Read Abstracts for 2022

To search for a specific abstract, use your browser's search function (Ctrl+F for Windows & Command+F for Mac)

Abstract Title & Poster All Authors Department affiliation Abstract
Best Practice Advisory for IV Antimicrobials

Rebecca Linfield, MD
Sean Carlton
David Epstein, MD
David Svec MD, MBA
Lisa Shieh MD PhD
Medicine Department Title: Best Practice Advisory for IV Antimicrobials


Stanford Health Care has been focused on increasing value-based care. We examined whether a Best Practice Advisory (BPA) “ i.e., an automatic flag “ in our electronic healthcare record (EHR), increased conversions from intravenous antimicrobials to oral equivalents. BPAs have proven effective in other clinical scenarios such as identifying sepsis or decreasing unnecessary antibiotic prescribing. We aimed to save costs, decrease nursing burden, and increase patient mobility though decreased connections to intravenous lines.

Intervention Detail:

Our prior work suggested that initiating cost transparency of medications to providers led to ~25% switch from intravenous to oral equivalents. We thus sought to extend this improvement through another intervention, a BPA. In collaboration with our colleagues from the Infectious Diseases and Healthcare Information Technology divisions, we created a workflow so that our EHR identified patients with enteral access who could receive oral antimicrobials, but were instead receiving these medications intravenously.

Outcomes and Impact:

Over the first three months, 44% (552/1195) of flagged antimicrobial orders were converted from intravenous to oral equivalents. At this current conversion rate, Stanford Health Care will achieve at least $30,000 in cost savings for the initial conversion for the year, with ultimately higher savings when factoring in the total antibiotic course. Our Best Practice Advisory is a low-intensity, high-yield intervention easily applicable to other hospitals with the potential for broad savings.
Clinical Deterioration Prediction Pathway

Robert Gallo
Margaret Smith
Kay Hung
Wilson Alobuia
Jerri Westphal
Ron Li
Rahul Nayak
Pranjal Gupta
Lisa Shieh
Medicine Department Background: Recent studies of inpatient deterioration prediction model pathways have been shown to decrease length of stay and are even associated with decreased mortality. One such model, the Epic Clinical Deterioration Model, estimates the risk of the composite of RRTs, codes, or ICU transfers in the subsequent 6-18 hours using demographic, flowsheet, and laboratory data. The model has been validated at Stanford Healthcare (SHC) and has since been piloted on several nursing units (B3, C3, B2, C2, and K7). Once a patient is alerted as having a high risk of deterioration, care team members (physicians and nursing) are instructed to huddle in-person or by phone to discuss potential reasons for deterioration and interventions to reduce risk. Despite evidence that similar pathways improve outcomes, the SHC pathway is adhered to only about 50% of the time.

Methods: Using survey data, feedback was collected from both nursing and physician users of the Clinical Deterioration Model. Feedback was then incorporated into the pathway to improve usability and adherence. Adherence to the workflow was tracked through flowsheet documentation rates. The composite outcome of RRTs, codes, and ICU transfers were monitored before and after implementation.

Results: From November 29th, 2020 to August 28th, 2021, average workflow adherence was 51% (n=259 encounters; 99.7% CI 42-61%). Based on feedback from users, the Best Practice Alert was updated to directly link to documentation flowsheets and to provide prompts guiding users through the workflow. Subsequently, adherence increased to 61% (99.7% CI 55-69%) on average from August 29th, 2021 to April 3rd, 2022 (n=469). Feedback also indicated misunderstandings about the purpose of the model and associated evidence base. Information on the pathway was incorporated into resident orientation sessions. Additionally, there has been a trend towards improved outcomes since model implementation, with 2.0 events per week after implementation compared to 2.5 events per week prior to implementation (p=0.06).

Conclusion: Improvements to the Clinical Deterioration Pathway have increased adherence. This demonstrates the importance of model usability in implementation efforts. Future efforts will focus on creating more reliable systems processes, with plans to incorporate the unit nurse manager into standard huddle workflows. While documentation is a surrogate for workflow adherence, future studies could additionally evaluate the quality of huddles.
Decreasing Hospital Acquired C. Difficile Infections

Bronson Sansoni
Sophia Gibbs, Suleman Qamar
Medicine Department Background: SHC hopes to reduce the amount of Hospital Acquired Infections (HAIs) in order to mitigate complications that arise during the administration of healthcare. SHC is presently ranked in the bottom quartile among its peers in C. Diff. prevention. SHC is at present at a significant disadvantage amongst its peers due to testing stewardship. Many of the cases are community driven rather than hospital acquired. In order to combat C. Diff. in a cost effective manner, one must have a greater understanding of the magnitude of the problem. One effort at testing stewardship through a hard stop decreased the amount of C. Diff. cases by 100 and 75% in two hospitals within the system without increasing the amount of C. Diff. cases.

Objective: In this study, we will be primarily occupied with laxatives as one of the key symptoms of C Diff. is the presence of loose stools. We intend to arrange sharper temporal constraints in the administration of C. Diff. tests to reduce the number of false positives as current C. Diff. testing is prone to high false positive rates. Alongside such measures, we also intend to prescribe additional suggestions for testing stewardship.

Methods: C. Diff. data from patients at Stanford Hospital in the last quarter of 2021 was analyzed for patients who had been administered laxatives within 1-2 days, 3 days, and 4 days. To supplement this data to gain a greater understanding of the problem at hand, an expansive literature search was undertaken, a survey on C. Diff. prevention was sent out to the nurses at SHC, and medical personnel from several other institutions and SHC were interviewed.

Results: In the last quarter of 2021, 677 patients were tested for C. Diff. 65 patients tested positive, and 612 patients tested negative. 61 patients out of the 677 were prescribed laxatives on the day of testing, the day before, or two days before. 44 patients out of the 677 were prescribed laxatives 3-4 days before testing. Out of the 65 patients that tested positive, only 42 of those were true positives. If testing was eliminated for patients on laxatives four days prior, the PCR testing burden would be reduced by 15.5% and EIA burden would be reduced by 7.69%. 7.1% of true positive cases would be missed.

In the non-laxative population, 14.67 PCR tests would have to be performed per true positive. In the laxative population 35 PCR tests would have to be performed per true positive. The overall odds ratio comparing C. Diff. rates between patients on laxatives versus patients not on laxatives for the last quarter of 2021 was shown to be 0.2899.

Conclusions: Eliminating tests for patients on laxatives in days 0, 1, and 2 prior to testing provides the greatest value. This population consumed 9 percent of PCR tests in the last quarter of 2021 but only accounted for 2.4% of true positives. Patients on laxative regimen days 3 and 4 that were tested consumed 6.5% of PCR tests but accounted for 4.8% of true positives.
Delays In Initiating Adalimumab Treatment in Patients with Hidradenitis Suppurativa: A Retrospective Study

Ghida EL-Banna
Maria Alexandrovna Aleshin
Dermatology Department Importance: Hidradenitis suppurativa (HS) is a chronic inflammatory condition characterized by recurrent nodules, abscesses, and tracts occurring in the intertriginous areas. HS severely affects patients™ quality of life. Prompt initiation of treatment may stop the progression of the disease and improve patient outcomes.

Objective: To determine the length of and reasons for the delays in initiating adalimumab treatment in patients with HS, in addition to HS flares while awaiting adalimumab initiation.

Design: A cross-sectional retrospective study.

Setting: Initiation of adalimumab treatment and follow-up care were conducted at the Department of Dermatology at Stanford University School of Medicine from September 2015 to December 2021.

Participants: 68 patients with HS. 

Exposure: Diagnosis of HS and initiation of adalimumab treatment with current FDA-approved dosing between September 2015 and December 2021.

Main outcomes and measures: total duration of the adalimumab initiation process, the duration of the various phases of the process, the causes of those delays, and HS flares associated with these delays. 

Results: The adalimumab initiation process of 68 HS patients, who were mostly female (76.5%), non-Hispanic (60.1%), English-speaking (92.6%), and between the ages of 20 and 44 years old (70.6%), was evaluated. Most patients had private insurance (63.2%) and required a prior authorization for adalimumab (81%). The median waiting time for initiating adalimumab treatment was 44 days. The longest delay occurred between specialty pharmacy processing of adalimumab orders and receipt of the first dose of adalimumab. The reasons behind delays in initiating adalimumab treatment were multifactorial and involved multiple stakeholders. About a quarter of patients in our cohort experienced flares while waiting adalimumab initiation, and these were associated with potentially avoidable healthcare utilization.

Conclusions and Relevance: HS patients experience delays while waiting for adalimumab initiation. Meanwhile, these delays worsen patient outcomes and increase healthcare utilization, they may be avoided, in some cases, by optimizing the adalimumab initiation process in dermatology clinics.
Diagnostic CT planning for palliative inpatient treatments: a cost-savings and clinical analysis of a pilot program to reduce the need for CT simulation scans

Erik Blomain
Nabil Alnajjar, Alex Chin, Jonathan Lewis, Nataliya Kovalchuk, Kate Horst
Radiology Department Background

Palliative radiotherapy is the most common inpatient consult received by the department of radiation oncology. The current state of treatment for these patients is that they undergo radiation simulation CT scan which allows for radiation treatment planning. A plan is then generated and the patient is treated at some later point (typically one or more days later). However, it has been anecdotally observed that a majority of the patients being seen as inpatient consults already have recent diagnostic CT imaging, some even have such imaging during their current ED/admission encounter.

Recently, advances in software and expertise have allowed dosimetrists and physicists to plan palliative 3d conformal radiation treatments from these prior diagnostic CT scans, thereby eliminating the need for a simulation CT scan. Despite these advances, most patients being treated palliatively as inpatients end up having both a prior diagnostic CT and a (now redundant) radiation simulation CT scan. The current study reports preliminary results of a pilot program in our department to promote the use of diagnostic CT planning for palliative inpatient treatments. Although previous studies have identified the feasibility of this approach, data regarding associated cost-savings and clinical outcomes of this paradigm remain to be refined.


We hypothesize that radiation planning from diagnostic scans will facilitate faster, more efficient treatment for palliative cases, and ameliorate the need for an extra CT scan which exposes the patients to additional ionizing radiation as well as increased discomfort, incurs costs and results in delays of delivering care with negligible benefit.


We developed a planning workflow which was developed collaboratively with all stakeholders involved (i.e. physicians, physicists, RTTs, schedulers and dosimetrists) to facilitate communication and process improvement for diagnostic scan planning.

Diagnostic scans were screened for quality and suitability via established criteria. Planning principles included simple beam arrangements, low modulation, large margins and density overrides. Staff communication and documentation was optimized for the new workflow.
During the interventional period, we collected clinical data from both patients receiving the alternate workflow (diagnostic scan planning) and those being planned conventionally with a CT simulation. Specifically, time from consult to beam-on, opioid use and maximum pain scores were quantified and compared between these patient cohorts. Cost savings were quantified with the help of the hospital finance team as part of the Cost-Savings Reinvestment Program.


Here we qualitatively report our planning workflow which was developed collaboratively with all stakeholders involved (i.e. physicians, physicists, RTTs, schedulers and dosimetrists) to facilitate communication and process improvement for diagnostic scan planning.

During an interim analysis of 6 months of a planned 12 month intervention, we observed a direct cost savings of $2556 associated with the elimination of inpatient CT sims in favor of diagnostic scan-planning. Of note, this does not include indirect costs such as machine time and staff time saved, which are likely unquantified sources of additional savings.

In terms of clinical outcomes, a pilot analysis was conducted for the first 10 patients (5 diagnostic scan patients and 5 conventionally-planned patients). Unexpectedly, we observed a longer time-lapse between consult to beam-on using the new workflow, although this may be reflective of different distribution of the treatment target and indication. There was no observed difference in time-lapse between consult and discharge.


Diagnostic CT planning for palliative inpatients is a clinically-tractable and cost-saving intervention with potential to improve clinical outcomes for this patient population . Further study is needed to evaluate the clinical findings reported here with more granularity.


Wong et al. Diagnostic Computed Tomography Enabled Planning for Palliative Radiation Therapy: Removing the Need for a Planning Computed Tomography Scan. PMID: 33186781.
Greening the Clinics: A Resident-run Effort to Reduce Medical Waste at Stanford 

John Chan, MD
Claire Baniel, MD
Melissa Frick, MD
Michelle Lee, MD
Paige Fox, MD
- -
How Many Future Lives Will You Save? Gamification Is An Effective Tool For Patient Safety and Quality Education

Gabriela Ruiz Coln BA
Kambria Evans MEd, MA, LMFT
Mia Kanzawa MD
Laurence Katznelson MD
Lisa Shieh MD, PhD
Medicine Department Background
Every year, as many as 98,000 people die from medical errors within the hospital, making medical errors one of the top 10 leading causes of death in the United States1. While the Accreditation Council for Graduate Medical Education (ACGME) has incorporated healthcare quality and patient safety into the Clinical Learning Environment Review (CLER) program to determine institutional accreditation, medical trainees were found to have limited knowledge of basic quality improvement (QI) concepts in most clinical learning environments. Current QI and patient safety curricula most commonly involves discussions and lectures, but notably lack robust trainee engagement. Gamification is a strategy that has been shown to enhance learner engagement and improve perceived importance of material2,3. Here, we describe results from the deployment of a novel, gamified patient safety and QI course entitled Safety Quest (SQ).

The SQ curriculum was developed based on the ACGME™s CLER. The core curriculum is structured as four levels that engages learners in topics such as communication techniques, QI tools, root cause analyses, failure modes effect analysis, and statistical process control charts (Table 1). Both medical trainees (residents, fellows, medical students) and Continuing Medical Education (CME) learners participated. We assessed the impact of SQ on two levels: learner satisfaction (Kirkpatrick™s level 1) and changes in learner attitudes and measures of learning knowledge (Kirkpatrick™s level 2). CME learners were assessed via pre- and post-test performance and satisfaction, which were analyzed with Fisher™s exact tests. Meanwhile, trainees™ preferences were collected, and thematic analysis of their feedback was conducted using two independent coders and corroborated using Cohen™s kappa coefficients.

In 2018 alone, 1364 graduate trainees and 65 medical students completed a level of SQ, while 374 CME learners participated in 2019-2020. Among CME learners, 59% resided in North America, 19% in Asia, and 11% in Europe, and users included physicians and trainees (51%), non-physicians (15%), nurses (3%), and allied health professionals (3%). Upon completing SQ, 98.6% of CME learners passed the post-test, whereas only 59.2% passed the pre-test (p < 0.0001). Among trainees, SQ outranked all other independent learning modalities with 45% of graduate trainees preferring SQ over video, HealthStream, PowerPoint, and articles (Figure 1). Cohen™s kappa for rating overall sentiment of comments was 0.90. Raters identified three major content areas: engagement, ease of use, and effectiveness. 87% of comments addressing engagement were positive, whereas 52% and 41% of comments addressing effectiveness and ease of use were positive, respectively. Among CME learners, 93% agreed or strongly agreed that SQ was engaging and interactive, 92% believed it contributed to their professional growth, and 90% believed it covered content useful to their practice.

SQ is an effective tool to promote engagement in learning critical information on patient safety and QI. While lectures and articles can be effective methods to teach QI vocabulary or share QI project examples, gamification allows users to apply the critical thinking skills of patient safety and QI that are required to reduce medical errors in clinical practice. Moreover, a virtual tool such as SQ allows for greater flexibility for learners to complete the curriculum and allows for national and international dissemination.
How Quickly Can You Pace? Implementation of Transvenous Pacemaker Placement Educational Guide in the Emergency Department

Nicole Prendergast, MD Emergency Medicine Department  Background: The insertion and management of a temporary transvenous pacemaker can be a life-saving, yet complicated procedure in the emergency department. Due to its infrequent performance emergency physicians are less comfortable performing it compared to other procedures without specialist consultation. Despite the essential nature of this procedure and published success of its performance by emergency physicians there is no published real-time, bedside instructional guide for review prior to or guidance during placement.

Methods: A transvenous pacemaker placement checklist, with QR codes for video instruction, was developed and included in a box inclusive of the procedure kit and pacer control box after review with emergency department operations team. Residents and faculty were encouraged to utilize this prior to and during the completion of this procedure. Number of pacemaker placements and kits utilized in the department during the 8 months prior to implementation, 3/2021, to present, 13 months, were recorded.

Results: Transvenous Pacemaker placement in the Stanford Emergency Department increased from 37.5% in the 8 months (3 procedures) prior to guide implementation to 61.5% in the 13 months following implementation (8 procedures)

Conclusion: Transvenous pacemaker placement, although rare, is a critical procedure that emergency physicians must be competent and efficient in performing. Although this procedure is uncommon, the number performed by Stanford resident physicians in the year following implementation of a real-time educational guide led to a 24% increase in the performance of this procedure. Initial data on the small amount of procedures performed points to need for ongoing evaluation of this low cost tool in our department, including frequency of use, effect on patient outcomes, and specialist utilization.
Identifying False Negatives within an Automated Intelligence Clinical Deterioration Prediction Tool

Pranjal Gupta MD
Rahul Nayak MD MTM
Robert Gallo MD
Kenny Shum PhD
Ron Li MD
Margaret Smith MD
Lisa Shieh MD
Medicine Department Identifying False Negatives within an Automated Intelligence Clinical Deterioration Prediction Tool

Pranjal Gupta MD, Rahul Nayak MD MTM, Robert Gallo MD, Kenny Shum PhD, Ron Li MD, Margaret Smith MD, Lisa Shieh MD


Patients that experience unplanned transfers to higher levels of care experience greater mortality and hospital length of stay.(1) Numerous previous efforts have been made to develop early warning systems of such patients based on predictive modeling from lab values, vital signs and other data available from the electronic health record.(2) The EPIC Clinical Deterioration Model is one such tool that calculates the probability of a rapid response team alert (RRT), code or ICU transfer in the next 6-18 hours. This tool has been internally validated and has been implemented at Stanford Health Care amongst a specific pilot population. However, further reflection needs to be done regarding the accuracy of predicting deterioration events at Stanford. One technique is to evaluate false negatives, or instances of an RRT or ICU transfer in which a Clinical Deterioration Alert (CDA) was not fired. The goal of this study was to evaluate the frequency of false negatives and evaluate characteristics of false negative events to eventually improve the Clinical Deterioration Model and decrease the false negative rate.


This study was a single-center retrospective chart review identifying patients within a pilot population within 4 General Medicine wards from November 1, 2021 to December 13, 2021. All patients who had an RRT, code or ICU transfer (deterioration event) but were not flagged as high risk by the model were identified (these patients did not have a CDA label associated with their chart, aka CDA firing ). Several variables were collected, including admission time and date, event time and date, initial department, ending department, service at time of transfer, age, gender, last set of vitals prior to event (heart rate, blood pressure, temperature, oxygen saturation, respiratory rate), last set of lab values prior to event (hemoglobin, white blood cell count, potassium, pH), as well as characteristics of the event, including escalation of care (death event, intubation, non-invasive positive pressure ventilation, cardiac arrest, vasopressor initiation, or extracorporeal membrane oxygenation initiation) and a narrative summary of the event.


A total of 56 patients were identified that had a deterioration event without being identified by the model as high risk. 37 patients had an ICU upgrade, while 19 patients either had an RRT or code called. 5 patients within the RRT/code group were also included in the ICU upgrade group (3 of these 5 had ICU upgrades as a result of the RRT, while the other 2 upgraded to the ICU at an alternate time). Out of the 37 ICU upgrades, 34 events did not have a CDA (34 false negatives). Out of the RRT/code group, none had a CDA (19 false negative). In total, 91% of ICU grades and 100% of RRT/codes within the selected patient population were missed by the Clinical Deterioration Model, or 93% of total patients. Of these 52 false negatives, 8 were intubated during the event, 11 required vasopressors, 1 required ECMO, 4 required non-invasive positive pressure ventilation (NIPPV), and 2 sustained cardiac arrest. 31 patients had a deterioration event without any of these actions. A few of these events included ICU upgrade for severe agitation, seizure, frequent neurological checks, airway watch, respiratory failure, cardiopulmonary collapse, syncope, and brief resolving arrhythmia. Of the 4 events with appropriate CDA firing, 3 involved respiratory failure and 1 required vasopressors.


An exceedingly large majority of patients with adverse events during our selected time frame and patient population were not labeled as high risk by the model. Of note, many of these events did not have a clinical prodrome, i.e. gradually decreasing blood pressure, increasing heart rate, etc., but rather were sudden events with a prodrome, such as syncope, seizure, agitation or sudden arrhythmia. Although many variables are incorporated into the AI model such as cardiac rhythm or nursing neurological assessment, a hypothesis is that this tool may struggle with events that do not have certain changes in flowsheet data that would offer valuable input into the AI algorithm. Conversely, many events did have prodromic features including hypoxemic respiratory failure, and septic shock, requiring intubation, vasopressors or other care escalation, but did not lead to an appropriate CDA firing. This may provide an opportunity to recalibrate the model over time. Overall, it is quite difficult to draw large conclusions with a small dataset over a pilot population in 1 month. Planned future analyses will investigate model drift, including measures of negative predictive value and positive predictive value, evaluate individual CDA scores, and include larger sample sizes.

Escobar GJ, Greene JD, Gardner MN, Marelich GP, Quick B, Kipnis P. Intra-hospital transfers to a higher level of care: contribution to total hospital and intensive care unit (ICU) mortality and length of stay (LOS). J Hosp Med. 2011 Feb;6(2):74-80. doi: 10.1002/jhm.817. Epub 2010 Dec 17. PMID: 21290579.
Escobar GJ, Liu VX, Schuler A, Lawson B, Greene JD, Kipnis P. Automated Identification of Adults at Risk for In-Hospital Clinical Deterioration. N Engl J Med. 2020 Nov 12;383(20):1951-1960. doi: 10.1056/NEJMsa2001090. PMID: 33176085; PMCID: PMC7787261.
Improving Advance Care Planning and Reducing Readmissions Amongst High Risk Hospitalized Patients

Benjamin Catanese, Zoe Fullerton
Aditya Ullal
Julia Sun
Dr. Lisa Shieh
David Crichton
Medicine Department Introduction:
During our analysis of preventable sepsis readmissions from Oct 2019 through Sep 2020 it was noted that inappropriate goals of care discussions was a significant cause of readmissions. Advance care planning further stood out as area of intervention because through interviewing physicians and advance practice providers it was noted that many were not using Epic tools to identify hospitalized patients at high risk for readmission or morality that would most benefit from advance care planning. It was also noted that general surgery and surgical subspecialty residents, and surgical advanced practice practitioners do not receive formal training in carrying out advance care planning. The goal of this project was to utilize the Epic tools and Serious Illness Care Program training to increase advance care planning for some of the most vulnerable patients in the hospital.

In Epic two tools were identified, these included the ToC Readmission Risk machine learning tool that identifies patients as being low, medium, and high risk of readmission based off data gathered from two hospital sites. The Recommended for Advance Care Planning tool flags patients that are in the top 25th percentile in the estimated risk of 1-year mortality among patients discharged from the General Medicine Service at SHC. Data was gathered on these patients flagging with both these tools between 7/2020 and 3/2022. Information guides were distributed to internal medicine residents and hospitalists describing how to add the Epic tools to their patient lists and steps to take if the patient flags as high mortality risk under the ACP tool and high risk of readmission through the Readmission Risk tool. SICP training was coordinated for general surgery and surgical subspecialty interns, and trauma surgery advanced care providers.

Between 7/2020 and 3/2022, there were 2411 patients that were flagged by the High Risk of Readmission tool alone and 1466 patients bother the Recommended for Advance Care Planning tool and the Hisk Risk of Readmission tools. Patients that flagged for both were found to have higher rates of readmission within 30 days (18.5% vs 15.0%, p =0.003) and 90 days (28.8% vs 14.9%, p<0.00005), higher rate of mortality during the hospital stay (3.0% vs 1.9%, p = 0.035) and overall recorded mortality (20.8% vs 13.6%, p<0.00005), lower rate of discharge to home w/o services (54.5% vs 61.7%, p<0.00005), and higher rates of having a hospital charge >$500K (16.3% vs 12.6%, p = 0.001) and hospital net loss of >$250k (31.6% vs 27.2%, p = 0.003). The general surgery residency and advance care provider (APP) leadership were both open and excited to coordinate Serious Illness Care Program training for the interns and APPs. Lastly, given the positive feedback to information guide from internal medicine residents and hospitalists, a best practice alert proposal was formed utilizing the data gathered for the project in order to alert providers if patients flag with both Epic tools and provide directions to fill of the Advance Care Planning form and to coordinate discharge planning with the team case manager and pharmacist.

Patients who are flagged by both the High Risk and the ACP tools in EPIC constitute a population at high risk of readmission, morbidity and mortality, and low value health care and should be pursued as a first start for SICP expansion. The new training provided to surgical residents and APPs along with the Best Practice Alert will serve to increase the number and quality of ACP discussions that occur and are documenting with the goal of reducing unplanned readmissions.
Improving and Enhancing Collaboration, Safety Champions, Resident Safety Council (RSC), Stanford University

Anna Pendrey
Claire Baniel
Stephen Ma
Mark Keroles
Medicine Department Improving and Enhancing Collaboration, Safety Champions, Resident Safety Council (RSC), Stanford University
Anna Pendrey1, Claire Baniel1, Stephen Ma1, Mark Keroles1
1Stanford University

Through the years challenges in collaboration and communication have presented in terms of patient safety initiatives. The need for residency programs to be represented in a council to expose concerns and develop solutions to important system challenges within a variety of different healthcare practice settings such as health equity, safe reporting, hospital acquired infections and climate change has been needed.
This academic year the RSC created the safety champions committee to address this and other safety needs throughout the organization.

Resident Leaders from the Residency programs were elected with expectations to participate in the committee™s educational curriculum and share their acquired knowledge at their respective residency programs. Participants immediately became engaged and received education in SAFE reporting, Health Equity, Greening the Clinic Project, and Hospital Acquired Infections.

The goal of the Safety Champions Committee this academic year was to engage and promote representation from all Residency programs.
Engage participation of the residents in the Health Equity workshops, Greening the Clinic Workshops, and promotion of education in Hospital acquired Infections and SAFE reporting.
A survey was conducted so participants could document the preferred presentation method and preferred educational material format for distribution in their residencies, all this was done to improve and enhance collaboration between programs.
19 (68%) residency programs chose to be represented during this academic year through the Safety Champions Initiative.
Most of the participants preferred presentation method of educational materials through List Serv and didactics.
Most of the participants preferred educational material was PDF and Prewritten Emails.
Multiple challenges are faced when improving and enhancing communication between disciplines
As there is difficulty in communication between residents, inconsistency of representation and leadership between programs non standardized educational material formats and educational presentation modalities.
Key Learning points and Conclusions
Communication between programs is key to enhance patient safety through collaboration, innovation, and leadership.
Participation and representation of all residency and fellowship programs is encouraged for next academic year to promote Safety Champions throughout Stanford.
 Improving and enhancing collaboration through multiple disciplines ensures the highest quality of healthcare delivery to our patients and wellbeing to patients, residents, and fellows.
Improving Osteoporosis Screening in Rheumatology Clinic

Natalie Howlett
Liya Stolyar
Kaniksha Desai
Janice Lin
Medicine Department Undiagnosed and untreated osteoporosis leads to fragility fractures, which cause significant morbidity for our patients at a high cost to our healthcare system. Glucocorticoid use for three months or longer put patients at significant increased risk for osteoporosis, so early and regular osteoporosis screening in patients on long term steroids is critical for preventing fractures. Our goal was to increase DEXA scan order rate for rheumatology clinic patients on long term steroids from 23.7% to 35% by March 31. Through our root cause analysis, we found three main reasons why DEXA scans were not completed: scans were ordered but not completed, scans were completed outside of Stanford but not documented in our EMR, and physicians did not order scans for patients who were eligible for screening. Using this analysis, we developed key drivers and interventions. We implemented several interventions addressing physician education. We trained our MAs to upload DEXA scans in the imaging tab in Epic, as well as to pend a DEXA order for all patients on long term steroids prior to their clinic visit. Finally, we designed an Epic order set that includes DEXA orders with prednisone prescriptions which will soon be available to our division.
We have seen a steady increase in DEXA order rates since the start of our project, and although we did not yet reach our goal, we hope to exceed it after implementing our upcoming interventions. In the future, we plan to expand our project to address osteoporosis treatment through partnering with Endocrinology. The interventions we created can be adopted by any division that prescribes long term glucocorticoids so as an institution we reduce the morbidity and mortality associated with osteoporosis for our patients. 
Increasing Outpatient Goals of Care Discussions in the Resident Neurology Clinic

Trevor Rafferty MD & Jacqueline Summers MD
Carl Gold MD
Neurology Department Studies suggest that most patients want to have advanced care planning (ACP) discussions with their outpatient neurologist. However, neurology residents at Stanford do not routinely have outpatient ACP discussions with their patients. In this project, we have aimed to understand the current barriers that prevent residents from leading ACP discussions in clinic. By doing so, we have been developing tools to empower residents to have ACP discussions with their patients in clinic. Residents participated in a survey quantifying the number of ACP/goals of care discussions they had held in continuity clinic as well as the number of times they had used the Stanford EPIC ACP SmartPhrase to document these conversations. In a survey of 18 Stanford Neurology residents, 12 had never led an outpatient ACP conversation, and only 2 had used the ACP SmartPhrase. Through discussions with residents, we identified some of the key drivers for this finding which were: 1) lack of resident comfort with leading outpatient ACP conversations, 2) resident uncertainty regarding the proper documentation of such discussions, and 3) time constraints making such discussions challenging in a busy clinic. We have created a schedule for proposed interventions that includes dissemination of a Tip Sheet , posting a How-to document for the ACP SmartPhrase in continuity clinic workrooms, implementing Serious Illness Care Program resident didactic training for residents, and training attending physicians to facilitate resident involvement in ACP conversations. We will measure the efficacy of these interventions by tracking the resident usage ACP SmartPhrase. We aim to increase the usage of the ACP SmartPhrase by neurology residents to 60% by July 2023.
Management of New Pregnancy Diagnosis in the Emergency Department

Carl M Preiksaitis, MD
Andrea Henkel, MD, MS
Kate A Shaw, MD, MS
Monica Saxena, MD, JD
Emergency Medicine Department  Introduction:
Emergency Medicine (EM) providers are required to disclose the new diagnosis of pregnancy to patients presenting for care. Approximately 6% of emergency department (ED) visits among people of reproductive-capable people result in the new diagnosis of pregnancy, and this increases to 13% of patients presenting with abdominal or pelvic complaints [1]. Half of all pregnancies are unplanned, and half of these pregnancies end in abortion [2]. The management of most ED complaints in reproductive-capable people requires a screening pregnancy test, thus EM providers are often the first to diagnosis and disclose a pregnancy. However, EM education does not routinely cover how to explore if the pregnancy is desired or that patient™s intention for continuing the pregnancy. This counseling is desired by patients: a qualitative analysis identified that EM providers do not routinely provide adequate resources or support in what is often an emotionally difficult encounter [3]. We sought to examine deficiencies in provider knowledge and develop solutions to improve care for these patients.

Project Aims:
To identify provider education gaps in management of new pregnancy diagnosis in the ED, and to develop training and an interdepartmental protocol that improves patient experience and access to medical abortion.

Through collaboration between the Obstetrics & Gynecology section of Complex Family Planning and the Department of Emergency Medicine, we developed a protocol to standardize management of a new pregnancy diagnosis in the ED. This protocol includes the provision of performing medication abortion in the ED, if appropriate and desired. We performed a cross-sectional analysis of emergency physician comfort and familiarity with management of a new pregnancy diagnosis in the ED. Finally, we completed a didactic session during EM departmental Grand Rounds on options counseling. This session included skills for engaging with the patient, overview of options available, and details regarding administering medication for abortion.

Results/Planned Outcomes:
We surveyed a cross-section survey of academic EM physicians (attendings, fellows residents), exploring knowledge, counseling, options and desire for additional training. Ninety-seven EM physicians completed the survey (response rate 63.3%). Most physicians reported comfort with the new diagnosis of pregnancy (85%); only 39% reported that they always explore if a pregnancy is desired. Many physicians (69%) reported at least some familiarity with options counseling but only 10% report discussion options with the patient. 80% desired additional training in options counseling. As we implement this protocol in the ED, we plan to prospectively evaluate frequency of use and additional touchpoints in the healthcare system. To further understand patient experience during encounters that end in abortion, we plan to undergo a qualitative analysis of patient™s experience in the ED who elect to undergo medication abortion. Finally, we plan to solidify management of new pregnancy diagnosis as part of the regular EM curriculum at Stanford. Our long-term goal is to bring this protocol and educational materials to other institutions.
Multidisciplinary Team Interventions to Reduce Hyponatremia-Related Readmissions After Transsphenoidal Surgery

Gaby Ruiz Coln, BA
Erin Wipff, NP
Olivia Chu, NP
Victoria Gates-Barzabay, RN
Khristine Vinluan, RN
Shelli Chittum, NP
Juan Fernandez-Miranda, MD
Julia J. Chang, MD
Neurosurgery Department Over 100 patients undergo transsphenoidal surgery (TSS) at Stanford Health Care each year. Due to vasopressin hormone fluctuations from the pituitary gland and shifts in intravascular volume, delayed hyponatremia from SIADH commonly occurs between post-operative days (POD) 4-12 with varying levels of severity. Symptomatic and severe hyponatremia can result in readmission and averages $19,000 in readmission hospital costs at our institution. Between June-November 2021, our readmission rate was 8.5% from delayed hyponatremia and SIADH following TSS, slightly higher than those reported from other institutional studies (3.4%-7.0%).

In this project, our primary goal is to decrease the six-month rate of hyponatremia-related readmissions after TSS from 8.5% to a target of 5% (overall reduction of 40%) by June 2022. Our secondary goal is to decrease the six-month rate of POD8 hyponatremia (sodium ‰¤134 mEq/L) from 28% to 17% during the same time frame.

An inpatient and outpatient multidisciplinary team involving neurosurgery and endocrinology bedside nurses, nurse practitioners, residents, fellows, and attendings employed A3 methodology to perform root cause analysis to recognize issues affecting POD8 hyponatremia and readmissions and identify key drivers and potential interventions. Five key drivers were identified: (1) deliver consistent provider messaging to the patient, (2) ensure appropriate fluid intake for each patient, (3) allow for reliable outpatient fluid intake measurements, (4) decrease inappropriate DDAVP usage for diabetes insipidus, and (5) obtain reliable and timely outpatient labs.

Between January-March 2022, several interventions were deployed, including more uniform and simple post-operative discharge instructions; initiation of 1 liter/day fluid restriction for patients at the time of discharge; provision of a hospital pitcher, urinal or urine hat, and fluid log to patients for outpatient intake/output monitoring; implementation of a new inpatient DDAVP protocol for diabetes insipidus treatment; and inpatient provider education sessions. Since the beginning of our intervention period, the readmission rate is 0%, and the POD8 hyponatremia rate is 16% as of April 1, 2022. Such significant reduction in readmissions demonstrates significant value creation in the form of hospital cost savings, inpatient bed availability, and staff burnout. Future work will focus on sustainability of interventions and barriers faced by patients in addressing their post-operative fluid status.
Official Radiology Policies for Residents in a Hurry: Increasing the Accessibility of Official Hospital Policies for Radiology Residents on Call or Injection Shift

Sanna Herwald MD PhD
Jennifer Padwal MD MSc
Preya Shah MD PhD MSBE
Hugo Decker MD PhD
Amy Bui MPH
Erlina Haun MS MBA PMP
Marc WIllis DO
Gloria Hwang MD
Radiology Department Official Radiology Department policies ensure safe, high-quality, and consistent care for all patients undergoing imaging, including patients receiving CT or MRI contrast, patients for whom an MRI may be unsafe, patients receiving radioactive tracer material internally, and potentially pregnant patients. Radiology residents on call and injection coverage shifts are expected to independently answer policy-based questions about patient imaging which span all of the modalities of radiology and all of the radiology sub-specialties. Lack of knowledge or experience in these scenarios can lead to non-conformity with hospital policy, delayed patient care or adverse outcomes.

Our SMART goal was to increase the confidence of radiology residents answering policy-based questions while on call or injection coverage shifts, measured on a 1-5 Likert scale, by 10% above the 2/25/2022 baseline by 4/4/2022. We found that the baseline confidence of junior residents answering policy-based questions was relatively low compared to senior residents.

We instituted multiple interventions to increase the ease with which residents residents can access, review, and ask questions about hospital policies, empowering residents to add novel tools into their workflow. These interventions allowed us to meet our SMART goal, with the greatest improved confidence reported by junior residents. The resources created during this project will continue to be updated and improved based on feedback from residents and other resource users. 
Operation Waste Reduction: An Initiative To Decrease Skin Biopsy Tray Waste

Paige W. Wolstencroft, MD, MS
Natasha C. Zacher, MD
Sierra Centkowski, MD
Bernice Y. Kwong, MD
Dermatology Department Background
The high contribution of healthcare to climate change is well established and in 2019, Health Care Without Harm reported that as a sector, the healthcare system is responsible for 4.4 % of global net emissions. 1,2 The United States healthcare system if counted as a country would rank 13th globally in global net emissions.1,2

Skin biopsies are performed with high frequency in dermatology clinics across multiple clinical sites, and the number of supplies and materials used to set up biopsy trays varies, making a project focused on assessing and decreasing skin biopsy tray waste both impactful and scalable. During one representative week of our initial waste audit, 46 skin biopsies were logged by 15 residents across 4 clinical sites. On average each Stanford dermatology resident is performing 3 skin biopsies per week. When combined, this means as a residency program we are performing 184 skin biopsies every month and 2,392 skin biopsies every year “ and these are conservative estimates. According to the American Board of Dermatology, there are approximately 1590 dermatology residents in the United States. If each one of those residents is similarly performing 3 skin biopsies per week, there are almost a quarter of a million opportunities by dermatology trainees to decrease environmental waste every year through skin biopsy tray waste reduction.

We approached our skin biopsy tray waste reduction project by breaking it up into 4 broad steps: (1) determine the amount of waste from biopsy tray setups at each clinical site with as much granularity as possible; (2) analyze data collected during step 1 to identify areas for improvement and develop sustainable interventions; (3) implement interventions to reduce skin biopsy tray waste; (4) collect data on skin biopsy tray waste after interventions to determine our degree of success.
Step 1 began with a residency-wide waste audit to determine what items and clinical sites had the highest frequency of skin biopsy tray waste and where to begin targeting interventions. Waste audits were conducted at the Cutaneous Oncology Clinics & Cutaneous T-cell Lymphoma Clinics (Stanford Cancer Center), Stanford Medicine Outpatient Center Dermatology Clinics, Veterans Affairs Dermatology Clinics, and Santa Clara Valley Medical Center Dermatology Clinics. Data was collected on total amount of biopsy tray waste at each site and waste per tray based on supplies and setup. Data collected during this initial waste audit allowed us to generate our SMART goal: decrease the average number of wasted supplies per skin biopsy tray in the Cutaneous Oncology Clinics by 50% from a mean of 9.6 to 4.8 wasted items per tray by March 15th 2022. Recognizing that interventions would require approaches customized to each clinical site, we focused first on the Cutaneous Oncology Clinics given they had a higher number of wasted items per tray. Through discussions that included clinical team members (nursing, medical assistants, residents, attendings) and administrative management team members, we explored potential causes of significant biopsy tray waste and considered ideas from all stakeholders. We gained support for our initiatives by sharing data about the potential environmental and financial impact of making changes to biopsy tray setup. We ultimately implemented targeted interventions and continued to perform waste audits and modify our interventions until we reached our target goal.

As of March 30th 2022 we have decreased the average number of wasted supplies per skin biopsy tray from 9.6 to 1.6 in the Cutaneous Oncology Clinics at the Stanford Cancer Center, which exceeded our target goal. This represents 2,943 fewer pieces of waste per year that will be generated by these clinics. We have also replaced the biopsy tray coverings from chux to 2-ply paper which we estimate will decrease the amount of waste by 19.6 kg every year from the Cutaneous Oncology Clinics. Prior to our interventions, 22% of skin biopsy trays had >10 pieces of wasted gauze and 4% had 0 pieces of wasted gauze. After our final interventions, 0% of skin biopsy trays had ‰¥ 10 pieces of wasted gauze and 70% had 0 pieces of wasted gauze. Similarly, prior to our interventions 74% of skin biopsy trays had ‰¥ 1 wasted alcohol swab and after our final interventions 0% of our skin biopsy trays had any wasted alcohol swabs. Since implementation of our biopsy tray set-up protocols, we have received 0 concerns about the new process and everyone performing biopsies has had the supplies required to perform the procedure successfully and safely.

Skin biopsy tray setup can be optimized to reduce waste without compromising ability to safely perform skin biopsies. Clinical sites require individualized approaches to decrease biopsy tray waste, but we anticipate that many of the interventions successfully implemented at the Cutaneous Oncology Clinics at Stanford Cancer Center will be easily adaptable to other dermatology clinics and help to decrease biopsy tray waste across clinical sites at Stanford and beyond.

1. Josh Karliner, Slotterback S, Boyd R, Ashby B, Steele K. Health care™s climate footprint: How the health sector contributes to the global climate crisis and opportunities for action. Published online 2019.
2. Fathy R, Nelson CA, Barbieri JS. Combating climate change in the clinic: Cost-effective strategies to decrease the carbon footprint of outpatient dermatologic practice. Int J Womens Dermatol. 2021;7(1):107-111. doi:10.1016/j.ijwd.2020.05.015
Opportunity to Increase Utilization of General Inpatient Hospice at Stanford

Aditya J. Ullal MD
Lisa Shieh, MD PhD (Project Mentor)
Medicine Department Stanford™s general inpatient hospice (GIP) is less frequently utilized for patients with acute symptom crises at the end of life compared to historic GIP use at Stanford. Overall, GIP utilization remains below pre-COVID pandemic levels. This abstract presents the results of a prospective screening of "end-of-life" patients on Medicine, Oncology, Hematology, and Medical ICU that was conducted twice per week during the last two weeks of February 2022. Twenty-five EOL patients (average age 75.23 ± 13.2, average length of stay 18.8 days) were identified. Eight patients (32%) had end-of-life symptoms while in the hospital that could have benefitted from referral for GIP. Five patients experienced in-hospital death, of which up to 3 may have eventually qualified for GIP before death. Twice per week screening of DNR/DNI patients is feasible. Additionally, screening identified opportunities for improvement of EOL care including underutilization of comfort care nursing orders and potential to simplify screening by creating a standalone "comfort care" code status.
Personalized Scrub Caps for Improved Communication and Professional Wellness

Katie Speirs, MD
Sara Goldhaber-Fiebert, MD
Anesthesia Department Introduction:

Effective communication in perioperative settings at an academic institution can be challenging. Team members change frequently and each residency matriculation cycle brings new resident members to the operating room. While standard name badges are mandatory, they are likely printed in small font, may flip backwards, and may be covered by PPE. Communication problems were identified as the root cause of approximately 70% of adverse events reported to the Joint Commission and it has been shown that communication failures in the OR occur in approximately 31% of team exchanges [1]. Personalized scrub caps with visible name and role have been proposed as a potential method of improvement [2].
In 2021 at Stanford Healthcare a three-stage quality improvement project was completed with an attempt to demonstrate whether the introduction of personalized scrub caps can improve perioperative communication among OR team members and decrease depersonalization among anesthesiology residents, fellows, and attendings.


The research design of the Personalized Scrub Caps quality improvement project was a three-part design. All information gathered was subjective, anonymous, and collected via online surveys. The Stanford IRB considered this project to be part of quality improvement and non-human subject research with no written consent required, though all participation was voluntary.

1) Small Group Resident Pilot Project; six representative resident volunteers over 14 day work period. 2) Departmental Implementation; eight week trial period utilizing personalized embroidered scrub caps within Stanford general OR locations.
3) Follow-up and Feedback Collection; survey from those electively wearing personalized embroidered scrub caps.

- Small Group Resident Pilot Project: Assess effectiveness of a simple, visible cap name sticker on increasing appropriate use of anesthesia resident names and the impact on resident wellness, specifically feelings of depersonalization in perioperative settings.
First week: resident volunteers completed a pre-survey, no cap name stickers worn.
Second week: cap name stickers worn, completed post-survey.
Pre-survey responses indicated that appropriate name use during perioperative communication was an area that necessitated significant improvement. The post-survey responses demonstrated a positive response to name visibility on scrub caps.
- Departmental Implementation: Personalized embroidered scrub caps were provided at no cost to all residents, fellows, and attendings within the general OR pool. Name and role, cap color (solid), embroidery color (contrasting color), and cap style were included as customizable options. Four caps were provided at no cost to each individual. Total of 80 anesthesia professionals requested caps.
- Follow-Up and Feedback Collection: After eight weeks of voluntary personalized scrub cap use, survey data was collected with a total of 17 anonymous responses collected including attendings, fellows, and residents.


Small Group Resident Pilot Project: Given the positive response from the representative resident cohort, survey data was presented to department leaders who then proceeded with the approval process for funding of personalized embroidered scrub caps.
Post-Implementation Follow-Up and Feedback: Overwhelmingly positive feedback from voluntary users of the personalized scrub caps. Improvements of 52% to 88% were seen across nine separate areas of perioperative/intraoperative communication and workplace-related wellness. Nearly 70% of anesthesia personnel requested personalized scrub caps when offered. Voluntary follow-up survey response rate of 22% may decrease the representative accuracy of survey findings.


The simple implementation of name and role displayed on surgical scrub caps can improve multiple areas of perioperative communication and professional wellness for anesthesia attendings and trainees. Further areas of data collection could potentially include feedback collection from patients, study of the effect on communication and efficiency in simulated critical OR crises scenarios, and mixed-methods studies including interprofessional colleagues i.e. surgeons, OR nursing, technicians, and other staff.

[1] Lingard L, Espin S, Whyte S, Regehr G, Baker GR, Reznick R, et al. Communication failures in the operating room: an observational classification of recurrent types and effects. BMJ Qual Saf. 2004 Oct 1;13(5):330“4.
[2] Brodzinsky L, Crowe S, Lee HC, Goldhaber-Fiebert SN, Sie L, Padua KL, Daniels K. What's in a Name? Enhancing Communication in the Operating Room with the Use of Names and Roles on Surgical Caps. The Joint Commission Journal on Quality and Patient Safety. 2021 Apr 1;47(4):258-64.
Postpartum Chemical Venothromboembolism Prophylaxis in Patients Following Cesarean Section: An Evaluation of Adherence to Guidelines and Adverse Outcomes

Abbie Hageman, MD
Sanaa Suharwardy, MD
Yasser Yehia El-Sayed, MD
Deirdre Judith Lyell, MD
Alex James Butwick, MD
Gillian Abir, MD
Yolanda Winters, RN, MSN
Obstetrics & Gynecology Department and Anesthesia Department Background
While VTE (venothromboembolism) is rare in pregnancy, it contributes significantly to maternal mortality (9.6%).1-5 Guidelines such as the California Maternal Quality Care Collaborative (CMQCC) chemical VTE prophylaxis(ppx) guidelines (published February 2018, see Table 1) are intended to help decrease preventable maternal deaths from VTE.6 A recent large, retrospective study has raised concern for increased adverse outcomes associated with chemical VTE prophylaxis such as wound hematomas (aOR 2.34), unplanned surgical procedures (aOR 1.29) , and blood transfusions (aOR 1.34.).7 The rate of adherence to the CMQCC VTE ppx guidelines is unknown among patients delivering via Cesarean section at Lucile Packard Children™s Hospital. We anticipated that implementation of clinical decision-making tools such as memory aids, order sets and alerts in the electronic health care system could help improve adherence to guidelines. It™s uncertain if adverse outcomes reported in recent studies of patients receiving chemical VTE ppx are clinically replicated in our patient population.

1) To evaluate and increased adherence to postpartum chemical VTE prophylaxis per CMQCC guidelines comparing data from 2018 to data from 2021.
2) To evaluate rates of adverse outcomes including wound complication, blood transfusions postpartum and unplanned surgical procedures in patients who received postpartum chemical VTE prophylaxis compared with those that did not receive chemical VTE prophylaxis.

Retrospective data was obtained from an electronic health record report of patients meeting CMQCC criteria. We analyzed data from after CMQCC publication 2018 and 2021 after implementation of a memory aid with the CMQCC criteria in the Obstetric provider handoff. Statistics were run using Microsoft Excel using unpaired student t-test for continuous variables and X2 test for categorical variables.

In 2018, 24 (7.7%) of the 310 patients meeting at least one CMQCC criteria were on chemical VTE ppx. In 2021, 168 (42.6%) of the 394 patients meeting at least one CMQCC criteria were on chemical VTE ppx. Baseline demographics were similar except for delivery QBL/EBL which did significantly differ (mean 1277cc vs 1051cc )in patients who received and did not receive chemical vte ppx, respectively. There was no significant difference in adverse outcomes including wound complications (3.9% vs 4.2%), postoperative blood transfusion (mean units PRBCs transfused 2.2 vs 2) and unplanned surgical procedures (0.5% vs 0.2%) among patients who received or did not receive chemical VTE ppx, respectively. In the 2018 cohort, the most common CMQCC criteria met in patients on chemical VTE ppx was a pre-existing thrombophilia whereas the most common criteria met in the 2021 cohort was BMI.

In the 2021 cohort, the chemical VTE ppx regimens used were 153 (91%) Lovenox 40mg q24hours, 1 (0.6%) Lovenox 40mg q12hours, 6 (3.6%) heparin 5000units TID, 3 (1.8%) heparin 5000units BID, 5 (3%) heparin 5000units BID or TID transitioned to Lovenox 40mg q24hours. Chemical VTE ppx was started on average on postop day 1.56 (0-5). Among the reasons identified for no chemical VTE ppx ordered in patients meeting CMQCC criteria, the most common reasons were: patient inaccurately assessed as not meeting criteria followed by patient being correctly assessed as meeting criteria but not ordered for chemical VTE ppx. Other less common reasons included chemical VTE ppx being ordered but not given, patient refusing, order being canceled, patient discharging prior to administration and once due to medication not available .

Implementation of a provider handoff memory aid significantly improved adherence to CMQCC guidelines from 7.7% to 42.6%. Our primary outcomes including wound complications, blood transfusion and unplanned surgical procedures did not significantly differ between patients meeting CMQCC criteria who did or did not receive chemical VTE prophylaxis.

The conclusions we can draw from this study are limited by the retrospective nature of the data collection as well as sample size that may be under powered for the adverse outcomes of interest.

1. James, A.H., Prevention and management of venous thromboembolism in pregnancy. Am J Med, 2007. 120(10 Suppl 2): p. S26-34.
3. Creanga, A.A., et al., Pregnancy-related mortality in the United States, 2006-2010. Obstet Gynecol, 2015. 125(1): p. 5-12.
4. Source: The California Pregnancy-Associated Mortality Review. Report from 2002-2007 Maternal Death Reviews. Sacramento: California Department of Public Health, Maternal Child and Adolescent Health Division. 2017
5. Hameed AB, Montgomery D, Peterson N, Morton CH, and A Friedman. Improving Health Care Response to Maternal Venous Thromboembolism. Developed under contract #11-10006 with the California Department of Public Health, Maternal, Child and Adolescent Health Division. Published by the California Department of Public Health, 2017.
7. Lu, Michelle Y., et al. "Evaluation of a risk-stratified, heparin-based, obstetric thromboprophylaxis protocol." Obstetrics & Gynecology (2021): 10-1097
Quantifying length of stay barriers to inform an escalation pathway

Melissa Franco
Kavitha Subramanian,
Janet Wei
Saloni Ajay Shah
Adam Lida Zhao
Omar Shalakhi
Joshua Sae-Kim
Kevin Jaehoon Jung
Shriya Gampala Reddy
Lisa Shieh 
Medicine Department Introduction: Patient long length of stays (LOS) is a large issue facing many hospitals. Therefore, hospitals have been searching for solutions to reduce LOS, improve their care management and operations, and maximize their costs efficiency. At Stanford Health Care (SHC) the LOS index performance has been trending up since April 2021.

Methods: The Stanford Care Coordination (SCC) team is a multidisciplinary team that comprises of healthcare consultants, physicians, residents, nurses, and quality improvement (QI) stakeholders with a mission to identify opportunities to decrease LOS. The LOS QI project had three objectives: (1) quantify LOS barriers at SHC, (2) conduct a stakeholder analysis to understand an LOS escalation pathway at a satellite community hospital, and (3) develop an escalation pathway tailored to SHC. To quantify the most frequent barriers to LOS patients, electronic health record (EHR) data representing one-week inpatient stays (March 1 “ 7, 2022) were extracted. A resident manually code EHR care coordination notes to 22 barriers from a pre-specified list. A descriptive analysis in excel was conducted to identify the top six barriers. The SCC team then conducted two one-hour interviews with key leadership at the satellite community hospital to understand their LOS escalation pathway and common barriers to discharge patients. Results from these two objectives were then used to draft a LOS escalation pathway specific to SHC needs and institutional constraints.

Results: The top six LOS barriers include: on-going clinical care, delayed procedures, difficulty finding placement (e.g., rejection from multiple skilled nursing facilities), no bed availability, patient/family disposition concerns, and pending insurance authorizations. Lessons learned from our stakeholder interview included the need to have a robust communication system among the care team, automatic tracking tools, and a dedicated complex LLOS team for in-patient stays over 30 days. One major strength highlighted from our stakeholder was the strength of their local community partnerships to accept their patients. This was key to the success of their escalation pathway. These results were incorporated into the SHC escalation pathway. The SHC pathway will triage patients based on their identified barrier and will have a dedicated team lead to progress to discharge. To automate the process, health technology experts are building the Epic infrastructure to collect barrier data and triage patients to appropriate management team. Additionally, the team is also creating a SHC complex LLOS team and working on strengthening community partnerships.

Conclusion: We used a data-drive approach to inform the SCH escalation pathway. Our descriptive analysis and stakeholder interview was able to identify key opportunities to address LOS. The SCC team is gradually working with hospital leadership and stakeholders to develop EHR tools to facilitate key processes, create a complex LLOS team, and strengthen community partnerships to further develop the LOS SHC escalation pathway.
Reducing readmissions from skilled nursing facilities through improved communication

Julia Pangalangan
Arman Sharma
- Reducing Readmissions from Skilled Nursing Facilities through Improved Communication

Post-acute care (PAC) facilities, skilled nursing facilities (SNFs) in particular, serve as an intermediate between the hospital stay and home. PAC facilities can help hospitals to decrease length of stay, but approximately 18% of patients discharged to a PAC are readmitted within 30 days.
(MedPAC, 2021, McKesson)

Readmissions occur for a multitude of reasons, but ineffective or insufficient communication among administrators and care providers between hospitals and post-acute care facilities has been cited as a key issue. Ineffective communication leads to increased healthcare costs and financial risk, increased readmission rates, and hinders the experience of both the patients and providers.

Efforts should be made to reduce readmissions and improve the discharge process to support Stanford Healthcare™s mission of prioritizing patient outcomes, reducing costs, and enhancing provider and patient experience.
Ineffective communication between hospital systems and post-acute care facilities is a leading modifiable reason for hospital readmissions. This project aims to understand the communication barriers and facilitators between Stanford Healthcare and skilled-nursing facilities. The work aims to define design criteria for an effective communication pipeline to ultimately reduce Stanford Healthcare™s readmission rate from 11.8% to 11% by May 2023.

Methods: We began with a literature review to better understand the landscape. We then conducted interviews with key stakeholders including hospital administrators, hospital clinicians, SNF administrators, and SNF clinicians. A thematic analysis with inductive coding was used to analyze the qualitative data. We identified crucial key drivers in the current communication pipeline and developed potential interventions.

Results: After interviewing 10 stakeholders, we identified the following recurring themes: lack of patient education, insufficient provider-to-provider interaction, and continuity of care. Across the care process, improvements in provider and patient sense of accountability could help to resolve communication obstacles. Potential interventions to aid the organizational design of the process could include developing and disseminating patient education material, implementing a warm provider handoff, or having nurses review the after-visit summary. The aim is to optimize communication in order to reduce healthcare costs and hospital readmissions, improve patient outcomes and satisfaction, and enhance provider experience.

Conclusion: By defining the current state and meeting with relevant stakeholders, we were able to identify critical gaps in the communication pipeline between Stanford Hospital and SNFs. Accountability emerged as a key theme. The data provides critical insights that can be leveraged to develop concrete changes in the communication process. Communication and education interventions at discharge can lead to fewer hospital readmissions and increased patient satisfaction (Becker 2021). Potential interventions may include disseminating patient education material, implementing communication nudges, and encouraging a warm provider-to-provider hand-off. Future efforts in patient communication and organizational design will help to establish continuity of care for vulnerable patients and reduce costly readmissions at Stanford Healthcare. 
Reimagining the Future of the MA Role at an FQHC

Aditya Narayan
Jay Shah
Joel Bateman
Medicine Department Background: West County Health Clinics (an FQHC in Sonoma County serving primarily MediCal patients) aims to implement a radical new model of care proposed in partnership with Maximon Behavioral Health called CARE+. The program seeks to better cater to the complex behavioral health needs of the MediCal patient population by offering wraparound services in which patients with several comorbidities and behavioral health needs will be connected with multidisciplinary care teams to help address their varied clinical and social determinants of health.
Objective: 1) To understand drivers and barriers to implementation using an organizational behavior lens by analyzing the role of medical assistants (MAs), who would serve as the point person for this new program.
2) To create a mechanism by which to measure the efficacy of such a program.
Methods: We conducted 6 focused interviews with MAs and MA leads at throughout various sites of the organization to garner feedback on program implementation.
Results: We found that the primary barriers to implementing this program programs were time, inconsistent levels of training across involved providers, and challenges communicating care goals and needs across clinics/departments. Using this information, a score card was co-designed to evaluate future organizational changes in the clinic system which may be applied to analogous changes at other healthcare organizations.
Conclusions: Identifying barriers and potential obstacles is essential before the implementation of a new program. We helped identify obstacles that will need to be addressed and developed a scorecard to help evaluate future progress.
Screening for health-harming legal needs in a primary care setting

Max McClure
Diana Cervantes
Michelle deBlank
Sapna Dalal
Rakhi Singh
Medicine Department Background: Fair Oaks Health Center (FOHC) is a San Mateo County Health primary care clinic in the North Fair Oaks unincorporated area of California. Predicted legal needs in this population include those related to the workplace (wage theft, discrimination, harassment), housing (evictions, unsafe housing), and immigration, all of which may have significant effects on patients™ health and ability to attend primary care appointments. San Mateo County Health clinics already have a mechanism in place to provide pro bono legal referrals to patients who present with legal issues, but it is unclear whether all patients who would potentially benefit from preventive lawyering are provided these referrals.
The initial phase of this project is focused on assessing the performance of a proposed legal needs screening tool developed in coordination with members of San Mateo Legal Aid and FOHC social work.

Methods: A subset of patients seen by residents in FOHC primary care clinic from August 18, 2021 to March 18, 2022 were screened for concerns about rent and bills, immigration status, and their children™s ability to attend school. If a patient screened positive, they were offered a social work referral. Patients who agreed were then assessed during a follow-up visit with FOHC social work and, if appropriate, referred to legal assistance. Screening results and social work referrals were noted in patients™ San Mateo County Health medical charts, and secondary legal referrals were tracked by FOHC social work.

Results: 48 patients were screened, 17 (35%) of whom screened positive (7 for rent/bills, 4 for immigration, and 7 unspecified). 15 (31%) of these agreed to social work referral. 10 (21%) were ultimately deemed appropriate for referral to legal services, yielding a positive predictive value of an initial positive screen for subsequent legal referral of 59%.

Conclusion: The three-question survey is easy to administer and has a moderate positive predictive value for identifying legal needs appropriate for referral. Unfortunately, we are unable to evaluate either the sensitivity of the tool or the ultimate outcomes of each legal referral. The next step will be implementing a version of this screening approach throughout the FOHC residents' clinic, with the goal of assessing patient screening rates and feasibility of provider adherence.
Streamlining Pediatric Stroke Codes in the LPCH Emergency Department

Ria Pal, MD
Katherine Xiong, MD
Rachel Walsh, MD
Sarah Lee, MD
Neurology Department Purpose
Pediatric stroke codes are uncommon, with current incidence of arterial ischemic stroke estimated at 4.6-6.4 per 100,000 children annually, leading to delayed recognition by providers and infrequent opportunities to learn and apply knowledge of protocols.1 Current pediatric guidelines recommend urgent head imaging for stroke-like symptoms to help guide management, with a preference for MRI imaging over CT primarily due to superior sensitivity and specificity in a situation with frequent stroke mimics.2 However, it can be difficult to coordinate emergent neuroimaging, particularly in pediatric patients, where sedation is often required.
The pediatric stroke code process at Lucile Packard Children™s Hospital (LPCH) ED is unique as the LPCH ED is located within the affiliated adult hospital (Stanford Health Care - 300P [SHC - 300P]). ED stroke codes therefore require additional coordination between separate hospital systems and are predictably subject to communication breaks, logistic challenges, and perceived unclear division of responsibility. The purpose of this project is directed towards improving the time from presentation to MRI in the LPCH ED setting.
We performed a retrospective assessment of LPCH ED stroke codes from 11/26/2016-9/28/2021. RCA was performed using the 5 Whys methodology and fishbone diagram. Surveys were conducted across neurology, emergency medicine, and anesthesia MD providers regarding subjective sources for imaging delays within the pathway. Staged interventions were directed towards improving provider familiarity of the protocol, improving information and equipment availability (including modifying the pediatric stroke code order set), and creating a largely-automated report-out email to involved parties. These were implemented gradually since January 2022. Ongoing review post-implementation (such as audits of the order set and measurement of time-to-imaging for each stroke code) will be performed to determine the success of the intervention.

Results/Anticipated Results
Analysis of all LPCH ED stroke codes from 11/26/2016-9/28/2021 found that the time from presentation to MRI imaging was < 60 minutes (the current institutional goal time) for only 47% of all stroke codes, with a median time of 1 hour 6 minutes. Following the opening of 500P and subsequent separation of pediatric and adult emergency departments, there was a trend towards longer times to MRI. Of note, data to distinguish between MRIs with and without sedation was unavailable.
Our initial survey of MD providers across neurology, emergency medicine and anesthesia subspecialties identified three key drivers: provider familiarity with the stroke code protocol, protocol information and equipment (imaging, anesthesia) availability, and process/outcome transparency.
Interventions including updating the departmental protocols and increasing provider accessibility to them (through badge buddies and bedside guides at LPCH and LPCH ED), optimizing the stroke code order set, and the generation of post-stroke code reports are currently active. Additionally, interventions including anesthesia equipment accessibility and formal integration of the protocol into ED provider and MRI technician education are underway.
We anticipate that following these multidisciplinary interventions, overall institutional familiarity of the protocol will increase and time from ED presentation to MRI imaging will decrease, with a larger percentage of stroke codes achieving the goal code-to-first image time of ‰¤60 minutes for MRIs without sedation and ‰¤90 minutes for MRIs with sedation by July 2023.

Since partial implementation of the interventions above, there have been two LPCH ED stroke codes requiring MRI (both without sedation). Times to first image were 10 and 30 minutes respectively.
Current time from ED presentation with stroke-like symptoms to MRI for head imaging is greater than the recommended 60 minutes for most stroke codes at LPCH ED, likely due to providers™ limited familiarity with the pathway, logistic challenges to obtaining appropriate workup and management, and complex care coordination between two hospital systems. Efforts to increase provider education on the ED stroke code protocol are ongoing, as are efforts to improve inefficiencies within the protocol.

1. Khalaf A, Michael I, Fullerton H, Wintermark M. Pediatric Stroke Imaging. Pediatr Neurol. 2018;86:5-18. doi:10.1016/j.pediatrneurol.2018.05.008
2. Mirsky DM, Beslow LA, Amlie-Lefond C, et al. Pathways for Neuroimaging of Childhood Stroke. Pediatr Neurol. 2017;69:11-23. doi:10.1016/j.pediatrneurol.2016.12.004
The halo effect : Room impacts patient perception of overall hospital experience

Ella Atsavapranee
Paul Heidenreich, MD, MS
Lisa Shieh, MD, PhD
Medicine Department Background: To improve quality of care, it is essential to first assess factors that influence patient experience. This study investigated how room type influences different aspects of patient experience at Stanford Health Care (SHC).

Methods: Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey data from medicine patients across both SHC hospitals from January 2018 to May 2021 were analyzed (n=1259). Top box scores were calculated for 18 survey sections to represent the percentage of patients who gave the most favorable rating. A chi-squared test was used to identify differences in top box scores between patients in semi-private and private rooms. Differences between patients in private rooms in the new and original hospitals were also analyzed. A multivariate regression was performed to estimate the effects of room type and patient characteristics on overall top box scores. The association of a top box score as an odds ratio (OR) and adjusted odds ratio (aOR) with 95% confidence intervals (CIs) was calculated. A significance level of 0.05 was used for all statistical tests.

Results: Patients in private rooms rated the overall assessment and 10 survey sections (room, visitors/family, hospital environment, human caring, admission, personal issues, discharge, tests/treatments, physician, and nurses) significantly higher than patients in semi-private rooms. Patients in private rooms in the new hospital rated 7 sections (room, hospital environment, visitors/family, human caring, tests/treatments, personal issues, and nurses) significantly higher than patients in private rooms in the original hospital. Adjusting for patient characteristics (gender, marital status, religion, language, ethnic group, and insurance type), staying in a private room was associated with an overall top box score (aOR=1.28, 95% CI [1.24-1.33]). In addition to room type, patient characteristics were associated with likelihood to give a top box score.

Conclusions: The greatest top box score differences between patients in semi-private and private rooms were found for room, hospital environment, and visitors/family, which was expected given that these aspects are directly related to room accommodation. However, private rooms were also associated with significantly higher ratings for other areas (admission, discharge, tests/treatments, physician, and nurses). Similarly, private rooms in the new hospital scored higher than those in the original hospital on sections that were unrelated to room type (tests/treatments and nurses). This suggests a halo effect in which a patient™s positive impression of their room influences other aspects of their hospital experience. Thus, improving a patient™s room can lead to widespread increases in satisfaction.
The Stanford Airway Review Committee: Improving Emergency Airway Management Through Video Review

Jonathan Hootman MD
Wil Gibb MD
Carl Preiksaitis MD
Jason Teng MD
on behalf of the Stanford Airway Review Committee
Emergency Medicine Department  BACKGROUND:
Endotracheal intubation is one of the most critical procedural skills in emergency medicine. Patients have limited physiologic reserve, placing them at increased risk of severe hypoxemia, hemodynamic collapse, cardiac arrest, and death, especially with increasing number of intubation attempts. Consequently, first-pass success of intubation is of paramount importance, and it is often used as a benchmark of skill proficiency in emergency medicine, anesthesia, and critical care. However, the first-pass success rate of intubations performed in the Stanford Emergency Department is currently unknown.

Targeted interventions such as high-fidelity simulation curricula, implementation of a standardized airway bundle, creation of a continuous quality improvement database, and ongoing and frequent airway education have been shown to improve first-attempt success rates. Video-assisted debriefing and review has been shown to enhance learning as well as identify procedural errors not routinely identified through traditional assessments in a number of critical procedures. To our knowledge this has not been used to enhance resident education for intubation in the emergency department.

Consequently, we aim to enhance patient care and safety by measuring and improving first-pass success of intubation; increasing resident clinical skills, confidence, and knowledge of airway management; and reducing the frequency of peri-intubation complications through implementation of a targeted airway review curriculum and quality-improvement initiative based on video review of emergency department intubations. 

This study will be conducted at the Stanford Hospital Adult Emergency Department (ED), which is a Level 1 Adult Trauma Center and quaternary care academic medical center, located in Palo Alto, California. The Stanford ED treats over 80,000 patients per year.

Study Design:
This will be a pre-and post-intervention analysis of emergency department airway management in the months prior to implementation of the airway review curriculum (September 25, 2021 - February 16, 2022) and the months since implementation (February 17, 2021 “ Present).

The Intervention:
The intervention will consist of several key elements:
1. Video recording of all emergency department intubations. Starting September 2021, all emergency medicine residents will be encouraged to record each attempt at video laryngoscopic intubation in the adult emergency department.
2. Enhanced data capture of intubation attempts. After each intubation, operators complete a data collection form within the electronic medical record (EMR) that captures pertinent intubation information.
3. Standardized monthly review of recorded intubations. Each month, we collect recorded intubations and match them with their corresponding forms from the EMR. An airway review committee comprising emergency medicine residents and faculty reviews each video, form, and medical record using a validated and standardized checklist.
4. Provision of individualized feedback to operators. After review of each intubation, the airway review committee provides individualized feedback to each operator.
5. Monthly airway reviews. The airway review committee presents to emergency medicine residents and faculty once a month for approximately 10 minutes during regularly scheduled weekly didactics. These sessions focus on reviewing cases from the month which were associated with adverse events, cases requiring more than one attempt at laryngoscopy, cases demonstrating challenging airway anatomy or patient physiology, or cases show-casing masterful procedural technique.

The primary outcome will be first-pass success of intubation. The secondary outcomes will be resident clinical skills, confidence, and knowledge of airway management as measured by periodic standardized knowledge and skills assessments; and the incidence of peri-intubation complications, including hypotension (requiring intravenous fluids and/or vasopressors), hypoxemia (pulse oximetry decreased to < 92%), bradycardia (heart rate decreased to less than 60 bpm), esophageal intubation, aspiration (including any witnessed aspiration of gastric contents or secretions during the intubation attempt), airway trauma (including any lacerations, swelling, edema, or dental injury related to the intubation attempt), and cardiac arrest.

Before September 2021, the first-pass success rate of intubations in the Stanford Emergency Department is unknown. As a result of the targeted intervention, the first-pass success rate of intubations from September 25, 2021 - April 16, 2022 is now known and is approximately 89.3%. Qualitative feedback from residents and faculty shows overwhelming support for the initiative, with common themes supporting the added educational benefit of video review.

Implementation of an airway review curriculum and quality-improvement initiative based on video review appears to allow accurate and reliable capture of first-pass rates of intubation in the emergency department and is associated with increased resident satisfaction with airway management education. Continued work and data collection is needed to elucidate the impact of this intervention on first-pass success rates and adverse events.
Ultrasound-guided Lumbar Puncture: Trainee Education and Clinical Implementation

David Rogawski
Connie Wu
Kenneth Leung
Sarada Sakamuri
Neurology Department Point-of-care ultrasound is being increasingly used to guide bedside procedures. Ultrasound of the lumbar spine is helpful for revealing the anatomy and guiding site selection prior to lumbar puncture (LP). When landmarks alone are used for site selection, the lumbar puncture failure rate is 19%, increasing to 42% in obese patients. Failure of bedside LPs increases patients™ frustration, delays treatment or diagnosis, and increases the financial burden as patients are typically referred to radiologists for fluoroscopy-guided LP. However, ultrasound has been shown in randomized trials to improve LP success rates to >90% even in obese patients, while reducing the number of attempts and the number of traumatic taps. Despite the extensive evidence base supporting the benefits of ultrasound, most lumbar punctures at Stanford are currently performed without ultrasound. We developed and conducted an ultrasound-guided LP training session for neurology trainees on 4/13/22. As judged by pre- and post-training surveys, the workshop improved trainees™ knowledge of ultrasound and increased confidence in performing ultrasound-guided LP. Trainees identified poor availability of ultrasound and time scarcity as major hurdles to overcome in implementing ultrasound-guided LP in regular practice. Next steps are to provide two Philips Lumify ultrasounds dedicated for neurology use in the 500 Pasteur hospital and the Hoover neurology clinic. By August 1, 2022 we aim to have 60% of neurology residents complete an ultrasound-guided LP in the hospital or clinic and complete a post-procedure survey on their experience. This project will serve as a starting point for making ultrasound-guided LP the standard practice in our neurology department.
Voalt: Improving Ease and Fidelity of Identifying the Correct Provider

Mugdha Joshi, MD
Arjun C. Gokhale, MD
Kelly Darmawan, MD
Anoosha Moturu, MD
Anna Pendrey, MD
Nicholas U. Schwartz, MD
Lauren Wozniak, MD
Garrison Carlos, MD
Gretchen Brown RN, ACNIO
Austin Wilson, MBA
Clifford Schmiesing, MD
Medicine Department Introduction: The communication landscape at Stanford Healthcare, a 605 bed quaternary care academic hospital is complex and multimodal. Four years ago, the HIPAA secured patient messaging application Voalt was introduced to a system that already had a hybrid communication ecosystem, including personal mobile phone use and a legacy SMS paging system. One consequence of this complicated system has been the development of nonstandard provider assignment and communication workflows that evolved differently and independently for every service in the hospital. In addition, there has been an institutional push to move most clinical communication to Voalt mobile, for which provider assignments currently are sourced through assignments made through the electronic medical record Epic. Given the variability in workflows and complexity of the communication landscape, there are frequent issues with identifying the correct provider for patient care concerns due to redundant, nonexistent, and inaccurate provider assignments. Consequences include delays in patient care, user fatigue, and frequent interruptions due to challenges with identification and being incorrectly contacted while off shift or following patient care transitions.
Objective: The interventions sought to 1) elucidate the causes of incorrect provider identification, 2) pilot a standardized workflow to increase the rate and accuracy of provider assignments, and 3) facilitate correct provider identification for nurses and support staff.
Methodology: A multi-disciplinary working group consisting of house-staff from different service lines, project managers, nurse managers and informatics leadership was assembled. Key drivers were identified using comprehensive value stream mapping for multiple service lines detailing provider assignment transitions from patient admission to discharge to identify pain points. A fishbone diagram was then created. A standardized workflow was developed and piloted on a sample unit. Pre and post pilot surveys were administered to nursing staff and participating housestaff on the unit. An analytic report using a clinical data warehouse, Clarity, was developed to allow for pre and post pilot analysis of correct provider assignment.
Results: Root cause analysis highlighted the following key drivers: 1) The first contact provider needs to be clearly identifiable when correctly assigned 2) Provider transition workflows during shift change needs to be standard and reliable 3) Workflow for provider assignments during patient movement in the hospital needs to be standard and reliable 4) Voalt needs to be adopted and used by providers across the board. As a result, the task force created the standardized workflow described below to address the first two key drivers. This workflow was successfully piloted on a selected unit and demonstrated improvement in provider assignment rates. Analysis of survey data supports significant need for refinement and implementation of more standardized provider assignment workflow.
Standardized Workflow Description: A new treatment team role was created in Epic called Covering Provider which can only be held by a single provider at any given time. This role would automatically update the Voalte Patient Care Team through an interface. Additionally, a column was created in Epic to display the name of the covering provider for each patient. Housestaff on the pilot unit were instructed to assign themselves to their entire list of patients using this role at the start of every shift. Likewise, nurses were instructed to add this column to their patient lists and contact the provider holding this role for patient care issues.
Conclusion: This work demonstrates the application of structured root cause analysis with an interdisciplinary group of front-line users to develop a standardized workflow to improve fidelity and ease of correct provider identification. We hope that the results of our intervention support the use of this workflow to improve rates of assignment and ease of identification to make care delivery safer and more efficient ultimately. The development of tools and strategies to address the widespread and growing problems and burdens resulting from clinical communication failures is needed. 
Watching the Tape: Improving endotracheal intubation feedback through video review

William Gibb MD
Jason Teng MD
Carl Preiksaitis MD
Jonathan Hootman MD
The Stanford Airway Review Committee
Emergency Medicine Department  While endotracheal intubation is a critical skill in emergency medicine, practicing continuous quality improvement can be challenging. Real time feedback is limited due to physician focus on the intubation itself, and post-hoc feedback is equally limited due to imperfect recollection of the procedure. Thus, the Stanford Airway Review Committee was created in September 2021, composed of Stanford emergency medicine residents and faculty. The goal of this committee was to utilize the video record function on the video laryngoscope devices to significantly improve the ability to receive personalized feedback on intubation. The videos are collected securely, de-identified, and edited by the committee before being distributed for feedback. From September 2021-April 2022, there were 159 total intubations in the Stanford Emergency Department, with 94% being performed by residents. Almost all of the intubations are done with video laryngoscopy. First, second, third, and fourth-year residents performed 8%, 48%, 29%, and 14% of the intubations respectively. Our team also wanted to measure the impact of presenting our preliminary results to the entire department in a grand rounds format on the number of intubations that physicians remembered to record. Before the presentation, 38% (40) of the intubations were recorded, with an improvement to 50% (27) of intubations recorded after the presentation. To our knowledge, this is the first curriculum that aims to improve resident education through review of video intubation footage. This project has the potential to improve how intubation is taught and ultimately improve patient care.
What Matters: Eliciting Overarching Goals from Older Adults Admitted to an Inpatient Service

Ashok Krishnaswami, MD, MAS
Christine E. Gould, PhD
Marina Martin, MD
Matthew Mesias, MD
Medicine Department Background: In older adults, elicitation of robust overarching goals (OG), consistent with geriatric principles/5Ms focus, may improve the operationalization of matching treatment with patient-centered outcomes “ the what matters most.

Methods: As part of a quality improvement project, we interviewed consecutive patients and/or caregivers who received consultative geriatric (acute care of the elderly) services during their admission to a tertiary university hospital. The interviewer provided examples of OGs and asked Would you be able to tell me What Matters Most to You? Further elaboration was asked to clarify OGs with respect to mobility, mind, and medications. OG responses were rated as robust if qualitatively thought to be Specific, Measurable, Achievable, Relevant and Time-limited. Cueing was documented to be present if prompts were needed. We gathered clinical variables to characterize the cohort.

Results: Among 11 admitted patients (mean age: 80.2 ± 9.9 years, 55% women, 73% white, 36% with limited English proficiency), 82% accepted the interview. There were equal proportions (36%) of interviews with patients and caregivers; 1 interview was with both. The primary admitting diagnosis was a fall in 45% of participants. The majority (73%) did not have advance care planning (73%) or physician ordered life-sustaining therapy form filled (55%) with 73% documented to be a full code. A general theme of OGs elicited were about spending time with family and/or the quality of life/enjoyment of the activity. However, only 27% of OGs elicited could be considered robust. The majority (89%) needed cueing to improve OG specificity. Qualitatively, more robust answers were obtained in regards to mobility than cognitive function and medication/polypharmacy. Obtaining an OG for medications was the most difficult and non-robust.

Conclusions: We found that initiation of discussions of OGs during routine clinical care is possible. However, OG elicitation was often non-robust and needed cueing. Future iterations of OG elicitation will focus on being specific, measurable, achievable, relevant and time-limited. This may help to translate the what matters most into operational goals that can be acted on by the healthcare team.
Working Smarter: Development of a Concise-Format Automated Rounding List

Jonathan DeLong, MD
Arjun Gokhale, MD
Edna Shenvi, MD
Monica Dua, MD
Brendan Visser, MD
Surgery Department -