Leveraging Ethnic Backgrounds to Improve Collection of Race, Ethnicity, and Language Data

INTRODUCTION

Healthcare disparities may be exacerbated by upstream incapacity to collect high-quality and accurate race, ethnicity, and language (REaL) data. There are opportunities to remedy these data barriers. We present the Denver Health (DH) REaL initiative, which was implemented in 2021.

METHODS

Denver Health is a large safety net health system. After assessing the state of REaL data at DH, we developed a standard script, implemented training, and adapted our electronic health record to collect this information starting with an individual's ethnic background followed by questions on race, ethnicity, and preferred language. We analyzed the data for completeness after REaL implementation.

RESULTS

A total of 207,490 patients who had at least one in-person registration encounter before and after the DH REaL implementation were included in our analysis. There was a significant decline in missing values for race (7.9%-0.5%, p < .001) and for ethnicity (7.6%-0.3%, p < .001) after implementation. Completely of language data also improved (3%-1.6%, p < .001). A year after our implementation, we knew over 99% of our cohort's self-identified race and ethnicity.

CONCLUSIONS

Our initiative significantly reduced missing data by successfully leveraging ethnic background as the starting point of our REaL data collection.

The neighborhood built environment and COVID-19 hospitalizations

Research on the associations between the built environment and COVID-19 outcomes has mostly focused on incidence and mortality. Also, few studies on the built environment and COVID-19 have controlled for individual-level characteristics across large samples. In this study, we examine whether neighborhood built environment characteristics are associated with hospitalization in a cohort of 18,042 individuals who tested positive for SARS-CoV-2 between May and December 2020 in the Denver metropolitan area, USA. We use Poisson models with robust standard errors that control for spatial dependence and several individual-level demographic characteristics and comorbidity conditions. In multivariate models, we find that among individuals with SARS-CoV-2 infection, those living in multi-family housing units and/or in places with higher particulate matter (PM2.5) have a higher incident rate ratio (IRR) of hospitalization. We also find that higher walkability, higher bikeability, and lower public transit access are linked to a lower IRR of hospitalization. In multivariate models, we did not find associations between green space measures and the IRR of hospitalization. Results for non-Hispanic white and Latinx individuals highlight substantial differences: higher PM2.5 levels have stronger positive associations with the IRR of hospitalization for Latinx individuals, and density and overcrowding show stronger associations for non-Hispanic white individuals. Our results show that the neighborhood built environment might pose an independent risk for COVID-19 hospitalization. Our results may inform public health and urban planning initiatives to lower the risk of hospitalization linked to COVID-19 and other respiratory pathogens.

87. Effectiveness of a Best Practice Advisory in Reducing Inappropriate Urine Cultures

Background

Inappropriate urine cultures can contribute to overutilization of antibiotic treatment for asymptomatic bacteriuria. The objective of this study was to evaluate the appropriateness of urine cultures and the impact of a clinical decision support (CDS) intervention.

Methods

The CDS intervention involved embedding three questions in the urine culture order: whether the patient has fever, leukocytosis or urinary symptoms. When the answer to all three questions is no, a best practice advisory (BPA) alerts the provider that the patient may not meet criteria for a urine culture and suggests cancellation of the order. Cultures obtained in patients experiencing fever, leukocytosis, or urinary symptoms, and those who were pregnant, undergoing invasive urologic procedure, or &lt; 3 years old were classified as appropriate. We performed a quasi-experimental study assessing appropriateness of urine cultures before and after implementation of the BPA. The pre-intervention period was 5/9/19 to 7/31/20 and the intervention period was 2/3/21 to 4/27/21. Random samples of 100 cases from pre- and post-intervention were reviewed to assess appropriateness.

Results

There were 12,679 and 8,270 urine cultures performed pre-intervention and post-intervention, respectively. In 100 cases reviewed pre-intervention, 74% of the cultures were appropriate. Of these, 54% were ordered due to fever or leukocytosis, 50% due to urinary symptoms, and 12% in pregnant women.

Post-intervention, the BPA fired on 458 orders and 106 (23%) were subsequently discontinued. Of the 100 cases reviewed post-intervention, 5 orders were discontinued after the BPA fired. Of the remaining 95 cultures, 78% were appropriate. Of these, 41% were ordered for fever or leukocytosis, 69% for urinary symptoms, and 11% in pregnant women. The change in the proportion of appropriate cultures pre- and post-intervention was not statistically significant (74% vs 78%, respectively, p=0.906).

Conclusion

In nearly one quarter of urine cultures performed, there was not an appropriate indication. Our intervention led to cancellation of 23% urine culture orders and resulted in an absolute increase in 4% of the cultures being ordered appropriately. However, the change in appropriateness was not statistically significant.

Disclosures

All Authors: No reported disclosures

88. Impact of a Computerized Clinical Decision Support Tool on clostridioides Difficile Testing and Oral Vancomycin Utilization as a Balancing Metric

Background

Over diagnosis of hospital-onset Clostridioides difficile infection (HO-CDI) is directly tied to inappropriate C. difficile testing which does not distinguish between infected or colonized individuals. This can lead to inappropriate therapy. Multiple studies have utilized Computerized Clinical Decision Support (CCDS) tools to reduce inappropriate C. difficile testing. Our study looks at the impact of a Self-Assessment CCDS tools on C. difficile testing for HO-CDI and oral vancomycin utilization as a balancing metric.

Methods

Our institution utilizes a two-step test to diagnose HO-CDI that consists of toxin A/B enzyme immunoassay followed by a confirmatory PCR. We applied a self-assessment driven CCDS approach to reduce testing for HO-CDI. Our intervention was deployed in the 3rd quarter of 2018. It asked 3 questions about stool frequency, laxative use and previous C. difficile testing in the order itself. Inappropriate indications for testing included any of the following: &lt; 3 bowel movements within 24 hours, receipt of a laxative within the past 48 hours, or a previous C. difficile test within the previous 7 days. Ordering providers would self-answer these questions. A ‘yes’ response to any of the three questions prevented further test ordering; though respondents had the freedom to change the answer and still proceed with the test order. We evaluated 3 metrics that were all calculated per 1000 inpatient census days: oral vancomycin usage, HO-CDI rates and C. difficile testing rates.

Results

Compared to baseline, our intervention resulted in a significant reduction of C. difficile testing and HO-CDI rates (Figure 1, Table 1). Oral vancomycin usage also decreased significantly (Figure 2, Table 1).

Figure 1. C. difficile testing and Hospital Onset C. difficile Infection Rates by Month, Before and After Intervention

Conclusion

Our self-assessment driven CCDS-based diagnostic stewardship resulted in a significant reduction in inappropriate C. difficile testing for HO-CDI and HO-CDI rates. Oral vancomycin utilization as a balancing metric also decreased significantly. This was despite the use of a self-assessment driven approach with the freedom to change the answers in order to proceed with the test order.

Disclosures

All Authors: No reported disclosures

302. Peripheral IV catheters, a common source of healthcare-associated Staphylococcus aureus bacteremia

Background

Healthcare-associated S. aureus bacteremia (HA-SAB) has traditionally been attributed to surgical site infections (SSI) or central line-associated bloodstream infections. However, peripheral IV catheters (pIV) are increasingly recognized as cause of HA-SAB. This study evaluates risk factors for HA-SAB due to pIV.

Methods

This is a retrospective, case-control study of adult patients hospitalized at Denver Health Medical Center with HA-SAB (SAB presenting with hospital-onset [≥3 days after hospitalization] or community-onset attributed to recent hospitalization [discharge ≤7 days prior]). The time period ranged from Jan 1, 2016 to Nov 30, 2019. Cases were reviewed by an infectious diseases physician to determine the source of SAB. PIV-related SAB was defined as phlebitis, cellulitis, and/or drainage at the site of a previous pIV AND no other source or another less likely source based on progress notes and microbiology results. Three controls were matched to each pIV-related SAB case based on the age of the patient (±5 years) and the date the pIV was placed (±3 days). Patients who were admitted for elective procedures, to psychiatry, to obstetrics, and those who died within 2 days of pIV placement were excluded.

Results

There were 376 episodes of SAB during the study period; 313 were community-onset while 63 were HA-SAB (50 hospital-onset and 13 community-onset attributed to hospitalization). PIV was the most common cause of HA-SAB (n=20, 29.4%); other common causes were SSI (n=10, 15.9%), source present at admission (n=8, 12.7%), and pneumonia (n=7, 11.1%). The median age of patients with pIV-related SAB was 53 years (SD 15.6), and 85% were male. The median duration of pIV was 5 days (SD 2.8). Twenty percent was MRSA.

As compared to controls, pIV in immunocompromised individuals and those placed by emergency medical services (EMS) were more likely to develop SAB (OR 11.8, 95% CI 2.5–56.5 and OR 6.9, 95% CI 6.9–24.0, respectively). Age, gender, pIV location, and duration of pIV were not associated with development of SAB.

Conclusion

PIV placed by EMS are more likely to cause SAB than those placed in the hospital. Facilities should consider changing these pIV promptly upon admission to the hospital and work with EMS to improve pIV placement technique.

Disclosures

All Authors: No reported disclosures

406. Utility of Tracheal Aspirates in Guiding Antibiotic Use in Mechanically Ventilated Patients with COVID-19

Background

In critically ill patients with COVID-19 it is difficult to determine the presence of bacterial co-infection. Many patients receive antibiotics until a bacterial infection can be ruled out. To minimize aerosolization of SARS-CoV-2, non-invasive sampling, such as endotracheal aspiration (ETA), is preferred over invasive techniques. The purpose of this study is to determine the diagnostic yield of ETA and effect of ETA on antibiotic management in patients with COVID-19.

Methods

This retrospective analysis included patients admitted to the intensive care unit (ICU) from March 1 to May 31, 2020 who tested positive for SARS-CoV-2. Patients who did not receive mechanical ventilation were excluded. Data were extracted from electronic medical records. When ETA was performed, records were manually reviewed to determine diagnostic yield and effect on antibiotic management. Diagnostic yield was defined as ETA result with a plausible respiratory pathogen in a quantity of moderate or many. Plausible respiratory pathogens exclude normal flora, yeast, coagulase-negative Staphylococcus sp and Enterococcus sp. The primary outcome is the frequency of initiation, change, no change, or discontinuation of antibiotics based on ETA results.

Results

124 patients with COVID-19 were admitted to the ICU; 76 met inclusion criteria. The average age was 58 years and 75% were male. Hispanic or Latino ethnicity made up the majority of the patient population (63%). Antibiotics were administered to 97% of patients for a median of 11 days of therapy (IQR 7, 21). There were 100 ETAs performed on 55 patients for a diagnostic yield of 21%. ETA led to a change in antibiotic management 47% of the time it was performed. Antibiotic changes include de-escalation (29), discontinuation (7), escalation (6), and initiation (5).

Conclusion

The diagnostic yield of ETA in mechanically ventilated patients with COVID-19 was low. Furthermore, ETA results led to a change in antibiotics less than half of the time. The use of ETA to diagnose bacterial co-infection and guide antibiotic therapy in patients with COVID-19 should be weighed against the risk of using a more invasive sampling technique vs the benefit of potential for increased diagnostic yield. Another conclusion may be to forgo ETA if the result is unlikely to change management.

Disclosures

All Authors: No reported disclosures

Reducing Interruptive Alert Burden Using Quality Improvement Methodology

BACKGROUND

Increased adoption of electronic health records (EHR) with integrated clinical decision support (CDS) systems has reduced some sources of error but has led to unintended consequences including alert fatigue. The "pop-up" or interruptive alert is often employed as it requires providers to acknowledge receipt of an alert by taking an action despite the potential negative effects of workflow interruption. We noted a persistent upward trend of interruptive alerts at our institution and increasing requests for new interruptive alerts.

OBJECTIVES

Using Institute for Healthcare Improvement (IHI) quality improvement (QI) methodology, the primary objective was to reduce the total volume of interruptive alerts received by providers.

METHODS

We created an interactive dashboard for baseline alert data and to monitor frequency and outcomes of alerts as well as to prioritize interventions. A key driver diagram was developed with a specific aim to decrease the number of interruptive alerts from a baseline of 7,250 to 4,700 per week (35%) over 6 months. Interventions focused on the following key drivers: appropriate alert display within workflow, clear alert content, alert governance and standardization, user feedback regarding overrides, and respect for user knowledge.

RESULTS

A total of 25 unique alerts accounted for 90% of the total interruptive alert volume. By focusing on these 25 alerts, we reduced interruptive alerts from 7,250 to 4,400 per week.

CONCLUSION

Systematic and structured improvements to interruptive alerts can lead to overall reduced interruptive alert burden. Using QI methods to prioritize our interventions allowed us to maximize our impact. Further evaluation should be done on the effects of reduced interruptive alerts on patient care outcomes, usability heuristics on cognitive burden, and direct feedback mechanisms on alert utility.

416. Improvement in Syphilis and HIV Screening Rates at a Community-Based Emergency Department in Columbus, Ohio: Six Month Post-Intervention Analysis

Background

Sexually transmitted infections (STIs) disproportionally affect individuals living in underserved areas and Emergency Departments (ED) can play a major role in STI screening. Given the overlapping risk factors for STIs, patients screened for gonorrhea and chlamydia should also be screened for syphilis and HIV. Rates of syphilis/HIV screening in the ED are very low and barriers include lack of knowledge about the risk/prevalence, difficulty with results interpretation, and concerns about follow-up.

Methods

The main study objective was to improve rates of syphilis/HIV screening in a community-based ED in Columbus, OH. A team of clinical providers, case managers, and social workers was formed to address barriers to screening. Using root cause analysis and data feedback, a multistep intervention that included provider education along with expert review of syphilis/HIV results was implemented to ensure proper screening, treatment and rapid linkage to care. Syphilis/HIV screening rates in the ED were compared between two periods: 2012–2017 (pre-intervention) and November 2018–April 2019 (post-intervention).

Results

Between 2012 and 2017, there were 24,427 ED encounters where any STI test was ordered. There were 23,652 (97%) tests for chlamydia, 23,637 (97%) for gonorrhea, 254 (1%) for syphilis, and 466 (2%) for HIV. Twenty-four (0.1%) encounters had screening that included all four tests. Six months after starting the intervention, there were 1,590 encounters where any STI test was ordered. There were 1,444 (91%) tests for chlamydia, 1,446 (91%) for gonorrhea, 493 (31%) for syphilis and 591 (37%) for HIV. Four hundred thirty-eight (28%) of encounters had screening that included all four tests.

Conclusion

Collaborative and practical interventions aimed at improving syphilis/HIV testing have resulted in dramatic increases in syphilis, HIV, and comprehensive STI screening (31-, 19-, 280-fold, respectively) over a relatively short post-intervention period. Additional steps are planned with the goal to further increase screening rates and improve linkage to prevention and treatment programs.

Disclosures

All authors: No reported disclosures.

Multidrug-Resistant Bovine Salmonellosis Predisposing for Severe Human Clostridial Myonecrosis

BACKGROUND The overuse of antibiotics in animals promotes the development of multidrug-resistance predisposing for severe polymicrobial human infections. CASE REPORT We describe a case of spontaneous clostridial myonecrosis due to ulcerative colonic infection with multidrug-resistant Salmonella enterica subsp. enterica, serotype 4,[5],12: i: -. Serotyping of the colonic Salmonella isolate in the index case and the bovine farm outbreak isolates from where the patient worked indicated they were both serotype I 4,[5],12: i: -, which is linked with a multitude of large reported disease outbreaks. Further analysis revealed that they are highly genetically related and antibiotic susceptibility testing indicated that they are phenotypically identical. CONCLUSIONS Enteritis due to human acquisition of multidrug-resistant Salmonella from cattle led to the invasion and dissemination of Clostridium septicum resulting in devastating myonecrotic disease. This highlights the ramifications of co-existence and evolution of pathogenic bacteria in animals and humans and lends support to reducing the use of antibiotics in animals.

1492. Chlamydia, Gonorrhea, Syphilis and HIV Screening among Men Presenting with STI-Related Complaints at a Community-Based Emergency Department in Columbus, Ohio: A 5-Year Retrospective Study

Background

Sexually transmitted infections (STIs) disproportionally affect individuals living in poor and underserved areas of the United States. Emergency Departments (ED) are often the only point of healthcare access for these at-risk individuals. As such, the ED often serves a key role in STI screening. The purpose of this study was to review STI screening practices for men at an urban and community-based ED affiliated with a large academic medical center in Columbus, Ohio.

Methods

Retrospective review of all ED visits from January 2012 to December 2017. A total of 279,929 patient-visits were analyzed for male patients by (1) exposure to an STI (2) STI-related symptoms (penile discharge/pain, scrotal/testicular pain/swelling). We analyzed the demographic characteristics of men who presented to the ED with an STI-related complaint and compared those who underwent STI screening (chlamydia, gonorrhea, syphilis or HIV) to those that did not.

Results

Conclusion

Over a 5-year period, screening for chlamydia, gonorrhea for men presenting with STI-related complaints was adequate. However, syphilis and HIV screening was very low among men presenting to an urban and community-based ED with an STI-related complaint. A separate analysis for women is being done. There is an urgent need to identify and eliminate barriers to syphilis and HIV screening in ED’s that serve at-risk populations.

Disclosures

All authors: No reported disclosures.