Read the Abstracts

Title & Authors Abstract
Detection of oxygen delivery failure on cardiopulmonary bypass

Leigh Hickerson, MD; Natalia Martinez Acero, MD

Stanford, Department of Anesthesia - Cardiac Anesthesia
Background
Oxygen delivery failure due to vaporizer malfunction on cardiopulmonary bypass (CPB) has been reported twice in the literature. Differentiating this and primary oxygenator failure (incidence 0.04%) must occur quickly to prevent hypoxic neurologic injury. This case series describes 3 incidents of vaporizer malfunction resulting in failure to deliver oxygen (O2) and volatile agent to the patient during CPB at our institution.

Current State
1. To detect oxygen and/or volatile anesthetic delivery failure to CPB circuit as soon as possible to prevent hypoxemia-associated morbidity and awareness under anesthesia, respectively.

Analysis
Analysis: After extensive evaluation and testing of oxygen and volatile delivery mechanisms, all events with delivery of oxygen or volatile failure most likely resulted from a leak between the gas inlet system and the oxygenator.

Specifically, leaks were reproduced when the control dial release lever was activated prior to the vaporizer locking lever. Simultaneous activation of the safety mechanisms also allowed the vaporizer to be in the on position while not properly seated. Proximal and distal flows and isoflurane concentration were measured. Partial and massive leaks were reproduced, explaining the above events. It was not visually apparent that the unit was improperly mounted, as the components were off-set by millimeters.

Intervention: A cannula Luer lock adaptor was spliced between the oxygen tubing proximal to the micro-bacterial filter and oxygenator with a sample line to the ventilator gas analyzer. This allowed for real-time monitoring of oxygen and volatile agent with adjustments in isoflurane levels detected within 15 seconds.

Results
1. Education of proper installment of vaporizers on cardiopulmonary bypass machines by perfusion.
2. Advocating for gas inlet gas analysis as part of standard monitoring during cardiopulmonary bypass. 
Training for Success: Introducing Team-Based Resuscitation Training and Quality Assurance for Cardiac Arrest in the Emergency Department

Haley Manella, MD; Keiran Warner, Stefanie Syer, Kenton Anderson, William Mulkerin, Marc Gautreau, Katherine Staats

Stanford, Department of Emergency Medicine
Background
Several studies have looked at rates of survival after out-of-hospital cardiac arrest (OHCA) but few studies have looked at survival rates of in-hospital cardiac arrest (IHCA) and neurologic outcomes. Most studies on OHCA show a 10% rate of survival to hospital discharge compared to a rate of about 20% for IHCA . Patients experiencing ventricular fibrillation (VF) arrest have proportionately higher rates of survival reaching roughly 30% overall for IHCA. The Seattle based Resuscitation Academy utilizes a team centric approach to provide High Performance CPR (HPCPR) training to achieve greater than 60% survival for out-of-hospital VF cardiac arrest. We set out to evaluate the effect of implementing a HPCPR training program on the performance of program participants and on the rates of survival following cardiac arrest of patients in the Stanford Emergency Department (ED).

Current State
Of patients presenting to the Stanford ED who experienced OHCA during 2017-2018, 46% survived to hospital admission and 14.8% survived to hospital discharge. We aim to increase the survival of neurologically intact patients. Our goal is to achieve an improvement in survival from cardiac arrest to hospital admission to a rate of 60% and in survival to hospital discharge to a rate of 30% by September 2020 after implementation of a HPCPR quality assurance program.

Analysis
We began our implementation program by educating ED technicians, nurses, and residents using an adapted four-hour version of the Resuscitation Academy. The training included didactics on the benchmarks of HPCPR, and an intensive practical session during which real-time compression quality feedback was employed. Participants completed written tests on HPCPR principles before and after training, and individual pre-course and post-course compression quality data was collected. A quality assurance program was begun where team members receive real-time feedback on performance during CPR. In the upcoming months, full code review summaries will be provided to resuscitation team members.

Results
Written pre-course and post-course scores of trainees showed an improvement from 54% to 84%. Individual performance evaluations show that 78% of participants demonstrated improvement in HPCPR benchmarks after training. We are awaiting data on survival to hospital admission and discharge for cardiac arrests occurring after the implementation of our HPCPR program.
Training ED team members in HPCPR principles improves individual compression quality, knowledge of HPCPR principles, and team dynamics during codes. We are pursuing a larger scale utilization of this training within the Stanford ED and establishing a QA process for all cardiac arrest codes in our department. We will test knowledge and skill retention of our program™s participants at one year to evaluate the long-standing educational impact of our training. We will also compare the rates of survival in cardiac arrest before and after HPCPR implementation to demonstrate the clinical impact of our QA program.
Creating a 21st Century Dental Box

Keir Warner, MD; Crystal Chang, DDS Alexei Wagner, MD Jean Reyes, RN Laleh Gharahbaghian, MD

Stanford, Department of Emergency Medicine
Background
Historically dental emergencies treated in the ED usually require treatment for infections or trauma to teeth, usually at off hours.  The dental equipment dental box the is kept within the Emergency Department has been poorly stocked and recently identified to not meet current infection control guidelines as it contains multiple dose vials and reusable equipment rather than single dose/use equipment and contained equipment/materials past their expiration date. 

Current State
The state at the start of the project was a poorly stocked dental box not meeting infection control guidelines.  Our SMART Goal for the endpoint was to build a dental box with all single use equipment in collaboration with hospital dentistry to have the complete equipment to diagnose and treat the most common dental emergencies based off of data from one year™s prior data as recorded in EPIC.  In addition to achieve sustainability nursing leadership will manage ordering and evaluation for expiration dates.

Analysis
Over the course of a month our multidisciplinary team created a new dental box with all single use equipment and identified an ideal par stocking level.  There is a system for checking expiration dates put in place through nursing leadership and plans are made to duplicate this new dental box and stocking system for our Fast Track and Walk in Clinic areas.

Results
Over a month our multidisciplinary team created a new dental box with all single use equipment and identified an ideal par stocking level.  There is a system for checking expiration dates put in place through nursing leadership and plans are made to duplicate this new dental box and stocking system for our Fast Track and Walk in Clinic areas. 
Addressing Social Disparities in ED Hallway Bed Assignment

David A. Kim, MD; Ian P. Brown

Stanford, Department of Emergency Medicine
Background
Hallway beds in the ED produce lower patient satisfaction and inferior care. The determinants of patients™ assignment to hallway beds are not known.

Current State
We sought to determine whether socioeconomic factors influence which patients are assigned to hallway beds, independent of clinical characteristics at triage. We sought to develop a triage framework to reduce any impact of social factors on bed assignment.

Analysis
We studied 301,654 adult visits to two academic EDs from 2013-2016. In a series of logistic regression models, we estimated the effects of patients™ insurance status and race on the likelihood of being assigned to a hallway bed, controlling for patient demographics, time and day of arrival, and illness severity at triage. We estimated Cox proportional hazards models of hallway bed assignment on length-of-stay.

Results
Median patient age was 56. 54.4% of visits were by women. 56.6% of visits were paid primarily by private payors, 24.3% by Medicare, and 19.1% by Medicaid. 16.1% of patients were assigned to hallway beds. In logistic regression models controlling for age, sex, race, time and day, and triage acuity, Medicaid status was associated with 35% greater odds of assignment to a hallway bed (OR 1.35, [95% CI, 1.32-1.38]), compared to private insurance. Black patients were more likely than white patients to be assigned to hallway beds (OR 1.04 [95% CI, 1.02-1.07]). Race did not account for the effect of Medicaid status on hallway bed assignment, and exhibited complex interactions with insurance status. Patients assigned to hallway beds had longer lengths-of-stay than roomed patients of comparable acuity (hazard ratio for discharge 0.87 [95% CI, 0.86-0.87]). In ongoing work we are using simulation to develop triage algorithms that optimize for clinical appropriateness while minimizing potential social disparities. 
Improving Trauma Charting Accuracy by Leveraging the Electronic Medical Record

Keir Warner, MD; Ian Brown, MD Laleh Gharahbaghian, MD

Stanford, Department of Emergency Medicine 
Background
The chaotic environment of the trauma bay is managed by strict adherence to the principles of ATLS to evaluate and resuscitate in a systematic fashion to avoid missed injuries and streamline emergent care.  This is especially important when multiple trauma patients present at the same time or within minutes of each other™s arrival to the ED.  Frequently hours later the provider is left to sit down and recall their exam on multiple patients to update their documentation for that patient™s chart.  As part of ACS verification of trauma centers the Primary and Secondary survey are scribed verbatim in real time using the Trauma Narrator in the EPIC EMR.  Further, there are three different places the trauma Primary/Secondary survey results are charted: The ED provider note, Trauma H&P and the Trauma Narrator.  These three different notes are frequently discordant, and the most accurate chart is usually the narrator note that was entered in real time.  This presents a patient safety issue for missed injuries and a medical/legal issue having three different documentations of the same exam.

Current State
At the start of the project the three different charts were filled out by the ED, Trauma and trauma narrator.   Our goal with this project is to leverage the EMR to take advantage of SmartLinks to use the Trauma Narrator documentation as the base templet for the systematic recording of injuries identified during the survey.  To ensure an accurate accounting of the injuries at the end of the secondary survey the resident performing the survey will summarize the head to toe findings with the Trauma Chief,  ED Attending and Recording Nurse to ensure all injuries are recorded correctly and ensure there is a plan for each injury.

Analysis
Working closely with Trauma Nursing, Emergency Medicine and the EPIC informatics team to create SmartLinks in the new ED Trauma Provider Note to document the findings in a systematic ABCDE/Head to Toe fashion for consistent and comprehensive documentation.   This new note templet was trialed in the ED and feedback collected prior to full implementation.

Results
Over the course of a year the ED Trauma Note (.EDNOTETRAUMA) was built and modified based off of feedback collected in the preliminary phase.  The ED Trauma Note was presented to the EDMC and approved for the base templet documentation with additional findings added by the EM Resident once they are able to sit down to chart.  Specifically, the note is most helpful in situations where multiple sick patients present simultaneously and details of the exam are frequently not remembered.  
Measuring the Impact of the Three Good Things in Life™ Positive Psychology Intervention on Inpatient positivity and outcomes

Nicholas Zehner, MD; Laura Pudding, Valeria Sindra

Stanford, Department of Medicine
Background
Since the development of positive psychology by Martin Seligman and others in the late 1990s and early 2000s, there has been growing interest in a variety of fields as to the potential for positive psychology interventions to increase happiness, subjective well-being, and resilience in the face of challenging circumstances. Bryan Sexton, a positive psychology researcher at Duke, has studied the impact of the Three Good Things in Life positive psychology protocol on healthcare providers and burnout.

Current State
While Sexton focused his research on healthcare providers, this research project will focus on the impact and efficacy of the Three Good Things in Life protocol within inpatient populations, populations that regularly report lower levels of happiness and subjective well-being than the general population. The goal is to provide a low-cost intervention that improves reports of subjective well-being and wellness.

Analysis
The Three Good Things in Life protocol encourages participants to identify and record in a journal three things that went well that day right before they go to bed and to reflect on the role they played in making those good things happen. By focusing the participant™s attention on what went well that day and their agency in bring it about, participants reported higher levels of subject well-being and increased capacity to respond to life difficulties. 

Results
The data generated from this project is currently being analyzed.
Efficient Left Atrial Appendage Evaluation Using MicroTEE Transducer Yields High Patient Satisfaction

Cindy Wang, MD; Alberta Yen MD, David Liang MD PhD

Stanford, Department of Medicine - Cardiology
Background
Transesophageal echocardiography (TEE) is currently the modality of choice for evaluation of the left atrial appendage (LAA) for thrombus. These procedures are commonly performed under moderate sedation prior to elective procedures such as atrial fibrillation ablation. However, moderate sedation poses the risk of circulatory and respiratory complications and requires a monitoring period in the echocardiography lab after the procedure, limiting the number of outpatient TEE exams and utilizing nursing resources. A smaller transducer with comparable image quality could allow for examinations without moderate sedation, less patient discomfort, and quicker procedure times to create a more efficient lab.

Current State
A smaller TEE transducer (S8-3t MicroTEE transducer, Philips, Andover, MA) was used for focused evaluation of the LAA in eligible patients undergoing elective atrial fibrillation ablation or LAA occlusion (Watchman) procedures.  Examinations were assessed for procedural success, image quality, patient reported discomfort, and duration of study, measured as time from patient consent to discharge from the lab.  To determine feasibility and efficiency of a no-sedation protocol, a subset of 32 patients was consented to undergo TEE evaluation with only topical application of 2% viscous lidocaine to the oropharynx. These patients were also assessed for procedural discomfort (scale 0-10) and desire to undergo a no-sedation MicroTEE procedure in the future.

Analysis
45 patients (mean age 66.6 years, range 43-89) had a LAA evaluation using the MicroTEE transducer. 95.5% of examinations were successful, resulting in interpretable and sufficient image quality to determine the presence or absence of thrombus. Patients rated an average discomfort level of 2.3 out of 10. Two patients were unable to complete an examination with the MicroTEE due to excessive gag reflex and discomfort in the no-sedation protocol. Average study time was significantly reduced in patients not receiving sedation (24.8 minutes compared to 78.6 minutes in patients who received moderate sedation, p<3x10-6).

Results
Focused TEE evaluation of the LAA can be performed without conscious sedation using a MicroTEE transducer and produces sufficient imaging quality to evaluate for thrombus and LAA function. Studies were significantly shorter due to an expedited nursing workflow and decreased monitoring required post procedure.  There was an added benefit of avoiding the risks of moderate sedation, however low they may be. Patient perception was positive with a reported low degree of discomfort and the majority favoring a no-sedation protocol.  This small study suggests that selected patients can benefit from MicroTEE use and time savings and cost benefits may improve lab efficiency.
To pee or not to pee - Treating acute urinary retention with heating pads in the elderly

Aarthi Kannan, MD; Christine Gould PhD, Hee-Jeong Kim CNS, Pratyusha Narra MD, Meera Sheffrin MD, Matthew Mesias MD, Philip Choe DO, Julia Cremer MD, Joyce Tenover MD

Stanford, Department of Medicine - Geriatric
Background
Acute urinary retention (AUR) is a common inpatient elderly problem which is treated exclusively by catheterization. Catheters are associated with urinary tract infections (CA-UTI), patient discomfort, mechanical trauma and hematuria. We wanted to assess prevalence of AUR seen by our Mobile Acute Care for the Elderly (ACE) team and investigate the use of non invasive treatment - heating pad to treat AUR. On a 3 month retrospective chart review, 13 of 50 geriatric patients (26%) seen by our mobile ACE team over different units had AUR which was managed exclusively by urinary catheters as per standard of care at Stanford hospital. A survey showed that 18/20 patients and 50/50 nurses were in support of using heating pads over catheters to treat AUR.

Current State
To assess the efficacy of heating pads for treatment of AUR and reduce the number of urinary catheterization by 25% in patients aged 65 years and above between January 2019 and April 2019

Analysis
A fifteen minute heating pad application to supra-pubic area as a treatment for AUR was assessed by 2 PDSA (plan-do-study-act) cycles in the inpatient setting. Treatment was administered by nursing.

Results
Over a 4 week period we found that heating pads are a promising method in treating AUR thus reducing urinary catheterization rates and their complications. On 6 out of 8 occasions (75%), patients were successfully treated with the heating pad for AUR, thus avoiding the use of urinary catheters on those occasions. The first PDSA cycle helped us identify barriers in the identification of AUR in our patients and in heating pad application. The second PDSA cycle ensured successful identification of AUR and application of the heating pad in our patients.

Next steps :-
1. Extending heating pad use to to the ER, postoperative units, PACU and units with increased incidence of urinary catheter use for AUR
2. Investigating heating pads to treat AUR in all adult populations
3. D2 unit is obtaining a nursing legacy grant to continue implementation of this measure to reduce urinary catheter use and downstream complications
Developing a Skilled Nursing Facility Curriculum for Geriatric Medicine Fellows

Pratyusha Kishore Narra, MD; Can Chen , Marina Martin , Philip Choe , Meera Sheffrin

Stanford, Department of Medicine - Geriatric
Background
We are developing a year-long curriculum for geriatrics fellows to adequately prepare
them for a career in a skilled nursing facility (SNF) setting by the end of their 1-year fellowship. We wish to further refine the curriculum to better meet fellow needs.

Current State
Beginning July 2018, we developed and implemented a year-long SNF curriculum based on
core competencies defined by The Society for Post-Acute and Long-Term Care Medicine and common issues encountered in a SNF setting, as determined by an expert panel of 4 geriatricians on our faculty. We did needs assessment through a standardized survey sent out to all graduates of the Stanford Geriatrics Fellowship program in the past 5 years. The survey assessed their subjective comfort with skills and knowledge necessary to practice in a skilled nursing facility. Fellows rated their comfort level on a variety of topics, and responded to an open-ended question asking them to identify specific areas for improvement.

Analysis
On content analysis of our the open-ended survey question requesting past fellows to identify specific areas of improvement, many respondents have said that they would have benefited from having more responsibility, having more mentorship and didactics in a SNF setting, and more training on SNF regulations. Working in SNF is a common career path for geriatrics fellows; however less than half of past graduates felt very prepared to begin a career in a SNF setting at the end of their fellowship. We are actively incorporating changes to rectify the deficiencies into our curriculum, including developing dedicated didactics sessions to address these topics.

Results
We had a total survey response rate of 73.3% (11/15). Only 45% of fellows felt very prepared to
begin a career in a SNF setting at the end of their fellowship. When asked about their comfort levels
with certain SNF related topics, we identified several critical deficiencies:
- Only 27% of respondents were very familiar with federal and state regulations regarding practice in a SNF setting
- Only 18% of respondents were very familiar with medical director responsibilities
- None of the respondents were very comfortable managing spinal cord injuries
- Only 9% of respondents were very comfortable managing wounds
We plan to continue to collect data from current and future fellows as they graduate from our fellowship to further improve the SNF curriculum in our fellowship.
        Why are They Eating Dinner in Bed: Barriers to Mobility in Hospitalized Older Adults

Julia Cremer, MD; Nannette Storr-Street, CNS; Matthew Mesias, MD; Ankur Bharija, MD

Stanford, Department of Medicine - Geriatric
Background
Background: Mobility loss in hospitalized older adults is associated with poor qualitative outcomes like longer length of stay, readmission, and facility placement. The Quality and Performance Measurement Committee of the American Geriatrics Society supports greater focus on mobility as an outcome for hospitalized older adults. Although the low fall rates have been exemplary at our institution, poor attention to mobility is increasingly being recognized. As an academic medical center, with many disciplines and teams involved in care, attention to evaluating and promoting mobility starts from having a clear understanding of the barriers and opportunities. Our goal was to assess the knowledge and viewpoints of residents and charge nurses regarding barriers to mobility for hospitalized older adults.

Current State


Analysis
Methods: A nursing unit at an academic medical center with greater attention to mobility was chosen. A survey was administered to internal medicine residents working on the unit. Eight questions asked in multiple-choice format assessed views and knowledge of mobility. 17 internal medicine residents were invited to participate, and 11 completed the survey. Two charge nurses from the unit were interviewed individually.

Results
Results: 55% of residents surveyed thought our patients value mobility. The same amount thought mobility loss was associated with longer length of stay, readmissions, and facility placement. 100% thought mobility should be more highly prioritized by staff. 64% thought RNs on the unit assess mobility during their daily assessments. 18% thought mobility was a quality outcome measured in our healthcare system. None of the residents were aware of any validated means to assess patient mobility. The barrier to mobility most frequently cited was the wait time for physical therapy to see the patient, with 82%. 73% thought risk of falling was a barrier. 9% of residents thought nursing staff shortage and nursing workload was a barrier. None of the residents cited restraints or bed alarms as barriers but did report telemetry and lines as a barrier. Both charge nurses cited nursing staff shortage, workload, and waiting for physical therapy as the greatest barriers.
Conclusions: Both residents and charge nurses felt that waiting for physical therapy was a major barrier to mobilizing our patients. Only one resident was in-tune with nursing staff shortages and workload as a barrier. The majority of residents surveyed, and both nurses, did think mobility loss is associated with length of stay, readmission, and facility placement. In the upcoming months, more unit-based nurses, nursing aides, and physical therapists will be interviewed, so common barriers may be identified. Patients themselves will be interviewed about barriers from their points of view. The goal is to design and implement a mobility care path™ through an interdisciplinary approach to proactive mobilization for hospitalized older adults. 
Antimicrobial Stewardship in Immunocompromised Patients: Evaluating Management of First Episodes of Febrile Neutropenia in Patients with Hematologic Malignancies

Amy Chang, MD; Marisa Holubar, MD;  Aruna Subramanian, MD;  Bruno Medeiros, MD; Emily Mui, PharmD; Lina Meng, PharmD; Michaela Leidke, MD; Stan Deresinski, MD

Stanford, Department of Medicine - Infectious Disease
Background
Antimicrobial stewardship (ASP) is critical to the appropriate management of immunocompromised patients, but often difficult to perform in the immunocompromised population.  However, with the rising use of broad spectrum antibiotics comes a concern for collateral adverse effects.  National guidelines have been developed to direct appropriate treatment of febrile neutropenia (FN.) 

Current State
We performed a retrospective chart review including 211 inpatient adults diagnosed with acute myelogenous leukemia who received inpatient management of first episodes of FN between 2011 and 2017.  Each new antibiotic start during the hospitalization was evaluated for appropriateness based on criteria utilizing a combination of national guidelines and expert review.  The most common issues identified included inappropriate changing of empiric broad spectrum agent (38%), inappropriate initiation of resistant gram positive agents (17%), and failure to de-escalate resistant gram positive coverage at 72 hours (13%).  Our goal is to decrease inappropriate use of antibiotics by half by the end of the year 2019.


Analysis
Key drivers contributing to inappropriate antibiotic included lack of knowledge of national guidelines, lack of an internal protocol, physician hierarchy with attending-driven decisions, and erroneous use of antibiotics by cross-cover providers.  Lectures were provided to attendings, advance practice providers (APPs), and trainees on study results and appropriate antibiotic management.  An internal FN protocol is being created to guide antimicrobial use.

Results
Pre/post-tests given to APPs and trainees suggested an improvement of some knowledge after lectures were given, though misperceptions remained regarding appropriate usage of empiric vancomycin.  A video lecture is planned to sustain the educational intervention.  The internal febrile neutropenia protocol will be released in the near future.
Results from Realizing Improvement through Team Empowerment:  A Multi-Disciplinary Team-based Quality Improvement Initiative to Reduce Clostridioides difficile Rates and Promote Antimicrobial Stewardship on E3/F3

Amy Chang, MD; Hurley Smith; Katrina Sullivan, RN, PCCN; Joanne Meneses, MSN, RN; Natalie Kirilcuk, MD; Jocelyn Cooper-Sterling, MPH, RN, CIC; Jamie Kuo, PharmD; Amer K Karam, MD; Tammy Schaffner, RN, BSN, CIC; Lucy Tompkins, MD PhD; Paul Maggio, MD; Anita Girard, DNP, RN, CNL, CPHQ, NEA-BC; Emily Mui, PharmD; Lina Meng, PharmD; Stan Deresinski, MD; Marisa Holubar, MD MS

Stanford, Department of Medicine - Infectious Disease
Background
Clostridium difficile (C. Diff) is a spore bacterium that can survive for  months and is easily spread from person to person from hands that have touched contaminated surfaces. C. diff is the most common cause of healthcare associated infections in the United States and was estimated to cause 29,000 deaths in 2011. The healthcare costs related to C. difficile is estimated to be as much as $4.8 billion for acute care facilities alone. These estimates are expected to increase as C. difficile morbidity and mortality rises along with the increased costs.


Current State
The CDI Standard Infection Rate (SIR) at SHC from Fiscal Year 2017, Quarter 2 to Fiscal Year 2018 (FY18), Quarter 1 was 1.18 (104 total infections). The organizational Vizient target for C Diff in Fiscal year 2019 (FY19) is 0.65. 19 of the 51 infections that represent our gap to top decile performance occurred on E3, F3, C2, and D2G2S cumulatively which represent approximately 37% of our opportunity to improve.   Our goal was to reduce the average hospital-acquired CDI (HO CDI) rate from a baseline of 19 infections per year to 10 infections per year by August 31st, 2019.

Analysis
Our root cause analysis was delayed by the over-abundance of perspectives provided by experts spanning from multiple departments, including nursing, quality, infection prevention (IP), antimicrobial stewardship (ASP), the surgical subcouncil, and frontline providers.  Ultimately, independent investigation performed by IP, ASP, and pharmacy representatives was performed.  Review of FY18 HO CDI cases on E3/F3 revealed inappropriate antibiotic prescribing practices, missed community-onset CDI (CO CDI), and poor patient/family compliance with PPE as highest yield issues contributing to HO CDI cases.  Key drivers identified included: improved compliance with the CDI prevention bundle, appropriate antibiotic usage, and developing a reliable process for early identification of CO CDI.  Numerous interventions were performed, including but not limited to creating a new pathway for early identification of CO CDI, nursing/staff education regarding on CDI identification and stool documentation, nursing-led stewardship leading discussion on antibiotic plan/duration on daily rounds, twice weekly rounds between ASP and a colorectal attending representative, and mandatory infectious disease consultation for patients on piperacillin/tazobactam for >72 hours.

Results
Interventions were started mid-to-late fiscal year and as such have not yet demonstrated a significant change in HO CDI rates to date.  However, the multiple interventions aimed at antibiotic use did lead to a significant decrease in antibiotic days of therapy.  Ongoing reflection by nursing and ASP are being performed to determine feasibility of sustaining current interventions.
There is No Place Like Home: Physician, Nursing, and Patient Perceptions on the Impact of Co-Localization

Swati Yanamadala, MD; Surbhi Singhal, Ashish Sarraju, Lisa Shieh

Stanford, Department of Medicine - Internal Medicine
Background
Effective communication between nurses, physicians and patients has been extensively documented as critical to providing safe and quality patient care. While co-localization of patients and physicians into hospital home units may have the potential to facilitate this communication, co-locating inpatient providers and patients can be logistically limiting. Here we aim to assess the perceptions of nurses, physicians and patients on the impact co-localization has on communication between team members and quality of patient care.

Current State


Analysis
Electronic surveys were distributed to nurses who worked on internal medicine team home units and internal medicine resident physicians at a tertiary care academic center. The survey questions employed a 5-point Likert scale to assess health care provider attitudes regarding co-localization of patients and providers into inpatient home units. Patient satisfaction results were obtained from Press Ganey surveys of patients who were cared for by internal medicine teams.

Results
In total, 61 resident physicians, 36 nurses, and 214 patients completed the surveys. Resident and nursing surveys were conducted in August-September 2018. Patient surveys were conducted in January-March 2017. 39% of nurses rated clarity of communication with the primary team as a 4 or 5 when patients were co-located, compared to only 11% without co-localization (p<0.01). Similarly, among residents, 85% rated clarity of communication between the primary medical team and bedside nurse team as a 4 or 5 when patients were co-located, compared to only 26% without co-localization (p <0.01). Additionally, residents reported they were more likely to come to the bedside to respond to patient questions (p<0.01) and visit the bedside more frequently throughout the day (p<0.01) when patients were co-located. Patients also reported improved communication when on home units, with 81% of co-located patients reporting doctors always carefully listen to them, compared to 68% of non co-located patients (p=0.04). While not statistically significant, co-located patients were 1.2 times more likely to report doctors always listened carefully to them (81% vs 68%), 1.2 times more likely to report doctors always explained in an understandable way (79% vs 68%), and 1.2 times more likely to strongly agree that they had a good understanding of managing their health (56% vs 47%) compared to patients who were not co-located.

Co-localization of patients, physicians and nurses into home units had a significant impact on physician and nursing perception on clarity of communication and the care they were delivering, as well as patient satisfaction.
Clinical outcomes associated with encounter to balloon vs door to balloon: a retrospective study

Karolina Jaluba, MD; David Patrick Lee

Stanford, Department of Medicine - Internal Medicine
Background
Prior ST elevation myocardial infarction (STEMI) studies have focused on outcomes associated with shorter door to balloon (D2B) times. Little is known about clinical outcomes associated with encounter to balloon (E2B) times, which is defined as the time from initial evaluation by emergency medical services (EMS) in the field or by medical personnel in the clinic to percutaneous coronary intervention (PCI). The aim of this study was to evaluate E2B outcomes and compare these outcomes to patients who were evaluated initially in the emergency department, referred to as the door to balloon (ED-D2B) group.

Current State
A total of 189 patients with STEMI from 2012 to 2017 from our institution were evaluated. Patients who presented to the hospital in cardiac arrest, cardiogenic shock or who required pre-PCI intubation were excluded. All-cause mortality and major adverse cardiac cerebral events (MACCE) including subsequent unstable angina or non-STEMI, stroke and/or hospitalization for heart failure exacerbation were noted in each group and compared at hospital discharge, 30 days and at one year.

Analysis
There was a trend towards a lower incidence of a composite of all-cause mortality and MACCE in patients within the E2B group at one year compared to the ED-D2B group (6.8% vs. 13.9%, P=0.126). The median D2B time in the E2B group was significantly lower than in the ED-D2B group (51 min vs. 68.5 min, p<0.001). The E2B group had more Caucasians (65.8% vs. 50%, p=0.031) and lower incidence of a family history of coronary artery disease (CAD; 15.4% vs. 27.8%, p=0.039). In a subgroup analysis, there was a trend towards a comparatively lower incidence of a composite of all-cause mortality and MACCE at one year in the E2B group, particularly for non-diabetics (4.3% vs. 13%, P=0.067) and age <75 (3.4% vs. 10.7%, P=0.091).

Results
The results show a trend towards better clinical outcomes in STEMI patients who were initially treated prior to hospital arrival, which may have been influenced in part by shorter D2B times, race and family history of CAD. Furthermore, there may be particular patient populations that derive more benefit than others from this early interaction. However, given the small sample size, these results will need to be validated in a larger study.
Understanding Patients™ Perspective of Cost-of-Care: The Financial Implications of Observation Medicare Admission Status

Divya Gupta, MD; Apurva Yeluru MD, Teresa Liu MD MPH, Divya Parikh MD, Lisa Shieh MD PhD

Stanford, Department of Medicine - Internal Medicine
Background
Hospitalizations comprise a significant component of patients™ healthcare costs, especially for Medicare patients admitted under observation. Per the NOTICE act, Medicare patients admitted under observation are required to be informed that they are receiving services as outpatients and the financial implications of such status. While there is a growing body of literature on inpatient hospitalization costs and the various drivers of such costs, the financial impact of observation stays from a Medicare patient perspective has been understudied. Moreover, there is scant literature on interventions that assist patients with coping with the cost of care beyond physician-patient discussions.

Current State
Based off a prior study, only 50% of patients hospitalized at Stanford Hospital felt that they had enough money to cover their out-of-pocket costs for their hospitalization. Despite the prevalence of financial concerns, only 10% of patients had met with a financial counselor.

Specifically, Medicare patients admitted under observation that may have larger out-of-pocket cost implications do not necessarily meet with a financial counselor. Currently, observation patients at Stanford Hospital are informed of their observation status by a bedside coordinator upon admission in the ED. A financial counselor is called to speak with the patient and answer further questions only if the patient requests it. Under this system, only 50% of Medicare observation patients knew that they were admitted under observation.

SMART Goal: In Medicare patients, increase awareness of observation stay to 90% by end of June 2019.

Analysis
To increase Medicare observation patients™ awareness and understanding of the financial implication of their observation stay, we implemented the following quality improvement interventions:
Patient population: English-speaking Medicare patients admitted under observation in medical-surgical units at Stanford Hospital
Intervention 1: Increase patient access to financial counseling
Patients randomized to standard observation workflow vs standard workflow plus meeting with a financial counselor. Surveys administered to patients to assess efficacy of financial counseling in helping patients understand observation status and the financial implications.
Intervention 2: Optimize information disseminated by financial counselors with an easy to understand MOON brochure
Patients randomized to standard observation workflow vs standard workflow and patient education with updated MOON brochure. Surveys administered to patients to assess understanding of observation status after education with MOON brochure.


Results
Intervention 1
Results: Of the Medicare observation patients that received financial counseling (total 24), 71% were aware that they were admitted under observation vs 50% of those that did not receive financial counseling. 75% found financial counseling to be valuable, and 90% believed they learned new information as a result of financial counseling.
Sustain Plan: Now, all patients admitted under observation meet with a financial counselor up front and are presented with the MOON.

Intervention 2:
Results: 80% of Medicare observation patients that received the MOON brochure knew that they were admitted under observation vs 33% of those that did not receive the MOON brochure.
Sustain Plan: Financial counselors and/or ED registration team to hand out updated, self-explanatory, MOON brochure as part of standard workflow to Medicare patients admitted under observation
Improving the Pediatric to Adult Transition for Cystic Fibrosis Adolescents

Alicia Mirza, MD; Cissy Si

Stanford, Department of Medicine - Internal Medicine
Background
The transition from a pediatric to adult health system can be a vulnerable time for young adult patients with chronic diseases. Treatment advances have led to a growing number of these individuals, and thus there is a compelling need for specialized health care delivered by providers with expertise in both adult medicine and in chronic conditions that begin in childhood. Clinical guidelines suggest routine use of structured transition readiness assessments and development of transition care plans for these patients.

Current State
There is currently no standardized transition process for cystic fibrosis patients between Lucile Packard Children™s Hospital Pulmonary clinic and the adult Stanford Chest Clinic. By December 2019, we will have performed a standardized hand off meeting with associated documentation on 100% of the patients who transition from pediatric to adult pulmonary medicine.

Analysis
We have scheduled multidisciplinary quarterly meetings to review pediatric patients who are ready to transition to adult pulmonary. We will create a standardized note in both their LPCH and Stanford charts with important clinical and psychosocial information. 

Results
We will use surveys to gauge communication effectiveness by both providers and patients and adjust the document accordingly. We will perform chart reviews of all patients who transition over to ensure all have received a transition document. 
The Medicare Admission Status Conundrum: Assessing the Understanding of Observation versus Inpatient Status in Internal Medicine and Emergency Medicine Residents

Teresa Liu, MD; Apurva Yeluru, Lisa Shieh

Stanford, Department of Medicine - Internal Medicine
Background
Hospitalizations constitute a significant portion of patients™ healthcare costs, particularly for Medicare patients admitted under observation status.  Choosing the most appropriate admission status has financial implications for both the patient and the hospital as an institution.  While there has been growing interests in studying the various drivers of inpatient hospitalization costs, there is scant literature on how best to improve the rates of improved designations of the admission order status for Medicare patients.  One potential opportunity for improvement that has been understudied is whether formal education on admission criteria for admission versus observation status to trainees, namely internal medicine and emergency medicine residents, is of benefit to both patients and the hospital. 

Current State
As of now, there is no official education module or curriculum regarding Observation versus Inpatient admission guidelines for Medicare patients for either internal medicine or emergency medicine residents.  The current workflow is for EM residents to place the admission order without a formalized feedback system with the IM residents.   If IM residents do not proactively review admission status, then unit case managers may review the cases and request a change in admission status based on clinical criteria.
For internal medicine and emergency medicine residents, increase understanding of Medicare Observation versus Inpatient admission guidelines to 90% of residents by beginning of July 2019.
Specific: Improving knowledge about OBV vs inpatient admission guidelines
Measurable: Survey internal medicine and emergency medicine residents pre- and post- education module
Achievable: A lot of room for improvement given no formal curriculum currently on Medicare admission guidelines
Realistic: Variable factors of residency limiting 100% of audience reached
Timely: Deadline July 2019

Analysis
To increase awareness and understanding of Medicare Observation versus Inpatient admission criteria for frontline residents (Internal Medicine and Emergency Medicine), we implemented a two-step, educational, quality improvement intervention.
Step 1- Pre-education module survey to internal medicine residents, then short lecture was given during IM noon conference on Medicare Observation versus Inpatient admission guidelines, followed by post-education module survey to assess understanding.
Step 2- Pre-education module survey to emergency medicine residents, short lecture given during weekly EM Wednesday conference on Medicare Observation versus Inpatient admission guidelines, followed by post-education module survey to assess understanding. 

Results
Post-education module intervention, assessment of the residents™ understanding of Medicare admission status guidelines increased from 75% to over 90% in the internal medicine group and increased from 60% to 85% in the emergency medicine group.   Overall the composition of the residents polled during the IM conference and the EM conference sessions were predominantly PGY1 and PGY2 residents.   Given these results, the sustain plan moving forward is to incorporate a formal education session for rising PGY2 internal medicine residents, and continue post-education assessment of understanding, as well as incorporate an easy-to-reference checklist and guide for both internal medicine and emergency medicine residents to utilize during admissions to facilitate designation of Observation versus Inpatient admission status for patients.   
A high value care checklist for improving laboratory test, telemetry, and level of care utilization on general medicine

Neel Chari, MD; Lisa Shieh, MD

Stanford, Department of Medicine - Internal Medicine
Background
Overutilization of laboratory testing, cardiac monitoring, and inappropriate level of care are common across many institutions. These interventions can negatively affect the patient experience by worsening sleep quality, causing iatrogenic anemia, and leading to further unnecessary downstream testing. They also come at a significant cost to both institutions and patients. Nationally, the ABIM Choosing Wisely campaign has identified repetitive complete blood counts(CBC) and basic metabolic panels(BMP) and cardiac monitoring outside of the ICU as two of the top five issues that physicians within hospital medicine should examine.

Current State
Resident Survey: 52 total responses
-51/52 residents believe that some patients receive unnecessary CBC and BMP lab draws
-Most cited reasons are concern that resident may not have the data when asked for it, lack of expectations set by attending physician, lack of feedback regarding appropriateness of current ordering habits, and ease of ordering repeat lab orders via EMR
-After their initial order of daily lab tests, only 46% of residents re-assessed which laboratory tests were still necessary on a daily basis, most residents addressed this issue either every other day, whenever their initial order expired, or never.
-On attending rounds, 89% of residents reported that they only discussed which labs were necessary on 25% or fewer of their patients.

SMART goal:
Reduce the number of total CBCs+BMPs per patient-day by 20% over a period of two months
Reduce the number of patients on IICU level of care by 20% over a period of two months
Reduce the number of patients on telemetry by 20% over a period of two months


Analysis
Five significant contributing causes:
-Fear of not having lab results available on rounds if asked
-Varying attending expectations for laboratory testing and telemetry utilization
-No feedback on whether current ordering habits are appropriate
-No clear protocol for daily assessment of lab tests, telemetry, and level of care
-Ease of ordering standing laboratory tests via EMR

Intervention:
Provide checklist to the senior resident on 3 of the 6 General Medicine teams on Stanford Wards. Senior resident instructed to ensure that all three items on the checklist(which labs are necessary for patient, whether patient should be on telemetry, and whether current level of care is appropriate) are discussed on attending rounds for each patient on the team. The residents were informed that lab orders, telemetry usage, and level of care will be tracked for all patients.

Collect data via manual audit:
-Checklist completion
-CBCs+BMPs per patient
-Number of patients on IICU and telemetry


Results
Mean CBCs+BMPs per day:
1.63 in control arm
1.29 in intervention arm
29% reduction
Cost savings $120,000 when extrapolated to one year

Mean percent of patients on telemetry per patient-day:
14% in control arm
9% in intervention arm
35% reduction
Cost savings $450,000 when extrapolated to one year

No significant difference between days on IICU level of care vs Acute care between groups

Total checklist completion for intervention arm: 70%

Sustainability: Integrate checklist into handoff tool in EMR for senior resident to use
Reducing Inappropriate Utilization of the Intermediate Intensive Care Unit Amongst Cardiovascular Medicine Providers

Justin Slade, MD; Lisa Shieh, MD, PhD and David Svec, MD, MBA

Stanford, Department of Medicine - Internal Medicine
Background
Stanford Hospital now utilizes Acuity Adaptable Units in order to more efficiently transition patients between an intermediate intensive care unit (IICU) and acute or floor level of care. This reduced the burden of transferring patients between units when their status changed, yet there was an increase in inappropriate utilization of IICU beds. Audits revealed a 40% rate of inappropriate utilization of this resource, which in turn contributed to higher direct accommodation costs.

Current State
We aimed to reduce the percentage of IICU level of care hours amongst cardiovascular medicine providers by 25% within six months.

Analysis
A root cause analysis was performed and deescalation from the IICU to acute level of care was identified as a key driver in reducing IICU overutilization. Two Epic best practice advisories (BPAs), a non-interruptive and once-daily interruptive, with prompts to efficiently downgrade a patient™s level of care if indicated, were designed and implemented. Additionally, house staff were provided with real time feedback regarding IICU utilization appropriateness per case management review.

Results
From April 2018 to January 2019, the percentage of IICU level of care hours per admission decreased from an average of 46% to 30%. Overall there was an average decrease of 27 IICU level of care hours per patient admission. A non- or minimally-interruptive Epic BPA in addition to house staff utilization feedback can be an effective means of facilitating timely transfer of patients to the most appropriate level of care, in turn reducing direct costs associated with over utilization.
Utilizing novel two-way communication methods to understand peri-urgent communication.

Sammita Satyanarayan, MD, Ron Li, MD; Amy Chang, MD -2 Neel Chari, MD -2 Karen Li, MD-3 Lyndly Tamura, MD -4  Katie Kvam, MD- 1 Christopher Sharp, MD- 2  (1- Stanford Neurology  2- Stanford Internal Medicine 3- Stanford Psychiatry  4- Stanford Physical Medicine and Rehabilitation)

Stanford, Department of Medicine - internal Medicine
Background
Rapid Response Team events (RRTs) are coordinated multidisciplinary provider teams that respond to urgent medical changes. In the last 3 years, there has been an average of 36 urgent Rapid Response Team (RRT) events per month at SHC. Review of communication prior to the event is currently limited to Epic documentation or one-way communication review. As of October 2018, Stanford Hospital initiated a new method of real-time HIPAA compliant texting technology (Voalte) allowing a detailed two-way communication record between providers prior to urgent events.

Current State
We are currently undertaking a retrospective chart review to understand the current state of documentation and communication regarding RRT events at SHC, as well as to detail a methodology as to extract and efficiently utilize Voalte text data as a research measure.

Analysis
We obtained IRB exempt approval and subsequently acquired retrospective Rapid Response Team (RRT) event data documented through the MIDAS system as well as Epic reports of treatment team members for specified event days. We subsequently requested Voalte aggregate text data for RRT dates, extracted relevant variables, and organized available data by conversation. We identified variables collected through MIDAS and Epic for RRT events on medicine units (B3, C3), as well as measured baseline metrics in Voalte texting data identified prior to that event. Lastly, we compiled our methodology of linking existing clinical platforms (epic, MIDAS) to Voalte data, with the goal of making Voalte data an accessible outcome or process measure in the Stanford community.

Results
Real-time HIPAA compliant texting data through Voalte is a new and powerful means of healthcare provider communication and can be used as a novel measure of communication in inpatient emergency settings. We plan to detail our methodology of extracting related Epic and Voalte texting data, clarify relevant variables within collected Voalte data, and re-create identifiable conversations prior to RRTs. We anticipate identifying the current state of documentation with RRTs at Stanford with Epic and MIDAS data, as well as the current state of instant messaging prior to RRTs. 
  ICU-Based Palliative Care for Older Adults with Traumatic Head Injuries: A Feasibility Study

Andrea Jonas, MD; K. Staudenmayer, R. Aslakson

Stanford, Department of Medicine - Pulmonary and Critical Care
Background
Geriatric head trauma patients in the Intensive Care Unit (ICU) present a large unmet need for palliative care and advanced care planning. Over half of older adults with head trauma require ICU care and overall have in-hospital mortality ranging from 30-80%. Despite best practice recommendations emphasizing early goal-related communication, few to no evidence-based approaches have been tested in this population. Evidence from 18 clinical trials across diverse care settings supports the efficacy of a video-based decision support tool developed by ACP Decisions. The 5-minute video outlines treatment options including full care (full code), limited care (do not resuscitate/do not intubate), and comfort care. The video™s feasibility, tolerability, and efficacy has not yet been demonstrated in an ICU-based, geriatric head trauma population.

Current State
There is an unmet need for palliative care interventions in geriatric traumatic head injury patients. Our goal is to bridge the communication gap through use of a video-based decision support tool for patients and their families.

Analysis
After identification by the research team and oral consent, participants (patients and/or family) watch the ACP Decisions video and then complete a 9-question survey. Goal enrollment is 35 participants. Inclusion criteria are: (1) Patients must be >age 65, presenting with head trauma, ability to consent, admitted/being considered for admission to the ICU, and able to speak one of the 17 languages in which the ACP Decisions video is offered; and (2) Family member participants must be >age 18, able to consent, and able to speak one of the 17 languages in which the video is offered.

Results
Data collection is ongoing with 20 participants already enrolled in the study (11 patients, 9 family and/or friends) and 87% enrollment to date of patients approached. The majority of participants felt comfortable (55%) or very comfortable (25%) watching the video. 35% of the population studied found the video contents helpful, with an additional 25% finding the video "very helpful". 75% of the studied population would either recommend or definitely recommend the video to another.

Preliminary results of our feasibility study for implementation of a video decision support tool among older adults admitted to the Stanford ICU with traumatic head injury have been encouraging. Participants report comfort watching the video and would recommend its contents to another person. We anticipate that enrollment will complete by March 2019. This study is the first phase of a broader goal to implement an evidence-based advance care planning and palliative care intervention among older adults with traumatic head injuries.
ICE-QUIP: Inhaler Compliance Education-QUality Improvement Project

Andrea Jonas, MD; Ryan Anderson, David Condon, Lauren Eggert, Bita Ghafouri, Alicia Mirza

Stanford, Department of Medicine - Pulmonary and Critical Care
Background
Problem Statement:
Patients™ understanding of their inhaler regimen and appropriate inhaler technique is suboptimal and prevents patients from receiving maximal benefit from their medications.

Background:
Most patients with asthma or COPD do not use their inhalers properly, and most have not had their technique checked or corrected by a health professional.
Incorrect inhaler technique when using maintenance treatments increases the risk of exacerbations and excess health care utilization.
Poor symptom control  could be related to incorrect inhaler technique.
Incorrect inhaler technique when using inhaled corticosteroids increases the risk of side-effects like dysphonia and oral thrush.
The steps for using an inhaler device correctly differ between brands.
Checking and correcting inhaler technique can improve asthma outcomes.
Low confidence in inhaler usage was associated with lower adherence and poor COPD-related health status.

Current State
Goal is to improve patient's understanding of their inhaler regimen, and improve inhaler technique using a provider-lead initiative of patient education during the clinic visit.

Analysis
We have developed a paper survey that is administered to all patients seen in the Pulmonary Fellows clinic at Stanford University Hospital. The survey gathers information regarding the patient's inhaler regimen, whether they make use of an inhalational spacing device, whether they can identify which inhaler serves as a "rescue" inhaler, their confidence in inhaler use, and what teaching resources they would like to have available to them to assist in inhaler technique. Our interventions will include use of in-person demonstrations of inhaler and spacer use, use of educational videos to demonstrate inhaler and spacer use, and use of adhesive labels indicating which inhaler is a "rescue" inhaler. 

Results
The survey is continuously being administered with ongoing data collection, allowing for real-time assessment of the impact of our different interventions as detailed above. We anticipate iterative improvement in inhaler education, with incorporation of the inhaler survey as a routine portion of Pulmonary Fellow clinic care. 
Re-training the well trained: Assessing the barriers of using RAPID3 for rheumatoid arthritis

Anna Postolova MD, MPH and Brian Abe MD, PhD; Janice Lin MD, MPH, Kate Lorig DrPH

Stanford, Department of Medicine - Rheumatology
Background
Rheumatoid Arthritis is a chronic systemic disease affecting around 1.3 million US adults with a significant impact of physical, functional, and social well-being. Measuring disease activity in patients with RA provides a metric for physicians to accomplish the treat to target strategy. RAPID3 is a validated disease assessment tool comprised of 10 self-reported questions addressing functional status.  Despite the known benefits of using RAPID3 in clinical settings, there are barriers to its use.  We sought to assess these barriers to increase utilization in our clinic. 

Current State
Currently, the use of RAPID3 is limited.  We sought to evaluate the current use of RAPID3 in our division and the address the barriers to its use.

Analysis
We surveyed the use of RAPID3 among faculty and fellows in our clinic with a pre-analysis survey to discover the barriers associated with use.  We then implemented a workflow in the clinic to address these barriers by capturing RAPID3 scores prior to each patient encounter. RAPID3 questions were collected directly by the patient in MyHealth or entered by a medical assistant in the exam room prior to each physician encounter.  Physicians were trained on how to access this information in EPIC, and a second survey was obtained regarding use and further barriers.  Based on these barriers, further instructions and areas of improvement are planned and will be followed by a third and final survey.

Results
In our pre-analysis survey, 75% of providers did not use RAPID3, with 12.5% using it less than 25% of the time, and 12.5% using it between 51-75% of the time.  No respondents used RAPID3 more than 76% of the time.  The most common barriers to using RAPID3 were, in decreasing frequency, time taken to obtain scores, unlikeliness that scores will impact patient care, unfamiliarity with RAPID3, and preference for alternative disease scores.  Providers requested that RAPID3 scores be obtained prior to each encounter and that scores can be tracked over time. After incorporating these requests in our workflow, RAPID3 non-users decreased from 75% to 56.3%. Among users of RAPID3, 87.5% felt very comfortable with its use in EPIC and 85.7% found previous scores very or somewhat useful.  57% of users reviewed RAPID3 scores with their patients, with 100% of these respondents stating their patients were more engaged with their disease management. Time spent checking RAPID3 scores took on average less than 60 seconds, and total time checking and reviewing scores with patients took between 1 and 5 minutes on average.  We conclude that implementing an efficient workflow in our clinic to obtain RAPID3 scores increases the use of RAPID3 and can help patients be more engaged in the management of their rheumatoid arthritis.  We plan to incorporate additional interventions for providers by further educating on the clinical utility of RAPID3 in powerpoint presentation format as well as implementing a dot-phrase in EPIC that will calculate and track RAPID3 scores in an easily viewable and accessible format.  We will perform an additional survey at the end of the year to assess if this will intervention will further increase utilization of RAPID3.
If a tree falls in the woods... improving POLST completion rates to memorialize patient care preferences

Mukund Ramkumar, MD; Jessica Langston, Christine Henley, Tiffany Lee, Lindsey Grabek, Kimberly Hall, Lindsey Daudt, Lisa Shieh, Michael Gisondi, Stephanie Harman, Winnie Teuteberg

Stanford, Department of Medicine- Critical Care Medicine
Background
Resident Safety Council  -- Goals of Care project

POLST completion rates by providers for patients with code status limitations are low. These patients are at high risk of harm by receiving undesired care.


Current State
Increase POLST completion for patients with code status limitations on Internal Medicine to 80% by June

Analysis
* GOC Note to include documentation regarding POLST Completion

* Document with guidance on having POLST conversation and sample terminology


Results
In process
Building and implementing an interdisciplinary protocol for the Brain Code

JoJo Yang, MD; Anna Janas MD PhD, Karley Mariano NP, Lindsey Rasmussen MD, Sarah Lee MD

Stanford, Department of Neurology
Background
Acute brain herniation and sustained intracranial hypertension (elevated intracranial pressure, ICP) are brain codes signifying catastrophic neurologic events that require immediate recognition and treatment to prevent irreversible injury and death. As in cardiac arrest, a brain code mandates the organized implementation of a stepwise management algorithm. In our current clinical practice, such an algorithm does not exist, and at our institution there are varying degrees of comfort and experience across fields of neurology, neurosurgery, pediatric intensive care, nursing, and pharmacy. With this Quality Improvement project, we propose to gather an interdisciplinary team of stakeholders in the aforementioned fields, develop an evidence-based and standardized approach to the evaluation and management of patients with intracranial hypertension, and implement the protocol.

Current State
As acute brain herniation can occur via many different pathologies, we first chose to narrow the target population to patients with traumatic brain injury (TBI). In order to measure intracranial pressure, patients included in the current state had an EVD (external ventricular drain) placed. 7 patients with EVDs were identified. As a work in progress, baseline data for the current state is being collected. The goal of the project will be to implement the ICP guideline in LPCH PICU decrease elevated intracranial pressure by 15% in TBI patients in 6 months. A secondary goal will be to improve patient outcomes by PCPC (pediatric cerebral performance category) score at 12 months. 

Analysis
We created an intervention with a tiered management guideline graphic, a dotphrase for neurology, neurosurgery, and pediatrics residents to include in notes, as well as an order set for elevated ICP medical management. The order set was reviewed by nursing, neurosurgery, and trauma teams additionally. 3% hypertonic saline (HTS) was added to the Pyxis in the intensive care units to minimize time to HTS administration.

Results
As a work in progress, results are still pending after implementation. The order set and dotphrase were released in April 2019. The initiative will be sustained by ongoing education of PICU fellows, PICU APPs, and residents in neurology and pediatrics. 
Curriculum development for neurology residents during the epilepsy rotation

Yi Li,  Aditya Joshi, MD; Dominic A Hovsepian, Zahra Sadat-Hossieny, Kevin Graber, Scheherazade Le

Stanford, Department of Neurology
Background
Epilepsy and Electroencephalogram (EEG) comprise two of the milestones neurology residents are expected to be competent in by the end of residency. At Stanford, neurology residents receive didactic teaching from the epilepsy attending physicians, concentrated during the resident Wednesday lectures. They also have a mandatory one month rotation with the epilepsy division. The topics presented during Wednesday lecture are presented twice over the course of neurology residency. However residents may miss the lectures if they have to cover a service (such as consults), are on night float, or are on vacation.

Epilepsy fellows, who serve as senior trainees on the epilepsy team, will spend the most time with residents, and are in a unique position to supplement resident teaching. Fellow-driven teaching can cover the gaps left by residents being unable to attend lectures, and may afford more in-depth, focused coverage of practical topics.


Current State
The goal of this project is to develop a curriculum for the PGY2 residents, with the aim to cover the most common topics of EEG interpretation and management of epilepsy. The fellow-driven curriculum is intended to be complementary and parallel to the teaching done by the epilepsy attending physicians during lectures and while on service.

While it is known that neurology residents miss some Wednesday lectures, the exact proportion is unknown. A survey will be administered to the PGY2 neurology residents to assess how many lectures they have missed over the course of the academic year. Pre-intervention surveys will be used to assess resident comfort with consensus topics felt to be important to learn by the end of the epilepsy rotation. The goal of the curriculum will be to improve average resident perception of comfort by one point (out of a 5-point Likert scale) on each of the consensus topics, by the end of the academic year.


Analysis
The first-year epilepsy fellows developed a set of consensus topics that was felt to be important for neurology residents to know. Based on these consensus topics, a set of concise lectures, meant to be delivered for 15-20 minutes four days a week, was created. A 12-question pre-curriculum survey, with questions addressing resident comfort with the consensus topics was administered to each neurology resident. A post-curriculum survey including the same questions as the pre-curriculum survey, as well as 5 additional questions meant to assess the effectiveness of the teaching format, was given to the residents after they completed the rotation.

Results
A pilot study was performed over a 10 month period from 7/2018-4/2019 and included 10 neurology residents rotating through the Stanford epilepsy division. Feed analysis response was 100%, and post survey response was 90%. The residents uniformly felt more comfortable with the consensus topics after the fellow-driven curriculum was taught, and they uniformly agreed with the format of the curriculum. For the sustain plan, the second-year epilepsy fellow can take the responsibility to curate the curriculum, and the first-year fellows on the inpatient services will be responsible for teaching.
Improving delay in detection and treatment of intermittent focal and subclinical seizures

Samuel Spiegel, MD and Thuy Nguyen, MD; Katherine Werbaneth

Stanford, Department of Neurology
Background
Delays in recognition and treatment of seizures can lead to poor outcomes.1,2  When there is clinical concern for subtle or subclinical seizures, patients are evaluated by electroencephalogram (EEG). The timing of EEG recording review is dependent on the location and primary service caring for the patient. Few hospital rooms have remote EEG access. Real time review of EEG is only possible on the neurology floor (H1) and E2 ICU. Lack of access to real time review may lead to delays in seizure detection for patients admitted to beds without access to real time EEG review.

Current State
Current state analysis revealed that non-remote access EEG recordings are downloaded to a universal serial bus (USB) drives by epilepsy technicians for later review by the epilepsy team. Our pilot data reveals 25% of EEGs were recorded in non-remote access rooms. This raises concern for delayed detection and treatment of intermittent focal or subclinical seizures for admitted patients.

Our SMART goal is to increase real time review of EEGs of patients with clinical concern for subclinical seizures or focal status epilepticus in hospitalized patients by 10% in the next 6 months.

Analysis
We plan to perform a retrospective review of patients who were hooked up to EEG at Stanford from the last 6 months. EEG data will be reviewed for hospital location, timing of lead placement and timing of first interpretation to identify factors that may contribute to delayed detection and treatment of seizures. We will develop a multidisciplinary pathway for triage of focal and subclinical seizures.

Results
Preliminary review of EEG recordings over 1 month at Stanford reveals that about 25% of patients are hooked up to EEG remotely. We suspect that barriers of access to real time review of EEG contributes to delayed care for patients with ongoing seizures. We anticipate that implementing a protocol targeted at both nursing and resident education will reduce time to identification of focal and subclinical seizures. Rapid identification of seizures will allow for rapid treatment and improved overall outcomes. 
Continuous EEG for Seizure Detection in Neonates after Cardiac Bypass without Deep Hypothermic Cardiac Arrest

Rebecca J Levy, MD; Natasha Purington, Rajani Kaimal, Mehreen Iqbal, Elizabeth Mayne, Amanda G. Sandoval Karamian, Kathleen Ryan, Courtney J. Wusthoff

Stanford, Department of Neurology
Background
American Clinical Neurophysiology Society guidelines recommend consideration of continuous EEG (cEEG) for seizure detection after neonatal cardiac surgery involving cardiopulmonary bypass (CPB). Early reports described seizures in >20% of children after CPB, though recent work reports seizure rates of 3-12%. Previously determined risk factors include deep hypothermic cardiac arrest (DHCA), duration of bypass, and age.

Current State
It is our policy to provide cEEG to infants 3 months or younger after CBP, or at the discretion of the cardiology team. We sought to define seizure rates through prospective cEEG in a contemporary cohort of infants after CPB without DHCA in order to determine what risk factors predict seizure incidence best and to consider whether this would change cEEG protocols.

Analysis
Single center chart review of all cases from July 2017 through February 2019 where cEEG was performed for the indication of age ‰¤3 months old after CPB surgery. During this period, our clinical protocol was to perform cEEG for 48 hours as standard care in this group. Clinical and EEG data was collected for univariate and multivariate analysis.

Results
10/95 (10.5%) patients had at least one seizure on cEEG during monitoring; of those, 4 (40%) were in status epilepticus, all patients had subclinical seizures, and 2 (20%) also had clinical seizures. Median time between end of bypass and start of cEEG was 7.75 hours. Median time from end of bypass until first seizure was 30.9 hours, inter-quartile range 22-34 hours, range 8-52hr. Only 2/95 patients underwent DHCA and neither had seizures. We conclude that seizure prevalence in our cohort was analogous to other reports from large academic centers, despite minimal DHCA in our group. Among patients with seizures, the time to first detected seizure ranged from 8-52 hours after end of CBP, suggesting a window of highest yield for cEEG. We will expand the cohort to include additional infants, as well as older children added at discretion of cardiology team, to better understand seizure risk prediction factors.
A novel curriculum to improve neurology residents™ goals of care communication skills

Tarini Goyal, MD; Barbette Weimer-Elder, PhD, RN, Merisa Kline, MHA, Sylvia Bereknyei Merrell, DrPH, Carl A. Gold, MD, MS

Stanford, Department of Neurology
Background
Neurology residents care for patients facing life-altering and life-limiting diseases. It is crucial that they are proficient in the principles of palliative care, including counseling patients and their families regarding goals of care. However, national surveys have demonstrated that neurology trainees are inadequately prepared to care for patients at the end of life. In our own program, residents have expressed lack of confidence in their ability to lead goals of care conversations.

Current State
In 2018, we conducted a needs assessment with neurology residents to identify barriers to effective goals of care conversations, including patient, physician, and systems factors. We sought to develop a curricular intervention to increase neurology residents™ skill in leading effective goals of care conversations, as demonstrated by resident self-evaluation, direct observation of resident skills, and patient/family surveys.

Analysis
We used Kern™s six-step approach to curriculum development; the first two steps, problem identification and targeted needs assessment, were completed in our prior study. We next mapped specific goals and educational strategies based on key areas identified in the needs assessment. We developed didactics and roleplay exercises targeting fundamental physician-patient communication skills, to help residents overcome their stated concern that communication skills are inherent to the individual and cannot be taught. Next, neurology faculty led case-based exercises to guide residents through working with surrogate decisionmakers and learn tenets of prognostication. The goals of care curriculum was implemented in February 2019. 


Results
On preliminary analysis, the curriculum was well-received and residents rated their own skills in communication and prognostication more highly after the sessions. Our future analyses will include direct observations of resident-patient interactions as well as patient surveys to determine the effect of the curriculum on effective goals of care conversations. We plan to sustain the effects of this intervention by further developing the curriculum, as well as developing a faculty coaching program for individualized resident feedback. 
Reducing duration and cost of antihypertensive infusions in critically ill patients with intracerebral hemorrhage

Adam MacLellan, MD; Hans Ang, PharmD, BCCCP; Elizabeth Fox, CNS, AG-ACNP; Kassi Kronfeld, MD MS; Nick Murray, MD; Alicia Wilson, BSN; Karen Hirsch, MD; Zachary Threlkeld, MD

Stanford, Department of Neurology
Background
Guidelines indicate that intravenous (IV) antihypertensive infusions play a crucial role in the acute lowering of blood pressure after intracerebral hemorrhage (ICH).  However, they contribute to increased length of stay (LOS), may increase in-hospital morbidity, are more costly than oral formulations, and are administered solely in the costly intensive care unit (ICU) setting.  Previous studies have shown decreased LOS, and savings of $18,000/patient with optimized transition from IV to oral antihypertensives, though sustainability was poor.  We thus defined our problem statement as intravenous antihypertensive medications are critical in the treatment of patients with acute ICH, but are associated with significant pharmacy costs and extended (ICU) length of stay.

Current State
We retrospectively analyzed all patients admitted to our hospital with an ICD-10 diagnosis of non-traumatic ICH from 09/2017 to 11/2018, identifying 92 patients, 66 (71%) of whom required an antihypertensive infusion in the ICU.  Mean duration of infusion was 54.2 hours; of note, for this initial period of our project we chose to focus on a process measure that was easily measurable on a day-to-day basis.  Our SMART goal was thus to reduce mean duration of IV antihypertensive use from 54 hours to 30 hours in patients with intracerebral hemorrhage admitted to the ICU by March 22, 2019, the date of the CELT program graduation.

Analysis
Comprehensive chart review and focus group discussions with key stakeholders identified areas for potential exploration.  Fishbone analysis identified several common contributors to prolonged antihypertensive infusion duration, including: discordant blood pressure targets/orders, discordant documentation, insufficient as-needed medication orders, delay in oral medication initiation, and delay in obtaining enteral access for medication administration.  Key drivers were then identified: (1) consistent blood pressure goals for each patient, (2) clear interdisciplinary communication, (3) early initiation of non-infusion medication, and (4) consistent options for medication administration. 

Interventions introduced during the study period included: (1) interdisciplinary task force formation, (2) ongoing stakeholder and provider education, (3) updated Neurocritical Care service documentation to encourage medication reconciliation, (4) rounding checklist updated to encourage IV to oral transition, and (5) EPIC order debuted with clear blood pressure target, and goal to maintain or wean off infusion. 

Results
Over 5 months, mean antihypertensive infusion duration reduced from 54.2 hours (n=66) to 37.9 hours (n=29).  ICU LOS reduced from 3.35 to 3.0 days.  Proportion of cases with discordant blood pressure goal documentation reduced from 46.0% to 24.1%, while discordance in documented goals to actual orders reduced from 57.1% to 27.6%.  There were no significant increases in countermeasures (infusion restarts, ICU readmission, and AKI due to blood pressure lowering).  Extrapolating from finance data, and our baseline infusion duration and ICU LOS data, IV antihypertensive infusions cost ~$36/hour.  Our improvement suggested $17017 in estimated cost savings in 5 months.  ICU accommodation cost was approximated at $262/hour, for an estimated $63823 additional cost savings.  Sustainability measures include automating a biweekly data pull for task force review, engaging other related services (e.g. neurosurgery), introducing our new EPIC order into preexisting order sets, and providing a pharmacist-designed infusion weaning algorithm to bedside nurses.
Improving neurology patient knowledge of medications at discharge

Hilary H. Wang, MD; Phyo Aung, Leslie Bennet, Olga Goldberg, Victor Tamayo, Noah Wachtel, Brian Scott

Stanford, Department of Neurology
Background
When patients don™t understand the purpose or side effects of their medications, it has a negative impact on their quality of care, overall satisfaction, and safety. For this reason, communication about medications is one of the 9 key topics of the HCAHPS patients experience survey, but at baseline only 50-60% of patients leaving Stanford hospital agreed they understood the purpose of their medications. Our project focused on neurology patients, a unique population often suffering from cognitive impairment, and sought improve their knowledge of medications for their hospital stay.

Current State
As an interim outcome measure, we surveyed neurology patients admitted to H1, and asked for a confidence level (1-10) to their understanding of the purpose of their medications. By March 22, 2019, our goal was to improve the percentage of patients who report a confidence level of >8 from a baseline of 38% to >50% (relative increase of 25%).

Analysis
Medication teaching is a complex process involving multiple members of the care team (physicians, nurses, pharmacists, therapists), and it is made more challenging by frequent medication changes during an admission and variable health literacy/engagement within our patient population. Coordination and communication between frontline providers, with clear expectations of roles and responsibilities, as well as creation of personalized teaching materials to engage patients and their caregivers emerged as key drivers to our project success. We focused several interventions to improve staff communication (huddles, check-ins, and standardized .dotphrases for written messages within Epic) and focused our patient teaching (TOC pharmacy involved for all neurology patients).

Results
From January through March 2019, we piloted several early interventions, including post-discharge phone calls, new formats for 11am multi-disciplinary rounds and 2:45pm huddles, offering a phone number of post-discharge questions, and visual aids for medication teaching, and finally our key intervention of TOC pharmacy teaching. Our baseline of 38% of patients reporting high confidence in their understanding of medications jumped to 94% (a change in mean confidence score of 6.5/10 to 9.3/10). We plan to sustain our improved staff communication tools (huddles, .dotphrase) and TOC pharmacy teaching, and hope to expand our project to other units (starting with G1). 
Improving Scheduling and Efficiency in NeuroDiagnostic Lab

David Post MD; Sri Muppidi and Laurice Yang

Stanford, Department of Neurology
Background
The NeuroDiagnostic lab (NDL) at Stanford Neurosciences Health Center offers multiple neurological diagnostic procedures such as Autonomic testing, Nerve conduction and electromyogram (EMG) and electroencephalogram (EEG).  These multimodality procedures helps with diagnosis of both simple and complex neurological syndromes and patients need quick access to these procedures to help make a final diagnosis in a timely manner. 

EMG testing is one of the main testing methodology offered in NDL. Currently, there is an average of 4 weeks to the next new patient open slot and yet, there are several open new slots per week and a relatively low cancelation rate of 10% per week. Since these studies require active faculty involvement we are only able to offer about 40 studies per week.

This discordance between open slots and prolonged wait times is likely due to inadequate reminder to patients; need to have patients prepared for testing, lack of wait list to quickly schedule patients if there are openings in schedule, delay in scheduling patients because of need for insurance clearance and inadequate scheduling staff.


Current State
The goal of project is to schedule 80% of new referrals for EMG testing to NeuroDiagnostic lab within 3 weeks of referral.

Analysis
We will analyze the time to first phone call to patient and also time to scheduling for each patient after an EMG request order is placed in EPIC. Additionally, we also will review weekly list of number of patients waiting to be scheduled as a marker of improved efficiency.

We have implemented multiple interventions to help improve overall efficiency in NDL. These were 1) increase in number of scheduling staff, 2) change in insurance prior-authorization process for certain insurance carriers, 3) weekly and monthly reports to physicians to improve scheduling and to utilize the neurodiagnostic space to the maximum extent possible, 4) cleared up redundant list of patients in the to be scheduled list. Additional interventions currently underway include 1) request physicians to increase the number of testing slot for each session by 1 (20%) and 2) request physicians to add extra half days in the NDL.


Results
Results of the intervention are currently being analyzed and will be presented.

Multiple interventions have been implemented and sustained for the last 3 months. We are currently in the process of recruiting additional faculty to improve number of testing slots available each week. 
Reducing Stanford ED Utilization for Breakthrough Seizures in Patients with Epilepsy

Erica Von Stein, MD or Aditya Joshi, MD; Erica Von Stein, Aditya Joshi, Katherine Werbaneth, Dominic Hovsepian.  Faculty Mentor: Laurice Yang

Stanford, Department of Neurology
Background
Breakthrough seizures in patients with known epilepsy account for approximately 83,000 emergency department (ED) visits annually. The most common cause for breakthrough seizures is sub-therapeutic anti-epileptic drug (AED) levels, which may be from non-adherence, inappropriate dosing, or drug-drug interactions. Other common causes include illness, toxic/metabolic conditions, or lack of abortive medication. Once a patient has a breakthrough seizure, factors such as communication issues with healthcare providers, lack of access to care, and poor education regarding breakthrough seizures, may lead them to present to the ED.  Prior studies have shown that interventions including providing proper abortive therapies, improving access to urgent neurologic care, developing seizure action plans for patients, and providing assistance for those at highest risk, may decrease unnecessary ED visits.

Current State
To analyze the current state, we limited our study population to patients who have established care at the Stanford Comprehensive Epilepsy Center and who also presented to the Stanford ED for breakthrough seizure from December 2017 to December 2018.  A total of 120 patient visits met this criteria, and were reviewed for primary reason of breakthrough seizure and disposition of the patient.  Our SMART goal is to decrease annual rate of Stanford ED visits for our established Epilepsy patients from 120 to 96 (20%), and sustain this reduction for at least 1 year after implementation of intervention. 

Analysis
Retrospective review of these 120 patients revealed that the majority were discharged from the ED.  Only 29% (n=34) were admitted under inpatient or observation status. Non-adherence to medications accounted for only 10% of visits (n=12) and sub-therapeutic antiepileptic levels were found in only 8% of patients (n=9). The most common reason for presenting to the ED was refractory epilepsy in 39% of patients (n=46).  Key drivers for ED presentation included access to patient services and unclear contingency plan for a breakthrough seizure. Action plans are common for pediatric conditions, though less so in adult practice. A formalized seizure action plan will help address the gap in management of a breakthrough seizure. Furthermore, these guidelines will help to reinforce the notion that breakthrough seizure can be managed appropriately in the outpatient setting.

Results
We plan to implement a seizure action plan to aid patients and their family in home management of breakthrough seizures and hopefully reduce unnecessary ED visits. The form will be completed together by the attending epileptologist and the patient, and the patient will keep the form with them. Results are pending, and we plan to continue the project during the next academic year. The seizure action plan can be sustained by completing the form with every new patient with epilepsy and updating the form during each subsequent visit prior to leaving the clinic encounter. In the future, the form will be computerized, avoiding the necessity of repetitively writing information (such as seizure type) that typically does not change.
Developing an Outpatient Neurology Resident Curriculum and Entrusted Professional Activities (EPAs)

Katherine Werbaneth, MD; Prateek Thatikunta MD, Teresa Wu MD, Sarada Sakamuri MD, Rebecca Miller-Kuhlmann MD, Scheherazade Le, MD

Stanford, Department of Neurology
Background
Although the majority of the clinical practice of Neurology is outpatient-based, the Stanford Neurology Residency formal educational curriculum is predominantly inpatient-based. Bedside teaching occurs often, but direct observation with constructive feedback of resident history-taking, neurologic exam skills, and patient counseling skills is limited. Neurology Residents attend formal, weekly 3-hour didactic sessions that are lecture-based, but research suggests that knowledge retention after the traditional pedagogical approach is limited [Bonnes]. Those lectures are planned months in advance and lack the flexibility to be tailored to and thus reinforced by immediate resident clinic experiences and cases”a powerful educational opportunity that is currently untapped. An integrated diversity curriculum is also lacking and not taught in the educational series, despite that when surveyed, most neurology program directors believe training in culturally-responsive care is important [Rosendale].
All Stanford Neurology Residents have a weekly continuity clinic, with longitudinal supervision and mentorship by a core group of faculty members. Building a routine outpatient curriculum is an opportunity to address two gaps in the current training curriculum: (1) The tools and methods to diagnose and manage outpatient neurological cases and (2) Strategies for providing culturally-responsive care. The limited teaching time available during the residents™ busy continuity clinic can be optimized by creating a formalized, flipped-classroom curriculum.

Current State
Our primary aim is to create outpatient, case-based, learner focused, flipped classroom teaching modules that Neurology Residents will discuss as a group with their peers and continuity clinic preceptors on a regular basis.
Our secondary aim is to develop and assess neurology residents in the outpatient setting through an apprenticeship model of entrustable professional activities (EPAs) defined by the American Academy of Neurology (AAN) and ACGME [Horak; Supplemental Table e-1 through e-5].

Analysis
Implement flipped classroom discussion regularly into Resident Continuity Clinic, average of one session per month with a goal of 10 new cases of flipped classroom online modules.
Learning can be assessed through pre- and post-tests for the online modules.
Learner complete evaluations of the modules and EPAs.
Faculty give feedback of use of EPAs versus current ACGME Clinical Core Competencies to assess residents.


Results
Continue regular outpatient teaching with flipped classroom modules, apprentice-ship coaching model for achieving EPAs in Neurology Resident Continuity Clinic. 
These modules could be easily adapted to other Neurology Residency Programs nationwide. 
Improving Resident Communication with Patients about Medication Side Effects

Shefali Dujari, MD; Hilary Wang, MD; Ruth Hua, Anahita Dua, Benjamin Weia, Justin Slade, Lisa Shieh

Stanford, Department of Neurology
Background
Understanding the benefits and risks of medications is important for patients to take an active role in managing their own health, to facilitate communication with their providers, and to promote adherence. Additionally, according to post-discharge assessments, a patient™s understanding of their medications is directly related to their overall satisfaction with their care.

Current State
65.5% of patients currently state that they receive information about possible side effects before starting a new medication. On self-reports, residents state that they counsel patients on medication side effects about 54% of the time. Our goal is for residents to increase their rate of counseling from 54% to 75% by December 2019.

Analysis
The Resident Safety Council created a group of residents and nurses to focus on improving resident communication with patients about medication side effects. Within this group, an internal analysis was done to identify reasons for lack of education. A resident survey was then created and disseminated among Internal Medicine, Neurology, General Surgery, and Psychiatry residents to test our hypotheses and better understand the key drivers.

Results
A total of 67 responses were obtained, with a 28% response rate. Residents self-reported that on average they provided education regarding medication purposes and side effects before starting a new medication in the inpatient setting 54% of the time. While the majority of residents felt that it was important to teach patients about their medication side effects, the majority of residents felt that they did not have the time to counsel their patients. 22% of residents responded disagree or neutral when asked if they had the knowledge to counsel their patients about medication side effects. Residents generally responded that there was no one member of the team who should be responsible for education, and rather that it was a shared responsibility. 4 residents felt that it was exclusively the role of the nurse and/or the pharmacist to provide counseling.

Based on this data, our proposed intervention includes developing medication side effect EPIC dotphrases for commonly prescribed medications which can be included a patient™s After Visit Summary (AVS). We will review recent After Visit Summaries to understand the baseline rate of side effect education included in the AVS and to obtain a list of the most commonly prescribed medications by service. After implementation of our intervention, we will again review the AVS, as well as perform a repeat assessment of resident™s reported rate of side effect counseling and review the HCAPS data.
A Medical Legal Curriculum for Residents to Increase Physician Comfort in Patient Interactions

Kate Therkelsen, MD; Laurice Yang

Stanford, Department of Neurology
Background
Medical ethics has become a requirement of medical education; however, medical law is not required and is a gap in current residency training. In order for residents to develop into well-rounded, successful physicians, a medical law background is essential to better prepare for the multitude of difficult situations that arise to better serve patients.

Current State
There is not a current formalized medical legal education for residents. Our goal was to create a medical legal curriculum instructing resident physicians how to 1) understand malpractice law 2) recognize and manage medical errors 3) more effectively and compassionately communicate 4) obtain informed consent and 5) utilize resources available for physicians.

Analysis
The curriculum was created in multidisciplinary approach with Stanford™s risk-management department. Residents completed pre and post session surveys to assess their comfort and knowledge with medical legal issues.

Results
Pre and post session surveys were completed by internal medicine (n=23) and neurology (n=13) residents. Before the initial session, the majority of residents rated their comfort level regarding the basic concepts of medical law, as well as their comfort regarding handling unanticipated outcomes, as poor or fair (78%, 18/23, SE 0.09 for both). Following the session, the majority of residents agree or strongly agree that they were more comfortable with malpractice law (73%, 16/22, SE 0.09), that they learned new ways to handle unexpected outcomes (82%, 18/22, SE 0.08).  They also felt more comfortable mitigating risk in their own daily practice (77%, 17/22, SE 0.09). A large majority would want the lecture next year and would recommend the curriculum to colleagues (91%, 20/22, SE 0.06 for both).
A formal education curriculum allows residents to ask questions and learn strategies on how to navigate the medical legal system in a low-pressure practice environment. We plan on further developing this curriculum as well as expanding to other specialties. 
3D printed models of brain and intracranial electrodes for epilepsy education and surgical planning

Pooneh Memar Ardestani, MD; , Kyle Gifford, Kristen K Bogart, Shannon Walters, Michael Zeineh, Dominik Fleischmann, Babak Razavi 

Stanford, Department of Neurology
Background
Approximately …“ of patients with epilepsy have seizures that are resistant to medication. Accurate pre-surgical workup, which may include imaging and intracranial recordings, is an important factor in the success of surgery in medication refractory epilepsies (MRE). In recent years, 3D printing has gained substantial grounds in many fields, including medicine. Practical applications include: 1) teaching, practicing and planning surgical techniques; 2) designing implantable prosthetics and 3) development of biological tissue-engineered implants. Furthermore, the possibility of customized 3D printed models makes them a perfect tool for complex case and a natural fit for epilepsy. Therefore, we propose using 3D printing to improve epilepsy care and education in this population.


Current State
Intracranial electrodes (IEs) can be used to identify the epileptogenic zone (EZ) for surgical treatment of MRE. Optimal placement and actual anatomical localization after implantation is critical. Thus, pre-operative MRI and a post-implant CT data are fused to identify the location of each electrode contact. We take this one step further and generate a 3D CAD model of the segmented structures to facilitate a personalized 3D printed brain with or without the IEs, or other anatomical regions of interest.


Analysis
The source imagine data can be used to produce multiplane reformations and 3D views of the anatomy, which is then transformed into a physical model. The workflow consists of 4 major steps: 1) Image acquisition: pre-implantation MRI, post-implantation CT, 2) Image post-processing: (CT / MR) segmentation and registration, 3) 3D design of CAD model, 4) 3D printing. TeraRecon Aquarius was used to segment and fuse the MR and CT images. The brain and electrodes were exported together as a single stereolithographic mesh file. Meshlab, Autodesk™s Netfabb, and Formlabs Preform softwares were used for building and refining the 3D model. They were printed using Flexible Resin V2 with Formlabs Form2.


Results
Three example models of the brain and ICEs were printed in Stanford 3D and Quantitative Imaging Laboratory. (1) Elastomeric: printed as two separated hemispheres in actual size. This is a simple model that is primarily for planning subdural grid placement, depth electrode data interpretation and surgical planning. (2) Rigid transparent: printed 60% of actual size in 3 pieces which were held together using magnets. The electrodes are printed with a dark material to highlight their path (some curved). The ventricles and subarachnoid space are blue. This model demonstrates well how the electrodes traverse through different tissues and positioned relative to different structures. (3) Rigid transparent: printed in actual size for 1 hemisphere. It includes colored white matter tracts derived from DTI, demonstrating their location relative to the lesion of interest. The model guided subdural electrode placement in the OR. 3D printing has become standard in our clinical practice: models can be ordered directly in Epic. We hypothesize that 3D printed models can help improve surgical planning and education in epilepsy as well as other specialities. As future direction, we plan to define metrics that capture improvements in surgical workflow and education as a results of using 3D printed models.
Quantifying the burden of unfilled clinic appointment slots created by late-notice cancellations

Hilary H. Wang, MD; Nicole Batino, Laurice T. Yang

Stanford, Department of Neurology
Background
Late notice, patient-initiated cancellations are an underrecognized cause of limited access to care. While the no-show literature is robust, late cancellations are a difficult metric to capture and lead to a paucity of data on their prevalence and impact.  Open slots due to late cancellations are difficult and time consuming to fill and are often left unoccupied. They lead to fewer total available appointments and increased wait time to the next open slot. This quality improvement (QI) project aims to 1) quantify the burden of late cancellations at a large academic institution™s neurology outpatient clinics 2) evaluate their key drives and 3) begin to intervene

Current State
By July 2019, we intend to decrease the percentage of open unfilled appointment slots at Neurology Outpatient Clinics from by 20% (from 14.5% to 11.5%).

Analysis
Our team extracted 6 months (Jan-June™18) of appointment booking data from the electronic health record (EHR) and our patient scheduling call centers, with slots retroactively subdivided into filled, cancelled, cancelled late, no-show. We also observed our cancellation call workflow and introduced an optional end-of-call survey to directly query our patients on factors leading to them not keeping their appointment. We introduced

Results
From our 6-month data, an average of 14.5+0.9% of clinic appointments (mix new/return) were left unfilled due to late cancellations This is as compared to the 5 + 0.4% average no-show rate in the same timeframe. Late cancellations represented almost half (46%) of all cancellation calls made. Our survey revealed many unalterable factors for cancellation, but also found transportation, inconvenient technology for cancellation (e.g. calling during business hours) as pain points, which our early interventions focus on.
Creating an Institutional Protocol for MRI in Patients with Legacy Cardiac Rhythm Devices

Collin J Culbertson, MD; Max Wintermark, Paul Wang, Angela Tsiperfal, Teresa Nelson, Connie Lund, Carl A Gold

Stanford, Department of Neurology
Background
Problem statement: Patients of Stanford Hospital and Clinics (SHC) with older pacemakers or implantable cardioverter defibrillators (ICDs) could not have MRI scans, leading to delay in diagnosis and treatment.

Background: MRI is an invaluable tool for diagnosing and managing a variety of medical conditions, especially neurological conditions. Patients with pacemakers or ICDs have historically been denied MRI unless their device system is labelled MR-conditional (first approved in 2011 and represents only ~25% of active devices). However, recent large studies showed that patients with nonconditional or legacy devices can in fact undergo MRI safely. Despite this data, many referral centers including SHC have not developed protocols to accommodate this practice.

Current State
Current state: While SHC has an established policy for MRI scans in patients with conditional pacemakers or ICDs, no such protocol has existed for patients with legacy devices. Creation of a protocol requires multi-disciplinary involvement by our Cardiac Electrophysiology (EP) and Radiology teams to ensure patient safety. While the number of anticipated legacy scans was previously unknown, a current state analysis by the authors found approximately 2.5 patients per month presented in 2017 with a neurological disorder necessitating MRI but could not undergo the scan due to a legacy device. Similar estimates for patients with other conditions requiring neurological MRI (such as CNS malignancy or spine disease) are not known but could be explored.

SMART goal: To improve timely diagnosis and treatment of patients with neurological disorders and legacy devices by launching a protocol for inpatient MRI at SHC starting July 2019.

Analysis
Intervention: Our intervention is implementation of a protocol for MRI in SHC inpatients with legacy devices. This protocol is adopted from the existing version for MR-conditional systems, with some additional safety measures and closer involvement of the EP team.

Analysis: We will create a prospective database of all inpatients with legacy devices undergoing brain or spine MRI. Three pilot cases have been completed and will be presented.
- Primary outcome: change in diagnosis/treatment based on MRI result.
- Secondary outcomes: change in device parameters (main safety outcome), time from order to scan, length of hospital stay.

Results
A committee comprised of the departments of EP, Radiology, Neurology, and hospital administration will use the aforementioned results to further revise and improve the protocol if necessary, including considerations of scanner availability and workflow. Ideally, the policy can eventually be extended to selected outpatient settings.
Improvements in clinical, laboratory, and Epic workflow to eliminate errors in diagnostic lumbar punctures.

Nina Bozinov, MD; Lucas B. Kipp, MD2, William Flores, CLS4,  Josephine Cabrera Aranda, CLS5, Rahul Guli6, Victoria Fitch4, Marites Lazzareschi, RN, CRRN, ACM3, Sujinder Johal, CLS7,  Esterlita Bengil, CLS7, Pablito Ang, CLS7, Isabelle Yi, RN3,  Stephanie Ogamba, MS-HCA3, Rocio Serrano, MA3, Dana Gonzales, BSN, RN, CNOR3, Amanda Johnson, RN3, Marlianne Asuncion, Supervisor3, Olga F. Goldberg, MD2

Stanford, Department of Neurology
Background
Lumbar Puncture (LP) obtains cerebrospinal fluid (CSF) to aid diagnosis of neurologic conditions. Collection is performed bedside in Neurology Resident Continuity Clinic (RCC) or by a radiologist with fluoroscopy guidance. CSF tests often require special processing including: simultaneous serum collection, send outs to specialized laboratories nationwide, and large volume cytology. Four CSF tubes are sent to the hospital lab, aliquoted, then processed in different laboratory sections. The referring provider orders CSF tests of interests and PT/INR (within 30 days of procedure) prior to LP scheduling. While considered a low-risk procedure, complications can include post-LP headache, CSF leak requiring blood patch, back pain, infection, bleeding, nerve damage, and brain herniation.

Current State
Over 100 lumbar punctures were performed on neurology patients in RCC and 80 in radiology over the past 12 months. Due to incomplete CSF processing, some patients required a repeat procedure.

SMART Goal: By September 1, 2019, eliminate cancelled/not-processed CSF tests that lead to delays in neurologic diagnosis or repeat lumbar punctures.


Analysis
After lack of sustained improvement in workflow utilizing electronic and telephone inquiries, an in-person working group was created. Relevant stakeholder representation included: Epic information analysts, clinical management, nursing, patient care coordinators, main and ambulatory clinical lab specialists, and physicians.

Cases of cancelled or not-processed CSF labs were reviewed by the group to determine the underlying causes, including:
(1) Non-functional LP referral orders in EPIC
(2) Complex work flow for referring patients to RCC
(3) Prerequisite PT/INR not completed
(4) Failed procedures due to difficult anatomy inappropriately referred to RCC
(5) Specialized CSF tests not processed correctly:
a. Serum collection not done for oligoclonal bands comparison.
b. Certain orders not visible to clinic and laboratory staff
c. Insufficient sample for cytology due to small aliquot passed to pathology

Planned EPIC changes include a new LP order set addressing referral/scheduling issues, pre-requisite/simultaneous serum studies, and CSF orders. A signed and held feature will allow orders to be placed by the referring physician in advance, be visible to clinic and lab staff, and be released at the time of procedure.

A laboratory and clinic documentation guide was also created:
(1) Specifying minimal CSF volumes per test
(2) Designating each test to 1 of 4 tubes
(3) Separating cytology tube prior to aliquoting


Results
We are still in the implementation stage. Once Epic changes are implemented, there will be provider education on the new ordering process/smart set. We plan to look at the rate of cancelled/not-processed labs, repeat lumbar punctures, missing serum studies in the 6 months following our implemented changes. The expectation is that this new work-flow will simplify the process and minimize confusion for the ordering providers, residents performing lumbar punctures, patient care coordinators, SNHC ambulatory lab, main clinical lab, and most importantly, our patients. 
Improving door-to-needle times in acute stroke care with pre-hospital intervention

Lironn Kraler, MD; Eric Bernier MSN RN, Nirali Vora MD

Stanford, Department of Neurology
Background
Rapid administration of thrombolytic therapy in acute stroke is associated with better stroke outcomes. Last-known-well (LKW) times are used by providers to timestamp stroke onset and mark the start of a 4.5 hour window during which thrombolytic therapy can be safely administered. Beyond 4.5 hours, the therapeutic effect is overshadowed by hemorrhagic complications and thus it is imperative to confirm an accurate LKW to reduce these risks.  The door-to-needle (D2N) times are important for capturing the time to initiation of thrombolytic therapy from when the patient arrives at the hospital.  Delays in D2N times are encountered when there is difficulty confirming an accurate LKW.  A current goal in Santa Clara County is a D2N time of less than 30 minutes in greater than 50% of cases. From 9/2018 to 2/2019, at Stanford we are reaching this goal only 41% of the time.

Current State
There are often discrepancies between EMS reported LKW, and LKW ultimately decided by the stroke physician and we predict that this contributes to delays in initiating thrombolytic therapy. There are several factors that compound this delay: witnesses/family are advised against traveling with the patient in the ambulance and it is not part of standard EMS protocol to collect a contact™s phone number. This makes it challenging and time-consuming to confirm collateral history once the patient arrives to the ED.
Using the county GFAST database, we will collect LKW times recorded by EMS providers and compare LKW times as charted by the stroke physician at Stanford over the past 6 months. We will assess whether there are discrepancies between the two reported LKW times, and analyze whether this correlates to delays in D2N.


Analysis
If a correlation exists that demonstrates discrepancy of LKW times between EMS and physician providers, then we plan to present this data to the county EMS systems. Interventions proposed to close the gap in the discrepancies will be: provide educational session for EMS service providers on best strategies for assessing accurate LKW, propose changes to EMS protocol to advise family to accompany patient in ambulance to assist in collateral history confirmation, and if unable to accompany EMS will obtain phone number from a reliable witness.

Results
Results are pending review of data. Plan for sustainability of intervention will be to reassess delays in D2N and correlation with discrepant LKW. From the time of intervention, will reassess every other month for three cycles then plan on biannual review with county EMS services and Stanford stroke team. 
Improving Bedside Care of Pediatric Epilepsy Monitoring Unit Patients

Sara Pavitt, MD; Alison Carley, RN and the LPCH Brain and Behavior Nursing Group, Gustavo Chavarria-Ortiz MD, Brenda Porter, MD Juliet Knowles, MD, PhD

Stanford, Department of Neurology 
Background
Lucile Packard Children™s Hospital (LPCH) has a dedicated epilepsy monitoring unit (EMU).  As part of standard pre-surgical workup, during Phase 1 studies, daily anti-seizure medications are routinely weaned/discontinued, for the purpose of allowing seizures to be recorded on EEG. However, such patients are at risk for status epilepticus and in rare cases, death. Thus, it imperative that we provide prompt, standardized and thorough safety measures during seizures (referred to as seizure first aid). This includes turning the patient on his/her side, assessment of vital signs and providing appropriate rescue medications. Neurologic assessment during/after the event further aids epileptologists in surgical evaluation. However, the LPCH EMU lacks a standardized protocol for bedside seizure management, leading to variation of practice that could impact patient safety.

Current State
The current state of bedside care for acutely seizing patients admitted to LPCH EMU for Phase 1 monitoring was determined by review of individual video EEG records. All video EEGs from patients admitted for Phase 1 studies (n=21) from 11/2017-06/2018 were examined. Nursing was notified in 12 seizures (9 generalized and 3 focal) which were fully reviewed. Complete seizure first aid was provided in 42% seizures. 17% were associated with a complete, relevant neurologic assessment. On average it took 3 min to begin to obtain vital signs and 3.3 minutes to begin neurologic assessment.

SMART GOAL “ Increase complete seizure first aid and neurologic assessment from 40% to 80% for patients admitted to LPCH EMU for phase 1 monitoring by February 2019.


Analysis
We developed a standardized bedside seizure management protocol based on the National Association of Epilepsy Centers guidelines. Nurses were taught the protocol in September 2018 in person and/or by videotaped teaching session. Multimedia interventions included a newsletter and posting of protocol to all bedside EEG machines.

Measures:
Process measure “ Use of seizure management protocol
Secondary measures “ Time to seizure first aid and time to neurologic assessment, Nursing confidence for caring for seizure patients
Balancing measure “ Nursing satisfaction


Results
Video EEGs were again reviewed from 11/2018 “ 2/2019 which included 13 patients admitted to the EMU for Phase 1 monitoring. Nursing was notified in 10 seizures (6 generalized and 4 focal) which were fully reviewed. Complete seizure first aid was completed in 90% of seizures over this period (p=0.03). 90% were associated with a complete, relevant neurologic assessment (p=0.003). The protocol was followed in 90% of all seizures. The average time to seizure first aid decreased to 1.2 min (p=0.02) and average time to neurologic assessment decreased to 1.6m (p=0.03) Results on nursing satisfaction will be available by the end of April.

We successfully implemented a standardized bedside seizure care protocol which increased the frequency of complete seizure first aid and neurologic assessment. It also led to decreased time to complete vital signs and neurologic assessment. This protocol has improved both patient safety and Phase I study utility in the LPCH EMU. 
Improving the Inpatient MRI Experience Through Pre-procedure Patient Education

Sara Pavitt, MD; Amanda Sandoval, MD, Prateek Thatikunta, MD, Theodora Bruun, Max Wintermark, MD, Carl A. Gold, MD, MS

Stanford, Department of Neurology 
Background
Problem statement: Patients undergoing first-time MRI feel unprepared for the experience of the test leading to poor patient satisfaction and potential time delays.

Background: To capture high-definition magnetic resonance images, patients must lay completely still for prolonged periods inside a tube with a width of 24 inches and a sound level of up to 110 decibels”as loud as a rock concert. This environment has led to MRIs being feared by patients young and old. Within the pediatric population, general anesthesia has been commonly used for school-aged children who cannot lay still. However, recently many children™s hospitals have started to offer kid-friendly MRI scanners and utilize Child Life Specialists to prepare patients to reduce the need for anesthesia, increase patient satisfaction, and improve MRI throughput. Critically thinking about the experience of MRIs for adults may reveal similar opportunities for improvement.

Current State
The current state of inpatient MRI practices was determined by patient and provider interviews and review of patient experience surveys.

We found that hospitalized patients tolerated MRIs, but many felt the experience was difficult due to claustrophobia and/or inability to hear audio prompts during the scan. Patients were typically informed by nurses about the MRI process while completing the MRI screening form, but nurses do not receive formal training about the MRI process at SHC. In satisfaction surveys, patients positively commented about the friendliness of the staff. However, they were dissatisfied with the inability to hear or absence of verbal cues which caused confusion and anxiety during the scan. They also commented about the intense volume of the MRI and offered suggestions such as music to relieve anxiety and claustrophobia.

To understand the MRI process in its entirety, a Swimlane diagram was developed outlining each step of the process. Based on this process map, we found if patients were not able to tolerate the scan there would be a large delay while obtaining anxiolytic medications. This could delay scans for the individual patient and for subsequent patients, potentially impacting patient satisfaction and throughput.

SMART GOAL:
To increase inpatient MRI patient satisfaction and decrease time of individual MRI scans in patients who are undergoing first-time MRI in the Clinic Decision Unit (CDU) at SHC by December 2019.

Analysis
Our planned intervention will be to develop an audio/video preparatory patient guide to be distributed by nurses to patients. We will seek input from the patient advisory council on the content and format of the video and will prepare versions in multiple languages.

Measures:
Primary Outcomes: Patient anxiety levels around MRI, MRI scanning times for individual patients
Process Measure: Use of the patient guide
Balancing measures: Patient satisfaction


Results
To assess our outcomes, we plan to develop a patient satisfaction survey with guidance from Patient Experience. This will also assess for patient anxiety levels. We will use times stamps of completed MRIs to determine length of scans.  

Based upon the results of this first cycle we will work to develop further interactive cycles to target our primary outcomes.
CREATION AND IMPLEMENTATION OF A STANDARDIZED TOOL TO IMPROVE INPATIENT CONSULT COMMUNICATION

Sara Pavitt, MD; Anne McHugh, MD, Kevin Chi, MD, Kim Hoang, MD, Elizabeth Lippner, MD, Kiran Mudambi, MD, Katelyn Saarela, MD, Hannah Basset, MD, Rachel Goldstein, MD, Jennifer Tsai, MD, Nivedita Srinivas, MD

Stanford, Department of Neurology 
Background
Communication failures are the most common root cause of patient safety events.  Consults provide a significant opportunity for miscommunications to occur; however, providers at all levels receive little education on how to effectively convey consult requests. Based on literature review and needs assessment of pediatric residents and fellows at Stanford, we identified essential consult elements (ECE) for optimal consult communication which include consult urgency, consult question, background, differential diagnosis and proposed plan.

Problem Statement:
There are significant variations among inpatient consult communication which lead to miscommunication and has the potential to affect patient safety.



Current State
54 consult requests were audited in real time by pediatric subspecialty fellows at LPCH from Feb-June 2018. Our baseline data showed that only 11% of consults included all ECE. At our institution, 98% of residents (n=57) and 95% of fellows (n=55) reported miscommunication around consults, 85% of which led to perceived patient safety errors. Only 51% of consult audits had clear/very clear consult question.

Aims/SMART
1)Develop a standardized consult communication tool that includes all ECE
2) Increase consult requests containing all ECE by pediatric residents from 9% to 40% within a 6-month period



Analysis
A consult communication tool containing ECE was developed using modified SBAR (Situation, Background, Assessment, Recommendations) and taught to pediatric residents from July-Aug 2018 through a series of workshops. Multimedia interventions (flyers, screen savers, name badge cards) were disseminated throughout the hospital. During the initial PDSA cycle, residents and fellows were queried to inform subsequent cycle changes, such as targeting faculty engagement, engaging fellows with consult feedback, and monthly reminder emails. An EMR intervention has now been developed to integrate the intervention into standard work.

Measures:
Primary outcome measure - Percentage of inpatient consult requests containing all ECE
Process measure - Adherence to the modified SBAR tool
Secondary outcomes - Clarity of consult question, perceived miscommunication, and perceived effects on patient safety.
Balancing Measures - Resident and fellow satisfaction


Results
Post intervention, from Aug-Feb 2018, 87 consults were audited. Consults including all ECE increased from 11% to 61%. SBAR was used correctly in 61% of consultations. With SBAR, 96% of consult questions were clear/very clear compared to the 52% baseline (p<0.01). 55 (63%) residents and 55 (71%) fellows completed post intervention surveys. Fellows perceived miscommunication frequency as most of the time™ 50% during the pre-intervention phase compared to 15% (P<0.01) during the post intervention phase. 76% of fellows and 71% of residents felt that SBAR positively/very positively affected patient care. Prior to the intervention, 51% of fellows reported being satisfied/very satisfied with communication which increased to 100% with the use of SBAR (p<0.01).  Prior to the intervention 96% of residents reported being satisfied/very satisfied with consult communication compared to 92% in the post intervention phase (p=0.32).

Use of the SBAR consult communication tool increased inclusion of essential consult elements from 11% to 61%, improved consult question clarity, decreased the perception of miscommunication frequency, and was viewed by users to positively affect overall consult communication and patient safety. To ensure sustainability we have embedded SBAR teaching into yearly orientation for both residents and fellows at LPCH and into a rounding checklist.
Reduction of Wait Time from Central Check-In to Destination in Outpatient Neurology Clinic

Ashley Rawls, MD; Laurice Yang, Christopher Locke, Jong-Mi Lee, Morgan Pennington, Amy Alcantara

Stanford, Department of Neurology - Movement Disorders 
Background
Prolonged patient wait times in outpatient clinic can lead to delays for those receiving and providing care.  At the Stanford Neuroscience Health Center, the average wait for the patient from central registration to clinic check-in was 22 minutes within the past year.   Increased wait times decreases overall patient satisfaction and decreases clinic efficiency.

Current State
Our goal is to reduce patient wait time from central registration to clinic check-in by 10 minutes from the average of current check-in of 22 minutes.  This goal will be completed by September 2019.

Analysis
Our analysis would include calculating average time from central registration to check-in, patient satisfaction scores, and provider satisfaction scores.  Our current run chart shows a monthly average check-in time from March 2018 to February 2019 is 22 minutes, with standard deviation of 1.6 minutes.  We have also written out a process map from patient's flow from central registration to clinic check-in.  We have identified crucial steps in the process flow and have done 2 Gemba round as part of our analysis. Possible interventions include streamlining paperwork prior to patient arrival, sending a co-pay reminder prior to appointment, and valet parking.

Results
At that time, we have not implemented our interventions.  However, we have put together a team (NP, MD, clinic manager and front desk manager) who will engage in the RITE program from May 2019 - Sept 2019 to further analyze key drivers and implement interventions for this project. 
Improving the Peri-Operative Imaging Workflow for Pediatric Epilepsy Surgery at Lucile Packard Children™s Hospital

Jonathon J. Parker MD PhD; Brenda Porter MD PhD, Chris Lee-Messer MD PhD, Emily Spelbrink MD PhD, Juliet Knowles MD PhD, Fiona Baumer MD, Liz Tong MD, Kristen Yeom MD, Shannon Walters, Kristen Bogart, Kyle Gifford, and Gerald Grant MD

Stanford, Department of Neurosurgery
Background
Background:
Pediatric epilepsy surgery is a multi-disciplinary enterprise which requires collaboration of neurosurgery, neurology, radiology, social work, nursing, and the allied rehabilitation specialties. Critically, integration of disparate streams of radiologic data (anatomic, functional, and connectome imaging) is required for surgical planning, provider discussion, and patient/family education. Further, three-dimensional modeling, visualization, and manipulation of this data is now required for surgical planning.


Current State
Prior State
Each patient undergoes several imaging modalities including anatomic MRI (ie T1/T2), functional MRI, magnetoencephalography (MEG), and patient-specific tractography as part of pre-surgical workup. Previously, these images were accessed via the PACS system in sequential fashion, reviewing each set of images separately. Additionally, fusion of images of implanted intracranial electrodes (CT) with brain anatomical images (MRI) was performed on a single surgeon-run laptop, not easily shared with the team members via PACS.


Analysis
Intervention:
We incorporated MEG imaging directly into the SEEG planning workflow using the ROSA platform software. Synaptive Bright Matter software was utilized to fuse these imaging modalities enabling 3D preoperative modeling to design safer surgical approaches. The Stanford 3D lab has established a SEEG Workflow initiated after each pediatric surgical case which involves 3D CT/MRI fusion and production of a customized 3D print of the surgical implantation. Lastly, a joint weekly Epilepsy Surgery clinic staffed by epileptologist and neurosurgeon has streamlined discussion of medically refractory patients to expedite surgical management.


Results
Results/Sustain Plan:
Beginning in July 2018, 18 patients have undergone SEEG utilizing this new imaging workflow. Pediatric neurosurgery team members have been trained on imaging upload and fusion protocols for ROSA/Synaptive planning. The Radiology Department 3D lab has established a sustainable SEEG pathway and has cross-trained several Radiology Assistants and 3D Printing technicians in this workflow. Ongoing efforts are underway to prospectively evaluate improvement of epilepsy surgical outcomes at LPCH.
Comprehensive Enhanced Recovery After Surgery (ERAS) Program for Elective Spine Surgery: Results from Anterior Cervical Discectomy and Fusion Patients

Allen L Ho, MD; Michael R Jensen MD PhD, Ping Ho Wei RN MPH, Yohan Lee MBOE, Arjun V Pendharkar MD, Karl Zheng MD, Jon Park MD, Serena S Hu MD, Anand Veeravagu MD, Atman M Desai MD, John K Ratliff MD, Amy C Lu MD.

Stanford, Department of Neurosurgery
Background
Enhanced Recovery After Surgery (ERAS) programs are a patient-centered, multidisciplinary perioperative approach designed to optimize the recovery process from surgery. These programs have begun to be applied towards spine surgery to improve pain management, promote earlier mobilization, prevent post-operative complications, and, ultimately, to reduce the in-patient length-of-stay and improve post-operative patient outcomes.

Current State
We present a summary of our comprehensive, institution-wide ERAS program and the pre- and post-ERAS implementation outcomes of our anterior cervical discectomy and fusion (ACDF) patients.

Analysis
Beginning in January of 2016, the Stanford Enhanced Recovery After Surgery (ERAS) Program was implemented by the spinal surgery programs at our institution. The program was targeted towards patients undergoing elective spinal surgery. The pathway encompassed the continuum of the patient™s perioperative care and all aspects of patient care were scrutinized and targeted with evidence-based interventions. A Program implementation goal of a 10% reduction in length-of-stay was established. A prospective patient registry synced to the electronic health record (EHR) was utilized for data collection and analysis. Variables included in this database included patient demographics, comorbidities, procedural data, quality metrics and outcomes.

Results
A total of 323 consecutive elective ACDF patients were included in this study from January 2014 to March 2018. 193 patients underwent surgery following ERAS program implementation in January of 2016. There were no significant changes in time to completion of specific process metrics after ERAS implantation. There were no significant differences in post-operative complications between the two cohorts. Length-of-stay (LOS) was decreased from a mean of 1.98 to 1.76 days (P = 0.649). There were also no significant differences in proportion of patients meeting discharge target date, discharge disposition status, 30- and 90-day readmissions, and 30-day emergency department visits.

This study demonstrates that implementation of a comprehensive, multidisciplinary ERAS program is feasible and potentially effective in lowering overall LOS in patients undergoing ACDF. Significant gains in traditional clinical outcomes with ERAS program implementation may be difficult to achieve with less invasive spinal procedures that already enjoy favorable outcomes, and more nuanced evaluation metrics for ERAS programs may need to be devised.
Implementation of ERAS in Posterior Lumbar Fusion: Prospective Results from an EHR-Linked Patient Registry

Allen L Ho, MD; Michael R Jensen MD PhD, Ping Ho Wei RN MPH, Yohan Lee MBOE, Arjun V Pendharkar MD, Karl Zheng MD, Jon Park MD, Serena S Hu MD, Anand Veeravagu MD, Atman M Desai MD, John K Ratliff, Amy C Lu MD

Stanford, Department of Neurosurgery
Background
Enhanced Recovery After Surgery (ERAS) programs are a patient-centered, multidisciplinary perioperative approach designed to optimize the recovery process from surgery. These programs have begun to be applied towards spine surgery to improve pain management, promote earlier mobilization, prevent post-operative complications, and, ultimately, to reduce the in-patient length-of-stay and improve post-operative patient outcomes.

Current State
We instituted a comprehensive ERAS program that spanned our entire spine care continuum and report on outcomes in our posterior lumbar fusion (PLF) patients.

Analysis
We report on our institutional outcomes of all PLF patients from 2013-2018 surrounding the implementation of our comprehensive spine ERAS program. Our program encompassed specific interventions geared towards standardizing care pathways including: pre-operative work-up and patient education, anesthesia protocols, surgical order sets, patient flow and nursing care process metrics, pain management, discharge disposition planning, and outcomes evaluation tools. Statistical analysis was performed on process measures, outcomes (LOS, readmission, return to ED, mortality), and complications in our patient cohort before and after ERAS implementation.

Results
713 patients underwent PLF during the study period with 311 patients receiving surgery prior to ERAS implementation and 402 patients post-ERAS. In terms of general outcomes, there was a statistically significant decrease in the LOS (4.09 to 3.71 days, P < 0.001) with an increasing proportion discharging home with or without home health services. Analysis of complications revealed an overall complications rate of 6% with no significant changes in complications after ERAS implementation. In terms of process metrics, patients had significant reductions in foley reinsertion, time to PO pain control (4.89 to 3.51 hours), and time to foley and IV removal after ERAS implementation.

Implementation of spine surgery ERAS program led to decreased length of stay, and improvements in ERAS-associated in-patient process metrics in patients undergoing PLF. These changes may reflect improved perioperative care coordination made possible by ERAS implementation.
Preventing Nosocomial Exposure Keratopathy in Pediatric Critical Care

Cassie Ann Ludwig, MD, MS; Jessica Bode, MSN, RN, NP,  Critical Care Float; Rachel Chapman, MSN, RN, PCNS-BC  PICU; Erica Barnum, BSN, MSN, CNS; Douglas Fredrick, MD; Diana Do, MD

Stanford, Department of Ophthalmology
Background
Sedated patients in the critical care unit with poor blink and with lagophthalmos are at risk for developing exposure keratopathy (ocular surface disease resulting in corneal abrasions, bacterial keratitis, and conjunctivitis). Order sets for intubation and critical care unit admission at LPCH historically included scheduled every two-hour ophthalmic lubricating ointment, but recently changed to every two-hour as needed ophthalmic lubricating ointment. With this change, the Ophthalmology service saw a rise in the number of cases of exposure keratopathy.

Current State
On review over 6 months, at least 2 new cases of exposure keratopathy arose at LPCH. On internal audit over 3 months, we found that the percentage of intubated patients in the pediatric critical care units with the appropriate level of ophthalmic intubation ranged from 60.0% - 69.2%.

By December 31st, 2019, we aim to increase the number of intubated children who meet performance criteria for placement of ophthalmic lubrication (presence of lagophthalmos) by 50%. We believe we can achieve this by increased adherence to a new protocol.


Analysis
A request for an EPIC build for assessments of the eyelids and cornea moisture every shift and subsequent trigger to adjust ophthalmic lubrication orders has been placed. We are working to educate all pediatric critical care staff on the nature of exposure keratopathy and on the new protocol.

Results
Regarding the new EPIC build, we intend for the nurse manager to review monthly to make any adjustments deemed necessary. Nurse shift leads will review the nurses™ assessment on the time of admission and when intubation/sedation status changes. The attending physicians will review the residents™ assessment on the time of admission and when intubation/sedation status changes. 
Clinical Care Redesign of the Geriatric Hip Fracture Pathway “ A Before and After Implementation Study

Noelle L Van Rysselberghe, MD; Lauren M Shapiro, MD, Blake J Schultz, MD, Robin N Kamal, MD, Amy C Lu, MD, Michael J Gardner MD

Stanford, Department of Orthopedic Surgery
Background
Geriatric hip fractures are associated with significant morbidity, mortality, and economic burden. Mortality rates within one year after a hip fracture can be as high as 58% and about 30% of patients experience significant functional decline. Overall costs for osteoporosis-related fractures are climbing, and expected to exceed $25.3 billion annually by 2025. With an annual incidence of approximately 330,000, hip fractures represent about 14% of all osteoporosis-related fractures, but account for 72% of the costs. Given the scope of this issue, a vast amount of research has been conducted and clinical practice guidelines have been set forth to encourage high-quality, evidence-based care. Despite this, significant variation exists in the care of geriatric hip fractures. 

Current State
Our goal is to create and implement a high-value, evidence-based clinical care pathway for patients sustaining geriatric hip fractures.

Analysis
We have created a multi-stakeholder team that will begin by creating a current-state process map of Stanford™s pathway for geriatric hip fracture management. We have identified quality measures and clinical practice guidelines from a literature review and are applying these to the current-state process map to identify areas for quality improvement and value-added interventions. Following our pathway redesign, we will measure outcomes based upon those same quality measures. Primary outcome will be length of stay. Secondary outcomes include time to surgery, length of time to orthopaedic consult, length of time to medicine consult, length of time to pain consult, length of time to regional anesthesia, location of discharge, use of regional and/or multimodal anesthesia, diagnosis of osteoporosis, completion of osteoporosis labs, completion of DEXA scan, prescription of osteoporotic medications, subsequent fragility fracture, and direct medical cost. The case-to-case variation of each outcome measure will be calculated.
 
This study will be historically controlled and data will be collected from the medical record. Patients ‰¥65 years with an ICD10 code representative of a femoral fracture in the neck, intertrochanteric, or subtrochanteric region will be identified. Patients with high mechanism of injury, isolated greater or lesser trochanteric fractures, or femoral head fractures will be excluded. Descriptive statistics for each response variable will be calculated and either unpaired t-tests or chi-squared tests will be used to compare each variable from pre- to post-implementation (significance set at p<0.05).

Results
(anticipated results) *** patients met inclusion (***% pre-implementation and ***% post-implementation). There was significant differences in the following variables: length of stay (p=***), time to surgery (p=***)¦ There was no difference in the following variables: length of time to pain consult (p=***). The total cost of care per episode decreased by ***%. The case-to-case variation decreased by ***%. This high-value, evidence-based hip fracture care pathway 1) improved benchmarking of clinical practice guidelines, 2) reduced cost of care, and 3) decreased variation of care episodes. 
Understanding the Patient Experience: Analysis of Two-Word Assessment and its Relationship to Likelihood to Recommend in Outpatient Hand Surgery

Lauren Shapiro, MD; Thomas KA, Eppler SL, Behal R, Yao J, Kamal RN

Stanford, Department of Orthopedic Surgery
Background
Likelihood to recommend (LTR) is a measure of patient loyalty and satisfaction which is correlated with patient referrals and increasing revenue to a health system. A two-word physician assessment gives a quick and efficient method to evaluate the patient experience and has been suggested as a replacement to lengthy surveys (e.g. H-CAHPS) in inpatient medicine. The two-word assessment may provide qualitative and potentially actionable feedback to drive improvement.

Current State
We sought to test the use of the two-word assessment in an outpatient hand clinic, the relationship of LTR and two-word assessment, and determine which words patients used to describe their experience.

Analysis
We enrolled patients from the outpatient hand and upper extremity clinic of two fellowship-trained orthopaedic surgeons at a suburban academic medical center. We used consecutive sampling to enroll new patients eighteen years or older demonstrating English literacy who presented to clinic. Because this study was considered quality improvement based on the criteria of our institution, it did not require institutional review board approval. Consent was obtained and the study was completed after the physician visit. 
Patients were asked to provide a two-word assessment of 1) their physician, 2) their overall experience, 3) recommendations for improvement, and 4) a likelihood to recommend scale (1 being extremely unlikely™, 5 being extremely likely™). LTR data was made binary (1-4: not likely to recommend, 5: likely to recommend), based upon implications from hospital administration and in accordance with quality improvement efforts. Sentiment analysis was used to categorize the two-word surveys into positive, neutral, or negative sentiment. Recommendations for improvement were classified into physician issue, system issue, unable to determine.
As no patients expressed a negative two-word sentiment, a Fishers exact test was used to determine whether the LTR status differed by two-word sentiment (positive or neutral). A chi-squared test was used to determine if LTR status differed between those who expressed improvement related to the system, related to the provider, or neither. A Fisher™s exact test was used to determine the difference in constructive improvement opportunities between those LTR and those not LTR.


Results
Sixty-nine patients completed the questionnaires (44.9% female, 42.3 years mean age (SD 17.6 years), 78.8% of patients were likely to recommend their provider). Sixty-seven (97.1%) patients noted positive sentiments toward their provider. Sixty-seven (97.1%) patients noted positive sentiments toward their overall visit experience. 31.4% of improvement recommendations were system based, 5.9% were provider based, and 62.7% were indeterminable. There was no difference in LTR between cohorts expressing positive or neutral sentiment in the two-word survey (no patients noted negative sentiment), nor was there a relationship between LTR and patients noting either physician or systems issues. Patients not LTR were more likely to leave meaningful opportunities for improvement than those LTR (p = 0.01). This insight may be used in future practice to obtain actionable, real-time patient feedback that can inform operational change and improve the patient experience.
Shared Deicison Making in Operative versus Nonoperative Displaced Midshaft Clavicle Fractures

Malcolm DeBaun, MD; Blake Schultz, MD; Cara Lai, BS; John Costouros, MD; Julius Bishop, MD; Robin Kamal, MD; Michael Gardner, MD

Stanford, Department of Orthopedic Surgery
Background
Management of displaced midshaft clavicle fractures (DMCF) is controversial. Patients can be treated operatively with open reduction and surgical fixation or nonoperatively with sling immobilization. Both treatments are reasonable with different expected risks and benefits. Consequently, shared decision making between the patient and the provider is the preferred approach to inform the management of these fractures.

Current State
It is currently the responsibility of the individual surgeon to lead the discussion with the patient when deciding between operative versus nonoperative management of their DMCF. At our institution, these patients are triaged and counseled by fellowship-trained orthopaedic surgeons with extensive experience in trauma or shoulder and elbow to facilitate a personalized treatment approach.  The goal of this study was to determine the level of decisional conflict for the patient after their choice has been made to pursue operative or nonoperative treatment.

Analysis
A validated decisional conflict survey was administered to 16 consecutive patients who sustained a DMCF.  The level of decision conflict was calculated for each patient regardless of their decision, with a score above 37 demonstrating a significant level of conflict experienced by the patient. Descriptive statistics were performed. 


Results
The average decisional conflict for all patients was low (avg=11.4) after the shared decision making process. The vast majority of patients felt informed, supported, with their values clarified and confident they made an effective decision with a low level of uncertainty.  Patients who sustained a DMCF may benefit from management by an orthopeadic specialized experienced in shared decision making and fellowship trained in trauma or shoulder and elbow.
Patients Find Videorecording Instructions for Post-Operative Ankle Fracture Care Useful

Sam Jiang, MD; Blake J. Schultz MD, Arlene J. Garcia, Michael J. Gardner MD, Julius A. Bishop MD

Stanford, Department of Orthopedic Surgery
Background
Optimal outcomes after ankle fracture surgery require that patients understand and adhere to post-operative fracture care instructions.  However, patients often have difficulty comprehending and retaining this information. Patient-recorded fracture care instructions at post-operative clinic visits offers a solution to this. We initiated a prospective study to assess patient comfort level and perceived utility of an audio-video recording option. We hypothesized that patients would find recordings of their care instructions practical and useful.

Current State
SMART goal:
-S: Do patients find videorecording their fracture care instructions useful
-M: 6 week post-op survey
-A: Very achievable
-R: Relevant to patients especially given the ubiquity of smartphones and potential benefit for minimal cost
-T: Completed

Analysis
Twenty-five consecutive adult patients who underwent surgical treatment for an ankle fracture were invited to record their fracture care instructions using their smartphones at the 2-week post-operative visit. These patients were administered a survey at the 6-week post-operative visit to assess their experiences.

Results
Patients rated their comfort level with their fracture care instructions at 9.6/10 (SD=0.6) and usefulness of audio-video recording at 8.7/10 (SD=2.1). All patients (n=25) indicated that recorded instructions made them feel more confident with their fracture care. 88% (n=22) reviewed their instructions after their 2-week visit. 60% (n=15) of patients shared their instructions with someone else and 88% (n=22) believe physicians should allow patients to record fracture care instructions.

Patients were comfortable recording their instructions with their surgeons and felt it was useful, increasing their confidence in their ability to properly follow their postoperative fracture care plans. Given the importance of patient comprehension and retention of post-operative instructions, surgeons should consider incorporating patient-generated audio-video recordings into their practice as an effective, efficient, and inexpensive way to improve patient care and satisfaction.
Assessing the Patient Perspective on Patient Reported Outcome Measures Following Elective Hand Surgery: A Convergent Mixed-Methods Analysis

Lauren Michelle Shapiro, MD; Sara Eppler, Allison Roe, Robin Kamal

Stanford, Department of Orthopedic Surgery
Background
Patient-reported outcome measures (PROMs) are increasingly used to evaluate quality of care from the patient™s perspective, in contrast to physician-measured or objective outcomes, which may be less important to the patient. It is possible that standardized PROMs may have disparate abilities to assess a patient™s outcome based on their individual goals in seeking care. Understanding which PROMs best reflect the patient™s perspective on his/her function and why can inform efforts to improve patient-centered, goal-directed care.

Current State
The aims of this prospective, exploratory, cohort study are four-fold: 1) to evaluate which PROM [the short version Disabilities of the Arm, Shoulder and Hand (QuickDASH) or the Patient-Specific Functional Scale (PSFS)] patients believe better assesses their goals in treatment, 2) to describe the responsiveness of each instrument, 3) to understand why patients prefer an instrument using qualitative methods, 4) to combine the quantitative and qualitative analyses for the evaluation of patient preference in PROMs using a convergent mixed-methods approach.

Analysis
Patients ‰¥18 years undergoing elective hand surgery displaying English literacy were recruited prior to surgery. Patients were limited to those who would be immobilized for less than four weeks. Patients received the two questionnaires pre-operatively and at or after their six-week visit. Questions included on the follow-up questionnaire were, 1) Which questionnaire better addressed your goals during this clinical experience? and 2) Please explain why you chose that questionnaire. Descriptive statistics were calculated for questionnaire preference, pre- and post-surgical scores and score change. Responsiveness was measured by change in score. A paired t-test was used to detect a difference between changes in PROM score. The frequency and percent of patient changes above the minimal clinical importance difference (MCID) of each questionnaire was reported. A Fishers exact test was used to evaluate the difference in those reaching MCID between groups. We qualitatively analyzed the patients™ responses to why they chose a specific questionnaire using content analysis. Results from the quantitative and qualitative analyses were brought together and arrayed side by side with a convergent mixed-methods analysis.

Results
Thirty-eight patients completed pre-operative questionnaires; twenty-five (65.79%) completed the post-operative questionnaires. Mean patient age was 54.8 years (SD 18.9), 60% were female. Seventeen (77%) patients preferred the PSFS, three (14%) had no preference, and two (9%) preferred the QuickDASH; three patients did not complete this question. The average change from pre- to post-operative QuickDASH was -10.26 (SD 20.4) while that of PSFS was -27.24 (SD 26.2) (p=0.0058). Ten (40%) patients reported a change in QuickDASH above the MCID, seventeen (68%) patients reported a change in PSFS above the MCID. (p=0.004). Content analysis revealed four themes: survey simplicity, personalization, goal-directed, distinct items. Most patients preferred their assessments of function to be based on specific, self-identified tasks and felt the PSFS better measured their goals. Employing PROMs that address patient-specific goals, opposed to using generic PROMs, may improve patient-centered and goal-directed care. Our future work will validate the PSFS in the field of orthopaedic surgery to promote its use as a patient-centered outcome measure. 
Implementation of an External Fixator Reprocessing Program and Associated Cost Savings

L Henry Goodnough MD PHD; Michael J. Gardner MD

Stanford, Department of Orthopedic Surgery
Background
External fixators used to temporize fractures are frequently encountered and removed upon patient transfer to a tertiary referral center, and yet the ability to reprocess certain external fixator components and potential cost savings may be under-appreciated. Rather than being discarded upon removal, certain undamaged external components can safely be reprocessed, re-sterilized, and repurchased. Up to 75% of external fixators can be recycled (1), reducing mean hospital costs up to $1200 per case and hospital charges $600 per case (2). Furthermore, level I evidence suggests such savings are realized without risk of complications (3). However up to 30% of orthopaedic surgeons interested in external fixator reprocessing have not implemented programs for because of administrative barriers to entry (4).

Current State
We set out to initiate an external fixator reprocessing program within our orthopaedic trauma service using an FDA-approved reprocessing program, to measure cost savings thereafter. Our purpose was to implement an evidence-based measure to improve cost-effectiveness of orthopaedic health care delivery. Our goal was to fully implement the program immediately.

Analysis
We used the current procedural technology code 20694 (Removal, under anesthesia, of an external fixator system) to identify patients who had external fixator removal performed at Stanford using the Stanford Research Repository (STARR) cohort finder tool. We identified 65 cases of ex-fix removal in 2018. We began external fixator reprocessing in October 2018. Cost savings data were tabulated and calculated by our Quality Consultants-High Value Care program.

Results
During the initial period of implementation October 2018 to February 2019, fifteen cases of external fixator removal were performed. The external fixator reprocessing program saved $31,344 during that time. External Fixator reprocessing continues to be reliable, safe, and cost-effective upon implementation at our institution. Rates of implant complications are low after external fixator reprocessing, however we will track our reprocessed implants to identify if any failures do occur.  Future directions will focus on percent adherence to reprocessing, as well as identifying and addressing any barriers to adherence.

References
1. Horwitz DS, Schabel KLS, Higgins TF. The economic impact of reprocessing external fixation components. J Bone Joint Surg Am. 2007 Oct;89(10):2132“2136.
2. Dirschl DR, Smith IJ. Reuse of external skeletal fixator components: effects on costs and complications. J Trauma. 1998 May;44(5):855“858.
3. Sung JK, Levin R, Siegel J, et al. Reuse of external fixation components: a randomized trial. J Orthop Trauma. 2008 Feb;22(2):126“130; discussion 130-131.
4. Thamyongkit S, Bachabi M, Thompson JM, et al. Use of reprocessed external fixators in orthopaedic surgery: a survey of 243 orthopaedic trauma surgeons. Patient Saf Surg. 2018;12:10.

Needs Assessment to Promote Capacity Building in Hand Surgery Outreach Trips

Lauren Shapiro, MD; Shapiro LM, Park MO, McGoldrick R, Kamal RN

Stanford, Department of Orthopedic Surgery
Background
The surgical burden in low- and middle-income countries (LMIC) is immense. It is estimated that about 95% of people in LMIC do not have access to basic surgical care and that surgically treatable diseases account for at least 11% of the total global burden of disease. Injuries (traffic injuries, falls) alone account for about five million deaths annually and about 11% of the global loss of disability-adjusted life years. Injuries and congenital malformations, while debilitating, may be treatable thereby reducing morbidity and mortality. Historically, surgical intervention has been a low global health priority, however recently, there has been an acknowledgement of surgery as a key component of healthcare and an interest in strengthening local surgical capacity.  An increased effort in surgical outreach trips focused on training and collaborating with local providers to support this endeavor has resulted. 

While the exact burden of hand and upper extremity (UE) pathology has not been reported, organizations  (eg: Touching Hands Program, ReSurge International) have been established to address hand and UE injuries, malformations, and burns in LMIC. Despite the increase in funds and resources invested, there has been a lack of standardized data collection regarding outreach trips. This paucity of data is a barrier to assessing and improving the quality, safety, and efficacy of care. Orthopaedic (trauma, pediatric) needs assessments exist, however these are primarily aimed at identifying educational deficits and needs of practicing surgeons at CME conferences in Haiti or children undergoing plastic surgery. Such a tool does not exist for hand surgery outreach trips. 


Current State
Given the importance and lack of a comprehensive tool, we aim to create a survey for site administrators that assesses the needs of a site prior to hand surgery outreach trips. We aim to optimize site selection, set appropriate expectations, and improve care delivery in LMIC.

Analysis
As a QI initiative, this study was IRB exempt. This study will be conducted in five phases. 1) Two standardized WHO tools will be evaluated by three research team members with orthopaedic surgery, hand surgery, and anesthesiology training and adapted for hand and UE surgery, 2) We will conduct field testing at two pilot facilities in which members of the research team will pilot the tool during a hand surgery outreach trip with the goals of surveying local host institutions on their needs and capacity to assess the importance of each domain/subdomain, 3) We will refine the tool based upon these experiences, 4) We will interview 15 expert hand surgeons with extensive hand and UE outreach experience to identify approaches to needs assessment and gaps in the tool 5) We will use the expert panel feedback to complete the tool. 

Results
A needs assessment tool will be created with eight domains and associated subdomains relevant to hand and UE surgery. This investigation will provide a novel and standardized tool for evaluating site needs for capacity building in LMIC for hand and UE surgery outreach programs. We anticipate that this tool will allow for improved site selection and expectation setting for outreach trips. 
A Novel Preference Elicitation Tool for Anesthetic Choice in Hand Surgery

Allison Roe, MD; Sara Eppler, Lauren Shapiro, Robin Kamal

Stanford, Department of Orthopedic Surgery
Background
Monitored anesthetic care (MAC) has long been the anesthetic choice for minor hand procedures in the US. This approach avoids epinephrine use, which has been theorized to cause digital necrosis. Despite this, there have been no modern reports of digital necrosis caused by epinephrine use in the literature. Wide awake local anesthetic with no tourniquet (WALANT) is an anesthetic technique that uses epinephrine at the surgical site, negating the need for a tourniquet and therefore sedation. While it is standard of care in Canada, its use is still gaining momentum in the US. Studies have demonstrated that WALANT costs less, requires less patient time in the OR and facility, results in fewer preoperative tests, and less postoperative narcotic use. However, not all patients are comfortable being awake during surgery. As such, the decision of which anesthetic option to use should be made based on patient preference. Preference elicitation tools enhance shared decision making by providing education about different treatment options and help patients define their values. The development of a ranking preference elicitation tool could be beneficial for patients deciding between WALANT and MAC.

Current State
The aim of this study was to develop a preference elicitation tool for anesthetic choice in patients undergoing minor hand surgery. A ranking preference elicitation tool was created using Qualtrics software that presented the anesthetic options and described attributes of each. The tool then allowed patients to rank their preferences for important care attributes. The preference elicitation tool met most of the quality criteria set by the International Patient Decision Aids Standards (IPDAS).

Analysis
We will use a mixed-methods approach for our preference elicitation tool development and testing. Patients undergoing elective hand surgery who are fluent in English and over the age of eighteen will be recruited to this study prior to their operation. Patients undergoing carpal tunnel release, De Quervain™s release, trigger finger surgery, dupuytren™s contracture excision and ganglion cyst removal will be included. Patients with contraindications to either anesthesia type will be excluded. Patients will be presented with the preference elicitation tool on an iPad and they will rank their preference of each attribute. We will conduct cognitive interviews for a qualitative analysis to determine how the tool can be improved. The interviews will be transcribed, and content analysis will be completed using the constant comparative method. We will use tool completion time and a validated System Usability Scale to evaluate the feasibility and usability of the tool, respectively.

Results
The preference elicitation tool has been developed and is currently undergoing cognitive interviews for patient feedback. We will plan to have twenty patients conduct cognitive interviews. After the tool is modified based on the qualitative patient feedback, we will conduct our quantitative analysis. We will have thirty patients use the tool to evaluate its feasibility and usability. We hope that after our testing, we will have developed a usable and feasible tool that can be used to help patients decide between anesthetic types for their hand procedure.
Does Melatonin Affect Sleep Disturbance and Outcomes in Orthopedic Trauma Patients- A Prospective, Randomized Control Trial

Blake J Schultz, MD; Natalie Tanner BS, Arlene Garcia BS, Andrew Fithian MD, Julius A. Bishop MD, Michael J. Gardner MD

Stanford, Department of Orthopedic Surgery
Background
Trauma patients are at high risk for sleep disturbance. Physical pain, stress, and the psychological impacts of the trauma, hospitalization and surgery cause sleep changes. Poor quality sleep can lead to slower recovery, chronic pain, delirium, and cardiac events and is a hallmark of PTSD, depression, and anxiety. Orthopedic trauma patients specifically have a high prevalence of sleep disturbance that is often under-treated or not addressed at all. To achieve optimal rehabilitation, orthopedists traumatologists should address both the physical and psychological disability of their patients

Current State
Orthopedic trauma patients are often prescribed sleep aids while in patient, but not routinely on discharge. Sleep disturbances secondary to trauma and surgery are known to be long lasting, so these patients may be undertreated. Melatonin is a hormone secreted to promote sleep, and is specifically altered following major surgery. It does not have the sedative and delirium side-effects of the standard opioids and hypnotics used in the post-operative period, and therefore may be a safe and effective way to treat the sleep disturbances experienced by orthopaedic trauma patients.
We aim to determine:
1) Are melatonin and sleep education effective in improving sleep quality in Orthopaedic trauma patients?
2) Is improved sleep quality correlated with improved post-operative functional outcomes and/or psychologic well-being?
3) Are injury severity or self-reported pain scores correlated with sleep disturbance?

Analysis
We are conducting a double-blinded, prospective, randomized control trial on all adult operative orthopedic trauma patients. Exclusion criteria include documented head trauma, cognitive impairment or previously diagnosed sleep disturbance. Eligible patients are enrolled at the two-week post-operative clinic visit and randomized to either the control group, which receives glucose tablets, or the experimental group, which receives 5mg melatonin and sleep hygiene education. Patients take the designated medication daily for six weeks. Patients and treating physicians are blinded to treatment group. The primary outcome measures are Pittsburgh Sleep Quality Index, VAS Pain Score and SF-36 surveys completed by all subjects at the 2, 6 and 12-week clinic visits.

Results
Enrollment began in March, with a goal of 100 patients based on a power analysis.  If effective, melatonin has the potential to be a safe, cheap and easily accessible sleep aid to address both the physical and psychologic stresses that affect orthopedic trauma patients. 
Prospective Validation of a Novel Serum mRNA Gene Panel for Diagnosis of Acute Septic Arthritis

Blake Schultz MD; Tim Sweeney MD, Malcolm DeBaun MD, Michael Gardner MD

Stanford, Department of Orthopedic Surgery
Background
Septic arthritis is an orthopedic emergency requiring immediate surgical intervention to prevent irreparable damage to the joint cartilage. Clinically, septic arthritis is difficult to distinguish from aseptic arthritis, such as gout, inflammatory arthritis or transient synovitis. Because aseptic arthritis can be managed medically, without surgery, correctly diagnosis the etiology of acute arthritis quickly is important to avoid unnecessary emergency surgery.

Current State
Current diagnostic standard of care for acute septic arthritis is serum labs and an invasive joint aspiration. Serum labs and synovial aspirate provide information about the inflammatory cells in the sample within a few hours, but there is often ambiguity about whether the source is infectious (e.g. bacterial) or non-infectious (e.g. gout). Synovial cultures are the gold standard of diagnosis, but can take days to result, so decisions about surgery are often made with incomplete data. Additionally, aspirations carry risk and require technical skill and potentially advanced imaging. Novel diagnostics are thus needed.

The Sepsis MetaScore™ (SMS) is an 11-mRNA host immune blood signature that can distinguish between infectious and non-infectious acute inflammation. It has been validated in multiple cohorts across heterogeneous clinical settings from outpatient acute infections to patients in ICU suspected of sepsis. Here we studied whether the SMS may hold diagnostic validity in determining if acute arthritis was due to an infectious cause.


Analysis
We conducted a blinded, prospective, non-interventional clinical study of the SMS. All patients undergoing work-up for a septic primary joint were enrolled. Patients proceeded through the normal standard-of-care pathway, including joint aspiration and inflammatory labs (White Blood Cells (WBC), erythrocyte sedimentation rate (ESR) and c-reactive protein (CRP). At the time of the lab draw, venous blood was also drawn into PAXgene RNA-stabilizing tubes and mRNAs were measured using NanoString nCounter„¢. SMS was calculated blinded to clinical results.

Results
Overall, 20 samples were included, with 10 patients ultimately diagnosed with septic arthritis based on positive synovial fluid cultures. The SMS had an area under the ROC curve (AUROC) of 0.87 (95% CI: 0.71-1) for separating infectious from non-infectious conditions. Notably, this is very similar to its diagnostic accuracy in multiple other cohorts, lending credence to the stability of the metric. For comparison, the AUROCs for serum ESR=0.58 (95% CI: 0.87-0.29), CRP=0.6 (95% CI: 0.87-0.34), and WBC=0.59 (95% CI: 0.85-0.33) and synovial WBC=0.54 (95% CI: 0.81-0.28) and PMN=0.51(95% CI: 0.79-0.24). At 100% sensitivity for infection, the specificity of the SMS was 40%. This suggests that a substantial fraction of non-septic patients could potentially be safely ruled out for further surgical intervention.

In this study, the SMS showed a high level of diagnostic accuracy in predicting septic joints compared to other diagnostic biomarkers.  This quick blood test could be an important tool for early, accurate identification of acute septic joints and need for emergent surgery, improving clinical care and healthcare spending.
Resident Safety Council (RSC): Culture of Safety Project 2018-2019 - A Quality Improvement Initiative to Optimize Process and Access to Resident SAFE Reports

Erna Forgo, MD; Katie Kan, Anne McHugh, Dexter Louie, Ioana Baiu, Jenny Chen, Jason Powell, Emily Larimer, Swati Yanamadala, Patricia Britt, Karen Frush, Thomas Caruso

Stanford, Department of Pathology
Background
Stanford Alert for Events (SAFE) reports initiate a reliable and timely process that captures events that effect patient safety, ultimately resulting in improved patient care. Although some single events propel process improvements, the majority of SAFE reports provide aggregate data that are used to forecast areas of quality drift and demonstrate areas of improvement.

Current State
Residents play a significant role in patient care at Stanford. However, they submit the lowest number of SAFE reports when compared to other providers (~2% of all reports). Culture of Patient Safety (COS) survey results reveal this low rate is due to a lack of feedback, fear of retaliation, and lengthy submission process. These results provided the framework for developing effective interventions to optimize the SAFE submission process. The goal of this quality improvement project is to provide Stanford residents with increased access to aggregate data trends of resident-submitted SAFE events and to improve the submission system process, by September 2019, measured by a 10% increase in resident submitted SAFE reports and improvement in responses on the COS survey.

Analysis
The multidisciplinary team developed 5 key drivers to increase resident SAFE reporting: (1) ease of use; (2) modeled reporting; (3) transparent reporting without retaliation; (4) data-driven feedback; and, (5) consistent feedback. Consequently, 4 interventions were proposed: (1) reporting website revision; (2) Stanford Graduate Medical Education (GME) website™s Quality Improvement (QI) section revision; (3) access to aggregate data on Stanford MedHub; and, (4) housestaff orientation optimization.

Results
Multiple interventions were initiated to improve the SAFE event reporting system. The reporting website (https://shcconnect.stanfordmed.org/depts/QPSED/SAFE/Pages/PhysicianonlyReporting.aspx) is undergoing major revisions to improve data capture, a crucial part of the system for trending and identifying root causes that in turn allow for feedback. The GME website™s QI section (http://med.stanford.edu/gme/quality_improvement.html) was revised and now serves as the central hub for SAFE reports, RSC, QI grant, Resident Quality Improvement/Patient Safety Symposium, and provides link to various QI training programs and modules. The MedHub website (https://stanford.medhub.com/index.mh) has been updated to display resident-submitted report submission rates and trends, providing feedback. A housestaff onboarding presentation has been prepared for the next academic year detailing the SAFE report system. Multiple access points are available for submission of reports, and the online locations where aggregate data and trends for resident-submitted reports are easily accessible.

Multiple avenues are in place for sustaining the interventions to ensure that the changes we have made will be continued in perpetuity and help change the culture of Stanford™s SAFE reporting. The aggregate data and trends will be tracked to assess their effectiveness. The resident-submitted report submission rates and trends will be updated quarterly on GME QI and MedHub websites by the GME project coordinator. The housestaff onboarding material will be presented at the Program Director and Chief Resident meetings early in the academic year to facilitate dissemination of how to place SAFE reports. RSC Safety Champions will also be active participants in disseminating the information.
52-in-52 Project: Improving the Gross Room Work Station Schedule after Sub-Specialization

Erna Forgo, MD; Erika Dobo Forgana, Quintina Herrera, Francis De Leon, April Young, Christina S. Kong

Stanford, Department of Pathology
Background
With increasing complexity of pathology diagnosis, there has been a move across the country to subspecialize pathology by organ system, e.g. gastrointestinal, gynecologic, breast, head & neck, thoracic, soft tissue/bone, etc. The change to subspecialized sign-out started at Stanford in 2011 and became fully subspecialized in 2018. A change in the resident schedule from a 3-day cycle (day 1 prosect (gross) specimens, day 2 preview slides, day 3 sign-out cases) to a 1-day cycle (gross, preview, sign-out all in one day) led to a bottleneck in the Gross Room.   Residents were expected to gross daily between the hours of 13:00-16:00. One of our pathology assistants (EDF) identified this problem and as part of our 52-in-52 program (adapted from the Stanford Department of Radiology quality improvement strategy) initiated a project to identify solutions.

Current State
The Pathology Gross Room is staffed by 7 pathologist assistants and 7 surgical pathology technologists who work in shifts from 07:00-21:30. Additionally, there are 6 Anatomic Pathology (AP) residents who typically gross specimens between 13:00-16:00. Given a fixed total of 13 grossing stations, there were insufficient stations in the afternoons for the residents to gross. The goal of this project is to ensure availability of grossing stations for six residents between the hours of 13:00-16:00 by 7/1/2018.

Analysis
Two Gross Room schedules were proposed, both focusing on the extension of the gross room hours of operation. The first proposal centered around the extended hours of 05:00-00:00. However due to the inability to fill the late-night shift, Proposal #2 was suggested which spanned the hours of 06:00-21:30. Both proposals went live for trial periods of 14 days. Gross Room turnaround time (TAT) was defined as the time a specimen is accessioned to the time the cassettes were received at the off-site Stanford Histology Laboratory. The number of non-grossed specimens carrying over to the next day was also monitored.

Results
Implementation of Proposal #1 generated work station availability for all residents, while Proposal #2 saw an improvement in work station availability. Both proposals shortened the average turnaround time of 10 hours for all specimens: Proposal #1 decreased to <6 hours; Proposal #2 decreased to 8 hours. The number of non-grossed specimens at the end of the day was lowest for Proposal #1, intermediate for Proposal #2 and highest at baseline. While proposal #1 yielded a faster turnaround time and more efficient specimen volume handling, the hours of operation it required was not immediately feasible. For this reason, Proposal #2 went into effect on 2/25/2019 and has shown sustained results. Plans are currently underway to implement a hybrid of both proposals on 5/20/2019, with the anticipated hours of operation from 06:00-23:00. 
Reducing Length of Stay and Red Blood Cell Transfusion by Implementing an Anesthesia Anemia Management Clinic

Muharrem Yunce, MD; Tho Pham, Anil K. Panigrahi

Stanford, Department of Pathology - Transfusion Medicine
Background
Preoperative anemia is a very common entity with proven subsequent adverse outcomes and is reported to be present in one third to half of patients scheduled for surgical procedures (1-3). Mitigating preoperative anemia is one of the three pillars of patient blood management (4), however, at our institution there currently is no systematic screening nor a robust infrastructure to offer treatment for patients with preoperative anemia. In order to address this problem, we sought to build a pathway for screening, evaluation, and treatment of anemia in elective surgery patients. In order to better understand the scope of preoperative anemia at our institution and best focus treatment efforts, we first evaluated the prevalence of preoperative anemia (defined as hemoglobin ‰¤10 gm/dL), transfusion rates, and hospital length of stay (LOS) among several surgical services.
We retrospectively reviewed 89,372 surgical cases performed between January 1, 2013 and June 30, 2016 at Stanford Health Care. After exclusion of emergent, non-invasive procedures, and cases with incomplete hemoglobin data, 12,473 cases were included in the final analysis. Procedures were categorized into eight major surgical services. Patients were then divided into anemic and non-anemic groups and rates of transfusion, as well as, LOS were compared. Statistical analysis utilized one tailed t-test and Fischer™s exact test.

Current State
The prevalence of anemia (as defined Hgb ‰¤10 gm/dL) was 11% among all patients included the analysis. Transfusion rates were 22% in the anemic group compared to 10% in the non-anemic group (p<0.00001). The LOS was significantly different between anemic and non-anemic groups at 10 days vs. 4 days, respectively. (p<0.00001)
Preoperative anemia was associated with longer LOS compared to non-anemic patients in all eight surgical services analyzed. RBC transfusion rates were also increased among anemic patients in gynecology, general surgery, neurosurgery & orthopedics, genitourinary, and thoracic surgery services. However, the transfusion rate did not differ between anemic and non-anemic groups in cardiovascular transplant, coronary artery bypass graft/valve repair, liver & kidney transplant services.


Analysis
Given the significant association between preoperative anemia and LOS and RBC transfusion rates in almost all surgical services, a clinical pathway was developed focusing on efficient characterization and treatment of preoperative anemia. Once anemia is identified in the surgical clinic, patients are referred to the Anesthesia Anemia Management Clinic (AAMC). As iron deficiency can be found in up to 60% of anemic surgical patients (5), prior to their visit patients undergo additional testing to identify nutritional deficiencies which may contribute to their anemia. During the clinic visit, patients will be counseled regarding the risks of preoperative anemia and, if amenable, will be treated with intravenous iron and/or erythropoietin as appropriate.

Over the course of one year patients treated at AAMC will be monitored for pre-treatment, day of surgery, and discharge hemoglobin values, in addition to hospital length of stay and RBC transfusion rates. The reduction in LOS and transfusion rates will be compared with our retrospective data for similar patient population.


Results
The Anesthesia Anemia Management Clinic will provide patients higher value care through reductions in hospital length of stay and reduction or avoidance of allogeneic blood transfusion. Additionally, functional status may be improved with reductions in anemia at time of discharge. The standard intervention of IV iron infusion has been thoroughly tested, is well-tolerated, and has very low risk of adverse events. With an efficient and well-coordinated clinical pathway, such treatment can take place with minimal delay in elective surgery scheduling. 
Improving Safe Sleep on Maternity at LPCH

Maya Landan, MD; Christine Walravens, MD

Stanford, Department of Pediatrics
Background
In the 1990s a successful Back to Sleep campaign improved safe sleep practices in the United States and decreased the rate of Sudden Infant Death Syndrome (SIDS) by 50%. Unfortunately, since 2001 there has been a plateau in adherence to safe sleep recommendations and an estimated 3500 infants still die unexpectedly during sleep. Research has shown that caregivers tend to model sleep patterns observed in the hospital setting. In 2017, LPCH adopted a newborn safe sleep protocol which was based on the 2016 AAP policy on safe sleep. Despite this, an audit of 70 sleeping infants on the Maternity unit in September 2017 found that 52 (74%) did not adhere with the AAP safe sleep guidelines. Additionally, there have been multiple infant falls since that time which presents a critical patient safety issue.

Current State
From GEMBA, several factors contributing to the current state were identified: 1) Lack of awareness of the LPCH safe sleep protocol 2) Variations in how physicians and nurses address observed unsafe sleep behaviors and 3) No standardized safe sleep teaching to families throughout admission or safe sleep focused materials.

Our goal was to increase the rate of adherence to the LPCH safe sleep protocol on the Maternity post-partum unit from 25% to >75% by May 2019.


Analysis
Based on key-driver analysis, interventions were planned through the support of the Maternity Local Improvement Team (LIT) to focus on training, educations, and materials. The first was raising awareness among the staff about the recent falls and state of unsafe sleep along with disseminating the protocol and focusing on role modeling safe sleep practices. This was done via nursing huddles and an email sent out to all the nurses in June 2018. Safe sleep was again highlighted during nursing huddles in January 2019 after more infant falls. Further nursing education about the importance of role modeling safe sleep behavior to families was provided via emailed info-sheets and potty press. Patient directed materials in the form of crib cards created by nurse residents were circulated between July and December 2018.

Results
After these initial series of interventions, an audit of 42 sleeping infants in January 2019 found that 30 (71%) were adherent with safe sleep guidelines, improved from the 26% in 2017. To continue to sustain and improve upon this increased compliance with safe sleep, we plan to use iterative PDSA cycles. Future possible interventions include: updating the after visit summary information on safe sleep that is reviewed with each family prior to discharge, designing permanent crib cards to visually exemplify safe sleep practices as a visual reminder to both families and staff, as well as nurse-initiated teaching and role modeling on safe sleep with parents during standard periods of care such as bathing or weight checks. 
Implementation of schedule-based family centered rounds on an inpatient pediatrics team.

Katherine Helmuth, MD; Marie Wang, Marissa Hutauruk, Shirley Cheung, Shanna Perales, Joseph Kim, Amit Singh

Stanford, Department of Pediatrics
Background
Family-centered rounds (FCR) on the inpatient Pediatric Hospital Medicine service are unpredictable and inefficient, contributing to inconsistent nursing participation, delays with the Patient Progression Huddle, and low satisfaction with the rounding process among residents.  Since 2003 the American Academy of Pediatrics has recommended that rounds include the patient and the patient™s family to promote their participation in medical decision-making. Studies have shown FCR increases both staff and family satisfaction and improves families™ confidence in the medical team. Multidisciplinary rounds which include nursing staff have been shown to improve communication, decrease costs and lengths of stay, and improve job satisfaction. Previous implementation of schedule-based family-centered rounds (SBFCR) on an inpatient cardiology service at our hospital improved nursing participation, adequately allowed for addressing family concerns, and was strongly recommended by staff who participated.

Current State
By June 30, 2019, we will increase nursing presence on daily rounds to >90% (baseline 81%), increase proportion of rounds ending before the Patient Progression Huddle at 11:20am to >80% (baseline 17%), and improve residents™ satisfaction with the rounding process (baseline 25%) on the inpatient pediatric hospital medicine service.

Analysis
A multidisciplinary workgroup developed new workflows to optimize coordination of rounds with two services, nursing, case management and interpreters A tool was developed within Epic that allows the senior resident to create a daily rounding schedule. The schedule was then disseminated to team members and nursing staff each morning.  The tool was implemented on November 26, 2018 and data collection including rounding times, nursing presence, and family presence, is ongoing.  Resident satisfaction was measured with a survey.

Results
Data through April 7, 2019 shows average nursing presence has increased to 91%.  The percent of rounds ending before the Patient Progression Huddle has also increased to 45%.  Resident satisfaction has increased to 81% since initiating the rounding schedule.  SBFCR will become standard of practice starting July 1, 2019, with future plans to account for high team census (ï‚10 patients) and to notify patients and families of their rounding time.
Demystifying antibiotic allergy: delabeling to optimize antibiotic therapy and reduce costs

Elizabeth Lippner, MD; Anne Liu, MD

Stanford, Department of Pediatrics
Background
National organizations including the CDC recognize the importance of antibiotic allergy evaluations in optimizing antimicrobial therapy. Goals of antibiotic allergy evaluations include decreasing surgical site infections, shortening hospitalizations, improving treatment efficacy, improving adherence to treatment guidelines, decreasing colonization with multi-drug resistant organisms, and reducing healthcare costs.  Blumenthal et al. estimated that using a computerized guideline to risk-stratify beta-lactam allergic patients would result in a savings of $8.9-13.7 million annually in drug costs alone across five hospitals, mirroring findings from other hospitals after implementation of antibiotic allergy evaluations.1
Patients with a reported beta-lactam allergy have longer hospital stays, increased surgical site infections, and increased infections with drug-resistant organisms. Yet 90% of patients with a reported penicillin allergy are able to safely tolerate penicillin after an allergy evaluation.  We lack SUH/LPCH specific data on the potential cost-savings benefits or improved patient outcomes  associated with increased beta-lactam allergy delabeling.  Furthermore, SUH/LPCH providers do not have educational resources to risk-stratify patients with a reported beta-lactam allergy that would benefit from allergy delabeling.

Current State
In a prior chart review of 150 SUH patients, only 56% complete reaction history documentation including description of the reaction and reaction type.2 A survey of 67 SUH nurses and physicians identified the following barriers to effective medication allergy documentation in the EHR: (1)Insufficient time (e.g. busy with other responsibilities/tasks and/or allergy documentation takes up time) (2) other tasks have higher priority (3) insufficient knowledge/comfort level to document allergies.2  These data support targeting medical education on medication allergy evaluation and documentation as an intervention. 
By June 2019 we will create an educational online video to increase provider knowledge about how to take an antibiotic allergy history, the logistics and utility of penicillin skin testing, and the improved patient outcomes associated with beta-lactam allergy de-labeling.  We will also implement a series of interactive educational sessions aimed at medical learners and trainees at SUH and LPCH.

Analysis
An interactive, educational online video to enhance/improve knowledge about antibiotic allergies, specifically which patients are at high risk for severe reaction upon re-challenge with the medication in question, and which patients are at low risk for a severe reaction upon re-challenge. 
Following the video the learner should be able to:
a. Identify essential elements of a complete medication allergy history including date of reaction, name of medication, time from ingestion to clinical reaction, organs involved in clinical reaction, severity of the clinical reaction for red flag symptoms, what medical intervention was given, the time course of resolution, and whether said medication or class of medications has been taken since.
b. Based on a clinical reaction history, identify patients that are at low risk of a severe allergic reaction upon re-challenge.
c. Identify the negative impact of carrying a beta-lactam allergy (increased medical costs, increased surgical infection risk, increase risk for resistant organisms, etc).
d. Be familiar with drug allergy skin testing and the difference between graded challenge and desensitization.

Results
Using the Electronic Medical Record (EMR) to Reduce CLABSI Rates in the PICU by Improving Maintenance Bundle Compliance

Anisha Mazloom, MD; Katherine Steffen, MD, MHS; Andrew Palmquist, MSN, RN, NEA-BC; Margie R. Godin, MS, BSN, RN-BC; Derek B. Garnholz, MA

Stanford, Department of Pediatrics
Background
Central line associated blood stream infections (CLABSIs) prolong hospitalizations, increase exposure to infectious disease and antibiotics, and increase costs. This results in unnecessary physical, emotional, and financial burdens for patients and families. Several studies have suggested that improved bundle compliance can lead to a reduction in CLABSI rates.

Current State
Our institution™s PICU adopted a central line prevention bundle in October 2015 to reduce CLABSI rates. However, average monthly compliance with the maintenance bundle was only 6% and CLABSI rates remained higher than desired (2.3 per 1000 central line days). Our aim is to achieve 80% adherence to the CLABSI maintenance bundle in the PICU by June 2019.

Analysis
Daily CLABSI bundle rounds at our institution assessed compliance with ten central line maintenance bundle elements and documented missed elements and reasons for missed elements on process check cards. Physical bundle compliance cards from October 2015 through September 2017 were reviewed. A root cause analysis and identification of key drivers formed the basis for interventions targeted at the most commonly missed bundle elements and reasons for missed elements. Interventions were modified using iterative monthly PDSA cycles. Outcome measures include maintenance bundle compliance and CLABSI rate, as quantified through continued analysis of monthly bundle rounds data. Our institution™s IRB determined this to be a QI project and deemed full review unnecessary. This QI project was supported by the Stanford Children™s Health QI Mini Seed Grant.

Results
The most common missed maintenance bundle elements were IV tubing/fluids dated, not expired (35.3%), Line necessity reviewed (15.7%), and Daily CHG bath (15.6%). The most common reasons for missing items excluding free text special cause responses were Forgot (32.2%), Knowledge deficit (14.1%), and Not a priority (12.7%). Based on this data, the EMR system was utilized to create visual reminders prompting line necessity review (September 2017) as well as a visual icon to easily identify completion status of individual bundle elements and the entire bundle (October 2017). Three months after implementation (December 2017), average compliance was 7% with an average of 1.9 CLABSIs per 1000 central line days. The degree of change was not as expected, and as part of iterative PDSA cycles, bedside interviews were conducted in January 2018. Interviews pointed to a need for incorporation of EMR visual reminders into daily nursing workflow, and revisions to the content and appearance of the EMR icon. Initiatives to implement these changes began in February 2018. Since February 2018, average compliance has been 40% with an average of 1.4 CLABSIs per 1000 central line days. Data continues to be collected and reviewed and iterative PDSA cycles continue on a monthly basis.

While numerous studies have proven the effectiveness of implementing a bundle strategy for reducing CLABSI rates, these bundles only enhance patient safety if there is adequate compliance with the bundle. With improving compliance rates, our institution has seen a 39% decrease in CLABSIs since initiation of this project. Based on our institution™s results, EMR-based tools can aid in improving compliance rates.
Improving Vision Screening in a Primary Care Adolescent Clinic

Brandii Criss, MD; Megan Wong, MD, Jonathan Avila, MD, and Megen Vo, MD

Stanford, Department of Pediatrics - Adolescnet Medicine
Background
While the prevalence of vision disorders in adolescents has not been quantified, it is known that uncorrected vision impairment can lead to permanent vision loss, as well as problems in development, education, and social interaction. Given the significant impact impairment can have on quality of life and improved outcomes with early intervention, AAP recommends selective screening at any age if risk assessment suggests concern, and universal screening at ages 5, 6, 8, 10, 12, and 15 years. Both the American Academy of Ophthalmology (AAO) and American Association for Pediatric Ophthalmology and Strabismus (AAPOS) recommend even more frequent universal screening-every 1-2 years after age 5.

Current State
The Center for Adolescent Health does not routinely screen vision unless requested by patient or family. Our goal is to perform vision screening in 80% of adolescents presenting for well-child visits or new establish care visits in 6 months.

Analysis
At the start of this project, some of the barriers were: provider familiarity with screening guidelines, plan for timing/workflow, vision screen requiring order to be placed, unfamiliar location of vision screen results in EHR, and lack of standardized follow-up for positive vision screens.

Our interventions included: training providers and staff regarding recommendations for vision screening, establishing a universal protocol for obtaining vision screening for adolescents and young adults who present to clinic for a well-child visits or establishing care, creating a workflow for providers and staff, and incorporating vision screen results in EHR.

Results
During the pre-intervention period (9/1/2018 to 1/9/2019), 4% of patients who presented for physicals or to establish care received vision screens (N=2 completed visions screens out of 43 medical visits).  After our first intervention of educating providers and staff on vision screening  (1/10/2019 to 2/19/2019),  14% of patients who presented for physicals or to establish care received vision screens( N=2 completed visions screens out of 14 medical visits).  The second intervention was creating a workflow (2/20/2019 to 3/29/2019) which resulted in vision screens in 25% (N=1 completed vision screen out of 4 medical visits) of patients who presented for physicals or to establish care.  The third intervention of adding vision screen tab results to the EHR and eliminating need for vision screen order (3/20/2019 to 4/19/2019), resulted in 33%of patients who presented for physicals or to establish care (N=2 completed vision screens out of 6 medical visits).

The project is ongoing and is Adolescent Medicine division™s quality improvement project.  Our next steps will be to have continued education and training with provides and staff at quarterly intervals, create vision screen reminder flyers at the nurses station and provider workroom, and having the vision screen results be pulled into the progress note.
Thyroid Diary: A Tracking Document for Monitoring Infants and Toddlers with Congenital Hypothyroidism

Ananta Addala, MD; Laura Bachrach

Stanford, Department of Pediatrics - Pediatric Endocrinology
Background
CH is a common and preventable cause of neurocognitive disability. 35% of children with CH served at LPCH are below the age of 3 and are most vulnerable to permeant damage if inadequately treated. Current guidelines recommend screening labs as frequently as every 1-2 months. However, many children are not getting the required number of screening labs due to distance, the frequency of lab draws, co-management with pediatric endocrinologist and primary care pediatricians. This delay in lab draw can result in inadequate or over supplementation of thyroid hormone, both of which are deleterious to cognitive and physical health.

Current State
To design an easy to understand and aesthetically pleasing document that allows families to remember when lab draws are due and a means of tracking results.

Analysis
It is very important that infants and toddlers with CH are treated are carefully followed by serial labs to prevent neurocognitive deficits. Because of the frequent lab draws, families find it easiest to coordinate labs between their endocrine visits with their PCP or local lab. Despite providing educational material to patients, videos for patients and PCPs and detailed letters to PCPs, lab screening is inconsistently meets current guidelines. Because the screening labs vary with age and it can be difficult for patients to know when they are due for labs. The target population for this QI project will be parents of children, ages 0-3 years, with CH. Each parent will be given an easy to use lab tracker that will tell the results of the prior labs and when the next labs are due. Ease of measure use and perceived improvement in lab management through surveys and focus groups.

Results
At this time, 150 Thyroid Dairy booklets have been printed in English and Spanish and was designed in an iterative and collaborative fashion with providers and our graphic designer. The booklet have now been placed in our Sunnyvale and Palo Alto clinic locations and are being distributed for use. There is a pre-survey that has been collected for each of the participants who receive the booklet.
Assessing and Documenting Code Status among Psychiatric Inpatients: Addressing the Gap

Meenakshi Denduluri, MD; Wilmarie Cidre Serrano, Kelli Smith

Stanford, Department of Psychiatry
Background
Although assessing and documenting code status has been recognized as an essential and required component of the hospital admissions process for over twenty years, the medical literature consistently documents a deficit in code status assessment and documentation for patients admitted to the hospital (Hakim et al, 1996; Anderson et al, 2011).
 
While limited, the literature on the assessment and documentation of code status in psychiatric units suggests an even more concerning deficit.  A 2015 study of 593 psychiatric and medical inpatients found that the rates of code status documentation among psychiatric inpatients were significantly lower (65 vs 96%, p<0.001) compared to medical inpatients (Warren et al, 2016). Currently, there is no formal training in code status discussion and documentation for psychiatry residents.

Our project aims to characterize and improve psychiatry residents™ code status assessment skills and documentation practices with particular focus on geriatric psychiatry inpatients (age >60), whose medical comorbidities make these skills particularly salient.

Current State
Using chart review, we have identified a deficit in code status documentation across all psychiatric inpatient services at Stanford Hospital, including patients with advanced age and multiple medical comorbidities. Given lack of formal training, we have surveyed Stanford psychiatry residents regarding current practices and barriers to proper documentation.
 
Our goal is twofold: to increase the rate of code status documentation for geriatric inpatients by 50% over 3 months following an educational intervention and to improve resident comfort with code status assessment by 50% post-intervention.

Analysis
Of 25 patients, 16 (64%) had no code status ordered, 6 had a prior code status carried forward, and 3 were full code. There were discrepancies between code status order entered and the code status documented in the H&P: 6 patients had no code status documented and 19 were documented as full code with no associated code status order, raising concern for copy-forwarding of code status from the admissions note template.

Based on a survey of psychiatry residents (n=23), primary barriers to code status documentation include: lack of comfort in assessing a patient™s capacity to decide code status, time limitations, lack of integration into current work flow, and lack of training in assessing code status. All survey respondents felt they would benefit from having more training in assessing and documenting code status.

Results
We are designing an educational intervention of didactic, role-play, and in-vivo practice with feedback to improve resident comfort with code status discussions. Pre- and post-intervention surveys will be used to assess effectiveness in improving resident comfort with code status discussion. We will repeat assessment of code status documentation among Stanford psychiatric inpatients 3 months post-intervention.
Management of Suicidal Callers at SHC

S Dina Wang-Kraus, MD; Diane Wu, MD; Filza Hussain, MD; Jose Maldonado, MD

Stanford, Department of Psychiatry
Background
SHC experiences high psychiatric outpatient call burden. Currently, these calls are not screened for suicidality and are routed to the on-call Psychiatry resident, who is untrained in telephone-based interventions for suicidal callers. In a 12 month period, ~60% of On-Call Residents managed suicidal callers. While the gold standard is to train call responders via Applied Suicide Intervention Skills Training (ASIST), which crisis hotlines utilize, on-call residents currently receive no training in managing such acute situations. Furthermore, as the only on-site psychiatrist, the on-call psychiatry resident is responsible for managing all psychiatric needs in the hospital. Ensuring care of a suicidal caller may detract from the care of patients who are currently admitted to the hospital, which may result in delays in patient care and poor patient outcomes. Furthermore, if suboptimal care is provided to a suicidal caller, this may present a risk to the hospital.

Current State
Current State: After hours and on weekends, Page Operators receive calls from patients who may request to speak with a psychiatrist. Currently 0% of page operators report universal screening of callers to psychiatry for suicidality. Callers who request to speak with a psychiatrist are directed to the on-call psychiatry resident, which may involve significant delays. Psychiatry residents currently do not receive training in management of suicidal callers.

SMART Goal #1: Starting in April 2019, 100% of callers endorsing desire to speak with on-call psychiatrist will be screened for acute mental health crisis, specifically suicidality.

SMART goal #2: By July 2019, >90% of callers with active suicidal ideation will receive the gold standard-of-care by being routed to trained call receivers at the National Suicide Prevention Hotline (NSPL). Routing of crisis callers is to occur within a minute of the patient speaking of the page operators.

Analysis
In coordination and collaboration with Page Operators, we have created a workflow that screens every outpatient caller requesting to speak with a psychiatrist for mental health crisis, including suicidality. Callers who screen positive for suicidality will be directed to crisis-trained call receivers through the NSPL. Our workflow has been reviewed and approved by multiple stakeholders including Psychiatry and Page Operator Leadership, Legal, Psychiatry Residents and Page Operators. Piloting of the workflow began in April 2019. Additionally, the principal investigators will become ASIST-trained and will create a resident-specific curriculum with the goal of training all psychiatry residents by June 2020.

Results
Approximately 60% of on-call residents and 90% of Page Operators report they have managed acutely suicidal caller patients in the past 12 months. Over 50% of residents report screening all outpatient callers for suicidality, while 100% of Page Operators report crisis screening is not a part of their current workflow.  No Psychiatry residents have received training in management of suicidal callers at the same quality as ASIST. About 70% of on-call residents and 67% Page Operators report experiencing significant levels of stress and requiring much more clinical time when managing crisis calls. Over 70% residents report these calls impact care efficiency of hospitalized psychiatric patients. 
Design and implementation of a call checklist for psychiatry interns

Csilla N. Lippert, MD; Diana Willard, Sallie DeGolia, Divy Ravindranath

Stanford, Department of Psychiatry
Background
As the night psychiatrist on duty (NPOD), the intern on call is usually the only psychiatrist physically present in the Palo Alto Veterans Affairs hospital during a given call shift. The NPOD is responsible for acute patient care concerns on the inpatient psychiatric and substance recovery units (that have 40 and 18 beds, respectively); psychiatric evaluations in the emergency department; consultations in medical and surgical units; admissions and irregular discharges; outpatient suicide calls; and transfer requests from outside hospitals. While the supervising attending is available by phone at all times, the NPODs previously only had a 22-page attending-developed NPOD handbook as a reference while on call.

Current State
Using the plan-do-study-act framework, we developed and implemented a preliminary call checklist to help NPODs organize their workflow. The development of the checklist was resident driven and has been available for NPOD use since August 2017.

Analysis
To study use of the checklist versus handbook, we conducted a Qualtrics survey of first year, second year, and senior residents in October 2018. In the survey, residents were asked about their frequency and pattern of use of the checklist compared to the NPOD handbook.

The survey response rate was 69%. Nearly 80% of respondents stated they would recommend the checklist to peers. Residents reported saving time when using the call checklist (average of 3.11 hours per resident) and the vast majority reported having fewer questions for their supervising attending because of the checklist. When how the call checklist has improved their ability to provide quality patient care, the most common resident responses were that it taught essential tips and tricks, decreased mistakes, enabled more comprehensive care, and increased confidence. Several residents also indicated that the checklist made call more enjoyable and gave them more time to talk to patients. While residents reported using the checklist for at least half of their shifts and more residents indicated they favored the checklist over the handbook, many residents still reported that they used both resources, including some who preferred the NPOD handbook.

Results
Based on the results that both the checklist and NPOD handbook had high utility for interns and residents on call, we are redesigning both resources so that they work together in a more integrated fashion. A major feature of the redesign was to ensure the checklist includes only brief action items (procedural), condensing it to two pages, while the handbook provides more details (informational).

We are planning a 1-hour intern training session with case scenarios that simulate experiences on-call to help interns practice for more independent work in the middle of the night. We hope that the improvements in the resources available to interns as well as training on how to use the resources will increase their knowledge of and comfort with call procedures by 50% on pre-/post-surveys of their training experience.
Improving electronic delivery of patient information during transfer between hospitals

Jenny H Pan, MD; Jenny H Pan MD, Mainak Chowdhury PhD, Marc Melcher MD PhD, Larry Katznelson MD

Stanford, Department of Surgery (General)
Background
The centralization of medical technology resources and subspecialty expertise in high-volume referral hospitals has resulted in an increase in patients who are transferred between hospitals during the course of their care. These interhospital patient transfers create an additional patient hand-off between hospitals that often do not share the same electronic health records (EHR). This process is not standardized and results in communication breakdown, delays in patient care and duplicate imaging and lab testing. In addition to affecting patient care, these failures of care coordination and administrative complexity cost the US healthcare system over $400 billion annually.

Current State
Although communication breakdown during the interhospital transfer of patients is well-studied and even reported as a leading cause of sentinel events, what is less known are physician attitudes about this process. For this study, we wanted to better understand the frontline provider perspective about this process to identify key problems to target future interventions.

Analysis
We administered a 15 question survey to all general surgery residents querying physician attitudes about the current interhospital transfer process and perceived problems. Included physicians were emailed a link via Qualtrics and all respondents received a $25 Amazon gift card.

Results
We achieved a 93.8% response rate out of 81 residents. On average, respondents reported having an adequate amount of information from the referring hospital 39.1% of the time at the time of writing admission orders. When asked if their understanding of their patient™s clinical status was limited by the availability of information present in the transfer packet of patient information, 85.5% of residents agreed with this statement. Furthermore, 86.8% of surgery residents reported being somewhat dissatisfied or extremely dissatisfied with the transfer process. The most significant problems from the perspective of frontline providers in the interhospital transfer process were incomplete information at the time of arrival, not being able to access patient information before their arrival and information siloing. In conclusion, most surgery residents are dissatisfied with the current workflow of interhospital patient transfers and most of the problems are related to having access to patient information.
Improving Care for Pediatric Trauma Patients: PTSD Screening & Secondary Prevention

Julia M. Chandler, MD, MS; Melanie S. Stroud, RN, BSN, MBA, Stephanie D. Chao, MD

Stanford, Department of Surgery (General)
Background
Up to 30% of children who experience traumatic injury are at risk of developing post-traumatic stress disorder (PTSD). PTSD is diagnosed at a minimum of 30 days after the traumatic event, at which point most trauma patients are no longer following up with their trauma provider. The American College of Surgeons, who certifies trauma centers, recommends screening trauma patients for PTSD, yet Lucile Packard Children™s Hospital (LPCH) had no formal screening program in place.

Current State
We aim to implement a PTSD screening program for children admitted to LPCH following a traumatic injury, with a goal of screening all children ages 7-17 while they are inpatient. Patients will be followed up at 1 and 6 months post-discharge to evaluate their mental health and functional status.

Analysis
Eligible patients fill out two validated screening questionnaires: the Screening Tool for Early Predictors of PTSD (STEPP) (inpatient) and the Child Trauma Screening Questionnaire (CTSQ) (2 weeks post-discharge). Outcome measures include 1- and 6-month scores on the Child PTSD Symptom Scale (CPSS) and the PedsQL, a pediatric quality of life measure.

Results
We have screened 11 patients since recruitment began in January. The program has been well received by patients, parents, and nursing staff. Four children (36%) were found to be at risk for developing PTSD based on the results of the inpatient screening tool. Eight children have been contacted for 2 week follow-up, 4 of whom (50%) have responded. Of the 4 children, 2 (50%) were identified as at-risk of developing PTSD. Interestingly, these 2 children were not the same ones who were identified as at-risk on the inpatient screening tool. Two children have been successfully contacted for 1 month follow-up, with neither being identified as having PTSD. We plan to enroll 100 children in the pilot phase of the study, and we are collaborating with social work to implement PTSD screening as a routine part of care for our pediatric trauma patients.
Attitudes of medical trainees towards Safety Quest: a novel, online, gamified Quality Improvement course

Mia Kanzawa, MD; Mariposa Garth-Pelly, BSN; Kambria Evans, MEd; Laurence Katznelson, MD; Lisa Shieh MD, PhD

Stanford, Department of 
Background
It has been estimated that as many as 98,000 people die in any given year from medical errors that occur in hospitals. Even without accounting for error outside of hospital settings, this number falls within statistics attributed to the top 10 leading causes of death in the US. Reduction in medical errors begins with enhanced knowledge about safety, robust errorreporting systems, increased expectations for safety, and implementation of systems for safe delivery of care1. In particular, quality improvement (QI) and patient safety education for medical trainees on the front-line of care is critical to addressing and preventing medical error. However, implementation of QI curricula is hindered by learner engagement and time constraints from competing educational demands2. Gamification is gaining traction within education and has been shown to improve engagement of learners, including health professionals3–5. Moreover, online learning continues to grow and allows for increased flexibility for learners, as well as widespread distribution. The Stanford School of Medicine therefore developed Safety Quest, an online game for medical trainees that teaches core tenets of QI and patient safety. The purpose of this study was to assess attitudes of medical trainees

Current State


Analysis
1364 residents and 65 medical students at Stanford School of Medicine completed one case in Safety Quest (http://safetyquest.stanford.edu) and provided a ranking in comparison to other independent learning modalities. Learners were also surveyed for free-text feedback. 918 free-text responses from residents were examined using a content analysis.

Results
Safety Quest outranked other independent learning modalities with 45% of residents and 48% of medical students preferring Safety Quest over video, Healthstream modules, powerpoint, and articles. 3 areas of content emerged from resident comments: engagement, ease of use, and effectiveness. Overall, 72% of resident comments were positive, 6% were neutral, and 22% were negative. Engagement was the most frequently addressed content area, contained within 52% of resident comments, and 87% of comments referencing engagement were positive.                                                                                                                                                                                                                                                                                                   Conclusion - Safety Quest, an online gamified approach to QI and patient safety education, is engaging and preferred by medical learners compared to other independent learning modalities. 72% of comments were positive. Future work can build upon themes discovered within this qualitative analysis such as applicability to certain specialties, interface design and clarity of instructions, and depth of content.