The 2021 Quality Improvement & Patient Safety Symposium (QIPSS) will be done digitally.

Event Information

Monday, May 17, 2021

Abstract presenters will be presenting on May 17 from 3:30pm-5:00pm.


45 total submissions

  • 6 education
  • 17 high value
  • 10 innovation
  • 13 Research





Judges & Planning Committee

The judges and planning committee dedicated their time and effort into organizing our third successful symposium.


Download the full presentation schedule for complete details.

Download all presentation posters below

Read Abstracts f0r 2021

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

Abstract Title & Poster

All Authors



Standardization of referrals to facilitate and improve follow up care for Emergency Department consults of the Ophthalmology service

Link to Presentation
Natacha C Villegas, MD
Kathryne Oruna
Andrea Shows, RN
Mariya Levina
Benjamin Erickson, MD, MHS
Heather E. Moss, MD, PhD
Ophthalmology Department Background: Patients with emergent and urgent ophthalmic symptoms and diagnoses frequently receive care in the emergency department setting.  Effective transition of care to the outpatient setting is critical to optimizing patient visual outcomes and satisfaction.  However, this transition can be delayed and sometimes does not occur.  In addition, there is often confusion and frustration on behalf of patients, new patient coordinators, eye center nurses and providers when trying to implement recommended outpatient ophthalmic follow up. 

Purpose: This project aims to ensure that the population of ER patients needing urgent outpatient follow-up in the opinion of the ophthalmology consultant have scheduled outpatient appointments in the recommended window in order to provide continuity of care and optimize ophthalmic outcomes.  Improvements in the appointment making process are expected to have additional positive impact on urgent ophthalmology referrals from other settings.  Streamlined workflows will also increase patient, provider and ancillary staff satisfaction with the process of transitioning ophthalmology care from emergency department to outpatient settings. 

Methods: We began by identifying the current state of our system, by creating a fish bone and swim lane of the process. Through this exercise we became more aware of our current scheduling process and the inherent barriers. We defined our goal as reducing the average time between referral placement and appointment scheduling for ophthalmology referrals from the ER to less than 2 days. Subsequently, we determined key drivers, or subgoals, and the interventions required to attain these goals. Finally, we designed a long-term plan to sustain our goal.

Results: We identified the following key drivers: design an easily accessible and reliable common reference for all the patient scheduling information to be available to all interested parties, passing of responsibility to specific parties with assigned roles, expand clinic schedule availability, improve patient education to include more specific and easily accessible scheduling information. We developed several interventions pertaining to these key drivers, and we were able to decrease the average turnaround time to schedule outpatient appointment from 2.4 days to 1.9.

Conclusion: A multidisciplinary team was key to understanding the process. Defining current state and performing analysis helped identify and motivate interventions. 
Development and Implementation of a Novel Musculoskeletal Ultrasound Curriculum for Physical Medicine & Rehabilitation Resident Physicians

Link to Presentation
Anne Kuwabara, MD
Christina Giacomazzi, DO
Sharlene Su, MD
Nolan Gall, MD
Anita Lowe, MD
Paige Dyrek, DO
David Bakal, MD
Emily Miller Olson, MD
Orthopedic Surgery Department Introduction: There is a rapidly growing interest in the teaching of musculoskeletal ultrasound (MSKUS) in Physical Medicine and Rehabilitation (PM&R) residency training. However, there are no current guidelines for a formal ultrasound curriculum, which inspired the creation of a longitudinal curriculum resource for musculoskeletal ultrasound training at our institution.

Methods: Core MSKUS topics were separated into nine units and arranged in order of difficulty. Appropriate supplemental learning resources were identified. A knowledge quiz and a thorough scanning checklist were developed and approved by multiple providers. Initiated in the spring of 2020, all 21 current residents were invited to participate on a voluntary basis. Participants were asked to complete a pre- and post curriculum intervention survey. Statistical significance was determined using a paired t-test analysis and defined as an alpha less than 0.05.

Results: 18 residents completed at least one session. 13 residents completed both the pre and post curriculum surveys. Survey results indicated that the majority of residents did not have any formal diagnostic or interventional ultrasound education prior to the rollout of the program and were interested in incorporating ultrasound in their practice. After the MSKUS curriculum implementation, all respondents reported improved knowledge including statistically significant (less than 0.05) improvements in self proficiency, confidence performing diagnostic musculoskeletal ultrasound for common injuries, confidence performing interventional ultrasound safely and effectively, and having knowledge of where to find resources to supplement ultrasound competence.

Conclusion: Ultrasound diagnostics and interventions are becoming increasingly prevalent in musculoskeletal medicine and literature supports its use for both diagnostic purposes and procedural guidance. Implementing a formal MSKUS curriculum improves resident physician self-proficiency and confidence with MSKUS. This curriculum may provide a structure for other PM&R residency programs to implement a MSKUS didactic curriculum.
Clinical Care Redesign of the Geriatric Hip Fracture Pathwy

Link to Presentation
Noelle Van Rysselberghe
Lauren Shapiro
Tanmaya Sambare
Blake Schultz
Andrea Fox
Alicia Wilson
Robin Kamal
Amy Lu
Ankur Bharija
Michael Gardner
Orthopedic Surgery Department Background: 
Geriatric hip fractures are associated with significant morbidity, mortality, and economic burden, yet treatment remains variable. The purpose of this work is to create and implement a high-value, evidence-based clinical care pathway for geriatric patients sustaining hip fractures and to measure established quality metrics pre and post implementation.
We began by creating a current-state process map of our institution’s pathway for geriatric hip fracture management. We identified quality measures and CPGs from literature review and applied those to the map to identify areas where we could add value. We then utilized a multidisciplinary team from geriatric medicine, orthopaedics, nursing, and anesthesia to redesign our protocols and order sets prioritizing geriatric-friendly medication regimens, mobility assessments within 24 hours post-op, documentation of DPOA/family caregiver as well as What Matters Most within 24 hours of admission, among others. We also launched a dedicated fracture liaison service to capture patients in need of ongoing osteoporosis care post-discharge.

Following pathway launch in February 2021, we compare outcomes between a pre- and post-implementation cohort of patients with geriatric hip fractures. Primary outcome is length of stay, discharge location, and 30 day readmissions. Secondary outcomes to be compared at a later date include time to surgery, time to medicine evaluation, time to pain consult, use of regional and/or multimodal anesthesia, mobility scores at 48 hours post-op, completion of osteoporosis work up, subsequent fragility fracture, cost, and variability.

Length of stay over the five years prior to implementation (n = 955 patients) averaged 5.9 +/- 6.3 days compared to 5.2 +/- 2.6 days in the post-implementation cohort (n = 31) (p = 0.89). There have been no 30 day readmissions thus far (pre-implementation rate 4%). The Fracture Liaison Service captured >80% of all hip fracture patients for osteoporosis work-up and treatment as indicated. Other high value interventions implemented include daily tracking of AMPAC mobility scores, creation of a patient/family education flier “What to Expect” after hip fracture, and ongoing bi-monthly multi-disciplinary working groups to continue to improve our protocol.

Multi-disciplinary teams can add significant value for geriatric patients with hip fractures. Our early data suggests that while length of stay may be similar during the implementation phase, variability in length of stay may already be starting to decrease. Future work will hopefully report an improvement in mobility scores, readmission rates and percent of patients beginning osteoporosis treatment.
What time will you be (a)round?: A quality improvement project to increase family notification of rounding time for schedule-based family centered rounds

Link to Presentation
Jessica Moriarty
Jessica Moriarty
Marie Wang, MD, MPH
R. Marissa Hutauruk, RN
Heidi Platfoot, BSN, RN
Danielle McFayden, BSN, RN
Jenina Chang, BSN, MSN, MBA
Amit Singh, MD
Pediatrics Department Introduction:
Schedule-based family-centered rounds (SBFCR) provides a rounding schedule to facilitate patient and family participation in shared decision-making with the healthcare team. Variable notification of families of the rounding time-frame leads to inconsistent family presence and involvement in rounds.

Increase the percentage of time patients and families are notified of the anticipated rounding time-frame on the pediatric hospital medicine (PHM) service from 0% to 80% by June 2021.

SBFCR began in November 2018 on the primary PHM service at a university-affiliated children’s hospital. Family notification of the anticipated rounding window by nurses began in June 2019 once the scheduled rounds process was well established. Our multidisciplinary team focused on the following key drivers: enhancing stakeholder knowledge of SBFCR process and standardizing and streamlining the notification process across units. Interventions included: 1) Presentations at nursing staff meetings, ongoing reminders at nursing huddles and weekly check-in meetings 2) Distribution of SBFCR badge cards to nurses with information about the process, 3) Standard weekly check-in meetings for the SBFCR family notification process and 4) Distribution of family notification cards with the rounding window (+/- 30 minutes of scheduled time) by unit staff. Family notification was measured by asking families if they were notified of the rounding time that day 1-2 times per week, and conducted by a case manager or physician beginning in September 2019.

As of February 2021, family notification of the rounding window increased from 0% to 53% (Figure 1). Weekly audits with healthcare team stakeholders revealed the following challenges: lack of standard process adherence among units and challenges adhering to the process during high census times. Patients and families shared appreciation for notification of the rounding time-frame which they perceived enhanced participation in rounds.

Developing a standardized patient and family notification process enhanced notification of rounding time-frame for SBFCR; however, variable notification persists particularly during high census times. Future interventions should work towards more automated messaging of families, possibly using in-room technologies or the electronic health record.
Barriers in Telemedicine Access in Senior Living Communities

Link to Presentation
Alice Mao, MD
Lydia Tam,
Audrey Xu,
Kim Osborn MPA,
Christine Gould PhD,
Marina Martin MD,
Matthew Mesias MD
Medicine Department Introduction:
During the COVID-19 pandemic, older adults are missing routine care appointments despite increasing availability of telemedicine video visits. We conducted a needs assessment of two Residential Care Facilities for the Elderly (RCFE) in Northern California as a first step to improving access to telemedicine visits for older community dwelling individuals.

We conducted voluntary surveys of the independent community dwelling adults of two RCFEs. Site A houses residents who are mostly Caucasian and middle and upper middle class. Site B provides subsidized senior housing and serves a large group of residents who are non-English speakers. Surveys ascertained residents’ preferred devices as well as comfort level, support, and barriers regarding telephonic and video visits.

Of the 700 surveys distributed, 249 surveys were completed and returned (36%). The average age of participants was 84.6 (SD = 6.6) and 77% were female. At site A, 89% of participants had a bachelor’s degree or beyond and 99% listed English as their preferred language. At Site B, 43% had a bachelor’s degree or beyond, and 13% preferred English while 73% preferred Mandarin. Regarding remote visits, 37% of all participants felt comfortable connecting with their healthcare team through video visits with computer being the most preferred device (23%) followed by smartphone (19%) and iPad/tablet (11%). Regarding perceived barriers, there were substantial differences depending on the site. Participants at Site A reported not knowing how to connect to the platform (24%), not being familiar with the technology (22%), and difficulty hearing (14%)as the top three barriers, whereas for the participants at Site B, the top three barriers were not being able to speak English well (55%), lack of interest in seeing provider outside of clinic (35%), and not knowing how to connect to the platform (35%).

Significant barriers exist for older adults in RCFEs with telemedicine visits with their care team. The largest barriers include difficulty with technology, language barriers, lack of devices or stable internet, and lack of desire to see provider outside of clinic. Due to site specific differences in reported telemedicine barriers, any intervention to improve access should be tailored to the specific needs of that site.
Implementation of a Standardized Antibiotic Prophylaxis Protocol for Acute Open Fractures

Link to Presentation

Taylor R. Johnson
Dr. Noelle Van Rysselberghe, MD

Dr. Divya Kurtian, MD

Dr. Michael Bellino, MD
Dr. Julius Bishop, MD

Dr. Jenny Aronson, MD

Dr. Lina Meng, PharmD

Dr. Edwin Lim, PharmdD

Dr. William Dixon, MD

Dr. Amy Chang, MD
Dr. Michael Gardner, MD

Orthopedic Surgery Department
Problem State/Background:
Open fractures are associated with significant morbidity due to high rates of post-operative infection and wound healing complications. Traumatic fractures are responsible for up to 19% of cases of osteomyelitis1 and infection rates range from 1 to 25% depending on the severity and location of injury. Standardized, evidence-based antibiotic protocols may result in decreased complication rates for open fractures by reducing times to antibiotic administration and simultaneously optimizing antibiotic prescribing and stewardship among providers. Despite available evidence-based prescribing guidelines, a standard antibiotic prophylaxis protocol for open fractures has not yet been initiated within the Stanford Healthcare System. Currently, antibiotic choice for open fractures, particularly those with a Gustilo-Anderson type III and above, remains variable according to provider preference. The purpose of this work is to implement and sustain an evidence-based antibiotic prophylaxis protocol for the treatment of open fractures with the goal of decreasing the morbidity of infection related complications over time and ultimately improving patient outcomes.

Project Aims:
To reduce the incidence of infection-related complications including septic nonunion, soft tissue infections requiring reoperation, and osteomyelitis by 10-15% in adult patients diagnosed with an open fracture via the implementation of a standardized, evidence-based protocol for antibiotic prophylaxis by January 1, 2022.

Interventions and Measures:
We have developed a pathway for standardized antibiotic administration within the Emergency Department and on the Orthopedic Surgery floor using current evidenced based guidelines from the literature. Educational sessions integrated into existing resident and faculty conferences and lectures and printed media displaying the new protocol will be used to incorporate the pathway into current clinical practice. Primary Outcome Measures will include rates of septic non-union, osteomyelitis, or any deep soft tissue infection in open fractures status post implementation of the protocol after a 12 to 24-month period. Additional primary measures include the average time from injury to antibiotic administration, rates of superficial soft tissue infections treated non-operatively, and reoperation rates status post implementation. Secondary measures include adverse reactions to antibiotics. Process measures include the percentage of patients who receive the correct first dose of evidence based antibiotic prophylaxis within 60 minutes of presentation to the emergency department based on the protocol and the utilization rate of a specified EMR order set which will include the protocol.

Plan for Statistical Analysis and Reporting of Results:
This study will be historically controlled and data will be collected from the medical record. Patients > 18 years of age with an ICD10 code representative of an open fracture will be identified in the pre-implementation and post-implementation period. The fractures will be further classified by the Gustillo Anderson Classification and by location and type of fracture. 
Descriptive statistics for each response variable will be calculated . Infection-related complication rates and re-operation rates for open fractures status post-surgical fixation, will be analyzed 24 months prior to implementation of a standard antimicrobial prophylaxis protocol and compared to the complication and reoperation rates 0-6 months (roll-out phase) and 6-24 months post implementation. Analysis of variance (ANOVA) testing will be used to determine the statistical significance of variables and determine the impact of the protocol implementation on the morbidity and mortality of infection related complications for open fractures overtime. 
Implementation of a Pediatric Status Epilepticus Pathway to Improve Quality Care

Link to Presentation
Katherine Xiong, MD, PGY-3
Prathyusha Teeyagura, MSBI
Amelia Sperber, MSN, RN, CNS, CPNP-PC
Courtney Wusthoff, MD
William Gallentine, DO
Neurology Department Purpose
Status epilepticus (SE) is associated with long-term consequences including cognitive impairment, irreversible brain injury, and death (1). Studies demonstrate that longer-duration seizures are more treatment resistant and associated with worse clinical outcomes (2), whereas successful early treatment of SE can result in less respiratory suppression/respiratory failure, decreased critical care utilization and subsequent charge mitigation (3). Inadequate dosing and delays in administration of antiseizure medications (ASM) in SE management occur frequently, and has been well documented by numerous studies (2). Furthermore, appropriate utilization of SE pathways have been shown to improve timing of AED administration (4).

Lucille Packard Children’s Hospital-Stanford (LPCH-S) implemented a clinical pathway for SE in June 2019, with only 35% of patients managed per the SE pathway prior to its distribution, and this rate actually decreasing following formal distribution of the clinical pathway to 18%. The purpose of this project is directed towards increasing adherence to the SE pathway, with hopeful secondary goals of decreased escalation to second line ASM agents and critical care utilization.

We performed a retrospective assessment of patients admitted to LPCH-S with SE 3 months prior to and 9 months following the implementation of a SE clinical pathway.  RCA was performed using a fishbone diagram. Assessment surveys were conducted across the RN and pediatric resident MD providers regarding knowledge of the SE pathway as well as subjective sources for medication administration delay within the pathway. Multiple staged educational RN-directed multi-modal interventions have been implemented since August 2020, with MD-directed interventions planned to begin in April 2021. Ongoing chart review post-implementation will be performed to determine the success of the intervention.

Results/Anticipated Results
Only 6/33 (18%) of patients admitted from June 2019-February 2021 followed the clinical pathway. 19/33 patients (58%) patients required admission to the PICU and all 19 (100%) of the patients admitted to the PICU did not follow the pathway. Deviation from the pathway occurred in several manners, including time to first, second and third line ASM administration and choice of second line ASM.

On MD provider survey, 50/102 residents (49%) across PGY1-3+ responded. Of the responders, only 24/50 (48%) reported familiarity with the LPCH-S SE pathway. Delays to ASM administration were felt by MD providers to be due to delays from pharmacy (22 responses), nursing administration (15 responses), delays in physician order entry (15 responses) and other delays including availability of medications in the Omnicell.

RNs across critical care and acute care units were also surveyed, with 85 total responses. They reported factors driving ASM administration including physician order entry (31%), pharmacy delays (25%), medication availability (in Omnicell) (22%), medication preparation (16%) and a small portion of other factors (6%).

Current SE pathway utilization at LPCH-S is poor, likely due to RN and MD providers’ limited familiarity with the SE pathway. Delays in ASM administration are perceived to be multifactorial. Efforts to increase provider education on SE pathway are ongoing, as are efforts to improve inefficiencies within the pathway

1. Ostendorf AP, Merison K, Wheeler TA, Patel AD. Decreasing Seizure Treatment Time Through Quality Improvement Reduces Critical Care Utilization. Pediatric Neurology. 2018;85:58-66. doi:10.1016/j.pediatrneurol.2018.05.012
2. Gaínza-Lein M, Fernández IS, Ulate-Campos A, Loddenkemper T, Ostendorf AP. Timing in the treatment of status epilepticus: From basics to the clinic. Seizure. 2019;68:22-30. doi:10.1016/j.seizure.2018.05.021
3. Ostendorf AP, Merison K, Wheeler TA, Patel AD. Decreasing Seizure Treatment Time Through Quality Improvement Reduces Critical Care Utilization. Pediatric Neurology. 2018;85:58-66. doi:10.1016/j.pediatrneurol.2018.05.012
4. Trau SP, Sterrett EC, Feinstein L, Tran L, Gallentine WB, Tchapyjnikov D. Institutional Pediatric Convulsive Status Epilepticus Protocol Decreases Time to First and Second Line Anti-Seizure Medication Administration. Seizure. 2020;81:263-268. doi:10.1016/j.seizure.2020.08.011
Improving Documentation of Seizure Events at Lucile Packard Children’s Hospital

Link to Presentation
Katherine Xiong, MD, PGY-3
Amelia Sperber, MSN, RN, CNS, CPNP-PC
Prathyusha Teeyagura, MSBI
William Gallentine, DO
Courtney Wusthoff, MD
Neurology Department Purpose
Appropriate documentation of clinical events is critical for accurate assessment of institutional performance on quality measures and patient safety events (1-3). Status epilepticus (SE) management is a key quality measure within child neurology, where metrics like time to first/second rescue anti-seizure medication (ASM) administration and seizure duration are followed (4).

In 2020, we attempted to assess the rates of adherence to a recently implemented SE clinical pathway at Lucile Packard Children’s Hospital – Stanford (LPCH-S). During the retrospective chart review, we noted that accurate seizure event documentation was infrequent and highly inconsistent in content. The purpose of this project was to investigate, improve and standardize current seizure event documentation practices. 

We performed a retrospective assessment of seizure event documentation in patients admitted to LPCH-S for video EEG (vEEG) monitoring from May -July 2020. RCA was performed using a fishbone diagram.  Assessment surveys were conducted across the acute and critical care RNs assessing knowledge and education of seizure clinical event documentation expectations and current documentation practices. Multiple RN-directed educational interventions have been implemented since August 2020, and a hospital policy for seizure precautions and response was established in February 2021. Ongoing chart review post-implementation will be performed to determine the success of the intervention on a bi-annual basis.

Results/Anticipated Results
On pre-intervention chart, of the 19 seizure events reviewed, only 9/19 (47.3%) events had appropriate clinical documentation of the seizure event.

Of 85 RN responses (56% PICU, 44% ACU), our assessment survey demonstrated that while nearly all (95%) RNs reported knowing how to document seizure events in the EMR, only 46/85 RNs (55%) reported receiving any training re: documentation, and this training was primarily informal through work experience. Our results also demonstrated that RN seizure documentation practices were inconsistent, with variability in documentation of essential seizure event details (duration, time of onset, semiology, time of medication administration)

Repeat chart audit of vEEG admissions is being conducted from April 09 onwards. Preliminary data (as of 05/01/21) demonstrates significant improvement in the percentage of seizure events that have appropriate clinical documentation following our interventions, increasing from 47.3% to 87.5%. We anticipate that with ongoing continuing education, clinical documentation will continue to improve.

Accurate documentation of seizure events details, including seizure start time, seizure duration, event description and medications administered, is essential for accurate assessment of institutional performance in SE quality metrics. RN education on clinical documentation practices is often limited and not standardized, leading to inconsistent documentation practices. Improving clinical documentation requires widely distributed, continuous formalized education initiatives as well as establishment of hospital policy guidelines.

1. Knightly JJ, Meyer SA, Weiss BB, et al. 117Accurately Dead or Alive: A Neurosurgical Review of Quality Patient Care and Outcomes. The Importance of Data Fidelity in Calculating Quality Metrics Utilizing University Health System Consortium (UHC) Clinical Administrative Database. Neurosurgery. 2013;60(CN_suppl_1):158-159. doi:10.1227/01.neu.0000432709.63380.89
2. Clinical Documentation for Intensivists: The Impact of Diagnosis Documentation | Ovid. Accessed April 27, 2021. https://oce-ovid-com.laneproxy.stanford.edu/article/00003246-202004000-00017/HTML
3. Loftus T, Najafian H, Pandey SR, Ramanujam P. The Impact of Documentation Training on Performance Reporting. Cureus. 7(7). doi:10.7759/cureus.283
4. Fountain NB, Van Ness PC, Bennett A, et al. Quality improvement in neurology. Neurology. 2015;84(14):1483-1487. doi:10.1212/WNL.0000000000001448
Establishing a sustainable CGM remote monitoring program for new onset type 1 diabetes patients

Link to Presentation
Ming Yeh (Mindy) Lee
Jeannine Leverenz, RN, CDE,
Brianna Leverenz, BS,
David Scheinker, PhD,
David Maahs, MD, PhD,
Priya Prahalad, MD, PhD
Pediatrics Department Background: 
LPCH diabetes clinic uses continuous glucose monitors (CGM) and telehealth platform to monitor new onset type 1 diabetes (T1D) patients and provide medical advice remotely. This service can improve patient glycemic control, and is valued by patients and providers. However, the program capacity is limited by intensive provider time requirement.
The objective is to reduce endocrinology provider time spent on CGM remote data review by 20% while sustaining patient glycemic control by Spring 2021.
QI methodology was applied to study and improve the remote monitoring workflow. Data on provider work time is tracked using MyChart message, and patient glycemic control is collected via chart review. Provider satisfaction was assessed through unstructured interviews. An analytic dashboard was developed to facilitate remote monitoring. The dashboard uses glucose trend alert threshold set by expert consensus, it was validated using historic data, and iterative improvement cycles minimized missed opportunities to intervene on glycemic control while maximizing provider time savings.
The process workflow, key drivers, and countermeasures were identified through the QI process. Subsequent interventions over 3 study periods reduced the number of patient charts requiring provider review by >57% and average provider review time from 4.5 minutes per patient per week to less than 2.7 minutes per patient per week. Provider satisfaction is high with the newly implemented workflow. Patient glycemic control was not adversely impacted by the interventions, as demonstrated by similar mean glucose, time in range, time in hypoglycemia, and time in severe hypoglycemia throughout baseline and each study period.
We analyzed the CGM remote monitoring workflow and developed a clinical decision support dashboard that accurately flags T1D patients requiring medical advice, thus reducing provider work burden by >57%. Provider work time is reduced by >40% from 4.5 to <2.7 minutes per patient per week. Provider satisfaction is high with the new workflow, suggesting that expansion of the program would have buy-in from key stakeholders. Importantly, the interventions did not adversely impact the balance measure of patient glycemic control. Further studies are needed to evaluate long-term patient glycemic control and health outcomes, as well as the scalability and sustainability of the program independent of grant support.
Using a Micropuncture Introducer Kit to Facilitate Intensivist-initiated Peripheral ECMO Cannulation to Reduce Time to Full Flow

Link to Presentation
Alberto Furzan, MD
Robert Arrigo, MD
Erin Hennessey, MD
Anesthesia Department Project Summary

Peripheral venous-arterial Extracorporeal Membrane Oxygenation (pVA-ECMO) cannulation has historically been performed by cardiac surgeons, but increasingly, intensivists, cardiologists, and emergency physicians are cannulating patients for ECMO. Because cannulation is often done on an urgent or emergent basis, and because of limited pVA-ECMO cannulation sites, first-pass success of femoral artery cannulation is essential. Complications from this procedure can be serious including hematoma, infection, nerve injury, retroperitoneal bleed, and failure to cannulate the target. Micropuncture kits have shown a benefit decreasing incidental bleeding or vessel trauma,[1-3] and it provides a margin of safety given its small size and may help to reduce pseudoaneurysm formation.[4] Micropuncture Introducer Kits are routinely used by cardiac surgeons and vascular surgeons for cannulation of calcified or otherwise challenging arteries, including the common femoral artery for pVA-ECMO cannulation. 

By survey, Stanford Critical Care Medicine (CCM) fellows who have not previously completed an ACTA fellowship generally are not familiar with Micropuncture Introducer Kits. Our institution stocks the Cook Micropuncture Introducer Set with 21g finder needle, flexible 40 cm 0.018” guidewire, and 10 cm 5.0Fr Sheath in the main operating room supply cores and in central supply, but they are not readily available in any of our six ICU locations. Stanford CCM fellows generally have not cannulated a patient for pVA-ECMO. Educating CCM fellows on how to successfully place arterial and venous femoral lines specifically for pVA-ECMO will facilitate CT surgery upsizing for ECMO cannula placement.

This document describes the first part of a two-year project. During this fellow cycle (2020 – 2021), we aim to have Micropuncture kits stocked at all Stanford ICU sites including E2, J2, J4, K4, M4 and L4, and to increase CCM fellow knowledge of their availability, and familiarity with kit components to 100% by June 1st, 2021. We will achieve this by organizing an on-site workshop with CCM fellows to educate them about where to find kits and give them hands-on experience. The second part of the project (2021 – 2022 fellow cycle), Dr. Furzan et al. will incorporate Micropuncture kit experience into a pVA-ECMO cannulation workshop. 
Technical aspects

Micropuncture kit
Introducer Access Set Micropuncture® 5.0 Fr. X 10 cm Length Outer Catheter, 0.018 Fr. X 40 cm Length Wire Guide, 21 Gauge X 7 cm Length Needle (Cook Medical, Bloomington, IN).

The micropuncture needle is used to cannulate a vessel under ultrasound guidance, using the traditional Seldinger technique. No skin incision is necessary after wire placement. The 5.0 Fr Sheath is inserted and can later be upsized to a larger catheter or cannula as needed. The optimal arterial target for pVA-ECMO is generally the common femoral artery distal to the inguinal ligament (to reduce retroperitoneal bleeds) and proximal to its bifurcation into the superficial femoral artery and the profunda femoral artery (to ensure an adequately-sized target for pVA-ECMO cannulation, often 20 Fr in adults.)[5]
PDSA Cycle - Year One of Two-Year Project

-Survey fellows on familiarity and experience with Micropuncture kits and pVA-ECMO cannulation.
-Ensure adequate stock of Micropuncture kits in all Stanford ICU locations.
-Train CCM fellows in the use of micropuncture kits in patients who are considered difficult (challenging anatomy, peripheral arterial disease) or at high risk for complications (body habitus, coagulopathy, hematoma.)
-Hypothesis: fellows will gain experience and familiarity with Micropuncture kits that will prepare them for a future pVA-ECMO workshop.

-Coordinate a workshop for fellows with hands-on access to Micropuncture kits and education about where they can be found and how to use them. 
-Carry out workshop using vascular access models (provided by the critical care department)
-If fellows are unable to attend the workshop, provide 1:1 session.

-Survey fellows on their knowledge of where Micropuncture kits are kept, whether they’ve used the kits in practice, and whether they feel adept in their usage.
-Survey fellows on whether they’ve participated in pVA-ECMO cannulation.

-Adopt: continue to expand Micropuncture kits to all ICU supply cores and hold additional usage workshop at the start of the next fellow cycle so fellows will gain more experience with the kits prior to a pVA-ECMO workshop (likely Spring 2022.)

State of the Project

Fellows nearly-universally expressed interest in the project. Only dual ACTA/CCM fellows (pre- or post-ACTA year) had experience with Micropuncture introducer kits. After working with several nurse managers, Cook kits are now stocked in E2 and L4 (in addition to the OR & Cath Lab cores) and will soon be stocked in the remaining ICU supply cores. A workshop was arranged with models, kits and educational material that was carried out in March 2021 and was well-attended. Few fellows have directly participated in pVA-ECMO cannulation, which is still driven by the Cardiothoracic Surgery on-call ECMO service. By repeating the Micropuncture workshop early in the coming fellow year (2021 – 2022, we hope that a subsequent pVA-ECMO workshop in the Spring of 2022 (Part Two of this project) will be widely-attended and beneficial. 
1.Ambrose JA, Lardizabal J, Mouanoutoua M, Buhari CF, Berg R, Joshi B, El-Sherief K, Wessel R, Singh M, Kiel R. Femoral micropuncture or routine introducer study (FEMORIS). Cardiology. 2014;129(1):39-43. doi: 10.1159/000362536. Epub 2014 Jul 9. PMID: 25012707.
2.B0rg0 E, Durmaz H. Placement of hemodialysis catheters with the help of the micropuncture technique in patients with central venous occlusion and limited access. Turk J Med Sci. 2021 Feb 26;51(1):95-101. doi: 10.3906/sag-2006-11. PMID: 32892538; PMCID: PMC7991868.
3.Ben-Dor I, Sharma A, Rogers T, Yerasi C, Case BC, Chezar-Azerrad C, Musallam A, Forrestal BJ, Zhang C, Hashim H, Bernardo N, Satler LF, Waksman R. Micropuncture technique for femoral access is associated with lower vascular complications compared to standard needle. Catheter Cardiovasc Interv. 2020 Oct 16. doi: 10.1002/ccd.29330. Epub ahead of print. PMID: 33063926.
4.Gutzeit, A., Schoch, E., Reischauer, C. et al. Comparison of a 21G Micropuncture Needle and a Regular 19G Access Needle for Antegrade Arterial Access into the Superficial Femoral Artery. Cardiovasc Intervent Radiol 37, 343–347 (2014). https://doi.org/10.1007/s00270-013-0669-0
5.Pavlushkov E, Berman M, Valchanov K. Cannulation techniques for extracorporeal life support. Ann Transl Med. 2017 Feb;5(4):70. doi: 10.21037/atm.2016.11.47. PMID: 28275615; PMCID: PMC5337209.
Special Acknowledgements
Thank you to Dr. Javier Lorenzo for supporting our project and providing vascular access models for the workshops. Thank you to Emily Watkins and Chris Hall, Cook Medical representatives, for assistance with setting up the workshop and providing kits for hands-on practice. 
Improving Controlled Substance Contract Compliance During a Pandemic at Stanford’s Senior Care Clinic

Link to Presentation
Jennifer Nguyen, MD
Silvia Tee, MD
Medicine Department Background: Controlled substance contracts are an important tool to help with regulation of patients with chronic opiate use. We describe a quality improvement effort to improve contract compliance rate at Stanford’s Senior Care Clinic through a stepwise intervention and during a global pandemic where telemedicine is a predominant modality for patient care.

Methods: Clinic providers were surveyed, and the preferred modalities chosen based on their responses were in-person visits, online messaging, mailed contracts, and individual chart reminders. An electronic medical records (EMR) dashboard generated a list of patients with chronic opiate use without an active contract. The following stepwise intervention was performed to improve compliance. Each provider and their medical assistant (MA) received a list indicating which patients had an upcoming in-clinic visit within the next month. For the remaining patients, the MA’s were asked to send the contract electronically through online messaging.  If after 30 days the contract was not returned online, the contract would be mailed. In addition, a manual chart reminder was placed in each patient’s chart indicating the date of their last completed contract. After 2 months, the data was then evaluated through chart review.
Results: The EMR generated a list of 17 patients without contracts. Out of the 17 patients, 1 patient was hospitalized so 16 patients were included in the project. 5 out of 5 (100%) patients with in-person visits were able to complete their contract in the clinic. Of the 11 remaining, 7 were sent the contract electronically. 3 out of 7 (<50%) patients completed the contract electronically. After no online response, a cover letter and contract were mailed to the remaining 4 patients along with 4 patients who did not have online access. 4 out of 8 (50%) mailed back the contract. With these efforts, we achieved an increase of compliance from 5% to 76% with in-person visits being the most effective modality.

Conclusions: Our interventions of using electronic outreach and mailed contracts improved compliance rate during the pandemic. We found that electronic contracts were technically challenging for the senior population. Mailed contracts achieved better compliance. We suspect that a telephone reminder and pre-stamped return envelope along with the mailed contract may yield a 100% return rate during a time when in-person visits are not as accessible.

Bite Injury Prevention in Patients Undergoing Motor Evoked Potentials

Link to Presentation

Shahla Moghbel, DO
Leslie Lee, MD
Scheherazade Le, MD
Sungho K Cho, MD

Richard Jaffe, MD

Sarah Stone, MD

Jaime Lopez, MD


Neurology Department Transcranial electrical stimulation (TES) is a standard and widely used technique for Intraoperative motor-evoked-potential (MEP) monitoring to prevent motor deficits in high risk surgeries. TES, however, increases risk of bite injuries due to high voltage and current stimulation causing forceful jaw clenching and possible shifting of the protective bite block. In our institution, we recently observed an increased number of bite injuries and tongue lacerations with MEP use. In this study, we propose a multidisciplinary approach to improve the quality of bite block placements and maintenance of their positions for the duration of the surgery. We developed a step by step checklist available for anesthesia colleagues and Intraoperative Neurophysiologic Monitoring (IONM) technologists to assure proper choice, placement of the bite block, and frequent inspection.
Greening the Clinics: Improving Environmental Sustainability at the Stanford Redwood City Outpatient Center

Link to Presentation
Anita Lowe Taylor
John Chan
Josh Levin
Paige Fox (faculty lead)
Orthopedic Surgery Department Background and Problem Statement: The World Health Organization declared climate change to be "the greatest threat to global health of the 21st century.” Pollution associated with healthcare causes the loss of 405,000-470,000 disability-adjusted-life-years annually in the United States. The healthcare system contributes 10% of carbon emissions and 9% of harmful non-greenhouse air pollutants in the United States.
SMART Goal: Our SMART goal was to reduce the amount of per-patient clinical and administrative waste in selected Stanford Physical Medicine and Rehabilitation (PM&R) and Orthopedic outpatient clinics by 30% by May 2021 and to create a template for implementing these waste-reduction strategies in other clinics.
Methods: Root cause and key driver analyses were performed on the PM&R Sports and Spine Clinics at the Stanford Redwood City Outpatient Center. A waste audit was conducted for traditional outpatient clinic visits, outpatient visits during which a procedure was performed, and visits at the Outpatient Surgical Center (OSC). 
Results: Despite representing the lowest number of patient visits by volume, visits at the OSC produced by far the most waste overall. In fact, an OSC visit created up to 6x the waste of a procedure visit in clinic.  Different clinic procedure set-ups produce dramatically different waste patterns. Due to recycling company policies, clinical waste is not recycled. Of non-clinical waste that is recycled, recycling costs 7x less than solid waste disposal. When looking at patient transportation to visits, an estimated 5,000+ video visits in 2020 saved an estimated 270,000+ driving miles equating to approximately 11,000+ gallons of gasoline.
Conclusion:  The PM&R Sports and Spine Clinics could save an estimated 1.17 tons of waste annually by adopting waste reducing clinic procedure practices. Further waste and emission reductions could be obtained by minimizing procedures performed in the OSC and by expanding telehealth.  Next steps include 1) Additional educational campaigns to providers, clinic staff, and building stakeholders, 2) Assessing if recycling can be implemented in select clinic areas, 3) Establishing “Green Teams” within clinics, and 4) Expanding to other departments.
Inpatient Process to Obtain Pathology Specimens from Outside Hospitals

Link to Presentation
Samuel Falkson
Lisa Shieh, MD, PhD
Medicine Department Purpose:
Stanford inpatient teams are not always familiar with the process to obtain pathology specimens from an outside hospital (OSH) when a patient transfers to Stanford. Before treatment based on pathology interpretation proceeds for these patients, Stanford pathology must confirm the patient’s diagnosis either by analyzing pathology specimens collected at the OSH or by re-biopsying the patient. Therefore, timely receipt of outside pathology specimens is an integral part of the care of inpatients transferring to Stanford. This project was initiated due to a common experience among inpatient service teams at Stanford regarding perceived inefficiencies with the process of requesting pathology specimens from an outside hospital. Unfamiliarity with the process of obtaining outside pathology specimens results in excessive clinician time being spent performing this task, delays in appropriate care of patients, and sometimes the need for re-biopsies in-house.

Multiple stakeholders involved in obtaining pathology specimens from outside hospitals were interviewed regarding their experiences with this process. The stakeholders included attending physicians from internal medicine, hematology/oncology, and pathology, fellows in hematology/oncology and pathology, residents in internal medicine, an oncology clinic new patient coordinator, and a case management liaison. Additionally, a survey of Stanford internal medicine residents on topics regarding obtaining outside pathology specimens was performed.

The survey of Stanford internal medicine residents yielded 16 response (31.2% PGY1, 25% PGY2, 43.8% PGY3). Residents report that they request outside pathology specimens 6.45 times per month on the inpatient oncology service and 1.25 times per month on the inpatient medicine service. 100% of residents surveyed reported having faced difficulty with the process of obtaining outside pathology specimens originating from unfamiliarity with a certain aspect of the process. Interviews suggested that people with previous experience obtaining outside pathology specimens are generally able to complete the process successfully and efficiently, while those without previous experience often face difficulties. Med8 and Med10 floors at Stanford have a role, the case management liaison, that is particularly helpful in the process of obtaining outside pathology specimens. Survey results suggest that 50% of residents are aware of this role. Residents that are aware of the case management liaison role and have relied on case management liaisons express that this service greatly facilitates the process of obtaining outside pathology specimens. Interviews additionally indicated that inpatient teams want guidance/resources to help navigate the process of obtaining outside pathology specimens.

Residents often lead the process of obtaining pathology specimens from outside hospitals when this is a required part of a patient’s care. Given that residents that are aware of the case management liaison role associated with Med8 and Med10 floors face less difficulty with this process, residents should be educated about the availability of this source of help. To provide requested guidance to residents and inpatient teams about the process of obtaining outside pathology specimens, a one-page guide to the process was created, as well as fax-templates that can be used to request specimens from outside pathology departments.
A retrospective assessment of level of care recommendations and nursing acuity scores following an appropriateness of care intervention

Link to Presentation
Gabriela Ruiz Colón, AB
Michaela Sullivan, MS
Moses Albaniel, BSN, RN
Patricia Britt, MSN, RN, CNS, NEA-BC
Lisa Shieh, MD, PhD
Medicine Department Background
As part of a multiprong intervention to eliminate waste in cost of hospital accommodations, the InterQual Level of Care (LOC) criteria – a clinical decision support tool – was deployed by our institution to assign patients to one of three LOCs: acute care, intermediate intensive care (IICU), or intensive care (ICU). In this intervention, which sought to decrease the number of patients in a higher LOC than what was clinically necessary, patient safety balancing metrics were stable in the pre- and post-intervention period. However, nursing workload, a key balancing metric, has yet to be examined. In this study, we examine nursing workload across departments before and after the intervention using a proprietary nursing acuity score. 

A retrospective, observational study was conducted analyzing admissions to participating departments at the study institution between 2017 and 2020. Patient’s LOC recommendation (as determined by InterQual), assigned (actual) LOC, and nursing acuity scores were collected and analyzed. Average nursing acuity scores were compared using Wilcoxon rank-sum tests based on whether they aligned with the InterQual recommendation (“Acute Match” or “IICU Match”) or if they were recommended to be in acute care but were actually in IICU care (“Mismatch”).

Following the intervention, percent of patients who were in the Mismatch cohort decreased across all departments from 13% to 7%. Average total nursing acuity score for Mismatch patients before the intervention was less than the IICU Match cohort and greater than Acute Match cohort in all eight departments analyzed with statistically significant differences to the IICU Match cohort in three departments, and statistically significant differences to the Acute Match cohort in seven departments. After the intervention period, the Mismatch cohort was statistically significantly distinct than the Acute Match cohort in all eight departments but only different than the IICU Match cohort in one department.

Collectively, this study demonstrates that our intervention successfully decreased inappropriate use of the IICU LOC, and that the residual Mismatch patients are a distinct entity with nursing needs that exceed that of the Acute Match cohort, and thus a higher LOC can be justified. This demonstrates that a nursing workload metric such as the nursing acuity score can be a valuable complement to clinical criteria such as the InterQual LOC criteria to objectively determine patient’s true necessary LOC and ensure nursing staff feels adequately staffed to care for patients.
Improving Prescription Refills in Outpatient Psychiatry

Link to Presentation
Michael Polignano, MD

Dexter Louie, MD
Brinly Garcia
Sonya Carten
Cindy Tse
Jake Mickelson
Psychiatry Department Background: In the summer of 2019, patients of the Stanford outpatient psychiatry clinic reported an average of 3 medication refill requests per week which were not filled in a timely manner. Meanwhile, providers reported that they were frequently unaware of pending refill requests until either the last minute, or they were overdue. These delays in care resulted in suboptimal treatment of patients’ conditions, withdrawal side effects, and dissatisfaction from patients and providers alike.

SMART Goal: Decrease the number of reported medication refill delays from an average of 3 per week to 1 per month (0.25 per week) by December 2020.

Key Learning: The advent of electronic medication refill requests has created multiple ways in which patients can request refills but has also unfortunately increased the number of ways in which refill requests can be lost. Our process mapping revealed two major issues: first, that refills were being routed incorrectly (whether by phone, fax, or electronically). Second, that the interface of Epic was leading to duplicate requests and other suboptimal processes. Our interventions focused on tracking and fundamentally analyzing reported refill delays, addressing incorrectly routed requests, and refining Epic support.

Through our efforts, we were able to reduce reported refill delays to the target of 1 per month (0.25 per week). Challenges for the future will include sustainability issues, most notably the protection of MA time.
Could it be Motor Neuron Disease? A Comparison of Clinical and Electrodiagnostic Impressions in Patients with Peripheral Nerve Hyperexcitability

Link to Presentation
Joy Lin, MD
Sarada Sakamuri, MD
Neurology Department Objective: When patients report cramps or fasciculations, clinicians have varying degrees of suspicion of motor neuron disease (MND) based on clinical history and exam. This study tested the hypothesis that electrodiagnostic findings consistent with MND on electromyography and nerve conduction studies (EMG/NCS) are unlikely unless there was already a high degree of clinical suspicion for MND.

Methods: A chart review was performed of adult patients seen at Stanford Health Care in 2019 who had muscle cramping, twitching, or fasciculations that contributed to the ordering of EMG/NCS. The patients’ pre-test history and exam were evaluated for features of MND, and their EMG/NCS were categorized based on electrodiagnostic impressions of MND. The ultimate diagnosis and management of the patients after electrodiagnostic testing was also reviewed to determine whether the diagnosis of MND was made by other means.

Results: Of patients suspected to have MND by clinical impression, 73% showed features consistent with MND on EMG/NCS. In patients not suspected to have MND based on clinical impression, none had an EMG/NCS consistent with MND or were subsequently diagnosed with MND. Instead, electrodiagnostic impressions in this group included neuropathy, polyneuropathy, radiculopathy, myopathy, neuromuscular junction disorder, and myotonic disorder. In summary, a reassuring history and physical against MND had a 100% predictive value for a benign EMG/NCS in our cohort.

Conclusions: History and exam may be sufficient in ruling out MND among patients with cramps or fasciculations, and electrodiagnostic studies may not provide additional value for this specific purpose. These findings suggest that providers may have confidence in their clinical impression when evaluating for MND in patients with cramps or fasciculations.
An Education and Electronic Health Record Intervention Increased Selection of Stress Tests Without Imaging in Primary Care Clinics

Link to Presentation
Neil M Kalwani
Sandra Tsai
Cindie Gaspar
Mircea Gafencu
Ria Paul
Zarrina Bobokalonova
Eleanor Levin
Medicine Department Background:  American College of Cardiology/American Heart Association guidelines recommend an exercise tolerance test (ETT) without imaging as the initial test for patients with suspected ischemic heart disease and low to intermediate pretest probability. ETT utilization by primary care providers (PCPs) at our institution was low.

Methods:  We identified 3 pilot primary care clinics. Beginning in April 2020, we conducted education sessions with PCPs on appropriate use criteria for ETTs and exercise stress echocardiograms (ESEs). On July 13, 2020, we implemented an electronic health record (EHR) alert highlighting these criteria when pilot clinic providers ordered ESEs. The pre-intervention period was August 1, 2019 to February 29, 2020, and the post-intervention period was July 13 to October 31, 2020. We used a logistic regression difference-in-difference model to measure the intervention effect on test selection.

Results:  From the pre- to post-intervention periods, the percentage of tests ordered as ETTs versus ESEs increased from 13% (19/152) to 52% (31/60) in the pilot clinics and from 14% (385/2,773) to 20% (177/899) in the non-pilot clinics. The effect of the intervention on test selection was statistically significant with a difference-in-difference odds ratio for ordering of an ETT versus an ESE of 5.4 [95% CI 2.6, 11.1].

Conclusions:  An intervention consisting of educational sessions and EHR alerts highlighting appropriate use criteria for stress testing modalities increased selection of stress tests without imaging. Low-cost interventions can align primary care provider selection of cardiac imaging studies with guidelines.
Evaluating process measures and identifying key drivers of mortality in patients with sepsis not present on admission at Stanford Hospital

Link to Presentation
Lauren Ammerman
Lisa Shieh
Medicine Department The mortality of patients with sepsis not present on admission at Stanford is higher than expected. Stanford uses an EMR-based Best Practice Alert (BPA) sepsis screen, with standard work to involve the critical care nurse and primary team, who may escalate to involve critical care fellows. With the objective of identifying drivers of suboptimal sepsis recognition and treatment, I reviewed the charts of patients who were coded with sepsis not present on admission and died while hospitalized (n = 27), and interviewed critical care nurses (n = 3) and fellows (n = 3). Of the 20 patients clinically validated to have sepsis not present on admission, the BPA triggered within 1 hour of sepsis onset in 70%, however critical care nurse validation was undetermined or not done in most (80%). Sepsis bundle management was notable for poor compliance with 30 cc/kg fluid bolus (7%). Blood cultures were not universally drawn (not done in 20%), however compliance was fair to good with baseline lactate (95%) and antibiotics (85%). Goals of care discussions were crucial in most (85%), however a timely advanced care planning note was documented in only 60%. Themes of role confusion and inadequate orientation for primary teams and fellows emerged in interviews. Overall, the following key drivers of sepsis not present to admission mortality were identified: sepsis bundle non-adherence, role confusion, inadequate goals of care conversations, and patient complexity. A key countermeasure includes alerting the primary team MD via the EMR and/or voalte when the sepsis BPA triggers with prompts to consider components of the sepsis bundle and documentation of goals of care. Additionally, timely orientation for critical care fellows prior to triage rotations, with an emphasis on goals of care and assessing for fluid responsiveness, as well as education targeted to services with a history of suboptimal collaboration with critical care nurses are indicated. Recommended interventions will be presented to the mortality committee for consideration.
Traversing the Tracheostomy: Improving the Elective Tracheostomy Pathway for Patients with Neuromuscular Diseases

Link to Presentation
Shefali Dujari, MD
Jacqueline Hayes Albarran, RRT-ACCS, RRT-NPS, RCP
Angie Murkins, NP
Tiffany Sun, BSN, RN
Sarah Stranberg, MA, CCC-SLP
Michelle Cao, DO
Jennifer Lee, MD
Neurology Department Introduction: Patients with neuromuscular diseases such as amyotrophic lateral sclerosis often develop progressive respiratory failure, leading to significant morbidity and mortality. A small percentage of these patients elect to undergo elective tracheostomy placement and are discharged with continuous invasive ventilation to home. This is a vulnerable patient population with a unique and variable pathway from ambulatory planning to intensive care unit stay to discharge. Living at home with invasive ventilatory support involves extensive pre-planning and a multidisciplinary team of providers with various expertise who can assist these patients. Given the complexity of this process, a multidisciplinary team of providers from Otolaryngology and Neurology participating in the Clinical Effectiveness and Leadership Training (CELT) course were tasked with improving upon this pathway. As much of the advanced discharge teaching for home care is done by the intensive care bedside nurse (RN) and respiratory care provider (RCP) and higher workforce confidence has been associated with higher patient rating of their experience in prior studies, we identified RN and RCP confidence as our primary outcome for this quality improvement project. Our SMART goal is to increase confidence in patient and caregiver training in home tracheostomy care for neuromuscular patients with tracheostomy on a ventilator from 48% of RNs and RCPs to 75% by June 2021. Confidence is defined as e3 on a 5 point Likert scale (1: extremely confident, 2: fairly confident, 3: somewhat confident, 4: slightly confident, 5: not at all confident).

Methods: Baseline data was obtained from 32 RCPs and 72 RNs on confidence in providing education to patients with neuromuscular disease and their caregivers with a new tracheostomy on a ventilator preparing for discharge from the ICU to home, as well as experience, prior training, specific educational needs, and qualitative comments. These comments were analyzed and a Pareto diagram was created to prioritize needs.  Data was also obtained form 17 caregivers of patients with neuromuscular disease who underwent tracheostomy placement at Stanford to understand sufficiency of the education they received, specific educational needs, and qualitative comments. A one question survey regarding confidence in patient and caregiver training is being administered weekly to understand the impact of our interventions.

Results: Survey results demonstrated that less than half of RNs and RCPs felt confident (extremely confident, fairly confident, or somewhat confident) in providing education to patients and caregivers with neuromuscular disorders, informing our SMART goal. Key drivers were identified including: 1) an efficient and standardized workflow, 2) clarity and accountability of roles in the workflow, 3) standardized and accessible tools for teaching and training, 4) creation of a feedback loop to avoid process failure. Based on these key drivers, several possible interventions have been identified and are being implemented. Thus far, in response to these interventions we have noted an increase in confidence in patient and caregiver training by RNs and RCPs.

Discussion: Our baseline survey demonstrated that over half of RNs and RCPs did not feel confident discharging a neuromuscular patient on a new tracheostomy with home ventilation from the ICU to home. Currently, we are continuing to develop and implement our proposed interventions, while measuring RN and RCP confidence with weekly surveys. Our goal is to show a sustained increase in confidence, with 75% of RNs and RCPs rating their confidence levels as extremely confident, fairly confident, or somewhat confident. Next steps will include obtaining additional patient level data regarding the impact of these interventions, such as length of stay, 30-day readmission rate, and patient and caregiver satisfaction. This project has highlighted the importance of a multi-disciplinary team, early involvement of key stakeholders, and a rigorous root cause analysis prior to intervention implementation. We hope that our effort in these areas will improve the sustainability of our interventions.
Improving the efficiency and safety of Neuro ICU to floor patient transfer process

Link to Presentation
Anna Janas, MD, PhD
Angie Murkins, NP
Christina Mijalski, MD
Zachary Threlkeld, MD
Neurology Department Background: Transfer of patients from the Neuro ICU to the floor is often delayed, even once an available bed is identified. Delays in transfer may result in unpredictable transfer timing that interferes with rounds and didactics, incomplete handoff between providers less familiar with the patient, occupation of ICU beds by patients who no longer require the ICU, increased length of stay, and most importantly compromised patient safety. Models at other services and institutions to address such delays include having the floor team round on floor-level patients who remain in the ICU, or having the ICU team be responsible for patients after transfer until provider handoff can be completed.

Current State: Our current state analysis revealed that between July 2018 and July 2019 the mean transfer time (from bed request to transfer completion) for Neuro ICU patients was 11.2 hours for all neurology patients and 10.5 hours specifically for stroke patients.

SMART goal: To reduce the total transfer time from bed request to completion to 3 hours.

Interventions: We identified 3 key drivers that play a role in making transfer process more efficient - prompt bed boarding, prompt signing of transfer orders, and prompt assessment of patient by the accepting floor team. To address these key drivers we implemented a new transfer protocol between Neuro ICU and Stroke floor that focuses on early (pre-rounding) bed boarding by the ICU fellow and attending, signing of transfer orders by the ICU team as soon as patient is bed boarded, and ICU team remaining the primary responsible team after transfer to floor until provider handoff can be completed. This new protocol was implemented to Stroke patients on March 2nd, 2020.

Results: A provider survey was administered 6 weeks after implementation of the new protocol. Among providers who worked in the Neuro ICU or stroke floor, 70% agreed that the new protocol improved efficiency of the transfer process. Monthly mean transfer time decreased from 10.5 hours to 8.5 hours for Stroke patients one year following implementation of the new protocol.

Future Directions: We will plan to extend the new protocol to all general neurology patients. Further analysis will include secondary measures including total hospital length of stay, ICU length of stay, and the percentage of patients transferring overnight.
An Evidence Based Approach to Imaging Dizzy Patients in the ED

Link to Presentation
Richard Baron
Shefali Dujari
Laurel Jakubowski
 Erica Von Stein
 Kirsten Fisher
 Diana Slawski
 Kristen Steenerson
 Brian Scott
Neurology Department Introduction: Dizziness is a common and nonspecific complaint that accounts for 3-5% of all ED admissions with a cost to the US health care system of $4 billion in 2011, led in large part by neuroimaging. Dizziness is a diagnostic challenge given patients’ difficulty describing their symptoms, and the wide differential that must be included— from benign reversible causes to posterior circulation infarcts and tumors.  Imaging practices vary by institution; between 20-48% have a CT head and 0.8-11% of patients have an MRI.  While neuroimaging is a powerful diagnostic tool,  in most cases a targeted history and neurologic exam, as outlined in the TiTraTE model (Timing, Triggers, And Targeted Exam), should guide decision-making for imaging. For example, patients with triggered episodic vestibular syndrome consistent with typical BPPV or orthostasis almost never have a central cause, while in acute vestibular syndrome the risk of stroke can be as high as 55% in patients with risk factors.  Still, small posterior circulation strokes may not be visible on early MRI, and a HINTS PLUS hearing exam has a greater sensitivity and specificity during the first 24 hours.  Multiple studies have found that head CTs without contrast have a diagnostic yield near 0% in patients presenting with isolated dizziness since bleeding, tumors, and bony abnormalities rarely present in this way.   In our experience at Stanford, the lack of confidence in exam maneuvers and presence of a clear diagnostic pathway leads to an overreliance on imaging prior to or in place of a careful history and neurologic exam, which in turn imposes financial costs and leads to suboptimal patient care.  

Methods: A root cause analysis was performed to understand the factors contributing to early patient imaging prior to or instead of important key history and physical exam diagnostics. Key drivers were identified and using this information, interventions will be developed and implemented. We will perform a current state analysis to gain a greater understanding of our current imaging practices and their yield rates, and how often an appropriate clinical exam is performed on dizzy patients so that we can quantify how Stanford compares to other regional and national samples. This will serve as baseline today, to allow us to assess the impact of our intervention.

Results: Our workgroup identified the following SMART goal: In the 3 months following our educational intervention and introduction of a diagnostic pathway, we will aim to achieve a 75% rate of documentation of diagnostic categories of dizziness (AVS vs EVS) and appropriate exam maneuvers (orthostatics and Dix Hallpike in EVS, HINTs plus hearing in AVS, all 3 if ambiguous) prior to neuroimaging.  With our root case analysis and the goal of using imaging more judiciously, we identified three key drivers: attending physician confidence that a change in imaging practice will not miss dangerous diagnoses, resident physician confidence in obtaining a dizziness history and performing a targeted exam, and the presence of a well-defined multidisciplinary workflow for these patients. 

Discussion/Next Steps: An overreliance on imaging can be caused a myriad of specific factors including knowledge gaps, diagnostic reasoning, interdepartmental communication, workflow, and hospital culture.  Through current state analysis we will gain a greater understanding of our current imaging practices and their yield rates, and quantify how Stanford compares to other regional and national samples.  We will compare the current practice to a theoretical best practice based on our proposed pathway and estimate a cost savings due to more targeted imaging.  Finally, post intervention, we can observe to what degree a change in practice patterns actually occurs, and how it affects imaging utilization, costs, and readmission rates. There will be opportunities for ongoing education, further review of outcomes, and revision of the diagnostic pathway as needed.
Time Well Spent: Improving Inpatient Drug Charting

Link to Presentation
Sierra Centkowski, MD
Elizabeth Wang, MD
Susan Weber
Srinivasan Boosi
Ron Li
Christopher Sharp, MD
Beth Martin, MD
Bernice Kwong, MD
Dermatology Department Title: “Time Well Spent: Improving Inpatient Drug Charting”

Authors: Sierra Centkowski MD, Elizabeth Wang MD, Susan Weber, Srinivasan Boosi, Ron Li, Christopher Sharp MD, Beth Martin MD, Bernice Kwong MD

Adverse cutaneous drug eruptions are the most common adverse event in hospital patients, affecting 2-3% of all inpatients. For patients with suspected drug eruptions, construction of an accurate and comprehensive drug history chart is critical to determine potential culprit medications for each patient. Inaccurate drug history can lead to unnecessary discontinuation of important medications, or late removal of a potentially life-threatening drug culprit. Constructing a drug chart takes significant time abstracting clinical data from the medical record. In partnership with Stanford University School of Medicine Engineering, Technology & Digital Solutions, a digital application, SEAL, was created to utilize technology to generate accurate inpatient drug charts and improve efficiency. Real time feedback by dermatology residents using the application and real time response and improvement of the SEAL app by the digital team led to both improved accuracy of drug history, and reduction of time spent and improvement of quality of life (QoL) for dermatology residents.

Our goal with the SEAL app was to decrease clinician time spent creating inpatient drug charts while maintaining or improving drug chart accuracy and dermatology resident confidence in use of SEAL for drug charts. More broadly, we hope that SEAL app will help other specialties at Stanford create reliable timelines of medication administration and improve clinician wellness.

Root Cause Analysis:
Via an anonymous survey of our dermatology trainees, we collected baseline data on the amount of time it takes to create an individual, handmade drug chart on paper or in Excel. In a single week on the consult service, residents report making about 3.5 drug charts. More than 2/3 (68.4%) of dermatology residents reported spending greater than 30 minutes per handmade drug chart, with over 20% of dermatology residents spending over 1 hour per chart.

Feedback was collected from dermatology trainees during real time use and provided to the digital team to continuously improve usability and function “live”.  With the goal of improving wellness/QoL  and reducing resident burnout through decreased time spent with patient charts, but maintaining accuracy of care, we surveyed residents at a baseline to understand the time burden and impact that handmade drug charts had on their QoL at work. We surveyed residents to understand the impact both on time and QoL at baseline (prior to consult months) and then at week 1 and week 4 of their consult months, as well as the confidence they had in accuracy of drug histories.

Residents were asked in real-time of SEAL development how many minutes it takes to handmake a drug chart vs. SEAL. Over 9 months of SEAL development,  residents consistently spent more time on a handmade chart. The maximum time it takes a resident to create a drug chart using SEAL is 30 minutes, compared to 60 minutes for a handmade chart. (Figure 1.)

Residents were asked about their confidence in the SEAL application over time, at week 0 of their inpatient consult month (baseline) and week 4 of their consult month. Confidence improved at week 4 compared to week 0, showing continuous use of SEAL improves the user faith and trust in the application. (Figure 2.)

Residents were asked how much time they saved using SEAL, on a scale of 1-9 (1= no time saved, 5=some time saved, 7+= a lot of time saved). 30% of residents endorsed saving “some time” while 70% of residents endorsed “a lot of time saved”. (Figure 3.)

Residents were asked about their QoL improvement on a scale of 1-9 with using SEAL. Over months of developing SEAL, residents at the start of their inpatient rotation (week 0) sequentially endorsed improvements in QoL (blue line). By week 4 of the rotation, all residents endorsed the highest level of QoL improvement. (Figure 4.)

Key Learning Points:
- Partnership between Stanford dermatology residents and Stanford Digital Solutions team to provide continuous real time “live” feedback and live app updates continuously improved the ease of usability and accuracy of an electronic drug app.
- Creation and improvement of the SEAL Drug App reduced the amount of time spent by dermatology residents to construct accurate drug charts, leading to decreased time spent and improved resident quality of life and physician job satisfaction.
- Future applications: continue to optimize the SEAL Drug app to include outpatient medications, medications from outside hospitals/admissions, and ultimately for use for other medical specialties including outside of SHC institution.
Penicillin Allergies in the EMR cause Unnecessary, Clinically Harmful Antibiotic Changes including Significantly greater exposure to Vancomycin

Link to Presentation
Kush Gupta
Jonathan Chen
Ron Li
Lisa Shieh
Medicine Department Background: Penicillin class antibiotics are the most frequently reported drug allergy in the electronic health record (EHR), with approximately 10% of the US population reporting allergies to penicillin. However, literature has consistently found that over 95% of patients documented to have penicillin allergy can tolerate penicillin-class antibiotics. The reasons for this discrepancy are well studied and include improper allergy documentation, poor education regarding types of allergic hypersensitivity reactions, and plethora other factors. Nevertheless, EHR clinical decision support (CDS) software produces prolific drug-allergy alerts based on documented penicillin allergies that ultimately affect physician prescribing behavior. It has previously been demonstrated that patients reporting penicillin allergy receive suboptimal therapy compared to those without penicillin allergy. However, there is a lack of discussion in literature regarding the specific and immediate deleterious effects of electronic drug-allergy warning alerts and their effects on antibiotic prescription behavior. We describe an approach using EHR audit trail data to characterize how drug-allergy warning alerts impact physician antibiotics orders for patients with documented penicillin allergies. We then describe the impacts of this analysis within the context of treatment for Methicillin-Sensitive Staphylococcus Aureus (MSSA) infection.

Methods: We analyzed EHR data containing inpatient medication orders and medication administration data from Stanford Health Care between 2008 and 2018. Chi-squared tests were used for significance. We first identified all orders for beta-lactam antibiotics (penicillin, cephalosporin, carbapenem, and monobactam class medications) amongst patients with documented penicillin allergies. We next cross-referenced drug-allergy alert data to discriminate between alerts that were ignored and alerts that resulted in cancellation of the order request. Medication administration record (MAR) records were used to confirm administration of ordered medications to the patient; these data further allowed identification of orders for which administration was stopped for reasons including allergic reaction. Finally, a retrospective cohort analysis of patients with culture-confirmed Methicillin-Sensitive Staphylococcus Aureus (MSSA) was performed. Bacterial antibiotic susceptibility laboratory data were analyzed, and antibiotic orders placed immediately thereafter were collected. These data were then stratified by patient allergy phenotype and pooled for analysis.

Results: Among the 133,795 patients with documented allergies to penicillins and/or cephalosporins, 149,631 drug-allergy warning alerts fired between 2008 – 2018 in response to prescription orders for beta-lactams. Approximately one-third of these orders (49,997, or 33%) were cancelled following the drug-allergy alert. Seventy percent (34,759) of these cancellations occurred despite the patient having previously safely tolerated a beta-lactam. See Table 1. When comparing the antibiotics initially requested (responsible for triggering the drug-allergy alert) to those ultimately ordered, in the majority of instances (20.8%) the medication was changed to vancomycin. This change was usually reflected a broadening of antibiotic spectrum of activity. See Table 2. Finally, when comparing treatment for MSSA infection amongst patients with vs. without penicillin allergy, a statistically significant difference (p < 0.0001) was noted. Patients with penicillin allergy were more likely to receive vancomycin (24.3% vs. 18.4%) and less likely to receive nafcillin (7.7% vs. 10.5%) or cefazolin (17.6% vs. 19.3%) than their non-allergic counterparts.

Conclusion: Inappropriate penicillin allergy documentation and subsequent drug-allergy warning alerts influence physicians to avoid prescribing first-line penicillin-class antibiotics when they otherwise should. The vast majority of antibiotic cancellations and changes due to penicillin allergy alerts are unnecessary: in nearly 70% of these scenarios, the patient has previously demonstrated tolerance to beta-lactams. Furthermore, approximately 18% of penicillin allergy alerts result in the broadening of antibiotic selection to vancomycin. These changes to antibiotic selection abound even in scenarios such as treatment for MSSA infection where antibiotic therapy should be highly standardized and regimented; penicillin allergic patients are significantly (p < 0.0001) more likely to receive vancomycin (24.3% vs. 18.4%) and less likely to receive nafcillin (7.7% vs. 10.5%) than their non-allergic counterparts.
Cost Transparency Affected Expensive Drug Prescribing and IV to PO Substitution

Link to Presentation
Rebecca Linfield, MD
 Bo Wang, MD,
Janjri Desai, PharmD, MBA
Lisa Shieh, MD, PhD
Medicine Department Introduction: Healthcare costs in the United States continue to increase, in part driven by the high cost of pharmaceuticals. Cost transparency, the act of showing the cost of drugs to providers, has been proposed as a method of reducing healthcare costs, but its effectiveness has not been studied on a large scale.
Methods: We tracked the volume of the top 50 medications in the Stanford Health Care network before and after the introduction of cost transparency in the electronic medical record on August 1, 2017, with a second group on June 27, 2018. We sought to determine if this intervention would generate at least a 10% decrease in the orders of expensive medications (greater than $200 per dose) and at least a 10% decrease in intravenous (IV) prescriptions when the oral (PO) substitute was available. We also surveyed house staff for their interest in cost transparency.
Results: 95% of house staff were interested in having cost transparency. We show a 5% decrease in the volume of expensive medications ordered in the 12 months following the intervention and a 25% decrease in IV medications during the same time frame, with the decrease in IV volumes roughly matched by increase in PO volumes. This led to ~$225K in cost savings.

Conclusion: We believe that cost transparency can be a low-intensity intervention to decrease unnecessary prescribing and increase the learning in our healthcare system.  
Development of a Prioritization Model to Compare Emergency Psychiatric Coverage Service Options

Link to Presentation
Jacqueline Sandling
Kathleen Carrothers, MS MPH
David Svec, MD MBA
Medicine Department Background: Reducing Length of Stay (LOS) is an important way for hospitals to improve emergency department (ED) costs and outcomes. Psychiatric patients represent a challenge to reducing LOS when the scarcity of psychiatric specialists leads to longer LOS. Previous literature describes the unique solutions different hospitals have employed across the US, but does not share methods for evaluating or selecting a solution that can be applied to other hospitals.
Methods: We conducted a review of hospital ED case data, market research on psychiatry services, and interviews with hospital staff. This information, along with projected return on investment, was aggregated to create a holistic model for evaluating different service options and selecting the one with the best fit.
Results: To develop a prioritization model that identifies the one psychiatric service improving psychiatric LOS and best fitting the hospital’s overall priorities and operations, our methodology identified 8 key factors that captured the overall difficulty of implementation and benefits associated with each service option.
Conclusion: The Prioritization Model created in this study was instrumental in selecting the solution for reducing LOS in a way that best meets patients’ and the hospitals’ needs. This model may be applied to other hospitals and service evaluations to provide a holistic review and direct comparison of opportunities.
Reducing Readmissions Following Radical Cystectomy

Link to Presentation
Lee White, MD, PhD
Caleb Seufert, MD
Bogdana Schmidt, MD
Peter Dy, MS, NP
Stacy Cox, RN, BSN
Mia Singh, RN, MSN, OCN, PHN, CPPS
Joanne L. Meneses, MSN, RN
Eila Skinner, MD
Urology Department Patients with bladder cancer undergoing radical cystectomy are readmitted to the hospital within 30 days of surgery at a rate significantly higher than other urologic procedures, and procedures in other fields of similar risk and complexity.  Whether reconstructed with ileal conduit or neobladder, these readmissions are a persistent challenge.

Bladder cancer is the sixth most incident cancer in the United States, and radical cystectomy (RC) is the standard of care for patients with refractory or muscle-invasive cancer. The removal of the bladder is a major operation with significant risk for post-operative morbidity and mortality. This patient cohort tends to be elderly, malnourished, frail, and medically complex – a situation exacerbated by preoperative chemotherapy. Early rehospitalization (within 30-days) has become a focus due to the impact on patient experience, high cost, and associated poor downstream health outcomes. RC has one of the highest readmission rates among all major surgeries, with a national average of 21 to 31 percent.

Common causes of readmission are urinary tract and abdominal infections, and failure to thrive. Risk factors associated with readmission include post-operative complications while inpatient, age, comorbid diabetes mellitus, obesity, and post-operative discharge to skilled nursing facility.

Reducing readmission following cystectomy may yield strategies and answers to reduce readmission following other complex and morbid surgeries. These answers can improve the experience for patients at Stanford and beyond.

Among patients undergoing radical cystectomy, we aim to reduce the 30-day readmission rate due to infection to 10%,  from the current 17.9% (FY 2018 15% + FY 2019 18% + FY2020 21%; n total = 217) to 10% by October 1, 2021.

We reviewed the literature and formed a fishbone diagram (Left) of interventions used beyond Stanford.  We proceeded to a chart review of surgeries Stanford submits to the American College of Surgeons National Surgical Quality Improvement Program, and records for radical cystectomy procedures from the billing department. During the chart review we discovered both the NSQIP data, and the billing data contained many surgeries that did not meet our inclusion criteria.  Our team resorted to a manual review of the surgical schedule for 2018-2020.  That review uncovered many surgeries not found in the NSQIP or billing data. These causes for readmission are summarized in a Pareto chart (Right).

Through our analysis we identified a set of interventions collectively termed ‘the bundle’. These include referral for consultation with nutrition and ordering of nutritional supplements during the initial patient consult appointment, enrolling the patient in a perioperative patient education platform called Seamless MD during the preoperative period, a set of standardized teaching for patients during their inpatient stay, removal of the ureteral stents by post-operative day 4, ordering of urinary tract infection prophylaxis with methenamine at the time of discharge, care coordination with the outpatient teams at the time of discharge, scheduling of a close follow-up visit within 3 days of discharge then weekly for 3 weeks following discharge.  Beginning on April 1, 2021 we initiated our intervention and are compiling data now.

Our analysis of the literature and a thorough chart review  of patients who underwent radical cystectomy at Stanford from 2018 to 2020 informed the design of a set of interventions summarized above.  We found that our surgeons perform 60-89 radical cystectomies per year.  28-32% of these patients were readmitted to Stanford or an outside hospital in the 30 days following their surgery. 15-21% of these readmissions were due to infection.  These data are summarized in the table (Left) below.  These data allowed us to construct a set of run charts (Right) to track the historical rate of cystectomy readmission.  We will chart the rate of cystectomy readmission going forward.  We use a 60-day rolling average and must wait for 30 days following surgery to determine if patients were readmitted.  For this reason, preliminary results will become available beginning 7/1/2021.

Radical cystectomy is a morbid surgery, during both in the immediate post-operative inpatient stay, and the period shortly after hospital discharge.  Stanford currently performs at the national average for this procedure.  We are optimistic that our set of proposed interventions, including both pre-operative, peri-operative, and post-operative efforts, will bend this curve.  Our goal is to reduce readmissions due to urinary tract and abdominal infections from 18% to 10% within 6 months. This represents 6-10 prevented readmissions, a savings of $1-1.5M.  Because this is a low volume surgery, averaging 1-2 cases per week at this hospital, our analysis is sensitive to data accuracy. For this reason, we opted for a manual chart review after finding NSQIP and billing data insufficient.
Addressing COVID-19 Immunization Disparities at a Community Health Center: Education and Targeted Outreach

Link to Presentation
Emmeline Ha, MD
Bridget Harrison, MD MPH
Family Medicine Department Background: Despite the availability of COVID-19 vaccine appointments at the Indian Health Center of Santa Clara Valley (IHCSCV), vaccine hesitancy and care gaps exist amongst patients. As of early March 2021, four racial/ethnic groups make up less of the vaccinated population compared to their prevalence in the total clinic population. Communities of disproportionate COVID-19 infection incidence (racial/ethnic groups or high-incidence ZIP codes) are potential areas of target outreach to address this issue. The purpose of this project was to create COVID-19 vaccine outreach resources to eliminate COVID-19 immunization barriers and misconceptions amongst our most vulnerable patients.

Methods: A process map of the immunization outreach protocol at IHCSCV - Family Medicine Center (Stanford Family Medicine Residency’s outpatient clinic in San Jose, CA) was created to identify key intervention steps for increasing vaccination rates. An anonymous survey was sent to IHCSCV FMC providers about their COVID-19 vaccine eligibility criteria knowledge and feedback for potential improvements in COVID-19 vaccination workflow. In March 2021, personal vaccine outreach calls were made with one provider’s clinic panel list, specifically targeting unvaccinated patients from zip codes of high COVID-19 case incidence (as reported by Santa Clara County data; N = 36). Patients were personally phoned by their primary care provider to offer vaccination appointments and answer any questions.

Results: The IHC FMC provider survey had a 85% response rate (23/27). At the time of the survey, 56.5% were incorrect in their knowledge of COVID-19 vaccine eligibility. A majority had positive interest in centralizing resources, information on counseling, and curated flyers/resources for patients. A “resource hub” website was then created with live IHC vaccine event updates, frequently asked questions for counseling, and patient handouts. A promotional video was also filmed featuring the IHC FMC providers in 10 different languages to represent the diversity of IHC FMC patients. Personal targeted outreach successfully reached 20 of 36 eligible patients, 15 of whom (75%) were referred to COVID-19 vaccination appointments.

Conclusions: Primary care providers can have a strong impact in increasing COVID-19 vaccination access. Inequalities in COVID-19 immunization can be addressed through culturally and linguistically sensitive resources and targeted outreach to  high-risk ZIP codes.
Exploring the Unknown: Understanding Readmissions and Health Equity Data Among Sepsis Patients at Stanford

Link to Presentation
Ben Catanese
Zoe Fullerton
Jonathan Hootman
Aydin Zahedivash
Chichen Qiu
Jason Block
Medicine Department Background:
30-day Hospital Readmissions for sepsis lead to worse health outcomes, higher health care costs, and increased health disparities for patients.  Sepsis survivors often experience cognitive, physical, and physiological sequela and a disproportionate risk for hospital readmissions. At Stanford, Severe Sepsis Related Conditions represents a significant source of readmissions with 13% of readmissions in 2020 being attributed to patients with sepsis. Quality organizations such as Vizient have set decreased unplanned readmissions as a priority for hospital improvement, however, the best practices for identifying and addressing readmissions within an individual hospital system are still being elucidated.
This multiyear project hopes to identify targetable causes for sepsis readmissions and design and implement interventions with the goal of reducing readmissions within the Sepsis DRG by 25% in three years.

Year 1 of this project was involved gathering and analyzing existing Stanford data around sepsis readmissions.  All patients with sepsis DRG code and readmission within 30 days between September 2019 and August 2020 were included. Quantitative analysis was preformed on the data set using Microsoft Excel. Demographic data including analysis of gender, age, insurance, relationship status, and zip code were analyzed and compared between readmitted and non-readmitted patients. The 3M 360 software tool was then was utilized to identify potential preventable readmissions.  In depth chart review was then performed on the identified preventable readmissions in order to assess for quality of tool as well as qualitatively characterize  the key drivers and key factors for readmission as dictated by the 3M tool.   

The 3M tool identified 24 patients whose readmissions were determined to be preventable.  Of these, 13 were deemed by two independent reviewers to have preventable readmissions of a mutually agreed up cause. Chart review revealed that many of the preventable readmissions were due to structural issues, including lack of an intermediate facility, access to ambulatory care, and underutilization of observation status. Issues of communication with team and inadequate advanced care planning for chronically ill patients were also common among preventable readmissions.  Larger analysis of data showed that Medi-Cal patients have a 4% higher rate of readmission, but that there was otherwise very little difference in the demographics of readmitted patients. Preliminary analysis of zip codes did show that the most common regions of admission were not proportionally responsible for readmissions.

Sepsis readmissions at Stanford are due to a multitude of factors and interventions will likely require multidisciplinary input and careful consideration.  Further analysis is needed to determine the significance of differences noted in geographic and insurance data for readmitted patients.
Improving Geriatric Trauma Care by Identification of Potentially Inappropriate Medications to Reduce Adverse Drug Events

Link to Presentation
Caroline Park, MD PhD
Matthew Mesias, MD
Kristan Staudenmayer, MD MS
Ankur Bharija, MD
Medicine Department [Background]
Adverse drug events (ADEs) are common in older adults. Use of the Beers criteria in evaluating prescribing patterns has become widely used as a measure of quality of care for older adults. The aim of this QI project is to: 1) assess for correlation between polypharmacy (medications on admission) and risk of falls, readmission, ED visits, mortality and discharge to skilled nursing facilities (SNF), and 2) identify the most commonly associated medications with the above-mentioned adverse outcomes.

Adults of age 65 years and above admitted under the trauma service at Stanford Hospital were included in our analysis. With chart review, we screened for potentially inappropriate medications (PIMs) (according to 2019 Beers criteria) on admission and discharge and analyzed discharge destinations, readmissions, ED visits, and mortality.

We reviewed  98 patients admitted between June 1, and Aug 31, 2020. The average number of medications in our studied cohort was 6.54 (minimum 0, to maximum 21) ± 8.75 and with 16% of the patients taking 1 or more PIMs which included analgesics, Z-drugs, and benzodiazepines and tricyclic antidepressants. Patients taking PIMs had an greater average number of medications (10.7 vs. 5.7), and 80% of patients (12 of 16) were admitted for fall-related injuries (ground level (GLF) and other falls), as compared to 58% (49 of 82) in patients who were not taking PIMs at baseline. 13 patients had ED visits of which 11 resulted in readmissions. 1 of 16 (6.25%) patient who was prescribed a PIM at baseline died. 1 patient was started on a PIM  (tramadol) inpatient and sustained a GLF. In 3 patients, PIMs were discontinued upon discharge. Discharges to skilled nursing facilities (SNF) were nearly double (40%) in the PIMs group compared to those who were not on PIMs (19%).

In our geriatric trauma patient cohort, we found that patients on PIMs were on greater number of medications on average, and were most commonly admitted for GLF, and were more likely to be discharged to SNFs. Given this preliminary result, pharmacy-driven interventions including discharge medication review and post-discharge follow-up may be beneficial in the high-risk cohorts to help prevent future ADEs including recurrent falls.
Improving Multi-directional, Real-time Feedback in Neurology Residency

Link to Presentation
Jingjing (Jenny) Chen, MD/MBA
Brian Scott, MD
Neurology Department Background: Residents are both learners and teachers. Effective feedback from all levels is crucial for improving clinical, interpersonal, and educational skills. Studies suggest residents are dissatisfied with the frequency and quality of feedback. We wanted to understand Stanford neurology residents’ current practices, perceptions, and satisfaction with feedback. From this needs assessment, we hope to develop tools to improve multi-directional, real-time feedback during residency training.

Methods: Residents participated in two live polls (n=21, 18 respectively) on giving feedback to their juniors (medical students and junior residents), peers (same level of training), and superiors (senior residents, fellows, and attendings). Residents also participated in a focus group discussion on barriers to giving/receiving feedback.

Results: About half of the residents are dissatisfied with the amount of feedback they receive. The majority of residents do not feel comfortable giving feedback to peers and superiors. 75% of residents never give feedback to peers; of those who do give feedback, they give feedback less than once/week to peers and superiors. Residents cited intrinsic hierarchy of medicine as the number one barrier for giving feedback. They also discussed lack of psychological safety, such as fear of retaliation, offending others, and damaging relationships. Other factors include lack of know-how for giving/receiving feedback and dedicated time/space.

Insights and conclusions: Residents experience inadequate, real-time feedback from all levels. Rather than personal factors, the biggest barriers to giving feedback are related to organizational cultural/structural issues (hierarchy, lack of psychological safety, lack of workflow process). Our SMART goal is to increase the amount of inpatient, real-time feedback given to residents from all levels by at least 50% by winter 2021. We created a resident workshop on giving feedback. We also plan on exploring how faculty can be involved to change the feedback culture through department-wide competency workshops, dialogues on feedback expectations, and standardized workflow processes.
Increasing Outpatient Goals of Care Discussions in the Resident Neurology Clinic

Link to Presentation
Trevor Rafferty MD, Jacqueline Summers MD
Jingjing Chen MD, MBA
Carl Gold MD, MS
Neurology Department Patients with serious neurologic disease are often hospitalized without having any prior advanced care planning discussions with their outpatient neurologist. Residents in particular do not routinely have outpatient goals of care (GoC) discussions with their patients.  We conducted a needs assessment survey of neurology attendings to help better understand the perceived barriers to having goals of care discussions in the outpatient clinic.  This survey identified (1) time constraints, (2) brief therapeutic relationship, and (3) desire to maintain a positive attitude as the three greatest barriers to having such discussions.  As part of our monthly resident QI workshop, we had a resident-focused discussion to further delve into root cause analysis as well as proposed interventions.  From this workshop, (1) time constraints, (2) lack of experience/formal training in GoC discussions, and (3) cultural/language barriers were identified as among the greatest barriers to these discussions. Proposed solutions included continuing Palliative Care’s Serious Illness Discussion Training annually and implementing the serious illness discussion guide and smartphrases into practice. Other logistical strategies for improved planning, including scheduling dedicated GoC discussion appointments with family and/or in-person interpretation when needed, were also discussed. Our planned intervention is to coordinate annual serious illness discussion training as well as distribution of a GoC “Tips Sheet”. Our SMART goal is: By the end of the academic year 2021-2022, we will increase the utility of the advanced care planning smartphrase, with the goal of at least 60% of residents using this with at least 1 patient. We will evaluate our progress with a survey at the end of the year.
Satisfaction with interpreter services in cardiothoracic surgery inpatients with limited English proficiency

Link to Presentation
Cayo Gonzalez BA
Trinie Harris RN
Pauline Regner RN
Jack Boyd MD
Natalie Lui MD
Cardiothoracic Surgery Department 62 million people who speak a language other than English call the U.S. their home, 26 million of whom are considered limited-English proficient (LEP). Hospitals and clinics often rely on ad hoc interpreters and underuse professional interpreters even though professional interpreters improve patient satisfaction and health outcomes in non-surgical settings. Studying the effect of interpreting on patient satisfaction and other outcomes has been trailing in surgical fields and has been completely absent to our knowledge in the surgical inpatient population. This single-institution cross-sectional survey study aims to fill this gap to ensure LEP patients receive equal quality care as non-LEP patients. A 5-point Likert scale questionnaire for cardiothoracic surgery inpatients was created and modeled after several validated patient satisfaction surveys. The questionnaire probes the following: general satisfaction, quality of information transmission, provider-patient language concordance, and suggestions. We administered this questionnaire to LEP and non-LEP patients and compared general satisfaction and surgical plan understanding between the two groups of patients. There was no significant difference in patient satisfaction or communication scores, although statistical analysis is limited by the small sample size (4 LEP patients and 3 English-speaking patients). Further administration of the survey is required. 
hitec.stanford.edu : Development of a Housestaff Resource Hub

Link to Presentation
Lee White, MD, PhD
Ivana Jankovic, MD
Hyo Jung Hong, MD, MBA
Benjamin Weia, MD
Jared Shenson, MD
Arjun Gokhale, MD
William Kethman, MD
Matthew Eisenberg, MD
Urology Department Housestaff, including residents and fellows, fill a vital role within Stanford and at the four external sites at which they rotate: Stanford Children’s Health & Lucile Packard Children’s Hospital (LPCH), Veterans Affairs Palo Alto Health Care System (PAVA), Santa Clara Valley Medical Center (SCVMC), and Kaiser Santa Clara. At each site, they interface with a distinct ecosystem of information technology (IT), including electronic medical records (EMR), picture archiving and communication systems (PACS), remote access, paging and messaging systems, dictation systems, and telephone systems to care for patients and complete clinical tasks.  Housestaff are uniquely responsible for keeping up to date with differences in IT as they seamlessly transition between sites, providing patient care despite the variation in technology. To help housestaff meet this challenge, the Housestaff Information Technology Enhancement Council (HITEC) was formed in February 2020.

A key achievement of HITEC was building and deploying hitec.stanford.edu, an IT resource website that provides housestaff with the resources to use technology to perform clinical tasks efficiently and safely.

Many departments within Stanford maintain resident resource websites. General Surgery maintains http://med.stanford.edu/scalpel.html and Anesthesia maintains http://ether.stanford.edu. Individual residents, including former General Surgery resident William Kethman, MD, and former Medicine resident Jonathan Chen, MD, PhD, maintained and distributed their own tools to assist residents using clinical technology resources. From these examples, HITEC identified requirements for hitec.stanford.edu, including: 1) a short, memorable URL, 2) ability to load on computers at any clinical site without requiring credentials, 3) quick links to often-used resources like remote access portals, 4) links to clinical apps such as Haiku, Voalte, Spok, Amion, Duo Mobile, Nuance, HospitalTree, ResConnect, Admit Guide, RSA Token, MobilePASS, MedHub, and UpToDate (Poster Column 1 Right), 5) Phone books for each clinical site, and 6) Dictation instructions (phone and Dragon).  Each clinical site has a homepage with a common structure: 1) Guides for new residents and EMR efficiency, 2) Information on account access, including setting up passwords and remote access, 3) Phone instructions to call out or reach the operator, 4) Tech support, 5) Mobile EMR access (Haiku), 6) Consult paging instructions, 7) Messaging support (e.g. Voalte), 8) Dictation setup, and 9) technology improvement requests.

Development progressed rapidly: 5/13/2020 – Site requested, 6/4/2020 – Preliminary approval and site development started, 6/29/2020 – Soft launch, 9/1/2020 – Formal launch and analytics added to site.  The site was developed using Adobe Experience Manager (AEM), a What You See Is What You Get (WYSIWYG) website editing platform used by Stanford Medicine.  Editing the website requires AEM training which limits the personnel who can make live edits. Additionally, as a public website, it cannot include screenshots or content directly from the Stanford, LPCH, or SCVMC Epic Hyperspace EMR instances due to copyright limitations.  For these reasons, hitec.stanford.edu links to a parallel password-protected website hosted on the MedWiki server.  The HITEC MedWiki site houses phonebooks and EMR-related content that can be both viewed and edited by anyone with a Stanford SUNet ID, allowing any enthusiastic housestaff to add content and correct errors such as stale phone numbers in real time. A key element of the site is an ‘ideas form’ where housestaff can request technology improvements. Direct links to hitec.stanford.edu have been implemented within weblinks in the Stanford instance of Epic Hyperspace (Poster Column 2 upper Right) and within PAVA Computerized Patient Record System (CPRS). LPCH and SCVMC requests remain pending.

From 9/1/2020 to 5/1/2020 hitec.stanford.edu saw an average of 198 unique visitors each month (Poster Column 2 lower Left). Pages views averaged 431 monthly during that period (Poster Column 2 lower Right).  Mobile visitors grew to a peak of 107. Analysis of the individual page traffic showed additional trends including an influx of traffic to LPCH pages and evidence of internal linking on SHC and LPCH intranet.

Housestaff use technology extensively to treat patients and are unique amongst hospital providers in the frequency with which they transition between clinical sites.  The hitec.stanford.edu website serves to provide a unified IT hub for housestaff. Ongoing work includes increasing adoption, content improvement, and EMR integration.  Additionally, to increase awareness,HITEC is partnering with onboarding resident educators and advertising  in monthly housestaff-wide HITEC emails.
Identifying Delays in Neurological Care due to COVID-19

Link to Presentation
Phuong Hoang MD PhD and Connie Wu MD
Sandra Mobley DNP
Lironn Kraler MD
Neurology Department Title: Identifying Delays in Neurological Care due to COVID-19

Authors: Phuong Hoang MD PhD, Connie Wu MD, Sandra Mobley DNP, Lironn Kraler MD

Patients who present with acute neurological emergencies often require urgent diagnostic tests such as CT and MRI imaging for stroke and electroencephalogram (EEG) monitoring for status epilepticus. Because of varying levels of hospital preparedness and changing guidelines on COVID precautions, it is unclear if this significantly impacted standard neurologic workflow during the early months of the pandemic. Due to strain on hospital resources, more stringent guidelines on disinfection and personal protective equipment (PPE), and limiting of hospital personnel to decrease the risk of spread of infection, the time to obtain and complete previously routine procedures such as stroke imaging and EEG may be less efficient. Delays in timely diagnosis and management in neurological emergencies such as stroke and seizures potentially contribute to worse patient outcomes.

There were a total of 533 and 617 stroke codes between March-September 2019 and 2020, respectively. There was a longer door-to-CT time (16 min vs 8 min, p=0.02) in August 2020 compared to August 2019. There was a longer median door-to-needle time (58 min vs 35 min, p =0.02) in April 2020 compared to April 2019. There was a longer median door-to-groin time (73.5 min vs 61 min, p =0.04) in July 2020 compared to July 2019.

There were a total of total of 452 and 552 EEGs (including Ceribell) between March-September 2019 and 2020, respectively. There were no significant differences in median latency between time of EEG order to time of hookup before and during the pandemic. There was also no significant difference in EEG hookup time for patients located in the ED, floor and ICU.

There was no consistent significant delay in door-to-CT, door-to-needle, or door-to-groin for stroke patients before and during the pandemic. There was also no significant difference in inpatient EEG hookup time before and during the pandemic, regardless of location of the inpatient. Overall, there were no consistently significant delays in neurological care during the pandemic which may reflect Stanford Hospital’s level of preparedness and ability to adapt during a time of crisis.

Next Steps: We plan to perform a more detailed analysis of patient subsets and assess whether their COVID status was a contributor. We also plan to perform similar analysis with patient cohorts at Santa Clara Valley Medical Center and Palo Alto VA.
Improving Resident Evaluations Through Directed Feedback

Link to Presentation
Connor Arquette, MD
David Perrault, MD
Paige Fox, MD
Surgery Department Introduction
Feedback is an essential part of any learning system. In surgical education, feedback is a highly valuable aspect of resident learning. In fact, Practice-Based Learning and Improvement is one of the six ACGME-mandated Core Competencies. The current ACGME required evaluation system requires biannual resident evaluation and is completed electronically. Given the importance of feedback directed learning, we aimed to create a structured in person, mid-rotation resident feedback system that would yield objective and actionable items for resident improvement.

An in person resident feedback system was implemented at the midpoint of each rotation based on pre-existing goals and objectives of each clinical site. Residents were scheduled to meet with a faculty member one-on-one and were evaluated based on the goals and objectives. Residents were surveyed before and after implementation to determine the preferred method of feedback and value of a formal midrotation evaluation.

Only 4/19 residents were very satisfied or completely satisfied with the previous evaluation system. 15/19 residents were either moderately satisfied, slightly satisfied, or not satisfied at all. All residents preferred receiving face-to-face feedback. Finally, 12/19 residents preferred to receive feedback midrotation. Using this feedback as substrate, the new system was designed and implemented. Midrotation evaluations were completed for all residents on 4-month rotations. Up-to-date results on resident perception of the new system is pending.

There was a perceived deficit in the current resident feedback system, prompting a redesign. The new system has been implemented and final results are pending.
Improving Effectiveness and Satisfaction of Communications with Voalte

Link to Presentation
Benjamin Weia
Garrison Carlos
Jared Shenson
Jill Evans
David Shively
Katrina Sullivan
Leean Rodolfich
Gretchen Brown
Austin Wilson
Medicine Department Communication between nurses and providers at Stanford Health Care is crucial to effective, satisfying, and timely patient care. Voalte is the primary communication software, with 30,000 texts exchanged daily by nurses on 26 inpatient units. This project examined commonly reported communication problems that include contacting the wrong covering provider, using ambiguous patient identification (ID), and not designating the urgency of messages. Root causes identified for these problems were non-optimal protocols, unnecessary HIPAA practices, and lack of provider relationship information in Voalte. To address these, the project’s goals from August 2020 to May 2021 were to (1) decrease provider messages indicating wrong contact by 10% and (2) increase nurse and provider usage of the following standardized practices by 10%: full-last-name patient ID, initiating messages by group-text for patient ID, and use of priority text messaging. Through interventions of interdisciplinary revision of protocols, educational campaigns, and piloting a software update in Voalte’s display, there was a 2% (80 texts a day) decrease in ambiguous patient identification, as well as a 0.1% increase in both priority texts (16 texts a day) and group-text-initiated messages (128 a day), sustained for 1 month to date. However, there was no change in provider messages indicating wrong contact. There is potential for further implementations of the revised protocols to reap benefits.
High-quality neurosurgeon communication and visualization during telemedicine encounters improves patient satisfaction

Link to Presentation
Adrian Rodrigues
Guan Li
Michael Zhang
Michael C. Jin
Melanie Hayden-Gephart
Neurosurgery Department Abstract
Introduction: While recent studies have focused on confirming satisfaction with telemedicine during the coronavirus disease 2019 (COVID-19) era, we leveraged a novel survey instrument to identify associations between patient experience and telemedicine-specific factors such as device selection, audio/visual resolution, and connection stability.

Methods: Telemedicine visit data were gathered from our institution between June 22, 2020 and February 14, 2021. Each patient indicated their overall visit score, likelihood-to-recommend (LTR) score, and device used for the encounter. Remaining questions were randomly distributed to patients to ensure equal distribution across respondents.

Results: Over 34 weeks, there were 901 unique neurosurgical telemedicine visits linked to a post-visit survey at our institution. The LTR top box score percentage showed no significant change across 34 weeks (p=0.218). After adjusting across available covariates, patients who experienced wait times exceeding 20 minutes were significantly less likely to report high overall scores (aOR: 0.12; 95% CI: 0.03–0.41; p=0.001). Patients who indicated they were less able to understand the provider (aOR: 0.22; 95% CI: 0.07–0.66; p=0.007), or who indicated the provider was not able to properly see them (aOR: 0.11; 95% CI: 0.03–0.43; p=0.002) were associated with substantially lower overall scores. Visits with interrupted connectivity or those forced to move to a regular phone call were not important predictors of overall score.

Conclusions: In the largest description of patient satisfaction with telemedicine in the neurosurgical setting during the COVID-19 era, we identified timely and high-quality physician-patient visualization and communication as among the most important predictors of patient satisfaction in virtual settings.
Safety of First Tracheostomy Tube Exchanges

Link to Presentation
Z. Jason Qian
Samuel Cohen
Noel Ayoub
Chloe Domville-Lewis
Flavio Olivera
Otolaryngology Department Tracheostomy is a common inpatient procedure performed by the OHNS service line. Tracheostomy tubes must be routinely changed postoperatively. The first exchange has the highest risk of airway loss, as the tracheostoma is fresh and the tract has not yet matured. This procedure is performed on the floor by residents. 

First tracheostomy tube exchanges can be a source of morbidity and mortality in service lines that manage tracheostomies.

Here, we develop a standard protocol for this procedure, in addition to a dedicated repository for supplies may streamline the workflow for residents and nurses, and improve patient safety. Improved communication between the residents and nurses can help ensure all necessary supplies are readily available for this procedure.

Decreasing Provider Discordance in Code Status Ordering and Interpretation at SHC

Link to Presentation

Emily Shearer
Jason Batten
Mukie Ramkumar
Kim Kopecky
Kim DeBruler
Stephanie Harman
Winnie Teuteberg
David Magnus
Anesthesia Department Background: Stanford currently has five code status orders available for patients: Full Code, Partial Code, DNR/DNI, DNR/DNE, and DNR/C. These code status orders are meant to convey patients’ wishes regarding limitation of life-sustaining treatments.

Local Problem: Despite the importance of correct ordering and interpretation of code status orders in achieving goal-concordant care for patients, there is currently significant discordance in a) code status ordering and b) code status interpretation at SHC.

Methods: In this quality improvement project, we used qualitative interviewing among 24 physicians and nurses at SHC across five departments to identify three domains in which SHC’s current code status options operate: interventions, level of care, and philosophy of care. From these insights, we devised eight hypothetical patient scenarios representing areas of confusion in the current code status options among providers. We then used these scenarios to assess provider concordance in code status 1) ordering and 2) interpretation for patients with given clinical scenarios using both the current code status ordering system and a new system devised to address these areas of confusion.

Results: Discordance in both code status ordering and code status interpretation across 8 patient scenarios decreased >90% with the implementation of a new code status ordering system.

Conclusions: Disaggregation of identified domains results in significant decreases in discordance in both code status ordering and code status interpretation at SHC.
The addition of “Did Not Discuss” box for interventions additionally clarifies whether appropriate goals of care discussions have been had with patient. The new code status ordering layout in Epic Storyboard is easy to use and popular among those responding emergently to code situations.

Engaging Families to Increase Accuracy of Admission Medication Reconciliation in Pediatric Patients

Link to presentation

Anna Brzezinski, MD
Stephanie Chen, PharmD
Sam Backus, MD
Whitney Chadwick, MD
Pediatrics Department Purpose/Objective:
Errors on admission medication reconciliation (AMR) can lead to adverse drug events and patient harm. We piloted the Med Rec Safety Check (MRSC) on a pediatric acute care unit to engage families in the AMR process as part of a quality improvement initiative aimed at reducing the rate of errors on AMR by 50% in 12 months.

In January 2020, a pharmacist-led pilot of the MRSC was trialed as a quality improvement initiative. We created the MRSC using patient-friendly language with the guidance of the Family Advisory Committee to provide an avenue for families to review their home medications. The MRSC is a document populated by the electronic medical record that is available in English and Spanish. After the provider completes the AMR, the MRSC is provided to the family for review and to suggest modifications. Then, a pharmacist performs a gold standard medication history in order to determine the accuracy of the suggestions. A total of 170 medications were reviewed for 33 patients.

Families detected 59% (36/61) of errors missed by the initial provider AMR that were subsequently captured on the pharmacist’s gold standard medication history. Families correctly confirmed 98% (107/109) of medications with no errors. Eighty-two percent (27/33) of caregivers reported wanting the MRSC at future admissions.

Accuracy of AMR can be improved with the implementation of the MRSC by allowing families an opportunity to review the home medication list. Families were able to identify errors on AMR that were initially missed by the provider. They were also able to identify medications that were correctly documented. Families expressed high satisfaction with the MRSC and wanted to use it again. Future work is expanding the MRSC to other hospital units and adapting for limited English proficiency families.
Increasing Resident Handoff Accuracy and Usability
Link to Presentation

Sofia Gomez, MD
Benjamin Weia, MD
Lisa Shieh, MD
Medicine Department Resident handoffs in Epic are often inconsistently populated and lack updates about interval patient events, making crucial information poorly accessible to cross-cover providers during clinical decision-making. Outdated handoffs therefore jeopardize patient safety and can lead to low-value care. Root causes contributing to inconsistent and inaccurate handoffs include lack of a standardized template for handoffs, lack of visibility of certain portions of the handoff tool, the time-consuming aspect of formatting handoffs, and differences in handoff usage for workflow among primary and cross-cover teams. To mitigate these issues, proposed changes to the Epic handoff tool were presented to the Internal Medicine (IM) residents at Stanford as well as the Resident Safety Council with favorable feedback: 81% of all survey item responses from 42 IM residents categorized the proposed interventions as “beneficial” or “extremely beneficial.” Interventions include making the patient "summary line" easily accessible in the handoff tool and auto-populating templates for significant day/night events as well as day/night team to-dos and a pertinent baseline physical exam. These interventions were designed to streamline and standardize handoff information as well as to encourage similar and more frequent usage of the handoff tool across primary and cross-cover team workflows for increased cohesiveness of information transfer. The primary outcome to best measure compliance with these interventions is inclusion of updated day events in handoffs. The specified goal is to increase the rate of day event inclusion in IM resident handoffs from 26% to 75% two months after implementation, with partial implementation having begun on 5/3/21. Prior resident handoff research has shown compliance upwards of 60%, making this goal attainable. The overarching objective of these interventions is to optimize patient safety and high-value care by increasing the accuracy and usability of handoffs.
Prevention of tracheostomy-related pressure injuries
Link to Presentation

Jared Shenson, MD
Chungmei Shih, RN, MSN
Aussama Nassar, MD, MSc
Douglas Liou, MD
Brian Phillips, RN
Brian Lee, RN
Christine Henley, RN
Jacqueline Hayes Albarran, RT
Ann Mitchell Ellsworth, RN, MSN, CNS
Jennifer Y. Lee, MD
Otolaryngology Department Tracheostomy-related pressure injuries (TRPIs) occur frequently in the post-operative period and are entirely preventable. By reducing their incidence, we can decrease resource utilization, patient pain, and medical care costs. Based on historical rates at SHC and in comparison with published rates from peer institutions, a goal was developed to reduce the rate of grade 2+ TRPIs from 2.2% to <1.0% of all patients receiving tracheostomy at SHC. Tracheostomy care is a multidisciplinary effort, including multiple surgical services (Otolaryngology, General Surgery, Cardiothoracic Surgery and Interventional Pulmonology), nursing, respiratory therapy, and wound care nursing. A cross-functional project team was convened to study current practice patterns and identified three key drivers for process improvement: reduction of pressure from the tracheostomy faceplate, protection of the peristomal skin, and standardization of post-operative care.

Based on the root cause analysis, a TRPI prevention care bundle was developed and deployed first as a pilot within the Otolaryngology service line (September 2020) and then rolled out across SHC and ValleyCare in November 2020. Process changes were supported with a combination of provider education, head of bed signage, technology enhancements (Epic order sets and trach care documentation), and supply chain optimization. In the five months after enterprise-wide implementation, there have been no reported TRPI events. Care bundle compliance data is being collected and will be available for reporting in the near term.
Querulous Queries: How to Improve Malnutrition Via Documentation

Link to Presentation

Yasaswi V. Vengalasetti, MS
Jason Hom MD
Mirvaldy Joseph, MD MBA
David Svec, MD MBA
Lisa Shieh, MD PhD
Medicine Department Anywhere from 20-50% of all hospitalized adult patients are impacted by malnourishment on admission. Malnutrition can complicate patient outcomes by increasing length of stay, infection rates, readmission rates, as well as treatment costs. By improving documentation for malnutrition, we have the potential to improve patient outcomes. Malnutrition can be a predictor for many other outcomes, and by documenting the predictor early we have an opportunity to act upon potential complications and improve patient outcomes.

At Stanford Hospital, Malnutrition was frequently the number one queried item in the medicine division of the hospital. Malnutrition was queried an average of 18-20 times over the past few years. By reducing the number of missed malnutrition diagnoses we should be able to reduce our query rate. We created an intervention that improves nutritional assessment and captures malnutrition-based diagnoses in an inpatient setting. . A survey was conducted to understand the comfort and confidence of physicians in diagnosing malnutrition. A macros was set up within a physician’s note that autopopulates from dietician’s note (group note).  We use a test of proportion to compare and contrast the query rate before and after the intervention. Our group note improves malnutrition documentation. Physicians can partner with dieticians to make the malnutrition diagnosis. The wide-spread adoption of the group note can streamline the documentation process reducing malnutrition missed-diagnosis, queries, and physician workload.
Improving Multi-directional, Real-time Feedback in Neurology Residency
Link to Presentation
Jingjing (Jenny) Chen, MD, MBA
Brian Scott, MD
Neurology Department Background: Residents are both learners and teachers. Effective feedback from all levels is crucial for improving clinical, interpersonal, and educational skills. Studies suggest residents are dissatisfied with the frequency and quality of feedback. We wanted to understand Stanford neurology residents’ current practices, perceptions, and satisfaction with feedback. From this needs assessment, we hope to develop tools to improve multi-directional, real-time feedback during residency training.

Methods: Stanford Neurology Residents participated in two live polls (n=21, 18) assessing feedback between themselves and their juniors (medical students and junior residents), peers (same level of training), and superiors (senior residents, fellows, and attendings). Residents also participated in a focus group discussion on barriers to giving/receiving feedback.

Results: About half of the residents reported receiving little or far too little real-time feedback. The majority of residents do not feel comfortable giving feedback to peers and superiors, but are comfortable with juniors and medical students. 75% of residents never give feedback to peers; of those who do give feedback, the frequency is less than once/week to peers and superiors. Residents cited intrinsic hierarchy of medicine as the number one barrier for giving feedback. They also described a lack of psychological safety, such as fear of retaliation, offending others, and damaging relationships. Other factors include lack of know-how for giving/receiving feedback and dedicated time/space.

Insights and conclusions: Residents experience inadequate, real-time feedback from all levels. Rather than personal factors, the biggest barriers to giving feedback are related to organizational cultural/structural issues (hierarchy, lack of psychological safety, lack of workflow process). Our SMART goal is to increase the frequency of real-time bidirectional feedback for residents from all training years by at least 50% by winter 2021. We created a resident workshop on giving feedback. We also plan on exploring how faculty can be involved to change the feedback culture through department-wide competency workshops, dialogues on feedback expectations, and standardized workflow processes.