrative coding is inappropriate for assessing trends in MRSA BSIs.

In the Seattle, Washington metropolitan area, where the first case of novel coronavirus 2019 disease (COVID-19) in the United States was reported (1), a community-level outbreak is ongoing with evidence of rapid spread and high morbidity and mortality among older adults in long-term care skilled nursing facilities (SNFs) (2,3). However, COVID-19 morbidity among residents of senior independent and assisted living communities, in which residents do not live as closely together as do residents in SNFs and do not require skilled nursing services, has not been described. During March 5-9, 2020, two residents of a senior independent and assisted living community in Seattle (facility 1) were hospitalized with confirmed COVID-19 infection; on March 6, social distancing and other preventive measures were implemented in the community. UW Medicine (the health system linked to the University of Washington), Public Health - Seattle & King County, and CDC conducted an investigation at the facility. On March 10, all residents and staff members at facility 1 were tested for SARS-CoV-2, the virus that causes COVID-19, and asked to complete a questionnaire about their symptoms; all residents were tested again 7 days later. Among 142 residents and staff members tested during the initial phase, three of 80 residents (3.8%) and two of 62 staff members (3.2%) had positive test results. The three residents had no symptoms at the time of testing, although one reported an earlier cough that had resolved. A fourth resident, who had negative test results in the initial phase, had positive test results 7 days later. This resident was asymptomatic on both days. Possible explanations for so few cases of COVID-19 in this residential community compared with those in several Seattle SNFs with high morbidity and mortality include more social distancing among residents and less contact with health care providers. In addition, early implementation of stringent isolation and protective measures after identification of two COVID-19 cases might have been effective in minimizing spread of the virus in this type of setting. When investigating a potential outbreak of COVID-19 in senior independent and assisted living communities, symptom screening is unlikely to be sufficient to identify all persons infected with SARS-CoV-2. Adherence to CDC guidance to prevent COVID-19 transmission in senior independent and assisted living communities (4) could be instrumental in preventing a facility outbreak.

Introduction Staphylococcus aureus is one of the most common pathogens in health care facilities and in the community, and can cause invasive infections, sepsis, and death. Despite progress in preventing methicillin-resistant S. aureus (MRSA) infections in health care settings, assessment of the problem in both health care and community settings is needed. Further, the epidemiology of methicillin-susceptible S. aureus (MSSA) infections is not well described at the national level. Methods Data from the Emerging Infections Program (EIP) MRSA population surveillance (2005–2016) and from the Premier and Cerner Electronic Health Record databases (2012–2017) were analyzed to describe trends in incidence of hospital-onset and community-onset MRSA and MSSA bloodstream infections and to estimate the overall incidence of S. aureus bloodstream infections in the United States and associated in-hospital mortality. Results In 2017, an estimated 119,247 S. aureus bloodstream infections with 19,832 associated deaths occurred. During 2005–2012 rates of hospital-onset MRSA bloodstream infection decreased by 17.1% annually, but the decline slowed during 2013–2016. Community-onset MRSA declined less markedly (6.9% annually during 2005–2016), mostly related to declines in health care–associated infections. Hospital-onset MSSA has not significantly changed (p = 0.11), and community-onset MSSA infections have slightly increased (3.9% per year, p

Introduction By 2007, all Department of Veterans Affairs medical centers (VAMCs) had initiated a multifaceted methicillin-resistant Staphylococcus aureus (MRSA) prevention program. MRSA and methicillin-susceptible S. aureus (MSSA) infection rates among VAMC inpatients from 2005 to 2017 were assessed. Methods Clinical microbiology data from any patient admitted to an acute-care VAMC in the United States from 2005 through 2017 and trends in hospital-acquired MRSA colonization were examined. Results S. aureus infections decreased by 43% overall during the study period (p

It has been hoped that rapid SARS-CoV-2 antigen testing could reduce the tragic toll of COVID-19 in nursing homes. The performance of the BinaxNOW antigen test in a nursing home during an ongoing SARS-CoV-2 outbreak was evaluated., Visual Abstract. Serial SARS-CoV-2 Rapid Antigen Testing During a Nursing Home Outbreak. It has been hoped that rapid SARS-CoV-2 antigen testing could reduce the tragic toll of COVID-19 in nursing homes. The performance of the BinaxNOW antigen test in a nursing home during an ongoing SARS-CoV-2 outbreak was evaluated., Background: To address high COVID-19 burden in U.S. nursing homes, rapid SARS-CoV-2 antigen tests have been widely distributed in those facilities. However, performance data are lacking, especially in asymptomatic people. Objective: To evaluate the performance of SARS-CoV-2 antigen testing when used for facility-wide testing during a nursing home outbreak. Design: A prospective evaluation involving 3 facility-wide rounds of testing where paired respiratory specimens were collected to evaluate the performance of the BinaxNOW antigen test compared with virus culture and real-time reverse transcription polymerase chain reaction (RT-PCR). Early and late infection were defined using changes in RT-PCR cycle threshold values and prior test results. Setting: A nursing home with an ongoing SARS-CoV-2 outbreak. Participants: 532 paired specimens collected from 234 available residents and staff. Measurements: Percentage of positive agreement (PPA) and percentage of negative agreement (PNA) for BinaxNOW compared with RT-PCR and virus culture. Results: BinaxNOW PPA with virus culture, used for detection of replication-competent virus, was 95%. However, the overall PPA of antigen testing with RT-PCR was 69%, and PNA was 98%. When only the first positive test result was analyzed for each participant, PPA of antigen testing with RT-PCR was 82% among 45 symptomatic people and 52% among 343 asymptomatic people. Compared with RT-PCR and virus culture, the BinaxNOW test performed well in early infection (86% and 95%, respectively) and poorly in late infection (51% and no recovered virus, respectively). Limitation: Accurate symptom ascertainment was challenging in nursing home residents; test performance may not be representative of testing done by nonlaboratory staff. Conclusion: Despite lower positive agreement compared with RT-PCR, antigen test positivity had higher agreement with shedding of replication-competent virus. These results suggest that antigen testing could be a useful tool to rapidly identify contagious people at risk for transmitting SARS-CoV-2 during nascent outbreaks and help reduce COVID-19 burden in nursing homes. Primary Funding Source: None.

Background Background. Understanding the viral load and potential infectivity of individuals in nursing homes (NH) with repeat positive SARS-CoV-2 tests ≥ 90 days after initial infection has important implications for safety related to transmission in this high-risk setting. Methods Methods. We collected epidemiologic data by reviewing records of a convenience sample of NH residents and staff with respiratory specimens who had positive SARS-CoV-2 rRT-PCR test results from July 2020 through March 2021 and had a SARS-CoV-2 infection diagnosed ≥ 90 days prior. No fully vaccinated individuals were included. Each contributed one repeat positive specimen ≥ 90 days after initial, which was sent to CDC and retested using rRT-PCR. Specimens were assessed for replication-competent virus in cell culture if Cycle threshold (Ct) < 34 and sequenced if Ct < 30. Using Ct values as a proxy for viral RNA load, specimens were categorized as high (Ct < 30) or low (if Ct ≥ 30 or rRT-PCR negative at retesting). Continuous variables were compared using Wilcoxon signed-rank tests. Proportions were compared using Chi-squared or Fisher’s exact tests. Results Results. Of 64 unvaccinated individuals with specimens from 61 unique NHs, 14 (22%) were sent for culture and sequencing. Ten of 64 (16%) had a high viral RNA load, of which four (6%) were culture positive and none were known variants of interest or concern (Figure 1). Median days to repeat positive test result were 122 (Interquartile range (IQR): 103–229) and 201 (IQR: 139–254), respectively, for high versus low viral load specimens (p=0.13). More individuals with high viral loads (5/10, 50%) reported COVID-19 symptoms than with a low viral load (1/27, 4%, p=0.003). Most individuals (46/58, 79%) were tested following known or suspected exposures, with no significant differences between high and low viral load (p=0.18). Conclusion In this study, nearly 1 in 6 NH residents and staff with repeat positive tests after 90 days demonstrated high viral RNA loads and viable virus, indicating possible infectivity. While individuals with high RNA viral load may be more likely to be symptomatic, distinguishing asymptomatic individuals who have high viral loads may be difficult with timing since initial infection, other test results, or exposure history alone. Disclosures John A. Jernigan, MD, MS, Nothing to disclose.

Background Although studies have shown blood culture rates have remained stable in recent years, understanding the variability in positivity over time and among hospitals may inform diagnostic and antimicrobial stewardship efforts. Methods We included all discharges from hospitals participating in the Premier Healthcare Database and Cerner Health Facts from 2012-2017 in months where a hospital reported at least one blood culture with antimicrobial susceptibility results. A blood culture episode was defined as one or more cultures drawn within 1 hour. Episodes on or before day 3 were defined as admission episodes (AE), and those drawn on day 4 or later were defined as post-admission episodes (PAE). Culture episodes yielding any organism were categorized as pathogen+ (i.e., at least 1 non-commensal organism identified) or commensal (i.e., only commensal organisms identified). Positive or commensal episode rates were calculated as the percentage of pathogen+ or commensal episodes among all blood culture episodes for AE and PAE. Logistic regression with generalized estimating equation models accounting for hospital-level clustering were used to measure time trends and facility level associations. Results Among 19.6 million discharges in 493 hospitals, 7.5 million blood culture episodes were identified; 336,102 (4.5%) were pathogen+, and 110,236 (1.5%) were commensals. The rate of pathogen+ AEs increased from 4.2% to 4.7% over the study period (p< .0001) and there was no significant temporal trend in the rate of pathogen+ PAEs (p=.7956) (Figure 1). AE commensals decreased significantly in 2016-2017 compared to previous years (1.6% in 2012 to 1.3% in 2017; p=.0092), and PAE commensals decreased significantly over the study period from 2.0% to 1.2% (p< .0001) (Figure 1). We observed substantial inter-hospital variability for each outcome (Figure 2). In addition, differences among hospital characteristics and seasonality were noted for AE and PAE pathogen+ rates and AE commensal rates (Figure 3), but not urbanicity or teaching hospital status. Monthly Positivity Rate of Blood Culture Episodes, Premier Healthcare Database and Cerner Health Facts, 2012-2017 Adjusted Odds Ratios and 95% Confidence Intervals of Blood Culture Episode Positive (Non-Commensal) and Commensal Rates and Associated Characteristics Conclusion While an increase AE pathogen+ rates and decrease in commensal rates could indicate improved culture ordering and collection practices, significant seasonal, regional, and facility-level variability calls for further investigation. Disclosures John A. Jernigan, MD, MS, Nothing to disclose

Background Treating patients with infections due to multidrug-resistant pathogens often requires substantial healthcare resources. The purpose of this study was to report estimates of the healthcare costs associated with infections due to multidrug-resistant bacteria in the United States (US). Methods We performed retrospective cohort studies of patients admitted for inpatient stays in the Department of Veterans Affairs healthcare system between January 2007 and October 2015. We performed multivariable generalized linear models to estimate the attributable cost by comparing outcomes in patients with and without positive cultures for multidrug-resistant bacteria. Finally, we multiplied these pathogen-specific, per-infection attributable cost estimates by national counts of infections due to each pathogen from patients hospitalized in a cohort of 722 US hospitals from 2017 to generate estimates of the population-level healthcare costs in the US attributable to these infections. Results Our analysis cohort consisted of 16 676 patients with community-onset infections and 172 712 matched controls and 8246 patients with hospital-onset infections and 66 939 matched controls. The highest cost was seen in hospital-onset invasive infections, with attributable costs (95% confidence intervals) ranging from $30 998 ($25 272–$36 724) for methicillin-resistant Staphylococcus aureus to $74 306 ($20 377–$128 235) for carbapenem-resistant (CR) Acinetobacter. The highest attributable costs for community-onset invasive infections were seen in CR Acinetobacter ($62 396; $20 370–$104 422). Treatment of these infections cost an estimated $4.6 billion ($4.1 billion–$5.1 billion) in 2017 in the US for community- and hospital-onset infections combined. Conclusions We found that antimicrobial-resistant infections led to substantial healthcare costs.

Skilled nursing facilities (SNFs) are focal points of the coronavirus disease 2019 (COVID-19) pandemic, and asymptomatic infections with SARS-CoV-2, the virus that causes COVID-19, among SNF residents and health care personnel have been described (1-3). Repeated point prevalence surveys (serial testing of all residents and health care personnel at a health care facility irrespective of symptoms) have been used to identify asymptomatic infections and have reduced SARS-CoV-2 transmission during SNF outbreaks (1,3). During March 2020, the Detroit Health Department and area hospitals detected a sharp increase in COVID-19 diagnoses, hospitalizations, and associated deaths among SNF residents. The Detroit Health Department collaborated with local government, academic, and health care system partners and a CDC field team to rapidly expand SARS-CoV-2 testing and implement infection prevention and control (IPC) activities in all Detroit-area SNFs. During March 7-May 8, among 2,773 residents of 26 Detroit SNFs, 1,207 laboratory-confirmed cases of COVID-19 were identified during three periods: before (March 7-April 7) and after two point prevalence surveys (April 8-25 and April 30-May 8): the overall attack rate was 44%. Within 21 days of receiving their first positive test results, 446 (37%) of 1,207 COVID-19 patients were hospitalized, and 287 (24%) died. Among facilities participating in both surveys (n = 12), the percentage of positive test results declined from 35% to 18%. Repeated point prevalence surveys in SNFs identified asymptomatic COVID-19 cases, informed cohorting and IPC practices aimed at reducing transmission, and guided prioritization of health department resources for facilities experiencing high levels of SARS-CoV-2 transmission. With the increased availability of SARS-CoV-2 testing, repeated point prevalence surveys and enhanced and expanded IPC support should be standard tools for interrupting and preventing COVID-19 outbreaks in SNFs.

Importance Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused epidemic spread of coronavirus disease 2019 (COVID-19) in the Seattle, Washington, metropolitan area, with morbidity and mortality concentrated among residents of skilled nursing facilities. The prevalence of COVID-19 among older adults in independent/assisted living is not understood. Objectives To conduct surveillance for SARS-CoV-2 and describe symptoms of COVID-19 among residents and staff of an independent/assisted living community. Design, Setting, and Participants In March 2020, public health surveillance of staff and residents was conducted on site at an assisted and independent living residence for older adults in Seattle, Washington, after exposure to 2 residents who were hospitalized with COVID-19. Exposures Surveillance for SARS-CoV-2 infection in a congregate setting implementing social isolation and infection prevention protocols. Main Outcomes and Measures SARS-CoV-2 real-time polymerase chain reaction was performed on nasopharyngeal swabs from residents and staff; a symptom questionnaire was completed assessing fever, cough, and other symptoms for the preceding 14 days. Residents were retested for SARS-CoV-2 7 days after initial screening. Results Testing was performed on 80 residents; 62 were women (77%), with mean age of 86 (range, 69-102) years. SARS-CoV-2 was detected in 3 of 80 residents (3.8%); none felt ill, 1 male resident reported resolved cough and 1 loose stool during the preceding 14 days. Virus was also detected in 2 of 62 staff (3.2%); both were symptomatic. One week later, resident SARS-CoV-2 testing was repeated and 1 new infection detected (asymptomatic). All residents remained in isolation and were clinically stable 14 days after the second test. Conclusions and Relevance Detection of SARS-CoV-2 in asymptomatic residents highlights challenges in protecting older adults living in congregate settings. In this study, symptom screening failed to identify residents with infections and all 4 residents with SARS-CoV-2 remained asymptomatic after 14 days. Although 1 asymptomatic infection was found on retesting, a widespread facility outbreak was avoided. Compared with skilled nursing settings, in assisted/independent living communities, early surveillance to identify asymptomatic persons among residents and staff, in combination with adherence to recommended preventive strategies, may reduce viral spread.

OBJECTIVE: To assess the variability in short term sepsis mortality by hospital among Centers for Medicare and Medicaid Services beneficiaries in the United States during 2013–2014. DESIGN: A retrospective cohort design SETTING: Hospitalizations from 3,068 acute care hospitals that participated in the Centers for Medicare and Medicaid Services inpatient prospective payment system in 2013 and 2014. PATIENTS: Medicare fee-for-service beneficiaries greater than or equal to 65 years old who had an inpatient hospitalization coded with present at admission severe sepsis or septic shock. INTERVENTIONS: None. MEASUREMENTS, MAIN RESULTS: Individual level mortality was assessed as death at or within 7 days of hospital discharge and aggregated to calculate hospital-level mortality rates. We used a logistic hierarchal linear model to calculate mortality risk-adjusted for patient characteristics. We quantified variability among hospitals using the median odds ratio (MOR), and calculated risk-standardized mortality rates (risk-standardized mortality rates) for each hospital. The overall crude mortality rate was 34.7%. We found significant variability in mortality by hospital (p

Background We examined the risk of sepsis within 90 days after discharge from a previous hospital stay by type of antibiotic received during the previous stay. Methods We retrospectively identified a cohort of hospitalized patients from the Truven Health MarketScan Hospital Drug Database. We examined the association between the use of certain antibiotics during the initial hospital stay, determined a priori, and the risk of postdischarge sepsis controlling for potential confounding factors in a multivariable logistic regression model. Our primary exposure was receipt of antibiotics more strongly associated with clinically important microbiome disruption. Our primary outcome was a hospital stay within 90 days of the index stay that included an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) discharge diagnosis of severe sepsis (ICD-9-CM code 995.92) or septic shock (785.52). Results Among 516 hospitals, we randomly selected a single stay for eligible patients. In 0.17% of these patients, severe sepsis/septic shock developed within 90 days after discharge. The risk of sepsis associated with exposure to our high-risk antibiotics was 65% higher than in those without antibiotic exposure. Conclusions Our study identified an increased risk of sepsis within 90 days of discharge among patients with exposure to high-risk antibiotics or increased quantities of antibiotics during hospitalization. Given that a significant proportion of inpatient antimicrobial use may be unnecessary, this study builds on previous evidence suggesting that increased stewardship efforts in hospitals may not only prevent antimicrobial resistance, Clostridium difficile infection, and other adverse effects, but may also reduce unwanted outcomes potentially related to disruption of the microbiota, including sepsis.

Background Previous studies suggest that duration of antibiotic therapy for community-acquired pneumonia (CAP) often exceeds national recommendations and might represent an important opportunity to improve antibiotic stewardship nationally. Our objective was to determine the average length of antibiotic therapy (LOT) for patients treated for uncomplicated CAP in US hospitals and the proportion of patients with excessive durations. Methods Records of retrospective cohorts of patients aged 18-64 years with private insurance and aged ≥65 years with Medicare hospitalized for CAP in 2012-2013 were used. Inpatient LOT was estimated as a function of length of stay. Outpatient LOT was based on prescriptions filled post discharge based on data from outpatient pharmacy files. Excessive duration was defined as outpatient LOT >3 days. Results Inclusion criteria were met for 22128 patients aged 18-64 years across 2100 hospitals and 130746 patients aged ≥65 years across 3227 hospitals. Median total LOT was 9.5 days. LOT exceeded recommended duration for 74% of patients aged 18-64 years and 71% of patients aged ≥65 years. Patients aged 18-64 years had a median (quartile 1-quartile 3) 6 (3-7) days outpatient LOT and those aged ≥65 years had 5 (3-7) days. Conclusions In this nationwide sample of patients hospitalized for CAP, median total LOT was just under 10 days, with more than 70% of patients having likely excessive treatment duration. Better adherence to recommended CAP therapy duration by improving prescribing at hospital discharge may be an important target for antibiotic stewardship programs.

Background Community-onset urinary tract infections (coUTIs) are one of the most common indications for antibiotic prescribing. It is important to understand patient demographic factors associated with microorganisms causing coUTI and their antibiotic resistance profiles, to tailor antibiotic prescribing practices. We analyzed microbiology data to understand factors associated with coUTI in the United States (US). Methods CoUTIs were identified in the Premier Healthcare Database and Cerner Health Facts among patients treated at participating healthcare facilities in the US between 2012-2017. Cases were defined by urine cultures yielding a bacterial organism and were collected in outpatient settings or within three days of hospitalization. Only the first specimen for each encounter was included in the analysis. Data on the organisms isolated, patient’s age, sex, and US census regions of the submitting facilities were described and compared using chi-square tests for associations. Encounters were classified as inpatient (INPT), observation (OBS), emergency department (ED), and outpatient (OTPT) based on the setting in which the culture was submitted. Results Using data from 637 acute care hospitals, urine samples from 3,291,561 encounters were included, with 776,653 (25.7%) INPT, 1,063,219 (34.8%) ED, 107,760 (3.5%) OBS, and 1,092,658 (35.8%) OTPT. The pathogens most frequently associated with coUTIs were Escherichia coli (57.3%), Klebsiella pneumoniae (9.7%), Enterococcus faecalis (5.1%), Proteus mirabilis (4.9%), and Pseudomonas aeruginosa (2.9%). Female sex, age < 65y and OTPT and ED settings were associated with higher relative frequency of E. coli (all p < 0.0001). Male sex, INPT setting and age >65 y were associated with higher relative frequency of P. aeruginosa, P. mirabilis and E. faecalis (all p < 0.0001, Figure). K. pneumoniae was found at higher relative frequency in those >45y, and in INPT and OBS settings (all p < 0.0001). Figure. Distribution of pathogens most frequently associated with community onset urinary tract infections Conclusion Understanding patient factors associated with the microbiology of coUTIs is an important step in developing treatment recommendations and antibiotic stewardship efforts. Further analyses will include assessing the impact of major antibiotic resistance phenotypes, geographic and healthcare settings. Disclosures All Authors: No reported disclosures

Background While discharge antibiotic prescriptions from emergency department (ED) visits have been reported, systemic antibiotic use during ED and hospital observation (OBS) visits have not been well assessed. We conducted a descriptive analysis of antibiotic use in these settings. Methods We identified ED and OBS visits not resulting in hospitalization, and systemic antibiotics administration charges during these visits from January 2012-December 2018 using the Premier Healthcare Database, representing at least 600 hospitals annually. Antibiotics prescribed after discharge were excluded. We reported the proportion of visits with antibiotic use, and described antibiotic use by class, agent and route stratified by location. We also examined trends in antibiotic use over time using a multivariable logistic model. Results We assessed 161,291,011 ED visits and 15,660,062 OBS visits from 2012–2018. Systemic antibiotics were identified in 9.0% of ED visits and 25.2% of OBS visits. Parenteral (IV) antibiotics were received in a high proportion of ED and OBS visits in which a systemic antibiotic was received (52.6% and 87.6% respectively). In the ED, 3rd/4th generation cephalosporins were the most commonly identified (32.7%) while in the OBS, 1st/2nd generation cephalosporins were most commonly identified (38.9%), Fig. 1. The most common agents in the ED were ceftriaxone, azithromycin, and cephalexin while the most common agents in the OBS were cefazolin, ceftriaxone, and levofloxacin. Any systemic antibiotic use in EDs declined slightly from 2012–2018 (9.2%-8.9%, p< 0.0001) while use in OBS settings saw the largest increase from 2017- 2018 (25.4%-30.4%, p< 0.0001), Fig. 2. Fluoroquinolone use decreased in both ED (41.3%) and OBS (31.2%) (both p< 0.0001) beginning in years 2012 and 2016 respectively. Figure 1: Antibiotic use in emergency departments and observation settings by antibiotic class, Premier Healthcare Database Hospitals, 2012–2018 Figure 2: Overall antibiotic use in emergency departments and observation settings by year, Premier Healthcare Database Hospitals, 2012–2018 Conclusion Hospital ED and OBS settings are uniquely positioned to improve appropriate antibiotic use across the spectrum of healthcare. Frequent use of IV antibiotics and recent increases in antibiotic use in observation settings call for evaluation of appropriateness of their use and presence of transition-of-care process. Further evaluation of diagnoses to evaluate the appropriateness of IV administration may highlight additional opportunities for optimizing prescribing practices. Disclosures All Authors: No reported disclosures

Background: Carbapenem-resistant Acinetobacter baumannii (CRAB) is an important cause of healthcare-associated infections with limited treatment options and high mortality. To describe risk factors for mortality, we evaluated characteristics associated with 30-day mortality in patients with CRAB identified through the Emerging Infections Program (EIP). Methods: From January 2012 through December 2017, 8 EIP sites (CO, GA, MD, MN, NM, NY, OR, TN) participated in active, laboratory-, and population-based surveillance for CRAB. An incident case was defined as patient’s first isolation in a 30-day period of A. baumannii complex from sterile sites or urine with resistance to ≥1 carbapenem (excluding ertapenem). Medical records were abstracted. Patients were matched to state vital records to assess mortality within 30 days of incident culture collection. We developed 2 multivariable logistic regression models (1 for sterile site cases and 1 for urine cases) to evaluate characteristics associated with 30-day mortality. Results: We identified 744 patients contributing 863 cases, of which 185 of 863 cases (21.4%) died within 30 days of culture, including 113 of 257 cases (44.0%) isolated from a sterile site and 72 of 606 cases (11.9%) isolated from urine. Among 628 hospitalized cases, death occurred in 159 cases (25.3%). Among hospitalized fatal cases, death occurred after hospital discharge in 27 of 57 urine cases (47.4%) and 21 of 102 cases from sterile sites (20.6%). Among sterile site cases, female sex, intensive care unit (ICU) stay after culture, location in a healthcare facility, including a long-term care facility (LTCF), 3 days before culture, and diagnosis of septic shock were associated with increased odds of death in the model (Fig. 1). In urine cases, age 40–54 or ≥75 years, ICU stay after culture, presence of an indwelling device other than a urinary catheter or central line (eg, endotracheal tube), location in a LTCF 3 days before culture, diagnosis of septic shock, and Charlson comorbidity score ≥3 were associated with increased odds of mortality (Fig. 2). Conclusion: Overall 30-day mortality was high among patients with CRAB, including patients with CRAB isolated from urine. A substantial fraction of mortality occurred after discharge, especially among patients with urine cases. Although there were some differences in characteristics associated with mortality in patients with CRAB isolated from sterile sites versus urine, LTCF exposure and severe illness were associated with mortality in both patient groups. CRAB was associated with major mortality in these patients with evidence of healthcare experience and complex illness. More work is needed to determine whether prevention of CRAB infections would improve outcomes.Funding: NoneDisclosures: None

Background: In recent years, the historic declines in the incidence of methicillin-resistant Staphylococcus aureus (MRSA) bloodstream infections (BSIs) in the United States have slowed. We examined trends in the incidence of community-onset (CO) MRSA BSIs among hospitalized persons with and without substance-use diagnoses. Methods: Using data from >200 US hospitals reporting to the Premier Healthcare Database (PHD) during 2012–2017, we conducted a retrospective study among hospitalized persons aged ≥18 years. MRSA BSIs with substance use were defined as hospitalizations having both a blood culture positive for MRSA and at least 1 International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) or ICD-10-CM diagnostic code for substance use including opioids, cocaine, amphetamines, or other substances (excluding cannabis, alcohol, and nicotine). MRSA BSIs were considered community onset when a positive blood culture was collected within 3 days of admission. We assessed annual trends and described characteristics of CO MRSA BSI hospitalizations, stratified by substance use. Results: Of 20,049 MRSA BSIs from 2012 to 2017, 17,634 (88%) were CO. Overall, MRSA BSI incidence decreased 7%, from 178.5 to 166.2 per 100,000 hospitalizations during the study period; However, CO MRSA BSI rates remained stable (152.7 to 149.9 per 100,000 hospitalizations). Among CO MRSA BSIs, 1,838 (10%) were BSIs with substance-use diagnoses; the incidence of CO MRSA BSIs with substance use increased 236% (from 8.2 to 27.6 per 100,000 hospitalizations) and represented a greater proportion of the CO MRSA rate over the study period (Fig. 1). The incidence of CO MRSA BSIs without substance use decreased 15% (from 144.5 to 122.4 per 100,000 hospitalizations). Patients with CO MRSA BSIs with substance use were younger (median, 40 vs 65 years), more likely to be female (50% vs 40%), white (79% vs 69%), and to leave against medical advice (15% vs 1%). Among patients not leaving against medical advice, CO BSI patients with substance-use diagnoses had longer lengths of stay (median, 11 vs 9 days), lower in-hospital mortality (9% vs 14%), and higher hospitalization costs (median, $22,912 vs $17,468) compared to patients without substance-use diagnoses. Conclusions: Although the overall CO MRSA BSI rate remained unchanged from 2012 to 2017, infections with substance use diagnoses increased >3-fold, and infections without substance use diagnoses decreased. These data suggest that the emergence of MRSA associated with substance-use diagnoses threatens potential progress in reducing the incidence of CO MRSA infections. Additional strategies may be needed to prevent MRSA BSI in patients with substance-use diagnoses, and to maintain national progress in the reduction of MRSA infections overall.Funding: NoneDisclosures: None

Background: Microbiology data are utilized to quantify epidemiology and trends in pathogens, antimicrobial resistance, and bloodstream infections. Understanding variability and trends in rates of hospital-level blood culture utilization may be important for interpreting these findings. Methods: We used clinical microbiology results and discharge data to identify monthly blood culture rates from US hospitals participating in the Premier Healthcare Database during 2012–2017. We included all discharges from months where a hospital reported at least 1 blood culture with microbiology and antimicrobial susceptibility results. Blood cultures drawn on or before day 3 were defined as admission cultures (ACs); blood cultures collected after day 3 were defined as a postadmission cultures (PACs). The AC rate was defined as the proportion of all hospitalizations with an AC. The PAC rate was defined as the number of days with a PAC among all patient days. Generalized estimating equation regression models that accounted for hospital-level clustering with an exchangeable correlation matrix were used to measure associations of monthly rates with hospital bed size, teaching status, urban–rural designation, region, month, and year. The AC rates were modeled using logistic regression, and the PAC rates were modeled using a Poisson distribution. Results: We included 11.7 million hospitalizations from 259 hospitals, accounting for nearly 52 million patient days. The median annual hospital-level AC rate was 27.1%, with interhospital variation ranging from 21.1% (quartile 1) to 35.2% (quartile 3) (Fig. 1). Multivariable models revealed no significant trends over time (P = .74), but statistically significant associations between AC rates with month (P < .001) and region (P = .003), associations with teaching status (P = .063), and urban-rural designation (P = .083) approached statistical significance. There was no association with bed size (P = .38). The median annual hospital-level PAC rate was 11.1 per 1,000 patient days, and interhospital variability ranged from 7.6 (quartile 1) to 15.2 (quartile 3) (Fig. 2). Multivariable models of PAC rates showed no significant trends over time (P = .12). We found associations between PAC rates with month (P = .016), bed size (P = .030), and teaching status (P = .040). PAC rates were not associated with urban–rural designation (P = .52) or region (P = .29). Conclusions: Blood culture utilization rates in this large cohort of hospitals were unchanged between 2012 and 2017, though substantial interhospital variability was detected. Although both AC and PAC rates vary by time of year and potentially by teaching status, AC rates vary by geographic characteristics whereas PAC rates vary by bed size. These factors are important to consider when comparing rates of bloodstream infections by hospital.Funding: NoneDisclosures: None

Background In 2019, CDC updated national estimates of antibiotic resistance. In this abstract we provide national estimates of and trends in proportion of bacterial pathogens expressing resistant phenotypes (%R), specifically: MRSA, VRE, CRE, ESBL, CRAsp, and MDR Pseudomonas, see Figure. Methods We measured incidence of clinical cultures yielding the bacterial species of interest among hospitalized adults in hospitals submitting data to the Premier Healthcare Database, Cerner Health Facts and BD Insights Research Database from 2012- 2017. Community-onset (CO) cultures were obtained ≤ day 3 of hospitalization; hospital-onset (HO) were obtained ≥ day 4. We determined hospital-specific %R for each species. We generated national estimates using a raking procedure to generate weighted adjustments to match the distribution for all U.S. acute care hospitals based on U.S. census division, bed size, teaching status, and urban/rural designation. We applied a weighted means survey procedure to calculate national estimates for each year. We used weighted multivariable logistic regression adjusting for hospital characteristics to examine trends. Results From 2012-2017, the overall number of hospitals contributing data was 890 (over 20% of U.S. hospital hospitalizations annually). National estimates and trends of %R are shown in the Figure. Between 2012-2017, significant annual decreases in %R were observed for MRSA, VRE, CRAsp, and MDR Pseudomonas. CRE %R did not change. Overall ESBL %R increased by 44% (CO=49% increase, HO=27% increase). Conclusion Reductions in %R were observed among MRSA, VRE, CRAsp, and MDR Pseudomonas, suggesting that prevention efforts focused in health care settings are having a disproportionate effect on resistant strains. However, %R remains unacceptably high for all pathogens we studied, and %R among ESBL-producing Enterobacteriaceae has increased, most prominently among CO infections. Continued focus on currently recommended intervention strategies as well as new ones for community onset infections is needed. Disclosures All Authors: No reported disclosures

Background Few data suggest that Clostridioides difficile infections (CDIs) detected by toxin enzyme immunoassay (EIA) are more severe and have worse outcomes than those detected by nucleic acid amplification tests (NAATs) only. We compared toxin- positive and NAAT-positive-only CDI across geographically diverse sites. Methods A case was defined as a positive C. difficile test in a person ≥1 year old with no positive tests in the prior 8 weeks. Cases were detected during 2014–2015 by a testing algorithm (specimens initially tested by glutamate dehydrogenase and toxin EIA; if discordant results, specimens were reflexed to NAAT) and classified as toxin positive or NAAT positive only. Medical charts were reviewed. Multivariable logistic regression models were used to compare CDI-related complications, recurrence, and 30-day mortality between the 2 groups. Results Of 4878 cases, 2160 (44.3%) were toxin positive and 2718 (55.7%) were NAAT positive only. More toxin-positive than NAAT-positive-only cases were aged ≥65 years (48.2% vs 38.0%; P < .0001), had ≥3 unformed stools for ≥1 day (43.9% vs 36.6%; P < .0001), and had white blood cell counts ≥15 000 cells/µL (31.4% vs 21.4%; P < .0001). In multivariable analysis, toxin positivity was associated with recurrence (adjusted odds ratio [aOR], 1.89; 95% confidence interval [CI], 1.61–2.23), but not with CDI-related complications (aOR, 0.91; 95% CI, .67–1.23) or 30-day mortality (aOR, 0.95; 95% CI, .73–1.24). Conclusions Toxin-positive CDI is more severe, but there were no differences in adjusted CDI-related complication and mortality rates between toxin-positive and NAAT-positive-only CDI that were detected by an algorithm that utilized an initial glutamate dehydrogenase screening test.

Key Points Question What are the estimated sample sizes needed to adequately power parallel cluster-randomized trials with common health care–associated infection outcomes, and how do the parameters affect these estimates in the field of hospital infection control and antimicrobial stewardship? Findings This cohort study found that large sample sizes were needed to appropriately power cluster-randomized trials in the field of hospital epidemiology, because the outcomes are rare. The expected effectiveness of the intervention and the strength of correlation within a cluster had the greatest association with the estimated sample size. Meaning These findings suggest that better-designed cluster-randomized trials in the field of hospital epidemiology and antimicrobial stewardship that appropriately account for clustering and realistic effect sizes will provide a more reliable evidence base for advancing recommendations and best practices., This study of national rates of health care–associated infections estimates the number of hospitals needed to power parallel cluster-randomized trials of interventions and evaluates how different parameters such as coefficient of variation and expected effect size are associated with the sample size estimates in practice., Importance An important step in designing, executing, and evaluating cluster-randomized trials (CRTs) is understanding the correlation and thus nonindependence that exists among individuals in a cluster. In hospital epidemiology, there is a shortage of CRTs that have published their intraclass correlation coefficient or coefficient of variation (CV), making prospective sample size calculations difficult for investigators. Objectives To estimate the number of hospitals needed to power parallel CRTs of interventions to reduce health care–associated infection outcomes and to demonstrate how different parameters such as CV and expected effect size are associated with the sample size estimates in practice. Design, Setting, and Participants This longitudinal cohort study estimated parameters for sample size calculations using national rates developed by the Centers for Disease Control and Prevention for methicillin-resistant Staphylococcus aureus (MRSA) bacteremia, central-line–associated bloodstream infections (CLABSI), catheter-associated urinary tract infections (CAUTI), and Clostridium difficile infections (CDI) from 2016. For MRSA and vancomycin-resistant enterococci (VRE) acquisition, outcomes were estimated using data from 2012 from the Benefits of Universal Glove and Gown study. Data were collected from June 2017 through September 2018 and analyzed from September 2018 through January 2019. Main Outcomes and Measures Calculated number of clusters needed for adequate power to detect an intervention effect using a 2-group parallel CRT. Results To study an intervention with a 30% decrease in daily rates, 73 total clusters were needed (37 in the intervention group and 36 in the control group) for MRSA bacteremia, 82 for CAUTI, 60 for CLABSI, and 31 for CDI. If a 10% decrease in rates was expected, 768 clusters were needed for MRSA bacteremia, 875 for CAUTI, 631 for CLABSI, and 329 for CDI. For MRSA or VRE acquisition, 50 or 40 total clusters, respectively, were required to observe a 30% decrease, whereas 540 or 426 clusters, respectively, were required to detect a 10% decrease. Conclusions and Relevance This study suggests that large sample sizes are needed to appropriately power parallel CRTs targeting infection prevention outcomes. Sample sizes are most associated with expected effect size and CV of hospital rates.

Although critical to prevent healthcare-associated infections, hand hygiene (HH) compliance is poor in resource-limited settings. In 2012, three Kenyan hospitals began onsite production of alcohol-based handrub (ABHR) and HH promotion. Our aim is to determine the impact of local production of ABHR on HH compliance and perceptions of ABHR. We observed 25,738 HH compliance opportunities and conducted 15 baseline and post-intervention focus group discussions. Hand Hygiene compliance increased from 28% (baseline) to 38% (post-intervention, p = 0.0003). Healthcare workers liked the increased accessibility of ABHR, but disliked its smell, feel, and sporadic availability. Onsite production and promotion of ABHR resulted in modest HH improvement. Enhancing the quality of ABHR and addressing logistical barriers could improve program impact.

Background Escherichia coli is a common cause of community-onset (CO) infections, including urinary tract and abdominal infections, and CO sepsis. Fluoroquinolones (FQ) are used in the empiric treatment of E. coli infections, but FQ-resistance may limit their effectiveness. We examined trends and regional differences in FQ-resistant E. coli clinical cultures among hospitalized adult patients in the U.S. Methods We measured the incidence of E. coli clinical cultures among hospitalized adults in a cohort of hospitals in the Premier Healthcare Database and Cerner Health Facts from 2012 through 2017. FQ resistance was defined as resistance to ciprofloxacin, levofloxacin, or moxifloxacin. Only cultures collected prior to day 4 of hospitalization, defined as CO, were considered. We extrapolated national estimates using a raking procedure to generate weighted adjustments matching the American Hospital Association distribution for U.S. acute care hospitals. Weights were based on U.S. census division, bed size category, teaching status, and urban/rural designation. We used a weighted means survey procedure to calculate national estimates and weighted multivariable logistic regression to examine trends and regional differences. Results In 2017, we estimated 949,393 CO E. coli infections with FQ susceptibility testing; 312,304 (33%) were due to E. coli resistant to FQ. Of FQ-resistant E. coli isolates, 76% were isolated from urine. We did not observe a significant trend in FQ-resistant E. coli from 2012 to 2017 (p = 0.85). Percent FQ-resistant varied significantly by region (p < 0.0001) with an estimated range of 19% (Mountain) to 42% (Southeast Central) in 2017. We also found variability by hospital (2017 Q1: 26% and Q3: 39%). FQ-resistance rates were higher in urine (36%: 95% CI 34-38%) than blood isolates (27%: 95% CI 26-29%) and higher for males (40%: 95% CI 38-42%) than females (33%: 95% CI 31-35%). Conclusion FQ-resistance is common in CO E. coli infections with significant variability by region and hospital. Empiric FQ treatment for infectious syndromes commonly caused by E. coli may need to be reconsidered. Clinicians should consult with local antibiograms and antibiotic stewardship programs to determine the most appropriate empiric treatment of E. coli infections in hospitalized adults. Disclosures All Authors: No reported disclosures

Background Decreasing inappropriate urine cultures in hospitalized patients has been a target of diagnostic stewardship to improve infection surveillance and antimicrobial use. The impact of such efforts has been largely unstudied. This study assessed temporal trends in urine culture rates in a cohort of acute care hospitals (ACHs) between 2012 and 2017. Hospital Level Variation in Admission Urine Culture Rates Hospital Level Variation in Post-admission Urine Culture Rates Methods We used microbiology data from ACHs participating in the Premier Healthcare Database and Cerner Health Facts to measure monthly urine culture rates. All cultures from the urinary tract collected on or before day 3 were defined as admission cultures (AC) and those collected on day 4 or later as post-admission cultures (PAC). Temporal trends in AC and PAC rates were estimated using general estimating equation models adjusting for hospital-level clustering, hospital size, teaching status, urban/rural designation, discharge month, and region. Results During the study period, ACHs had 20.8 million discharges and performed 4,946,717 urine cultures, of which 21% were PAC. In 2012 and 2017, the unadjusted AC rates were 18.7 and 18.4 per 100 discharges; the unadjusted PAC rates were 11.5 and 8.5 per 1,000 patient days (PD) respectively. The median annual hospital-level AC rate was 17.2 with inter-hospital variation ranging from 12.7 (quartile 1) to 24.1 (quartile 3) per 100 discharges, Figure 1. Similarly, the PAC rates varied among the ACHs with a median of 7.1 and inter-hospital variation ranging from 4.6 (quartile 1) to 10.5 (quartile 3) per 1,000 PDs, Figure 2. In multivariable analysis, no temporal trends in AC rates were detected (rate ratio (RR) 1.01; 95% confidence interval (CI): 0.99–1.02). However, PAC rates decreased 6.3% annually (RR 0.937; 95% CI: 0.93–0.95). Factors significantly associated (p< 0.02) with PAC rates were discharge month, teaching status, bed size, and region. For AC, significant associations (p< 0.0001) were discharge month and region. Conclusion Between 2012 and 2017, the rate of AC remained unchanged, but PAC rates decreased significantly. Factors driving this trend are unknown, but potential explanations include changes in culturing practices and/or decreases in hospital-onset urinary tract infections. Understanding factors related to the decrease and the impact on patient outcomes warrants further study. Disclosures All Authors: No reported disclosures

Background: Studies on the effectiveness of hospital-based interventions often measure hospital-onset infections as the outcome of interest. However, hospital-associated infections may manifest after patient discharge (classified as hospital-associated community-onset, HACO), and the epidemiology may vary by antibiotic resistance (AR) profile. We examined the epidemiology and trends of HACO infections of AR and non–antibiotic-resistant (non-AR) bacteria. Methods: We included clinical community-onset (CO) cultures (obtained sooner than or on day 3 of hospitalization) yielding the bacterial species of interest among hospitalized patients in 260 hospitals in the Premier Healthcare Database from 2012 to 2017. HACO infections were defined as CO cultures in a patient who had a previous hospitalization in the same hospital within 30 days. We examined methicillin resistance among Staphylococcus aureus (MRSA), vancomycin resistance among Enterococcus spp (VRE), carbapenem resistance among Enterobacteriaceae (E. coli, Klebsiella spp, and Enterobacter spp) (CRE), extended-spectrum cephalosporin resistance suggestive of extended-spectrum β-lactamase (ESBL) production in Enterobacteriaceae, carbapenem resistance among Acinetobacter spp (CRAsp), and carbapenem resistance among Pseudomonas aeruginosa (CRPA). We described the proportion of CO infections that were HACO, the proportion of HACO infections from sterile sites, overall HACO rates, and annual trends for sensitive and resistant phenotypes. Generalized estimating equation regression models that accounted for hospital-level clustering were used to estimate annual trends controlling for hospital characteristics and month of discharge. Results: The rate of HACO infections by pathogen ranged from 0.78 to 38.76 per 10,000 hospitalizations; 7%–34% were sterile site infections (Table 1). For each bacterial pathogen, a significantly higher proportion of AR CO infections had a previous hospitalization compared to non-AR CO infections (all χ2, P < .05). The annual trends for AR and non-AR HACO infections between 2012 and 2017 were significantly decreasing for most pathogens, except ESBL HACO infections. Conclusions: Even when using a definition limited to readmission to the same hospital, HACO infections occur commonly with differing rates by pathogen and antibiotic resistance profile. Although these rates are decreasing for most of the pathogens studied, improving surveillance and identifying prevention strategies for these infections are necessary to further reduce the burden of hospital-associated infections.Funding: NoneDisclosures: None

Background Unnecessary antibiotic use (AU) contributes to increased rates of Clostridioides difficile infection (CDI). The impact of antibiotic restriction on hospital-onset CDI (HO-CDI) has not been assessed in a large group of US acute care hospitals (ACHs). Methods We examined cross-sectional and temporal associations between rates of hospital-level AU and HO-CDI using data from 549 ACHs. HO-CDI was defined as a discharge with a secondary International Classification of Diseases, Ninth Revision, Clinical Modification code for CDI (008.45), and treatment with metronidazole or oral vancomycin > 3 days after admission. Analyses were performed using multivariable generalized estimating equation models adjusting for patient and hospital characteristics. Results During 2006–2012, the unadjusted annual rates of HO-CDI and total AU were 7.3 per 10 000 patient-days (PD) (95% confidence interval [CI], 7.1–7.5) and 811 days of therapy (DOT)/1000 PD (95% CI, 803–820), respectively. In the cross-sectional analysis, for every 50 DOT/1000 PD increase in total AU, there was a 4.4% increase in HO-CDI. For every 10 DOT/1000 PD increase in use of third- and fourth-generation cephalosporins or carbapenems, there was a 2.1% and 2.9% increase in HO-CDI, respectively. In the time-series analysis, the 6 ACHs with a ≥30% decrease in total AU had a 33% decrease in HO-CDI (rate ratio, 0.67 [95% CI, .47–.96]); ACHs with a ≥20% decrease in fluoroquinolone or third- and fourth-generation cephalosporin use had a corresponding decrease in HO-CDI of 8% and 13%, respectively. Conclusions At an ecologic level, reductions in total AU, use of fluoroquinolones, and use of third- and fourth-generation cephalosporins were each associated with decreased HO-CDI rates.

BackgroundAs the US population ages, the number of hip and knee arthroplasties is expected to increase. Because surgical site infections (SSIs) following these procedures contribute substantial morbidity, mortality, and costs, we projected SSIs expected to occur from 2020 through 2030.MethodsWe used a stochastic Poisson process to project the number of primary and revision arthroplasties and SSIs. Primary arthroplasty rates were calculated using annual estimates of hip and knee arthroplasty stratified by age and gender from the 2012–2014 Nationwide Inpatient Sample and standardized by census population data. Revision rates, dependent on time from primary procedure, were obtained from published literature and were uniformly applied for all ages and genders. Stratified complex SSI rates for arthroplasties were obtained from 2012–2015 National Healthcare Safety Network data. To evaluate the possible impact of prevention measures, we recalculated the projections with an SSI rate reduced by 30%, the national target established by the US Department of Health and Human Services (HHS).ResultsWithout a reduction in SSI rates, we projected an increase in complex SSIs following hip and knee arthroplasty of 14% between 2020 and 2030. We projected a total burden of 77,653 SSIs; however, meeting the 30% rate reduction could prevent 23,297 of these SSIs.ConclusionsGiven current SSI rates, we project that complex SSI burden for primary and revision arthroplasty may increase due to an aging population. Reducing the SSI rate to the national HHS target could prevent 23,000 SSIs and reduce subsequent morbidity, mortality, and Medicare costs.

Background Opioid misuse is epidemic in the United States (US), and persons who inject drugs are at increased risk for serious bacterial and fungal infections, including Candida bloodstream infections. Historically, candidemia has occurred almost exclusively among patients with severe underlying illness and extensive healthcare exposure. We examined whether the opioid crisis may be having an impact on the epidemiology of candidemia in the United States. Methods Using data from 200 US hospitals reporting to the Premier Healthcare Database (PHD) between 2012–2017, we conducted a retrospective study among hospitalized persons ≥ 18 years. Candidemia was defined by any blood culture yielding Candida species. Drug use-associated (DUA)-candidemia hospitalizations were defined as hospitalizations having both candidemia and at least one ICD-9-CM or ICD-10-CM diagnostic code for recreational drug use; drugs were classified as opioids, cocaine, amphetamines, or other drugs (excluding cannabis, alcohol, and nicotine). We described the characteristics and annual trends of candidemia hospitalizations, stratified by drug use. Results Of 7,590 candidemia hospitalizations during 2012–2017, 679 (9%) were DUA-candidemia. During this time, the rate of DUA-candidemia increased from 3.6 to 9.1 per 100,000 hospitalizations, while the rate of non-DUA-candidemia decreased from 64.7 to 55.6 per 100,000 hospitalizations. Patients with DUA-candidemia were younger (median 40 vs. 64 years), had a longer lengths of stay (median 14 vs. 13 days), and had lower in-hospital mortality (12% vs. 26%). Among DUA-candidemia hospitalizations, opioids accounted for 78% of substances identified. Among patients aged 18–44 years, the proportion of candidemia hospitalizations associated with drug use more than tripled from 13% in 2012 to 44% in 2017 (Figure 1). Conclusion DUA-candidemia hospitalizations increased almost 3-fold during 2012–2017, with drug use identified in nearly half of candidemia patients ages 18–44 years in 2017. These data suggest that the opioid crisis is having an impact on the epidemiology of candidemia in the United States, especially among young adults, underscoring an additional negative consequence of the ongoing crisis. Disclosures All authors: No reported disclosures.

Background Difficult-to-treat resistance (DTR) is a metric for clinically relevant “pan-resistance” to available high-efficacy, low-toxicity antibiotic treatment options at any given time. Previous DTR prevalence estimates in Gram-negative (GN) bloodstream isolates from 2009 to 2014 have ranged between 1 and 1.5%. We sought to estimate the national burden of DTR GN isolates and more recent trends by region, site, and species. Methods Clinical cultures with GN isolates were identified from inpatient encounters in hospitals reporting at least one culture with susceptibility testing for a given month to Premier Healthcare Database or Cerner Health Facts Database from 2012 to 2017. DTR was defined as intermediate susceptibility or resistance to all tested carbapenems, other β-lactams, and fluoroquinolones, but not including agents introduced 2014 onwards. For each year, a raking procedure generated weights to extrapolate the sample estimate to match American Hospital Association distributions based on US census division, hospital bed capacity, teaching status, and urban designation. A weighted means survey procedure was used to extrapolate the sample estimate to obtain national DTR burden. Trends in DTR incidence were examined by using weighted multivariable logistic regression. Results Extrapolating from a 373-hospital sample, the estimated 2017 US inpatient burden of DTR isolates was 3,315 (1.3%) among sterile-site and 31,509 (1.7%) among all cultures, ranging from 0.5% to 3.3% in Mountain and New England regions respectively. P. aeruginosa was the most common species overall (37%), while A. baumannii was most common among sterile sites (31%). Between 2012 and 2017, there was no annual percent change in DTR incidence for sterile sites [OR 0.99 (0.93, 1.06)] but for all cultures it decreased 4.1% annually [OR 0.95 (0.91, 0.99)], including 9% annually for A. baumannii [OR 0.905 (0.860, 0.953)] and K. pneumonia [OR 0.903 (0.824, 0.991)], respectively. Conclusion The US inpatient burden of GN isolates displaying DTR is relatively low, varies by region, and has remained stable or declined slightly in recent years. Periodic inclusion of emerging antibiotics in the DTR classification will allow for a dynamic index between resistance and available agents. Disclosures All Authors: No reported Disclosures.

Background Extended spectrum β-lactamase-producing Enterobacteriaceae (ESBLs) have been identified as a serious antibiotic-resistant threat. Studies have shown that ESBL infection rates were increasing through 2014. Our objective was to examine more recent ESBL trends and to evaluate differences across regions in the United States. Methods We measured the incidence of positive clinical cultures from inpatient encounters in a cohort of hospitals submitting data to the Premier Healthcare Database and Cerner Health Facts from 2012 through 2017. We included Escherichia coli and Klebsiella spp. cultures and defined ESBL as non-susceptibility to cefotaxime, ceftriaxone, ceftazidime, or cefepime. Cultures collected on days 1, 2, or 3 of hospitalization were considered community-onset (CO); cultures from day 4 or later were considered hospital onset (HO). We developed weights using a raking procedure to match the American Hospital Association distribution for acute care hospitals based on US census division, bed size category, teaching status, and urban/rural designation. We used weighted multivariable logistic regression adjusting for hospital characteristics to examine trends and regional differences in ESBL rates. Results In 2017, the estimated rate of ESBLs was 40.3 per 10,000 discharges for CO and 6.4 per 10,000 discharges for HO; 86% of all ESBLs were CO. The percent that were ESBLs among all included cultures increased for CO (8.2% in 2012 to 11.6% in 2017) and HO (13.1 to 16.8%) cultures. From 2012 – 2017, adjusted ESBL rates increased for CO (7.9% annually, P < 0.0001), while HO rates did not change significantly over time (P = 0.39, Figure 1). We found significant regional differences in the rates of ESBL (P < 0.0001) across US census divisions in 2017 (Figure 2). Estimated rates for 2017 varied 5-fold from 15.3 ESBLs per 10,000 discharges in the Northwest Central to 82.4 ESBLs in the Mid-Atlantic. Conclusion We estimated a 40% increase in the rate of CO-ESBLs among hospitalized patients from 2012 to 2017, but no increase in HO rates. ESBL rates varied greatly by region of the country and are estimated as much as 5× higher in some areas. A better understanding of factors contributing to community transmission and regional variation is necessary in order to inform ESBL prevention efforts. Disclosures All authors: No reported disclosures.

Background Despite overall progress in preventing Clostridioides difficile Infection (CDI), community-associated (CA) infections have been steadily increasing. Although the incubation period of CDI is thought to be relatively short, gastrointestinal microbial disruption from remote healthcare exposures (e.g., inpatient antibiotic use) may be associated with CA-CDI. To assess this potential association, we linked CA-CDI infections identified through CDC’s Emerging Infections Program (EIP) to Medicare claims data to describe prior healthcare utilization. Methods We defined an EIP CA-CDI case as a positive C. difficile test collected in 2014–2015 from an outpatient or inpatient within 3 days of hospital admission, provided there was no positive test in the prior 8 weeks and no admission to a healthcare facility in the prior 12 weeks. We linked EIP CA-CDI cases aged ≥65 years to a Medicare beneficiary using unique combinations of birthdate, sex, and zip code. Cases were included if they maintained continuous fee-for-service coverage for 1 year prior to the event date. To calculate exposure odds ratios for previous hospitalizations, each case was matched to 5 control beneficiaries on age, sex, and county of residence. We used logistic regression to calculate adjusted matched odds ratios (amOR) that controlled for chronic conditions. Results We successfully linked 2,287/3,367 (68%) EIP CA-CDI cases. Of these, 1,236 cases met inclusion criteria; the median age was 77 years and 63% were female. We identified 69 (5.6%) cases with misclassification of prior healthcare exposures, most of whom (48, 70%) were hospitalized in the 12 weeks prior to their event. Among the 1,167 true CA-CDI cases, 33% were hospitalized in the prior 12 weeks to 1 year. The median number of weeks from prior hospitalization to CDI was 27 (IQR 18–38, Figure 1). Cases had a higher risk of hospitalization than matched controls in the prior 3–6 months (amOR: 2.33, 95% CI: 1.87, 2.90) and 6–12 months (amOR: 1.43 95% CI: 1.18, 1.74). Conclusion Remote hospitalization in the previous year was a significant risk factor for CA-CDI, especially in the 3–6 months prior to CA-CDI. Long-lasting prevention strategies implemented at hospital discharge and enhanced inpatient antibiotic stewardship may prevent CA-CDI among older adults. Disclosures All Authors: No reported Disclosures.

Background US laboratories are increasingly using nucleic acid amplification tests (NAAT) to diagnose Clostridium difficile infection (CDI) due to their increased sensitivity over toxin enzyme immunoassays (EIA), but NAATs may be more likely than toxin EIAs to detect colonization rather than true disease. Limited data indicate patients positive by toxin EIA (toxin+) have worse outcomes than those positive by NAAT (NAAT+) only, suggesting toxin EIA detects true infection more often than NAAT. We used multisite CDI surveillance data from the Centers for Disease Control and Prevention’s Emerging Infections Program to compare clinical course and outcomes between toxin+ and NAAT+ only patients. Methods A case was defined as a positive C. difficile test in a person ≥1 year old with no positive tests in the prior 8 weeks. Cases detected during 2014–2015 by a testing algorithm using toxin EIA and NAAT were classified as toxin+ or NAAT+ only. Medical charts were reviewed. Death data were obtained from state death registries. Multivariable logistic regression models were used to compare CDI recurrence and 90-day mortality between the two groups, adjusting for age, sex, race, Charlson comorbidity index, and receipt of oral vancomycin. For the outcome of recurrence, we also adjusted for history of CDI in the prior 6 months. Results Of 4,878 cases, 2160 (44%) were toxin+ and 2,718 (56%) were NAAT+ only. Toxin+ cases were more likely than NAAT+ only cases to be ≥65 years old (48% vs. 38%; P < 0.0001), have white blood cells ≥15,000/µL (483/1,539 [31%] vs. 423/1,978 [21%]; P < 0.0001), and have received oral vancomycin ≤3 days of diagnosis (32% vs. 29%; P = 0.03). Comparing toxin+ to NAAT+ only cases, 21% vs. 11% had a recurrence (P < 0.0001), of which 71% vs. 33% had a toxin+ recurrence (P < 0.0001), and 10% vs. 9% died ≤90 days of diagnosis (P = 0.12). In multivariable analysis, a toxin+ result was associated with recurrence (adjusted odds ratio [aOR]: 1.89, 95% CI: 1.61–2.22) but not with 90-day mortality (aOR: 0.99; 95% CI: 0.81–1.22). Conclusion Toxin+ CDI is more severe by some markers and more likely to recur as toxin+. However, there was no difference in adjusted mortality, which may reflect an effect on mortality in NAAT+ only cases from mild CDI, receipt of unnecessary CDI treatment, or other factors. Disclosures G. Dumyati, Seres: Scientific Advisor, Consulting fee.

Background Carbapenem-nonsusceptible Acinetobacter spp. (CNAB) and carbapenem-resistant Pseudomonas aeruginosa (CRPA) are recognized causes of severe and difficult to treat healthcare-associated infections. This study estimated and compared the incidence of CNAB and CRPA among patients admitted to US acute care hospitals in 2012–2017. Methods We measured the incidence of positive clinical cultures from inpatient encounters in a cohort of over 300 hospitals submitting data to the Premier Healthcare Database and Cerner Health Facts in 2012–2017. We included clinical cultures from any body site yielding Acinetobacter spp./P. aeruginosa non-susceptible/resistant to imipenem, meropenem, or doripenem. Cultures collected on days 1–3 of hospitalization were considered community-onset (CO) and cultures from later were hospital-onset (HO). Duplicate isolates identified within 14 days of an incident culture and surveillance cultures were excluded. For each year, a raking procedure generated weights to extrapolate the sample estimate to match the American Hospital Association distributions based on US census division, hospital bed capacity, teaching status, and urban designation. We compared estimated rates in 2017 vs. 2012 using weighted multivariable logistic regression adjusting for hospital characteristics and hospital-level clustering. Results In 2017, the estimated rates of HO and CO CNAB rates were 0.77 and 1.39/10,000 discharges, and HO and CO CRPA rates were 3.14 and 6.57, respectively. Compared with 2017, rates of HO CNAB decreased 49% (Odds Ratio (OR) 0.51; 95% CI: 0.34–0.75) and rates of CO CNAB decreased 29% (OR 0.71; 95% CI: 0.54–0.92). For CRPA, the incidence of HO decreased (OR 0.66; CI: 0.49–0.88) with no change in CO rates (OR 0.93; CI: 0.79–1.11). Assessment of cultures from sterile sites alone showed similar results, but they did not reach statistical significance, Figure 1. Conclusion We estimate significant national decreases in the rates of HO and CO CNAB, and HO CRPA. Risk factors and effective interventions to reduce CO CRPA might differ from CNAB and HO CRPA. Additional prevention strategies are needed to address CO CRPA. Disclosures All authors: No reported disclosures.

Background The National Action Plan for Combating Antibiotic-resistant Bacteria calls for monitoring inpatient antibiotic use to inform stewardship efforts. We estimated national trends in inpatient antibiotic usage from 2012 to 2017 in a large cohort of US hospitals. Methods We utilized the Premier Healthcare Database, containing detailed administrative records available by census region, including inpatient drug utilization data based on billing records, for all patients discharged from a convenience sample of over 700 US hospitals annually, approximately 20% of US inpatient discharges. We retrospectively estimated days of therapy (DOT)/1,000 patient-days (PDs) by year overall and by antibiotic class. To examine trends over time, we created multivariable models adjusting for hospital-specific location of antibiotic use (ICU vs. other) and hospital-specific summary measures including average patient age, average co-morbidity score, case mix index, number of hospital beds, teaching status, urban/rural location, US census division, proportion of discharges with a surgical diagnosis-related code, and proportion of PDs with an infectious disease primary ICD-9/10-CM discharge code. Estimates and models were weighted to be nationally representative using projected weights from the database. Results 58% of patients had at least one antibiotic DOT, and the overall DOT for all hospitals was 810 DOT/1,000 PDs (interquartile range 701 to 913 DOT/1,000 PDs). Glycopeptides and third-/fourth-generation cephalosporins were the most common antibiotic classes (Figure 1). Overall antibiotic DOT did not change significantly over time, P = 0.9133. However, class-specific DOT varied with large decreases in fluoroquinolones from 2012 to 2017 (55% decrease, P < 0.0001), and large increases in third-/fourth-generation cephalosporins and tetracyclines (32% and 49% increase, respectively, P < 0.0001) (Figure 2). Overall antibiotic DOT significantly varied among US census divisions (Figure 3). Conclusion Estimated overall inpatient antibiotic DOT did not change in US hospitals from 2012 to 2017, but there were significant class-specific changes. The large decrease in fluoroquinolone use may reflect increased awareness of adverse events. Disclosures All Authors: No reported Disclosures.

Background Evidence suggests that interventions such as MRSA decolonization are useful in the prevention of MRSA bloodstream infections (BSI) both during hospitalization and post-discharge. However, decolonization may be costly and have diminishing effectiveness when used on all inpatients. Hospital length of stay (LOS) is a known risk factor for MRSA BSI. To determine whether LOS could be useful in prioritizing patients for intervention, we aimed to evaluate (i) distribution of time from admission to hospital-onset (HO) MRSA BSI, and (ii) frequency and LOS of hospitalizations that preceded community-onset (CO) MRSA BSI. Methods MRSA-positive blood cultures among adults admitted to New York hospitals from 2013 to 20s16 were identified in the Centers for Disease Control and Prevention’s (CDC’s) National Healthcare Safety Network (NHSN). We linked these data to admissions in New York’s hospital discharge dataset, the Statewide Planning and Research Cooperative System (SPARCS), where the NHSN blood culture collection date was between a patient’s SPARCS admission and discharge dates and there was an exact match for birth date, gender and facility. Time to MRSA BSI was defined as the number of days from admission (day 1) to collection of a blood culture positive for MRSA. We defined positive blood cultures collected on days 1–3 as CO, and those collected ≥day 4 as HO. Results We linked 10,425 (79%) MRSA BSIs from NHSN to SPARCS. 78% (8,147) of MRSA BSIs were CO and 22% (2,278) were HO. The median time to HO MRSA BSI was 10 days (IQR 6–21) (Figure 1), in contrast to the median LOS for all hospitalizations of 4 days (IQR 3–7). By definition, 35% of all hospitalizations were never at risk of HO MRSA BSI because their LOS was < 4 days. Among CO MRSA BSI, 48% were discharged from a hospital in the 90 days preceding their BSI (Figure 2). The median LOS of these prior hospitalizations was 8 days (IQR 5–14), and 87% were at least 4 days in length. Conclusion Over half of HO MRSA BSI occur on or after day 10 of hospitalization and a large fraction of CO MRSA BSI had a lengthy hospitalization shortly before their BSI diagnosis. Our results suggest that patients likely to have a long LOS could be evaluated as potential targets for prevention strategies (e.g., decolonization) to reduce both HO and CO MRSA BSI. Disclosures All authors: No reported disclosures.

Background The age-adjusted rate of drug overdose deaths in the United States tripled from 1999 to 2016. Public health surveillance data indicate that an increasing proportion of infections due to bacterial and fungal pathogens is associated with injection drug use (IDU). We describe healthcare encounters (HCEs) of PWID as potential opportunities to prevent infections related to IDU by identifying risks and treating SUD, including with medication-assisted treatment (MAT) for opioid use disorder. Methods At six hospitals in western New York, we abstracted medical records from hospital admissions and emergency department (ED) visits for PWID (i.e., IDU in the preceding year) who had positive cultures for Staphylococcus aureus (any clinical specimen, April–July 2017), group A Streptococcus (invasive specimens, all of 2017) or Candida spp. (blood specimens, all of 2017). We reviewed hospital admission and ED records for 1 year preceding the positive culture to identify visits during which opportunities to prevent infection and treat SUD by addressing SUD and IDU were missed. Results We identified 99 PWID with positive cultures. The median age was 33 years (range 19–68) and 61 were female. Sixty-nine had a skin and soft-tissue infection, 44 had a bloodstream infection, and 20 had both. Thirty-one PWID left against medical advice during a hospital admission or an ED visit. Seventy-nine PWID were hospitalized, of whom 4 died. Ninety-five used opioids and 71 used cocaine in the preceding year. Seventy-five PWID had an HCE in the 12 months prior to the index visit, with a median of two HCE per person (interquartile range 1–4); 53 of PWID had a previous HCE for infection and 28 for opioid overdose. SUD was documented during a prior HCE at the same hospital for 61 PWID, but only 10 (16%) were offered MAT during any prior HCE and for 24 (39%) there was no documentation that any form of treatment for SUD was offered. Conclusion In this cohort, PWID frequently had one or more healthcare encounters documented at the same hospital in the year prior to a serious bacterial or fungal infection. These prior HCEs were often for infections or overdose that signaled the need for MAT, demonstrating that there are critical missed opportunities to identify risks, prevent infection, and treat SUD. Disclosures All Authors: No reported Disclosures.