Changes in prospectively collected patient-reported outcomes among women with incident endometrial cancer

PURPOSE

We examined associations between patient and treatment characteristics with longitudinally collected patient-reported outcome (PRO) measures to provide a data-informed description of the experiences of women undergoing treatment for endometrial cancer.

METHODS

We administered National Institutes of Health Patient Reported Outcomes Measurement Information System (PROMIS) questionnaires at the preoperative visit and at 6 and 12 months after surgery. Anxiety, depression, fatigue, sleep disturbance, pain, physical function, and ability to participate in social roles were assessed. Analysis of variance (ANOVA) and linear mixed models were used to examine associations between patient characteristics and PRO measures at baseline and through time.

RESULTS

Of 187 women enrolled, 174 (93%) and 103 (69%) completed the 6- and 12-month questionnaires, respectively. Anxiety was substantially elevated at baseline (half of one population-level standard deviation) and returned to general population mean levels at 6 and 12 months. Younger age, Medicaid/None/Self-pay insurance, prevalent diabetes, and current smoking were associated with higher symptom burden on multiple PRO measures across the three time points. Women with aggressive histology, higher disease stage, or those with adjuvant treatment had worse fatigue at 6 months, which normalized by 12 months.

CONCLUSIONS

We observed a high symptom burden at endometrial cancer diagnosis, with most PRO measures returning to general population means by 1 year. Information on risk factor-PRO associations can be used during the clinical visit to inform supportive service referral.

IMPLICATIONS FOR CANCER SURVIVORS

These findings can inform clinicians' discussions with endometrial cancer survivors regarding expected symptom trajectory following diagnosis and treatment.

Exploring the Incorporation of a Novel Cardiotoxicity Mobile Health App Into Care of Patients With Cancer: Qualitative Study of Patient and Provider Perspectives

BACKGROUND

Cardiotoxicity is a limitation of several cancer therapies and early recognition improves outcomes. Symptom-tracking mobile health (mHealth) apps are feasible and beneficial, but key elements for mHealth symptom-tracking to indicate early signs of cardiotoxicity are unknown.

OBJECTIVE

We explored considerations for the design of, and implementation into a large academic medical center, an mHealth symptom-tracking tool for early recognition of cardiotoxicity in patients with cancer after cancer therapy initiation.

METHODS

We conducted semistructured interviews of >50% of the providers (oncologists, cardio-oncologists, and radiation oncologists) who manage cancer treatment-related cardiotoxicity in the participating institution (n=11), and either interviews or co-design or both with 6 patients. Data were coded and analyzed using thematic analysis.

RESULTS

Providers indicated that there was no existing process to enable early recognition of cardiotoxicity and felt the app could reduce delays in diagnosis and lead to better patient outcomes. Signs and symptoms providers recommended for tracking included chest pain or tightness, shortness of breath, heart racing or palpitations, syncope, lightheadedness, edema, and excessive fatigue. Implementation barriers included determining who would receive symptom reports, ensuring all members of the patient's care team (eg, oncologist, cardiologist, and primary care) were informed of the symptom reports and could collaborate on care plans, and how to best integrate the app data into the electronic health record. Patients (n=6, 100%) agreed that the app would be useful for enhanced symptom capture and education and indicated willingness to use it.

CONCLUSIONS

Providers and patients agree that a patient-facing, cancer treatment-related cardiotoxicity symptom-tracking mHealth app would be beneficial. Additional studies evaluating the role of mHealth as a potential strategy for targeted early cardioprotective therapy initiation are needed.

Emergency department patient and physician survey accuracy compared to chart abstraction in patients with acute respiratory illness

BACKGROUND

High-quality research studies in older adults are needed. Unfortunately, the accuracy of chart review data in older adult patients has been called into question by previous studies. Little is known on this topic in patients with suspected pneumonia, a disease with 500,000 annual older adult U.S. emergency department (ED) visits that presents a diagnostic challenge to ED physicians. The study objective was to compare direct interview and chart abstraction as data sources.

METHODS

We present a preplanned secondary analysis of a prospective, observational cohort of ED patients ≥65 years of age with suspected pneumonia in two Midwest EDs. We describe the agreement between chart review and a criterion standard of prospective direct patient survey (symptoms) or direct physician survey (examination findings). Data were collected by chart review and from the patient and treating physician by survey.

RESULTS

The larger study enrolled 135 older adults; 134 with complete symptom data and 129 with complete examination data were included in this analysis. Pneumonia symptoms (confusion, malaise, rapid breathing, any cough, new/worse cough, any sputum production, change to sputum) had agreement between patient/legally authorized representative survey and chart review ranging from 47.8% (malaise) to 80.6% (confusion). All examination findings (rales, rhonchi, wheeze) had percent agreement between physician survey and chart review of ≥80%. However, all kappas except wheezing were less than 0.60, indicating weak agreement.

CONCLUSIONS

Both patient symptoms and examination findings demonstrated discrepancies between chart review and direct survey with larger discrepancies in symptoms reported. Researchers should consider these potential discrepancies during study design and data interpretation.

Understanding comorbidities and health disparities related to COVID-19: a comprehensive study of 776 936 cases and 1 362 545 controls in the state of Indiana, USA

Objective

To characterize COVID-19 patients in Indiana, United States, and to evaluate their demographics and comorbidities as risk factors to COVID-19 severity.

Materials and Methods

EHR data of 776 936 COVID-19 cases and 1 362 545 controls were collected from the COVID-19 Research Data Commons (CoRDaCo) in Indiana. Data regarding county population and per capita income were obtained from the US Census Bureau. Statistical analysis was conducted to determine the association of demographic and clinical variables with COVID-19 severity. Predictive analysis was conducted to evaluate the predictive power of CoRDaCo EHR data in determining COVID-19 severity.

Results

Chronic obstructive pulmonary disease, cardiovascular disease, and type 2 diabetes were found in 3.49%, 2.59%, and 4.76% of the COVID-19 patients, respectively. Such COVID-19 patients have significantly higher ICU admission rates of 10.23%, 14.33%, and 11.11%, respectively, compared to the entire COVID-19 patient population (1.94%). Furthermore, patients with these comorbidities have significantly higher mortality rates compared to the entire COVID-19 patient population. Health disparity analysis suggests potential health disparities among counties in Indiana. Predictive analysis achieved F1-scores of 0.8011 and 0.7072 for classifying COVID-19 cases versus controls and ICU versus non-ICU cases, respectively.

Discussion

Black population in Indiana was more adversely affected by COVID-19 than the White population. This is consistent to findings from existing studies. Our findings also indicate other health disparities in terms of demographic and economic factors.

Conclusion

This study characterizes the relationship between comorbidities and COVID-19 outcomes with respect to ICU admission across a large COVID-19 patient population in Indiana.

2171. Clinical Characteristics and Outcomes of Immunocompromised Patients with Enterococcal Pneumonia

Background

Although Enterococcus species can colonize and cause infection in multiple body sites, it is considered a rare cause of pneumonia (PNA). The study objective was to describe clinical outcomes in immunocompromised patients with PNA and Enterococcus in respiratory cultures to better elucidate the pathogenic role played by Enterococci.

Methods

A retrospective, descriptive observational study was conducted on immunocompromised patients with Enterococcus spp on lower respiratory culture from January 1, 2015 to September 22, 2021 with clinical PNA. Data was collected on patient demographics, microbiologic details including bacteremia, other potential infectious causes of PNA and clinical outcomes.

Results

116 cases of immunocompromised patients with PNA and Enterococcus spp on respiratory culture were identified. Two patients also grew Enterococcus spp on pleural fluid and 1 on lung tissue. Vancomycin-resistant Enterococcus grew in 41/116 (35%) cases. Six (5%) patients had a bloodstream infection due to Enterococcus spp and a central line was present in all cases. In 52/116 (45%) cases there was growth of another organism, and 61/116 (53%) cases had an alternative infectious cause for PNA, whether it was a typical respiratory bacterial pathogen, viral illness, or invasive fungal infection. In 55 (47%) cases, there was no alternative infectious cause of PNA identified; 44 (80%) of these patients received an antibiotic active against Enterococcus spp and of these, 30 (68%) were clinically improved after 14 days. Overall mortality was high with 52/116 total cases (45%) in which the patient was alive after 90 days. In 25 cases patients did not receive antibiotic therapy active against Enterococcus and 28% were alive at 90 days.

Conclusion

Enterococcus spp should be considered a significant pulmonary pathogen in immunocompromised hosts in the appropriate clinical context. At our institution, the majority of immunocompromised patients with a monomicrobial pulmonary infection with Enterococcus were treated with targeted antibiotics. Further studies should evaluate outcomes of and indications for targeted treatment of pleuropulmonary Enterococcus.

Disclosures

All Authors: No reported disclosures.

480. Redefining the Epidemiologic Profile of Histoplasmosis from 2011 to 2021 in Central Ohio

Background

Histoplasma is a common endemic mycosis in multiple Midwestern states. Most of the data defining epidemiologic risk are from 20^th century outbreak investigations. A known exacerbating factor is immune suppression; but use of biologics, immunosuppressants (ISP), and chemotherapy has advanced since these investigations were conducted. This study sought to use clinical data from a tertiary care facility to address these gaps to define the evolving epidemiology of histoplasmosis in Ohio.

Methods

In a single-center retrospective cohort study, we identified adult patients with a histoplasmosis-specific diagnosis, a positive diagnostic test, and an itraconazole/amphotericin B prescription between 1/1/2012 and 10/1/2021. Prisoners were excluded. A random sample of 100 patients underwent in-depth chart review. Comorbidities, environmental risk factors, and ISP use data were collected. Cases were classified by the CDC 2017 Histoplasmosis case definition. “Confirmed” cases had culture growth, a confirmatory lab result, or histopathology plus signs and symptoms of disease, and “probable” cases had a non-confirmatory lab result plus signs and symptoms. Associations between variables were assessed using Fisher’s exact test, p < 0.05.

Results

Of 100 patients, 56 had confirmed and 22 had probable histoplasmosis (Table 1). Three had confirmed blastomycosis, 15 met definitions for both histoplasmosis and blastomycosis, and 4 met neither definition. These were excluded from analysis. No environmental exposure history was documented for 23 of 78 (29.4%) patients. Of those with documented history, 15 (27.1%) had recent renovation or construction exposure; 14 (25.4%) to soil, gardening, or outdoors; 12 (21.8%) to birds, bats, and farm animals; and 25 (45.4%) had no known exposure. There were 51 (65.4%) exposed to at least one dose of ISP in the year prior to diagnosis, including 41 (52.5%) to corticosteroids. Table 1

Impact of concomitant fluconazole on direct oral anticoagulant bleeding risk

STUDY OBJECTIVE

The aim of this study was to evaluate the effect on bleeding risk when fluconazole is administered concomitantly with direct oral anticoagulants (DOACs).

DESIGN

This was a retrospective cohort study including hospitalized adult patients prescribed a DOAC with or without fluconazole.

SETTING

The Ohio State University Wexner Medical Center, a tertiary care academic medical center with more than 1800 beds.

PATIENTS

Hospitalized patients ages 18-89 years who received apixaban or rivaroxaban with or without fluconazole from October 1, 2016, to September 30, 2021, were included. The minimum duration of DOAC or DOAC with fluconazole therapy was 48 h. Patients were excluded if they received fluconazole <400 mg daily or a DOAC at doses outside those recommended for atrial fibrillation or venous thromboembolism treatment or prophylaxis. Patients were matched based on DOAC received.

INTERVENTION

Patients who received a DOAC with fluconazole were compared with those receiving a DOAC alone. The primary outcome was a composite of major, clinically relevant nonmajor, and minor bleeding events at 30 days.

MEASUREMENTS AND MAIN RESULTS

There were 216 patients included, 108 in the DOAC with fluconazole group and 108 in the DOAC alone group. More patients in the DOAC with fluconazole group experienced bleeding at 30 days compared with the DOAC alone group [35/108 (32%) vs. 21/108 (19%), respectively; p = 0.03]; however, after adjusting for proven confounding variables (hemoglobin and concomitant carvedilol) this was found not to be statistically significant [adjusted odds ratio 1.71, 95% confidence interval 0.85-3.40].

CONCLUSIONS

Patients receiving a DOAC with fluconazole were not at significantly increased risk for bleeding at 30 days compared with those receiving a DOAC alone after controlling for confounding variables. As an increasing number of patients are prescribed DOACs, the results of this study may inform clinical decision-making on the safety of concomitant DOAC and fluconazole use.

Severe Monkeypox in Hospitalized Patients - United States, August 10-October 10, 2022

As of October 21, 2022, a total of 27,884 monkeypox cases (confirmed and probable) have been reported in the United States.§ Gay, bisexual, and other men who have sex with men have constituted a majority of cases, and persons with HIV infection and those from racial and ethnic minority groups have been disproportionately affected (1,2). During previous monkeypox outbreaks, severe manifestations of disease and poor outcomes have been reported among persons with HIV infection, particularly those with AIDS (3-5). This report summarizes findings from CDC clinical consultations provided for 57 patients aged ≥18 years who were hospitalized with severe manifestations of monkeypox¶ during August 10-October 10, 2022, and highlights three clinically representative cases. Overall, 47 (82%) patients had HIV infection, four (9%) of whom were receiving antiretroviral therapy (ART) before monkeypox diagnosis. Most patients were male (95%) and 68% were non-Hispanic Black (Black). Overall, 17 (30%) patients received intensive care unit (ICU)-level care, and 12 (21%) have died. As of this report, monkeypox was a cause of death or contributing factor in five of these deaths; six deaths remain under investigation to determine whether monkeypox was a causal or contributing factor; and in one death, monkeypox was not a cause or contributing factor.** Health care providers and public health professionals should be aware that severe morbidity and mortality associated with monkeypox have been observed during the current outbreak in the United States (6,7), particularly among highly immunocompromised persons. Providers should test all sexually active patients with suspected monkeypox for HIV at the time of monkeypox testing unless a patient is already known to have HIV infection. Providers should consider early commencement and extended duration of monkeypox-directed therapy†† in highly immunocompromised patients with suspected or laboratory-diagnosed monkeypox.§§ Engaging all persons with HIV in sustained care remains a critical public health priority.

Review of applied health informatics courses in a multidisciplinary biomedical informatics department

Introduction

Applied health informatics infrastructure is a requirement for learning health systems and it is imperative that we train a workforce that can support this infrastructure. Our department offers courses in several interdisciplinary programs with topics ranging from bioinformatics to population health informatics. Due to changes in the field and our faculty members, we sought to assess our courses relevant to applied health informatics.

Methods

In this paper, we discuss the three‐phase evaluation of our program and include the survey we developed to identify the skills and knowledge base of our faculty.

Results

We show how this assessment allowed us to identify gaps and develop strategies for program expansion.

Conclusions

A focus on workforce development can help to guide and focus curricular review in an interdisciplinary graduate program.

Concerns and frustrations about the public reporting of device-related healthcare-associated infections: Perspectives of hospital leaders and staff

BACKGROUND

Public reporting of healthcare-associated infections (HAIs) aims to incentivize improvement in infection prevention. The motivation and mechanisms of public reporting have raised concerns about the reliability of this data, but little is known about the specific concerns of hospital leaders and staff. This study sought to better understand perspectives of individuals in these roles regarding the identification and public reporting of HAIs.

METHODS

We conducted interviews with 471 participants including hospitals leaders (e.g., administrative and clinical leaders) and hospital staff (e.g., physicians and nurses) between 2017 and 2019 across 18 U.S. hospitals. A semi-structured interview guide was used to explore perspectives about the use of HAI data within the context of management strategies used to support infection prevention.

RESULTS

Interviewees described concerns about public reporting of HAI data, including a lack of trust in the data and inadvertent consequences of its public reporting, as well as specific frustrations related to the identification and accountability for publicly-reported HAIs.

CONCLUSION

Concerns and frustrations related to public reporting of HAI data highlight the need for improved guidelines, transparency, and incentives. Efforts to build trust in publicly-reported HAI data can help ensure this information is used effectively to improve infection prevention practices.

Implementing a Novel Interprofessional Clinical Informatics Curriculum

A novel interprofessional clinical informatics curriculum was developed, piloted, and implemented, using an academic medical record. Targeted learners included undergraduate, graduate, and professional students across five health science colleges. A team of educators and practitioners representing those five health science colleges was formed in 2016, to design, develop, and refine educational modules covering the essentials of clinical informatics. This innovative curriculum consists of 10 online learning modules and 18 unique imbedded exercises that use standardized patient charts and tailored user views. The exercises allow learners to adopt the role of various providers who document in EMRs. Students are exposed to the unique perspectives of an attending physician, nurse, radiological technician, and health information manager, with the goal of developing knowledge and skills necessary for efficient and effective interprofessional communication within the EMR. The campus-wide clinical informatics curriculum is online, flexible, asynchronous, and well-established within each college, allowing faculty to select and schedule content based on discipline-specific learner and course needs. Program modifications over the past 4 years have correlated with a positive impact on the students' experience.

Identifying management practices for promoting infection prevention: Perspectives on strategic communication

BACKGROUND

Engaging leaders to share information about infections and infection prevention across their organizations is known to be important in initiatives designed to reduce healthcare-associated infections (HAIs). Yet the topics and communication strategies used by leaders that focus on HAI prevention are not well understood. This study aimed to identify and describe practices around information sharing used to support HAI prevention.

METHODS

We visited 18 U.S. hospitals between 2017 and 2019 and interviewed 188 administrative and clinical leaders to ask about management practices they used to facilitate HAI prevention. Interview transcripts were analyzed to characterize practices involving strategic communications.

RESULTS

Sharing information to support infection prevention involved strategic communications around two main topics: (1) facilitators of success and best practices, and (2) barriers to success and lessons learned. In addition, the practice of storytelling reportedly allowed leaders to highlight impact and elicit emotion, provide education, and acknowledge success in infection prevention by providing examples of real events.

CONCLUSIONS

Our findings provide insight about how strategic communication of information around HAIs and HAI prevention can be used to support improvement. Organizations and leaders should consider the different opportunities to incorporate the practice of strategic communication, including using storytelling, to advance their infection prevention efforts.

Making the case for workforce diversity in biomedical informatics to help achieve equity-centered care: a look at the AMIA First Look Program

Developing a diverse informatics workforce broadens the research agenda and ensures the growth of innovative solutions that enable equity-centered care. The American Medical Informatics Association (AMIA) established the AMIA First Look Program in 2017 to address workforce disparities among women, including those from marginalized communities. The program exposes women to informatics, furnishes mentors, and provides career resources. In 4 years, the program has introduced 87 undergraduate women, 41% members of marginalized communities, to informatics. Participants from the 2019 and 2020 cohorts reported interest in pursuing a career in informatics increased from 57% to 86% after participation, and 86% of both years' attendees responded that they would recommend the program to others. A June 2021 LinkedIn profile review found 50% of participants working in computer science or informatics, 4% pursuing informatics graduate degrees, and 32% having completed informatics internships, suggesting AMIA First Look has the potential to increase informatics diversity.

600. The Effect of Medication-Assisted Treatment on Completion Rates of Outpatient Parenteral Antibiotic Therapy

Background

Injection drug use is a nationwide epidemic associated with an increased risk of invasive Staphylococcus aureus (S. aureus) infections. Medication-assisted treatment (MAT) is effective in reducing substance use and increasing adherence to inpatient medical therapy in persons with injection drug use (PWID). Studies assessing the impact MAT has on completion of outpatient parenteral antibiotic therapy (OPAT) are limited.

Methods

This was a single-center, retrospective, cohort study at The Ohio State University Wexner Medical Center in patients admitted from 12/1/2017 to 12/1/2019 with a diagnosis of S. aureus bacteremia who were identified as PWID either by ICD-9 or 10 code or chart review. A formal MAT program was established on 11/30/2018. Patients were assigned to the pre-MAT group if they were discharged prior to 11/30/2018 and to the MAT group with treatment after 11/30/2018. We evaluated a composite outcome of failure to complete OPAT, recurrence of S. aureus bacteremia during the OPAT period and readmission within 30 days. A multivariable logistic regression analysis was performed to examine the association between MAT therapy and the primary composite outcome, while adjusting for proven confounders.

Results

A total of 700 patients were identified with 644 patients omitted based on exclusion criteria. The study population included 27 in the Pre-MAT group and 17 in the MAT. Median age was 37 years (IQR 30.6 - 46.1). There was a higher number of females in the MAT therapy group compared to the pre-MAT group (82% vs. 33%, p=0.002). Patients in the pre-MAT group had a significantly longer length of stay (25 days vs. 17 days, p=0.01). The primary composite outcome was met if a patient did not complete their OPAT, if they had a recurrence of S. aureus bacteremia during their OPAT or if they were readmitted to the hospital within 30 days. In the pre-MAT group 14/27 (52%) met the composite outcome versus 6/17 (35%) of the MAT group (p=0.28).

Conclusion

Patients in the MAT group met the composite outcome 17% less than those in the pre-Mat group which is suggestive of the impact MAT has on completion of OPAT therapy; however, this study did not reach statistical significance as it was underpowered. Further longitudinal evaluation with greater sample size is needed to fully evaluate this intervention.

Disclosures

Mohammad Mahdee Sobhanie, M.D., Regeneron (Scientific Research Study Investigator)Regeneron (Scientific Research Study Investigator, Was a sub-investigator for Regeneron 2066 and 2069)

318. Description of Patients Readmitted within 30 Days from COVID-19 Hospitalization

Background

Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) has led to increased hospitalizations and utilization of critical care services. There are few studies describing co-morbidities and demographics associated with patients re-admitted within 30-days of discharge. The purpose of this study is to describe this patient population.

Methods

This was a single-center, retrospective study at The Ohio State University Wexner Medical Center to identify patients who were admitted secondary to SARS-CoV-2 and required readmission within 30 days due to complications that might be associated with COVID-19. Adults admitted between 3/15/2020 and 11/15/2020 were included in this study. Baseline demographics including age, gender and race in addition to select comorbidities were identified.

Results

250 patients were identified who were readmitted for various reasons. Readmitted patients had a median age of 55 years, 44% were male, and 41.2% were Black/African American. 62.4% of the population was obese (BMI ≥30 kg/m2) with 21.6% with a BMI ≥ 40 kg/m2. The top three co-morbidities seen included Diabetes Mellitus (DM) (32.2%), Hyperlipidemia (48.3%) and Hypertension (51.7%).

Conclusion

Though this study lacked a comparator group, it is clear that patients readmitted with all cause etiologies were disproportionally Black/African-American and obese, with a high prevalence of DM, hyperlipidemia, and hypertension. We recommend close monitoring of patients in these groups to reduce COVID19 readmissions. This is the first step in identifying which patients may be more likely to develop complications and required readmission, the next step is to compare these patients to those that were not readmitted to develop a risk model for readmission.

Disclosures

Carlos Malvestutto, M.D., Lilly (Scientific Research Study Investigator)Regeneron Inc. (Scientific Research Study Investigator)ViiV Healthcare (Advisor or Review Panel member) Mohammad Mahdee Sobhanie, M.D., Regeneron (Scientific Research Study Investigator)Regeneron (Scientific Research Study Investigator, Was a sub-investigator for Regeneron 2066 and 2069)

Emergency department disposition of non-neutropenic febrile patients with cancer

OBJECTIVES

National data reveal that 60% of the 4.5 million annual emergency department (ED) visits by patients with cancer result in admission. Many of these visits are due to a febrile illness. Current literature provides limited guidance on how to treat non-neutropenic febrile ED patients. This study characterizes clinical outcomes of non-neutropenic febrile patients with cancer presenting to an academic, Comprehensive Cancer Center affiliated ED.

METHODS

Retrospective chart review of 101 randomly selected adult patients with active cancer presenting with a chief complaint of fever or a documented fever in the ED and an absolute neutrophil count above 1000 between October 2015 and September 2016. Descriptive statistics were calculated.

RESULTS

The primary malignancies represented were hematologic (24%), gastrointestinal (13%), head and neck (13%), and genitourinary (8%). Sixty-two percent were on chemotherapy, 15% on radiation therapy, and 12% were on targeted therapy. Severe illness outcomes occurred in 39% and 83% were admitted with a median length of stay of 4 days. Among admitted patients, 24% experienced a length of stay ≤2 days. A return visit to the ED or an in-system hospitalization within 7 days of the index visit occurred in 10% and death occurred within 7 days of the index visit in 4%.

CONCLUSION

A majority of patients presenting to the ED with non-neutropenic fever are admitted (83%), of whom nearly a quarter experience a length of stay of ≤2 days with infrequent serious illness outcomes. Future efforts should focus on the development of risk stratification tools in this population to avoid potentially unnecessary hospitalizations.

Diagnosing Dyspneic Older Adult Emergency Department Patients: A Pilot Study

Study Objectives:

Pneumonia, chronic obstructive pulmonary disease (COPD), and heart failure (HF) exacerbations can present similarly in the older adult in the Emergency Department (ED), leading to sub-optimal treatment from over- and under-diagnosis. There may be a role for antimicrobial peptides (AMPs) in improving the accurate diagnosis of pneumonia in these patients.

Methods:

This pilot was a prospective, observational cohort study of older adults (aged ≥65 years of age) who presented to the ED with dyspnea or elevated respiratory rate. To identify biomarkers of pneumonia, serum levels of white blood cell count, procalcitonin (PCT), and antimicrobial peptides (human beta defensin 1 and 2 [HBD-1, -2], human neutrophil peptides 1–3 [HNP1–3] and cathelididin [LL-37]) were compared between those with and without pneumonia. Criterion standard reviewers retrospectively determined the diagnoses present in the ED.

Results:

Three hundred ninety-one patients were screened, 140 were eligible, and 79 were enrolled. Based on criterion standard review, pneumonia was present in 10 (12.7%), COPD in 9 (11.4%) and HF in 31 (39.2%) with a co-diagnosis rate of 10.1% by criterion standard review. Comparatively, emergency medicine attending physicians diagnosed pneumonia in 16 (20.3%), COPD in 12 (15.2%), and HF in 30 (38.0%) with co-diagnosis rate of 15.2%. Emergency physicians agreed with criterion standard diagnoses in 90% of pneumonia, 75% of COPD and 65% of HF diagnoses. Differences in leukocyte count (p<0.01) and two novel AMPs (DEFA5 (p=0.08) and DEFB2 (p=0.09)) showed promise for diagnosing pneumonia.

Conclusions:

Emergency physicians continue to have poor diagnostic accuracy in dyspneic older adult patients. Serum AMP levels are one potential tool to improve diagnostic accuracy and outcomes for this important population and require further study.

Management practices for leaders to promote infection prevention: Lessons from a qualitative study

BACKGROUND

Prevention of healthcare-associated infections (HAIs) is critical to reduce preventable deaths and healthcare costs. Variable success with HAI prevention efforts has suggested that management practices are critical to support clinical infection prevention practices. This study examined hospital leaders' management practices around the prevention of catheter-associated urinary tract infections (CAUTIs) and central line-associated bloodstream infections (CLABSIs) to identify actions that leaders can take to promote HAI prevention efforts.

METHODS

We conducted interviews with 420 key informants, including managers and frontline staff, in 18 hospitals across the United States. Interviewees were asked about management practices supporting HAI prevention. We analyzed interview transcripts using rigorous qualitative methods to understand how management practices were operationalized in infection prevention efforts.

RESULTS

Across hospitals and interviewees, three management practices were characterized as important facilitators of HAI prevention: (1) engagement of executive leadership; (2) information sharing; and (3) manager coaching. We found that visible executive leadership, efficient communication, and frequent opportunities to provide and promote learning from feedback were perceived to promote and sustain HAI prevention efforts.

CONCLUSIONS

Our findings provide insight into management practices for leaders that support successful HAI prevention. In practice, these tactics may need to be adjusted to accommodate the current restrictions caused by the COVID-19 pandemic, in order to maintain HAI prevention efforts as a priority.

367. Clinical Characteristics and Outcomes in Patients with Pneumonia secondary to Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2)

Background

Since discovery in December 2019, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) which causes the disease of COVID-19 has become a global pandemic. Little is known about which risk factors lead to more severe disease or increased mortality in patients diagnosed with SARS-CoV-2. We aimed in this study to compare clinical characteristics associated with disease severity and increased mortality in hospitalized patients with COVID-19.

Methods

This was a single-center, retrospective study at The Ohio State University Wexner Medical Center to compare clinical characteristics associated with increased mortality in hospitalized patients with confirmed SARS-CoV-2. Adults patients positive for SARS-CoV-2 between March 1, 2020 and April 20, 2020 were included in the study. Prisoners and pregnant women were excluded. Baseline demographics, clinical characteristics, and outcomes were collected, and then compared to determine association with mortality. Statistical analysis used univariate and multivariate logistic regression analysis to evaluate the relationship between patient characteristics and mortality.

Results

The cohort included 92 patients. Median age was 58 years (ranging from 25–93) and 47/92 were men (51%). 12 patients were admitted directly to the intensive care unit (ICU), with 22 additional patients transferred to the ICU. 23 patients required mechanical ventilation. Clinical characteristics significantly associated with mortality in univariate analysis included underlying coronary artery disease (CAD) (OR=7.8, p = 0.002), chronic obstructive pulmonary disease (OR=5.21, p=0.02), living in an extended care facility (ECF) (OR=4.2, p=0.025), and immunocompromised status (OR=4.2, p=0.025). Multivariate analysis showed a statistically significant association in patients with underlying CAD (OR=13.1, p=0.001) and those admitted from an ECF (OR=12.1, p=0.005), when adjusted for other variables in the model.

Characteristics Associated with Mortality in Patients with COVID-19 in Univariate Analysis

Conclusion

Our study found that CAD and admission from an ECF were associated with SARS-CoV-2 mortality, when adjusted for age and other comorbidities. Further studies are necessary to identify potential preventative strategies to mitigate mortality in this vulnerable population.

Disclosures

All Authors: No reported disclosures

Amplifying Domain Expertise in Clinical Data Pipelines

Digitization of health records has allowed the health care domain to adopt data-driven algorithms for decision support. There are multiple people involved in this process: a data engineer who processes and restructures the data, a data scientist who develops statistical models, and a domain expert who informs the design of the data pipeline and consumes its results for decision support. Although there are multiple data interaction tools for data scientists, few exist to allow domain experts to interact with data meaningfully. Designing systems for domain experts requires careful thought because they have different needs and characteristics from other end users. There should be an increased emphasis on the system to optimize the experts' interaction by directing them to high-impact data tasks and reducing the total task completion time. We refer to this optimization as amplifying domain expertise. Although there is active research in making machine learning models more explainable and usable, it focuses on the final outputs of the model. However, in the clinical domain, expert involvement is needed at every pipeline step: curation, cleaning, and analysis. To this end, we review literature from the database, human-computer information, and visualization communities to demonstrate the challenges and solutions at each of the data pipeline stages. Next, we present a taxonomy of expertise amplification, which can be applied when building systems for domain experts. This includes summarization, guidance, interaction, and acceleration. Finally, we demonstrate the use of our taxonomy with a case study.

ENHANCING USEFULNESS AND USABILITY OF A CLINICAL DECISION SUPPORT PROTOTYPE FOR ANTIBIOTIC STEWARDSHIP

Human factors engineering can enhance software usefulness and usability. We describe a multi-method approach to improve clinical decision support (CDS) for antibiotic stewardship. We employed a heuristic review to generate recommendations to improve the usability of a prototype CDS to support empiric antibiotic prescribing in the hospital setting. We then engaged in a design improvement cycle in collaboration with software programmers, which resulted in additional enhancements to our prototype. Finally, we used the revised prototype during three walkthrough demonstration interviews with physician and pharmacist subject matter experts. These walkthrough interviews generated recommendations to improve the interface, functionality, and tailoring for groups of users. We discuss common elements of the recommendations for models for using clinical decision support in general.

Abstract LB-275: Measuring concordance of CD8 and PD-L1 expression in non-small cell lung cancer between a novel multiplex immunofluorescence assay and a brightfield laboratory developed test

The quantitative assessment of PD-L1 protein expression has been widely used in the clinic to predict response to immune checkpoint therapy in non-small cell lung cancer (NSCLC) patients. The conventional method of assessment is the manual quantification (scoring) of PD-L1 positive tumor cells by a certified pathologist on samples stained by immunohistochemistry (IHC). To improve the accuracy of the PD-L1 scoring, we assessed the use of digital image analysis tools to accurately quantify the positive tumor cells and correlate the data with the conventional manual scoring method used in the clinical setting by both brightfield and multiplex immunofluorescence assays. Thus, the aim of this study was to evaluate the digital analysis scoring in NSCLC samples stained by IHC and by the ImmunoVUE™ PD-L1 multiplex assay and to correlate the results with manual scoring as well as assess the concordance in the digital scoring using the two assays.

The ImmunoVUE PD-L1 multiplex assay is currently being validated for clinical use by showing concordant results in manual scoring of PD-L1 and CD8 in NSCLC samples with the multiplex assay compared to the IHC laboratory developed test staining in terms of precision, accuracy, and reproducibility. Digital analysis was performed using the HALO software on the manually scored images stained with IHC and the ImmunoVUE PD-L1 multiplex assay. For the digital analysis, the samples were segmented into tumor and non-tumor regions using random forest classification. Positive cells were quantified using the CytoNuclear algorithm for IHC images and the HighPlex FL algorithm was used to quantify images stained with the ImmunoVUE PD-L1 multiplex assay. Total tumor cells and percentage of PD-L1+ tumor cells and CD8+ T cells associated with the tumor were obtained and compared to the manual PD-L1 and CD8 scoring.

Digital analysis of the individual samples stained by IHC or by ImmunoVUE PD-L1 multiplex showed no significant differences compared to the manual scoring in terms of total tumor cell count or the percent of PD-L1+ tumor cells and CD8+ T cells associated with tumor. Coefficients of correlation were greater than 0.9 when comparing manual scoring to digital analysis. Additionally, the digital analysis results obtained from IHC stained slide was concordant with an ImmunoVUE PD-L1 multiplex stained slide from the same sample showing correlation in terms of positive cell counts. The correlation for PD-L1 between brightfield LDT and mIF was measured at 0.8 while that for CD8 was 0.99.

This study shows that digital image analysis can provide an accurate method of PD-L1 and CD8 analysis and is comparable with the manual scoring method using IHC staining. Digital image analysis can be used on samples stained with IHC or with the ImmunoVUE PD-L1 multiplex assay and can be used for assessment of PD-L1 and CD8 positivity in NSCLC patient samples. Lastly, the ImmunoVue assay appears to be concordant with an established LDT.

Citation Format: Shravani Shitole, Monique Johnson, Courtney Hebert, Jamie Buell, Amy Ly, Sean R. Downing. Measuring concordance of CD8 and PD-L1 expression in non-small cell lung cancer between a novel multiplex immunofluorescence assay and a brightfield laboratory developed test [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr LB-275.

Repurposing Geographic Information Systems for Routine Hospital Infection Control

This chapter discusses the potential role of geographic information systems (GIS) for infection control within the hospital system. The chapter provides a brief overview of the role of GIS in public health and reviews current work applying these methods to the hospital setting. Finally, it outlines the potential opportunities and challenges for adapting GIS for use in the hospital setting for infection prevention. A targeted literature review is used to illustrate current use of GIS in the hospital setting. The discussion of complexity was compiled using the nonadoption, abandonment, scale-up, spread, and sustainability (NASSS) framework. Challenges and opportunities were then extracted from this exercise by the authors. There are multiple challenges to implementation of a Hospital GIS for infection prevention, mainly involving the domains of technology, organization, and adaptation. Use of a transdisciplinary approach can address many of these challenges. More research, specifically prospective, reproducible clinical trials, needs to be done to better assess the potential impact and effectiveness of a Hospital GIS in real-world settings. This chapter highlights a powerful but rarely used tool for infection prevention within the hospital. Given the importance of reducing hospital-acquired infection rates, it is vital to identify relevant methods from other fields that could be translated into the field of hospital epidemiology.

2208. Development and Evaluation of Predictive Models for Estimating Infection Susceptibility to Empiric Treatment Regimens Among Patients with Pneumonia in Intensive Care Units

Background

Predictive models for empiric antibiotic prescribing often estimate the probability of infection with multidrug-resistant organisms. In this work, we developed models to predict coverage of specific treatment regimens to better target antibiotics to high- and low-risk patients.

Methods

We established a retrospective cohort of adults admitted to the ICU in a 1,300-bed teaching hospital from November 1, 2011 to June 30, 2016. We included patients with a diagnosis of pneumonia and positive respiratory culture collected during their ICU stay. We collected demographics, comorbidities, and medical history from the electronic health record. We evaluated three penalized regression methods for predicting infection susceptibility to 11 treatment regimens: least absolute selection and shrinkage operator (LASSO), minimax concave penalty (MCP), and smoothly clipped absolute deviation (SCAD). We developed models for susceptibility prediction at two stages of the diagnostic process: for all pathogenic bacteria and for infections with Gram-negative organisms only. We selected final models based on higher area under the receiver operating characteristic (AUROC), acceptable goodness of fit, lower variability of the AUROCs in the cross-validation run, and fewer predictors.

Results

Among 1,917 cases of pneumonia, 54 different pathogens were identified. The most frequently isolated organisms were: Pseudomonas aeruginosa (16.6%), methicillin-resistant Staphylococcus aureus (16.1%), and Staphylococcus aureus (13.5%). Frequently selected variables included age, Elixhauser score, tracheostomy status, recent antimicrobial use, and prior infection with a carbapenem-resistant organism. All final models used MCP or SCAD methods. Point estimates for the AUROCs in the training set ranged from 0.70 to 0.80, and estimates in the internal validation set ranged from 0.64 to 0.77.

Conclusion

MCP and SCAD outperformed LASSO. For some regimens, models predicted infection susceptibility with fair accuracy. These models have potential to help antibiotic stewardship efforts to better target appropriate antibiotic use.

Disclosures

All authors: No reported disclosures.

1433. Predictive Models for Antibiotic Coverage of Gram-Negative Urinary Tract Infections

Background

Providers use institutional recommendations, national guidelines, and antibiograms to decide on empiric antibiotics. As local antibiograms are most effective after organisms are known, we sought to use local microbiology and clinical data to develop predictive models for antibiotic coverage prior to identifying the organism. We focused on Gram-negative organisms as they are common urinary pathogens and are often the cause of sepsis originating in the urinary tract. As such, they are important to cover in hospitalized patients with urinary tract infections (UTI).

Methods

Hospitalized patients, with a diagnosis of UTI and a positive urine culture in the first 48 hours were included. Gram-positive organisms, yeast, and cultures without susceptibilities were excluded. Unknown susceptibilities were filled in using expert-derived rules. Clinical information from electronic health record (EHR) data were extracted on each patient. Penalized logistic regression with 10-fold cross validation was used to develop final models for coverage for five antibiotics (cefazolin, ceftriaxone, ciprofloxacin, cefepime, piperacillin–tazobactam). Final models were chosen based on their discrimination, calibration, and number of predictors, and then tested on a held-out validation dataset.

Results

Included were 5,096 patients (80% training; 20% validation). Coverage ranged from 65% for cefazolin to 90% for cefepime. Positive blood cultures were present in 544 (11%) with 388 (71%), including a urinary pathogen. In the first 24 hours, 2329 (46%) were hypotensive, 2179 (43%) had a respiratory rate > 22, 2049 (40%) had a WBC > 12, 1079 (21%) were febrile, and 584 (11%) required ICU care. Final model covariates included demographics, antibiotic exposure, prior resistant pathogens, and antibiotic allergies. The five predictive models had a point-estimate for the area under the ROC on the validation set that ranged from 0.70 for ciprofloxacin to 0.73 for ceftriaxone.

Conclusion

In this cohort of moderate to high acuity hospitalized patients with Gram-negative urinary pathogens, we used EHR data to develop 5 models that predict antibiotic coverage which could be used to support empiric prescribing. These models performed well in a held-out validation set.

Disclosures

All authors: No reported disclosures.

197. Infective Endocarditis Over a Five-Year Period in an Academic Teaching Center: The Validity of ICD Codes vs. Manual Chart Review

Background

Opioid dependence and overdose are at epidemic levels in the United States. Ohio has the third highest rate of opioid-related overdose deaths. Infectious complications of intravenous drug use (IDU) include increased acquisition of hepatitis C, HIV and infective endocarditis. In this study, we aimed to characterize cases of infective endocarditis admitted to our healthcare system over a five-year period. We additionally sought to determine the validity of using ICD codes to identify infective endocarditis cases and IDU.

Methods

Patients with ICD-9 or 10 discharge diagnosis codes for infective endocarditis were identified from our institution’s electronic health record. ICD codes pertaining to substance abuse were used to classify patients according to IDU status. Readmissions during the same episode of infective endocarditis were excluded. We compared chart review to ICD code for the identification of infective endocarditis and IDU in a random sample of 296 of 1590 cases.

Results

Of 296 charts reviewed, 133 (44.9%) were excluded because they did not meet criteria for definite infective endocarditis by modified Duke’s criteria or because the episode was a readmission. A total of 163 (55.1%) cases met inclusion criteria, all of whom were seen in consultation by the inpatient Infectious Disease service. Of these, 52 (31.9%) had ICD 9 or 10 codes linked to substance abuse. Following manual chart review, we established that in fact 86 of these 163 cases (52.8%) had evidence of substance abuse.

Conclusion

Misclassification due to use of ICD codes is a well-established challenge to epidemiological research. However, the extent of misclassification in this analysis was greater than expected. If prior research on IDU and infective endocarditis has relied on medical record data alone without verification through manual chart review, the observed epidemiological trends may not be accurate.

Disclosures

All authors: No reported disclosures.

482. Time Series Analysis of Antimicrobial Consumption and Pseudomonas aeruginosa Resistance in an Academic Medical Center in the United States (2013–2018)

Background

Monitoring antimicrobial use and resistance are key components of initiatives to promote antimicrobial stewardship and prevent antimicrobial-resistant infections. In this surveillance study, we evaluated trends in resistance among healthcare-associated P. aeruginosa isolates and potential associations with antimicrobial consumption.

Methods

We established a retrospective cohort of P. aeruginosa isolates collected ≥48 hours after inpatient admission at a 1,300-bed academic medical center from July 1, 2013 to July 31, 2018. We included isolates from all clinical cultures and retained the first isolate for a patient encounter. We defined the multidrug-resistant (MDR) status in accordance with the phenotype definitions established by the Centers for Disease Control and Prevention. We calculated the monthly percentage of class-specific resistance and MDR status among isolates. We measured monthly antimicrobial consumption as days of therapy per 1,000 patient-days. To evaluate potential associations between identified trends in resistance and antimicrobial use, we constructed autoregressive integrated moving average models (ARIMA) with transfer functions.

Results

Of 1,897 isolates included in the analysis, 303 (16.0%) were classified as MDR P. aeruginosa. The rate of healthcare-associated P. aeruginosa infections and percent of MDR isolates remained stable over the five-year study period. However, we identified trends in resistance to specific antimicrobial classes: there was a significant increase in resistance to antipseudomonal carbapenems, while resistance to aminoglycosides and extended-spectrum cephalosporins decreased. Using the ARIMA modeling strategy, bivariable analyses of resistance and antimicrobial use revealed that carbapenem-resistant P. aeruginosa was positively correlated with the use of antipseudomonal carbapenems at a 1-month lag and ertapenem at a 5-month lag.

Conclusion

Risk assessments that only measure rates of MDR organisms may miss underlying trends in class resistance. Increasing carbapenem resistance despite a stable proportion of MDR isolates highlights a critical area for continued monitoring and antimicrobial stewardship initiatives targeted at carbapenem use in our hospital.

Disclosures

All authors: No reported disclosures.

Network analysis of intra-hospital transfers and hospital onset clostridium difficile infection

OBJECTIVES

To explore how social network analysis (SNA) can be used to analyse intra-hospital patient networks of individuals with a hospital acquired infection (HAI) for further analysis in a geographical information systems (GIS) environment.

METHODS

A case and control study design was used to select 2008 patients. We retrieved locational data for the patients, which was then translated into a network with the SNA software and then GIS software. Overall metrics were calculated for the SNA based on three datasets and further analysed with a GIS.

RESULTS

The SNA analysis compared cases to control indicating significant differences in the overall structure of the networks. A GIS visual representation of these metrics was developed, showing spatial variation across the example hospital floor.

DISCUSSION

This study confirmed the importance that intra-hospital patient networks play in the transmission of HAIs, highlighting opportunities for interventions utilising these data. Due to spatial variation differences, further research is necessary to confirm this is not a localised phenomenon, but instead a common situation occurring within many hospitals.

CONCLUSION

Utilising SNA and GIS analysis in conjunction with one another provided a data-rich environment in which the risk inherent in intra-hospital transfer networks was quantified, visualised and interpreted for potential interventions.

Abstract 1677: Understanding the TME: Advanced analysis and visualization of multiplexed fluorescence images

Background: Multiplexed immunofluorescence (mIF) has the potential to revolutionize immuno-oncology and pathology research as it enables the identification of complex cell phenotypes and their potential interactions in the tumor microenvironment (TME). But for a whole slide image with millions of cells, as we increase the number of biomarkers imaged for every cell, the complexity of the data analysis and visualization task increases exponentially. For n markers, a total of 2n phenotypes are possible (e.g. 10 markers have 1024 potential phenotypes). To address this problem and reveal the biologically relevant information embedded in the data, we have developed software tools to reduce the complexity, visualize, and quantify spatial distributions of cells across the full spectrum of possible phenotypes.

Methods: Here we present results using two different methods. The first is an image processing technique called Phenotypic Surface Density Mapping (PSDM), that produces not only true surface density images of each phenotype (cells/µm2), but also surface density images that quantify a variety of other statistics such as the level of expression (intensity) of key markers, inter-phenotype nearest neighbor distance maps, and maps of cell size/morphology. Some important features of these surface density maps are that they are quantitatively robust, have real physical units (e.g. cells/µm2 or intensity/µm2), and they are generated in an unbiased fashion to reveal information about every possible phenotype. The second analysis method, dimensionality reduction, exploits a new technique called Uniform Manifold Approximation and Projection (UMAP), reducing dozens or even hundreds of dimensions for millions of cells to a simple 2D scatter plot. We have developed interactive software that displays the UMAP for a slide and allows the user to select a given cell or region of cells to view closeup images of each cell and statistics about the collection.

Results: Examples of the surface density maps provide insights into mapping the complexity of the TME. We assess the results on deidentified samples by comparison with both human generated labels (pathology review) for individual cells and with automatically generated labels (software review). We show how these tools can be used to both identify tumors and quantify the level of activity in different tumor regions. We demonstrate how increasing the level of multiplexing allows one to differentiate subtle variability and separate subclasses of cells from each other.

Conclusions: Multiplexed data brings valuable information about the TME but much of this information is inaccessible by simply viewing the captured images or performing simple cell counting alone. To address this problem, we demonstrate two new software tools, PSDM and UMAP that preserve and quantify the spatial information of the underlying biology and provide this analysis for all possible phenotypes.

Citation Format: Douglas Wood, Bonnie Phillips, Courtney Hebert, Aditi Sharma, Jamie Buell, Sean Downing. Understanding the TME: Advanced analysis and visualization of multiplexed fluorescence images [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1677.

Abstract 4569: Intra- and inter-run assessment of reproducibility and quantification of UltiMapper™ I/O APC and T-act kits for tissue multiplexing

Background: The field of immuno-oncology research has enthusiastically adopted multiplex immunohistochemistry (IHC) techniques to establish the spatial relationships between various immune cell populations in a tumor biology in context. Multiplexing enables researchers to gain a deeper understanding and insight into the tumor microenvironment. Unfortunately, many multiplexing technologies face several challenges, specifically in generating highly robust, reproducible, and easily quantifiable data sets. Ultivue’s UltiMapper I/O APC and T-act kits utilize InSituPlex® technology, a new method of multiplexed IHC that utilizes a streamlined workflow with single antigen retrieval, staining, amplification, and detection steps, allowing for the completion of the assay in 5 hours. Here we assess these kits for intra- and inter-run reproducibility and quantification.

Methods: Intra-run reproducibility and quantification was accomplished by manually staining 5 serial sections from three tissue types (deidentified samples of tonsil, melanoma, NSCLC) with one set for each of the UltiMapper I/O APC (CD11c, CD20, CD68/CD163, and MHC Class II) and T-act (CD3, Granzyme B, Ki67, and pan-Cytokeratin/SOX10) kits. Inter-run assessment was determined by staining each slide from a set of five serial sections of each tissue type independently. Images were acquired using the ZEISS® Axio Scan.Z1, without the need for linear unmixing allowing for direct whole slide imaging. Analysis was accomplished using Indica Labs HALO® software. Coefficients of variation (CVs) were calculated based on resulting data.

Results: Analysis of intra-run serial section images revealed that cell counts from section to section had CVs of &lt;10% across all markers, in all tissues, for both the UltiMapper I/O APC and T-act kits. This included total cell counts and average signal intensity. Similar results were seen for inter-run comparisons for both kits (&lt;10% CV).

Conclusions: The results presented here indicate that InSituPlex technology is potentially much more reproducible than other tissue multiplexing techniques currently available. Histological standards for coefficients of variation in IHC based assays are typically &lt;15%. Data presented here falls well within that standard indicating the potential for future translational applications. In conclusion, InSituPlex is a highly reproducible and quantifiable multiplexing technology that is able to produce these results across a variety of tissue types and markers.

Note: This abstract was not presented at the meeting.

Citation Format: Bonnie Phillips, Katir Patel, Courtney Hebert, Aditi Sharma, Jamie Buell, Sean Downing. Intra- and inter-run assessment of reproducibility and quantification of UltiMapper™ I/O APC and T-act kits for tissue multiplexing [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 4569.

Development of a PD-L1 multiplex immunofluorescence assay with advanced visual analysis for understanding the tumor microenvironment

e14283

Background: Current PD-L1 Immunohistochemistry (IHC) assays utilizing conventional brightfield, chromogenic 3,3′-Diaminobenzidine (DAB) staining modalities are the norm in both translational and clinical research. With the rise of immuno-oncology, multiplexed immunofluorescence (mIF) has the potential to revolutionize pathology research as it enables the identification of complex cell phenotypes and their potential interactions in the tumor microenvironment (TME). Ultivue has developed a mIF assay for profiling PD-L1, CD8, CD68, pan-Cytokeratin, and Sox10 enabling whole-slide imaging and data analysis in a single workday. Methods: With an expanding number of biomarkers being interrogated, the complexity of the data analysis and visualization tasks increase exponentially. For N markers, a total of 2N phenotypes are possible. To address this problem and reveal the biologically relevant information embedded in the data, we have developed software tools to reduce the complexity, visualize, and quantify spatial distributions of cells across the full spectrum of possible phenotypes. In this study we demonstrate the utility, robustness, and ability to derive meaningful biological insights by validating the accuracy, precision, sensitivity, and specificity of the PD-L1 multiplex immunofluorescence assay. Results: Here we present results that identify all distinct binary phenotypes within cohorts of lung cancer and melanoma samples using qualitative pathology review and an image processing technique called Phenotypic Surface Density Mapping (PSDM). Important features of this quantitative technique are robustness, use of real physical units (e.g. cells/µm2 or intensity/µm2), and data generation in an unbiased fashion to reveal information about every possible phenotype. Conclusions: Applying the pathology review and PSDM to the cohorts of samples confirmed a high level of concordance demonstrating sensitivity and specificity with high degrees of confidence above 85%. Analysis of intra-run, inter-run, and inter-technician runs revealed that cell counts were within a CV less than 15% across all markers. Data presented here indicates the potential for use in a high throughput testing laboratory to aid in collection and interpretation of meaningful biological insights.

Changes in the use of effective and long-acting reversible contraception in Vietnam

OBJECTIVES

To calculate the prevalence and identify correlates of unmet need for contraception and to assess whether prevalence of use of effective contraception and long-acting reversible contraception (LARC) has changed over time among married or cohabiting, reproductive-age women in Vietnam.

METHODS

Study population was drawn from nationally representative Multiple Indicator Cluster Surveys conducted in 2000, 2006, 2011 and 2014. Unmet need for contraception was defined as occurring when a fecund, married or cohabiting woman is not using any method of contraception but either does not want children or wants to delay birth for at least 1 year or until marriage. Following the ranking of method effectiveness by the Centers for Disease Control and Prevention, we defined "effective contraception" as implant, intrauterine device, male and female sterilization, injectable, pill, patch, ring or diaphragm. We used multivariable logistic regression to identify correlates of unmet need for contraception in 2014 and Cochran-Armitage trend tests to assess changes in effective contraception and LARC use from 2000 to 2014. All analyses used survey weights to account for the complex sampling design.

RESULTS

In 2014, 4.3% of married or cohabiting, reproductive-age women had unmet need for contraception. Multivariable analysis showed that age, education and number of children ever born were statistically significant correlates of unmet need for contraception. Use of effective contraception statistically significantly declined from 53.0% in 2000 to 45.7% in 2014 (p<.0001). Similarly, LARC declined from 39.6% in 2000 to 30.0% in 2014 (p<.0001). After adjusting for age, education, residence and having at least one son, these secular trends remained.

CONCLUSION

Findings indicate that effective contraception and LARC use have decreased among married or cohabiting women of reproductive age in Vietnam. Correlates of unmet need for contraception should be used to inform interventions to prevent unintended pregnancy.

IMPLICATIONS

Although the prevalence of unmet need for contraception was low (4.3%) in 2014, the use of effective contraception and long-acting reversible contraception declined among reproductive-age, married or cohabiting women in Vietnam from 2000 to 2014. This finding is particularly striking given the economic growth in the nation during this time frame.

188. A Novel Inpatient Antibiotic Stewardship Assistance Program (ASAP) Using Real-Time Electronic Health Record Data, Prediction Modeling and Epidemiologic Data to Provide Personalized Empiric Antibiotic Recommendations

Background

Antibiotic prescribing varies amongst clinicians, which can result in inappropriate or overuse. Inappropriate antibiotics can increase the risk of adverse drug events and multi-drug-resistant organisms (MDRO). Decreasing variability and increasing alignment with guideline-based therapy may improve antimicrobial stewardship and outcomes.

Methods

We developed a point of care stewardship tool embedded in the electronic health record (EHR) that provides empiric antibiotic recommendations for four syndromes, urinary tract infection (UTI), abdominal biliary infection (ABI), pneumonia, and cellulitis. We identified key variables that alter antibiotic selection or need for infectious disease (ID) consultation such as allergy history, immunosuppression and risk factors for MDRO, and mortality. We created algorithms of preferred empiric antibiotic choices based on national and hospital guidelines using these risk factors. We used a weighted incidence syndromic combined antibiogram (WISCA) prediction model to recommend ID consultation when likelihood of coverage was below a defined threshold. We also incorporated a home-grown epidemiologic tool that takes real-time data from outpatient clinics on incidence of influenza-like-illness (ILI) to recommend influenza PCR testing during periods of high ILI risk. Data on risk factors and WISCA variables including demographics, allergy history, ICD10 codes, vitals, laboratories, and microbiology results were extracted in real-time from the EHR and sent via URL into a web server which has an embedded Windows ASP.NET C# web site and an SQL server database. The web server was then embedded back into the EHR. This tool stores recommendations into the database for stewardship auditing.

Results

Thirteen key and 20 WISCA variables are extracted from the EHR in real-time. There are eight distinct antibiotic recommendations for UTI and ABI, 12 for cellulitis, and 40 for pneumonia. An illustration of the ASAP tool is shown in Figure 1.

Conclusion

ASAP is an HER-embedded platform that provides clinicians access to personalized antibiotic prescribing tied to best practices and optimal stewardship initiatives. Future work will look into the tool’s effect on variation in care, antibiotic prescribing, and outcomes.

Disclosures

All authors: No reported disclosures.

1515. Patient-Based Stratification of Weighted-Incidence Syndromic Antibiogram (WISCA) for Empiric Antibiotic Prescribing

Background

Antibiograms are often stratified by location (e.g., ICU) to better assess resistance risk of patients in those locations (1). A weighted-incidence syndromic antibiogram (WISCA) may be more useful for empiric prescribing in that it stratifies on syndrome (e.g., urinary tract infection (UTI)) and calculates coverage over all organisms (i.e., weighted incidence). Here we explore the impact of stratification by admission location and patient-specific factors. We suggest that with the availability of patient data from EHRs historic microbiology data can be stratified by syndrome and patient-level factors, making them available for empiric decision support.

Methods

The cohort included patients admitted from November 1, 2011 to July 1, 2016, with a positive urine culture in the first 48 hours and a diagnosis of UTI. Data on admission from a nursing facility (SNF), intensive care unit (ICU) stay in the first 24 hours of admission and antibiotic use in the last 30 days were extracted from the local data warehouse. Expert consensus enriched the susceptibility information that was not reported for organism-antibiotic pairs. The most recent admission for each patient was included. Antibiotic coverage was compared between strata by a chi-square test.

Results

Of the 6,366 patients with UTI, 13% were admitted to an ICU; 8% were admitted from an SNF and 44% had exposure to antibiotics in the last 30 days. Antibiotic coverage did not significantly differ between ICU and non-ICU patient admissions. However, those admitted from an SNF and those admitted with antibiotic exposure in the past 30 days had lower levels of coverage to all antibiotics under study.

Conclusion

Our findings suggest that stratifying by patient factors, easily obtainable from the EHR, may provide more useful empiric prescribing information than stratifying by ICU location.

Research reported in this publication was supported by the National Institute of Allergy and Infectious Diseases of the NIH under Award Number R01AI116975.

Reference

1. Clinical Infectious Diseases 2007; 44:867–73

Disclosures

All authors: No reported disclosures.

ICARUS: Minimizing Human Effort in Iterative Data Completion

An important step in data preparation involves dealing with incomplete datasets. In some cases, the missing values are unreported because they are characteristics of the domain and are known by practitioners. Due to this nature of the missing values, imputation and inference methods do not work and input from domain experts is required. A common method for experts to fill missing values is through rules. However, for large datasets with thousands of missing data points, it is laborious and time consuming for a user to make sense of the data and formulate effective completion rules. Thus, users need to be shown subsets of the data that will have the most impact in completing missing fields. Further, these subsets should provide the user with enough information to make an update. Choosing subsets that maximize the probability of filling in missing data from a large dataset is computationally expensive. To address these challenges, we present ICARUS, which uses a heuristic algorithm to show the user small subsets of the database in the form of a matrix. This allows the user to iteratively fill in data by applying suggested rules based on their direct edits to the matrix. The suggested rules amplify the users' input to multiple missing fields by using the database schema to infer hierarchies. Simulations show ICARUS has an average improvement of 50% across three datasets over the baseline system. Further, in-person user studies demonstrate that naive users can fill in 68% of missing data within an hour, while manual rule specification spans weeks.

Abstract 5657: Use of a novel immunofluorescence multiplexing technology, InSituPlex™, for the simultaneous detection of immune cells in multiple cancer types

Introduction: The field of immuno-oncology has enthusiastically adopted multiplexed

immunohistochemistry (IHC) techniques to establish the spatial relationships between various

immune cells in tumor biology. Multiplexing enables researchers to gain a deeper understanding

and insight into the tumor microenvironment. Unfortunately, many of the multiplexing

technologies currently utilized in the immuno-oncology field face a number of challenges. These

challenges include complicated and difficult assay optimizations for high-level multiplexing and

the need of specialized imaging instrumentation. In addition, harsh stripping methods and

irreversible staining techniques compromise the integrity of precious tissue samples that ultimately

can no longer be used for additional forms of testing. InSituPlex™ is a new method of multiplexed

IHC that relies on the detection of a unique DNA barcode conjugated to primary antibodies. This

technology removes the need to perform multiple, harsh stripping methods to perform multiplexed

IHC and allows for higher order of multiplexing in less time than currently available multiplexing

techniques, all the while maintaining tissue integrity for use in downstream analyses.

Methods: InSituPlex was used to perform a multiplexed IHC assay on 5 samples from 4 tumor

types (breast, lung, colon, and melanoma). The assay consisted of the following markers: CD4,

CD8, CD20, CD68, PD1, Ki67, FoxP3, and a tumor marker; pan-cytokeratin (CK) for breast,

colon, and lung carcinomas and Sox10 for melanoma. Staining was performed using the Leica

BOND RX autostainer, imaging was performed on various platforms including the Leica VERSA

whole slide scanner, and image analysis was performed using HALO from Indica Labs.

Data: All markers were detected utilizing the InSituPlex technology across all tumor types. Data

indicates a range of phenotypes for infiltrating immune cells in tumor types from highly inflamed

to “immune deserts”. The panel provided for analysis of many phenotypic subtypes, (e.g.,

CD4+/PD1+, CD8+/Ki67+, and Ki67+/CK+), cell densities, proliferative indexes, and spatial

analysis (e.g., CD8+ to tumor cells).

Conclusion: InSituPlex technology is proficient in maintaining tissue integrity and detecting a

range of immune cell phenotypes in multiple tumor types. InSituPlex can be implemented by any

lab possessing a standard fluorescent microscope and does not rely on the purchase of, nor access

to, more expensive platforms.

Citation Format: Sean Downing, Katir Patel, Jamie Buell, Courtney Hebert. Use of a novel immunofluorescence multiplexing technology, InSituPlex™, for the simultaneous detection of immune cells in multiple cancer types [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 5657.

Abstract 5652: Validation of novel InSituPlex™ technology utilizing standard chromogenic IHC and multiplexed tyramide-based immunofluorescence

Introduction: The demands of today's immuno-oncology research has driven the development of

more robust tools to interrogate the tumor microenvironment, specifically using multiplex

immunofluorescence (mIF). Multiplex immunofluorescence allows investigators to visualize

multiple biomarkers in tissue sections while preserving spatial context. The current gold standard

of immunohistochemistry is “brown” staining using DAB with a hematoxylin counterstain. In

recent years, multiplex immunofluorescence has been accepted in the field using tyramide

molecules (TSA) that permanently fix fluorescent dyes to the tissue. Consequently, tissue integrity

is compromised and cannot be utilized in additional forms of analysis. In addition, high-level

plexing using TSA requires the use of highly specialized and costly imaging equipment. Here we

validate a novel technology, InSituPlex™, utilizing DNA barcoding of antibodies and the detection

of markers using simple DNA hybridization techniques. InSituPlex allows for high-leveling of

multiplexing, preservation of tissue integrity for additional analyses, and implementation on a

variety of commonly used imaging platforms.

Methods: Performance of the InSituPlex technology was assessed using 15 immuno-oncology

relevant markers (CD3, CD4, CD8, CD20, CD25, CD68, CD163, FoxP3, Granzyme B, Ki67, PD1,

PD-L1, pan-cytokeratin, Sox10, and TTF1). Each marker was benchmarked by performing a

single-plex DAB chromogenic IHC and an immunofluorescence assay. Antibodies to be used for

InSituPlex testing were first conjugated to unique DNA barcodes and the resulting conjugates were

assessed utilizing the previously described assays to ensure equivalent performance of the

InSituPlex conjugate to unconjugated antibodies. The performance of the complete InSituPlex

assay was evaluated using InSituPlex HRP labelled imaging strands and the newly barcoded

antibodies. InSituPlex vs. DAB/TSA performance was compared across platforms. All staining

was performed manually and on the Leica BOND RX autostainer. Images were acquired across

multiple imaging platforms, including the Leica VERSA whole slide scanner, and image analysis

was performed using HALO from Indica Labs.

Results: All antibodies performed equally well across all the staining methodologies as assessed

qualitatively. Quantitatively, cell counts were performed on serial sections. Statistical analysis

comparing InSituPlex to standard chromogenic and immunofluorescent assays across all 15

antibodies did not yield any significance (p-value &gt; 0.1).

Conclusions: Detection of relevant immune and cancer related markers utilizing InSituPlex

technology is equivalent to standard chromogenic IHC and immunofluorescent assays.

Citation Format: Katir Patel, Courtney Hebert, Jamie Buell, Sean Downing. Validation of novel InSituPlex™ technology utilizing standard chromogenic IHC and multiplexed tyramide-based immunofluorescence [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 5652.

A MIXED METHODS APPROACH TO TAILORING EVIDENCE-BASED GUIDANCE FOR ANTIBIOTIC STEWARDSHIP TO ONE MEDICAL SYSTEM

Our objective is to operationalize a novel antibiotic advisor, called the personalized weighted incidence syndromic combination antibiogram (pWISCA), intended to help physicians with initial antibiotic choice in hospitals. Clinical decision support tools are a promising technology for providing evidence-based guidance that incorporates data about patients from electronic health records. Nevertheless, congruence with policies and procedures and local experts' opinions, as well as taking into account local resistance data for the medical center's patient population, is needed when selecting and ordering the antibiotic medication options provided by pWISCA. This paper presents findings from applying a mixed methods approach to identify and prioritize antibiotic medications and associated contextual data to display in a CDS tailored to the local hospital. We discuss implications of these findings.

Development and validation of an automated ventilator-associated event electronic surveillance system: A report of a successful implementation

BACKGROUND

Surveillance is an important tool for infection control; however, this task can often be time-consuming and take away from infection prevention activities. With the increasing availability of comprehensive electronic health records, there is an opportunity to automate these surveillance activities. The objective of this article is to describe the implementation of an electronic algorithm for ventilator-associated events (VAEs) at a large academic medical center METHODS: This article reports on a 6-month manual validation of a dashboard for VAEs. We developed a computerized algorithm for automatically detecting VAEs and compared the output of this algorithm to the traditional, manual method of VAE surveillance.

RESULTS

Manual surveillance by the infection preventionists identified 13 possible and 11 probable ventilator-associated pneumonias (VAPs), and the VAE dashboard identified 16 possible and 13 probable VAPs. The dashboard had 100% sensitivity and 100% accuracy when compared with manual surveillance for possible and probable VAP. We report on the successfully implemented VAE dashboard. Workflow of the infection preventionists was simplified after implementation of the dashboard with subjective time-savings reported.

CONCLUSIONS

Implementing a computerized dashboard for VAE surveillance at a medical center with a comprehensive electronic health record is feasible; however, this required significant initial and ongoing work on the part of data analysts and infection preventionists.

High magnetoresistance at low magnetic fields in self-assembled ZnO-Co nanocomposite films

The solid phase growth of self-assembled nanocrystals embedded in a crystalline host matrix opens up wide perspectives for the coupling of different physical properties, such as magnetic and semiconducting. In this work, we report the pulsed laser growth at room temperature of thin films composed of a dispersed array of ferromagnetic Co (0001) nanoclusters with an in-plane mono-size width of 1.3 nm, embedded in a ZnO (0001) crystalline matrix. The as-grown films lead to very high values of magnetoresistance, ranging at 9 T from -11% at 300 K to -19% at 50 K, with a steep decrease of the magnetoresistance at low magnetic fields. We establish the relationship between the magnetoresistance behavior and the magnetic response of the Co nanocluster assembly. A spin-dependent tunneling of the electrons between the Co nanoclusters through and by the semi-insulating ZnO host is achieved in our films, promising with regard to magnetic field sensors or Si-integrated spintronic devices. The effects of thermal annealing are also discussed.

Searching for management approaches to reduce HAI transmission (SMART): a study protocol

BACKGROUND

Healthcare-associated infections (HAIs) impact patients' lives through prolonged hospitalization, morbidity, and death, resulting in significant costs to both health systems and society. Central line-associated bloodstream infections (CLABSIs) and catheter-associated urinary tract infections (CAUTIs) are two of the most preventable HAIs. As a result, these HAIs have been the focus of significant efforts to identify evidence-based clinical strategies to reduce infection rates. The Comprehensive Unit-based Safety Program (CUSP) provides a formal model for translating CLABSI-reduction evidence into practice. Yet, a national demonstration project found organizations experienced variable levels of success using CUSP to reduce CLABSIs. In addition, in Fiscal year 2019, Medicare will expand use of CLABSI and CAUTI metrics beyond ICUs to the entire hospital for reimbursement purposes. As a result, hospitals need guidance about how to successfully translate HAI-reduction efforts such as CUSP to non-ICU settings (clinical practice), and how to shape context (management practice)-including culture and management strategies-to proactively support clinical teams.

METHODS

Using a mixed-methods approach to evaluate the contribution of management factors to successful HAI-reduction efforts, our study aims to: (1) Develop valid and reliable measures of structural management practices associated with the recommended CLABSI Management Strategies for use as a survey (HAI Management Practice Guideline Survey) to support HAI-reduction efforts in both medical/surgical units and ICUs; (2) Develop, validate, and then deploy the HAI Management Practice Guideline Survey, first across Ohio hospitals, then nationwide, to determine the positive predictive value of the measurement instrument as it relates to CLABSI- and CAUTI-prevention; and (3) Integrate findings into a Management Practices Toolkit for HAI reduction that includes an organization-specific data dashboard for monitoring progress and an implementation program for toolkit use, and disseminate that Toolkit nationwide.

DISCUSSION

Providing hospitals with the tools they need to successfully measure management structures that support clinical care provides a powerful approach that can be leveraged to reduce the incidence of HAIs experienced by patients. This study is critical to providing the information necessary to successfully "make health care safer" by providing guidance on how contextual factors within a healthcare setting can improve patient safety across hospitals.