IT STRATEGY BOARD September 27, 2016 AGENDA > Call to Order - - PowerPoint PPT Presentation
IT STRATEGY BOARD September 27, 2016 AGENDA > Call to Order Welcome and introductions Proposed IT Strategy Board 2016-2017 agenda > Research High Performance Computing Strategy > Major Projects Update UW Medicine EPIC
—Welcome and introductions —Proposed IT Strategy Board 2016-2017 agenda
—UW Medicine EPIC Migration —HR/Payroll Modernization —Integrated Service Center Update > IT Project Portfolio Executive Review
2
Month IT Strategy Board Topic
September 27, 2016
February 6, 2017
May 8, 2017
3
4
5
Kelli Trosvig Vice President, UW-IT and Chief Information Officer Martin Savage Chair, Hyak Governance Board Professor, Physics
Martin Savage – Chair, Hyak Governance Board
6
Educating the next generation of leaders UWIT’s Hyak and Lolo
> In choosing Seattle and Washington State, our UW community deserves to have competitive Simulation/Scientific Computing/Data research capabilities and educational opportunities and, ideally, considerably more > A viable and significant growth strategy is required for UWIT’s research and education High Performance Computing and Data Ecosystem (HPCDE)
7
> 125 Petaflops > 10.6 Million compute cores > 1.3 PB ram > x5 any US machine
Wired, June 21, 2016
8
9
Addressing present and future challenges
10
SUPERCOMPUTING, THE CANCER MOONSHOT AND BEYOND
Designing the next generation of supercomputers for biomedical research
THE INNOVATORS
Lecture S eries G uest Lecture
Kusnezov
Chief scientist
U.S. Department of Energy National Nuclear Security Administration
How can the next generation of supercomputers unlock biomedical mysteries that will shape the future practice of medicine? Scientists behind the National Strategic Computing Initiative, a federal strategy for investing in high-performance computing, are explor - ing this question. Guest lecturer Dr. Dimitri Kusnezov serves as chief scientist in the U.S. Department of Energy National Nuclear Security Administration and advises the secre - tary of energy. His lecture explores how exascale computing—computing systems capable of a billion billion calculations per second—can open new doors in biomedical research. He discusses how supercom - puting could support federal research, technology, and policy initiatives, including cancer research and the quest to revolutionize understanding of the human brain.
September 20, 2016 4:00 p.m.
Allen Institute 615 Westlake Ave. North Seattle, WA
Hosted eception to follow
Register Today: innovators.wsu.edu
R eservations required by S eptember 15
11
> 2008 Physics/Astro Athenna cluster: ~10 users, a few Tflops > Phase 1: Factor of x20 growth in compute capability between 2010 and 2016 (185 Tflops) — kept up with #500 position — mainly CPU, but with an IBM GPU testbed > NSF MRI : Intel Xeon-Phi components to be added, ~$800K > 2016: Delivers >7 Million Core-hrs per month to UW community > Phase 2: > 10 Pflops in 2020, with accompanying data and communication infrastructure
12
2016 Hyak Support: > Stephen Fralich > Pramad Gupta > Chance Reschke
2017 … +
100% FTE 33% 50% 50% > UW HPC expertise imparted to faculty and students alike to take them from laptops or iPads to modern computing tools used to solve large-scale complex problems across disciplines > Prepares researchers to compete nationally in teams for DOE+NSF Leadership Class computing awards – INCITE, ALCC, Exascale Compute Projects
13
http://www.int.washington.edu/users/mjs5/HYAK/governance/
14
http://students.washington.edu/hpcc/
15
> SoS – The Speed of Science
– No proposal writing and hoping for resources – Immediate access to resources to try
> P4E – Preparing for Exascale
– Trying out new algorithms and codes with Exascale type architectures
> BDP – Big Data Pipelines
– Rapid access to large data sets and fast distribution – Local experiments, sensors, real- time processing
16
> Conversation with Researchers in 2009 > IT and Data Management Expertise > Data Management Infrastructure > Storage Infrastructure > Computing Power > Communication and Collaboration > Computational Leaders Meeting – 2007/08
— Mary Lidstrom/OR
> Kaplan Letter – 2007 > Chance Reschke is hired to deploy Hyak > Rob Fatland is hired to facilitate research use of cloud computing (2015)
17
18
> 2009 – 30 full racks > 2010 –Hyak creates consolidation, 10 racks → 1 Hyak rack > 2013 Biochem funds Hyak expansion
> Hardware footprint dropped from 30 racks to 12 racks > Dozens of research results published
0.00% 50.00% 100.00% 150.00% 200.00% 250.00% 300.00% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Month
UWB STEM Hyak % Utilization
UWB STEMhas averaged80% utilization
their two years using Hyak Hyak is an ELASTIC platform In periods
low use,
borrow from UWB STEM In periods
high use,
return the favor
100%
In 2014 UW Bothell STEM faculty demand for access to
HPC became critical.
initial response was to build a new data center costing as much as $20M. Because the STEMdean had the
investing in Hyak instead, the $20M went to start a new ME program, the STEM program spent $50k
central funds, and world class research results were being published within a year.
19
UWB STEM has averaged 80% utilization over two years Hyak is an ELASTIC platform
In 2014 UW Bothell STEM faculty demand for access to on-site HPC became critical. The
initial response was to build a new data center costing $20M. Because the STEM
dean had the option of investing in Hyak instead,the $20M started a new ME
program, the UWB invested $50k in Hyak, and world class research results were being published within one year.
UWB STEM Hyak % Utilization
20
Data from early 2015
5 10 15 20 25 IHME HHMI Biostat Statistics Chemistry Provisioned Hyak Equiv.
> Large colo deployments outside Hyak see low utilization, and fast growth, threatening a repeat of the past. > Bringing these groups into the Hyak fold is
> Others remain challenging.
21
22
23
— Unanimously voiced it to be a great success and vital to UW Research — Unanimously voted to embark on Next-Gen Hyak - Phase-2 — Unanimously voted not to terminate UW’s HPCDE
24
Hyak
Extra Large Medium Small 2,800 2,100 867
6,000 4,000 2,000
Cores per FTE at US HPC Centers
4,696
25
– Hyak HPCDE is embedded into UW-IT infrastructure, >7M core-hrs/month – Saves UW lots of money through economy of scales, power, people
> e.g., 20 racks at 20% usage 4 racks at 100% usage
– Brings in grant money and diversifies UW research capabilities – Attracts and retains scientists at all career stages
– Cost is carried by PIs (grants), students (STF), Colleges, Departments and UW-IT
26
> 2016-2020: Next-Generation Hyak being deployed
– UW-IT is adding more staff
> Mitigate system support risk
– Additional staff needed
> Support, educate, trouble-shoot, optimize
> HPCDE is important to education and research, and needs a sustainable funding stream to enable UW to remain competitive > Discussion
27
28
29
James Fine Chief Information Officer Information Technology Services, UW Medicine
30
Overview of Current State – EHR Mosaic > UW Medicine: Multiple EHRs Today
– Cerner – UWMC and HMC
– EpicCare
Primary & Specialty at UWMC, HMC and NWH
– Siemens Soarian
– Epic - Valley Medical(Separate instance)
– MIND data repository and MINDscape – Department Systems:
> Practice Management: Business Systems
– Epic – UWMC, HMC, NWH
– Epic – Valley (separate instance)
– McKesson
30
31
UW Medicine
UWMC HMC SCCA NWH Neighborhood Clinics Valley Medical Center
Overview of Current State – EHR Mosaic
Mindscape
31
32
History of UW Medicine EHR Strategy
2001-2 Changed strategy to Cerner EHR 2008 Decision to pursue Mosaic EHR Strategy 1996 - 1997 Emtek/Eclipsys as enterprise EHR MIND data / MINDscape Epic neighborhood clinics Consider Transition to Single EHR Cerner Inpatient and Epic Ambulatory
not ready
superior outpatient product; inpatient capability not viewed as mature
32
33
UWM Information Systems Diagram
Theory…
33
34
Current State Interfaces Diagram
…Reality
34
35
What We Heard - Drivers For Moving to One EHR
Improve Patient Safety
> ONE Medical Record for Inpatient AND Outpatient — Patient problem list — Orders — Allergies — Medications — Results > “One source of truth” across locations AND time
Improve Ease of Use for Clinicians & Staff
> Training and staying up to date in two systems a challenge > Simplify communication between providers that share a patient > Frustration around duplicative work. Looking for information in multiple places
Reduce Complexity & Potentially Cost
> Managing two of everything > Making EHRs “talk to each other” is costly > Duplicative costs of maintaining separate inpatient and outpatient EHRs
Stay competitive in the marketplace
> ACNs, Population Health require integration
External Drivers
> Meaningful Use > Sharing patient records outside of UWM
35
36
> EpicCare Inpatient > Inpatient Pharmacy > Ambulatory Pharmacy > Emergency Dept. > HIM Chart Tracking > Inpatient Portal > OB/L&D – Inpatient > Oncology - UWMC > OR - Procedures
One EHR Options – Epic Implementation Minimum Install Fully Integrated EHR
Minimum Plus: > Anesthesia + OR > Cardiovascular > Ophthalmology > Transplant > Lab > Radiology Pros: > Smaller scope & initial cost Cons: > Least integrated, more interfaces required > Additional future go-lives required Pros: > Fully integrated solution, minimal interfaces required Cons: > Largest scope & initial cost
36
37
Timing Considerations
> 3-5 years to plan, procure, and implement > Implementation timing considerations:
> NWH: Siemens Contract > ACN/ACO Projects, MU, HIE, MACRA and other govn’t mandates > Major system upgrades – eg. Cerner, McKesson … > UW Campus HR/Payroll project > Enterprise Financial System > Enterprise Pharmacy inventory mgmt. system (just starting)
37
38
Preliminary Estimate of Costs of Moving to One EHR
— Scope: includes UWMC, HMC, NWH only — Transition to Epic Inpatient — OR & Anesthesia in scope — Lab & Radiology not in scope — Valley Medical and SCCA not included in scope estimate
38
39
Original cost estimates with start 2017
Fiscal Year 2017 (NEW) > $22.5M
– $16.3M OpEx – $6.2M Cap
> Project Activities – Jul’17-Mar’18 – Planning, Ramp- Up – Apr-Jun’18 - Design Notes > Current project estimate = $176M
– $99M OpEx – $77M Cap
> Does not include operations staff costs for Subject Matter Experts (SMEs), training time, expert users > FY20 – FY22 costs reflect proposed vendor payment plan
FY17 FY18 FY19 FY20 FY21 FY22 OpEx $16.3 $23.4 $59.0 $0.5 $0.5 $0.2 Cap $6.2 $31.6 $23.7 $5.9 $5.9 $3.0 $- $10.0 $20.0 $30.0 $40.0 $50.0 $60.0 $70.0 $80.0 $90.0
Projected Cost
(Millions)
Cap OpEx 39
Given our cost estimates we commissioned a formal ROI study
Program Description: Implement an integrated EHR across inpatient, ambulatory, and procedure areas as well as selected specialty and ancillary areas of the UW Medicine entities (excluding Valley Medical Center) Current-State Challenges
Soarian for Inpatient) and numerous essential department-based ancillary systems (surgical services, cardiology, ED, etc.) Current state challenges also include additional automated systems and areas where paper orders and documentation are still in use.
the UW Medicine enterprise and causes challenges with scheduling, documentation, ordering, results, charging, and billing. It compromises our ability to deliver comprehensive patient care across the continuum of care and to more efficiently manage population health across the enterprise.
40
ROI analysis is focused on the costs and benefits with implementing the following integrated solutions:
> Inpatient EHR - including CPOE, Clin Doc, Med Admin, Interdisciplinary Care Plan, Clinical Pathways, ICU, Hospital Outpatient Depts. > Pharmacy - Inpatient, ambulatory/retail and others (to replace ETRBY) > Emergency department information system > Specialty information systems
– Oncology – Cardiovascular – Operating room management – Anesthesia – Obstetric – Ophthalmology functionality required for Procedure areas – Ambulatory transplant module to replace homegrown MAX
> Tablet-based system for patients (e.g., treatment team’s info, educational material, videos) > Patient kiosk system - key for Procedure Area workflows > Epic-approved Bedside Monitor Device Interface (BMDI) solution > Epic HIM Modules – deficiency and ROI required to replace Cerner * ROI Model can be adjusted for changes to in-scope apps
41
Readiness Program continues Approvals Q4 CY16 Program Planning starts Jan 2017 Requirements Gathering starts Mar 2017 Project Staff Training begins Mar/Apr 2017 Execution/ Design Kickoff begins Q4 CY17 Build begins Q1 CY18 Test begins Q4 CY18 User Training begins Q1 CY19 Go Live 1 Q2 CY19 Go Live 2 Q4 CY19
The rough timeline used for developing the cost and benefit models A critical part of the detailed program planning effort (targeted to begin Jan 2017) is to validate the project timeline, scope, and staffing assumptions developed during the program readiness and ROI activities
42
The model has been updated to include operations resource costs and consistent escalation; it will continue to evolve through the planning phase of the project
FY 2017 FY 2018 FY 2019 FY 2020 FY 2021 FY 2022 FY 2023 FY 2024 FY 2025 FY 2026 Total
TOTAL COSTS
4.6 54.5 70.5 49.1 9.5 9.5 3.2 3.3 3.3 3.4 210.8 Original Cost Model Adjusted for Escalation
173.2 $
Internal Resources
1.5 14.3 16.4 8.1
$
External Resources
1.5 24.6 38.4 23.4
$
Software
5.5 8.0 3.0 3.0
$
Software Vendor Support
1.0 3.3 3.4 3.4
$
Training
0.1 0.4 0.4
$
Supplies and Facility Costs
0.1 0.3 0.2 0.1
$
Project Related FTE Backfill
3.3 1.6
$ Incremental Costs Identified
37.6 $
Operational Resources*
0.2 3.3 4.3 1.6
$ Hardware 1.2 3.7
$ Software/Hardware Maintenance (previously classified as negative benefits)
1.0 2.9 3.0 3.1 3.2 3.3 3.3 3.4 23.4 $ TOTAL BENEFITS
28.3 31.3 32.3 33.2 34.2 35.2 203.9 $ Legacy Software & Hardware Maint. Fees
7.0 7.2 7.4 7.6 7.9 8.1 47.0 $ NWH Cerner Recurring & Support Fees
8.1 8.3 8.6 8.8 9.1 9.4 58.2 $ ITS Resources
3.50 3.60 3.71 3.82 3.93 4.05 23.5 $ Operational Benefits
9.7 12.2 12.6 12.9 13.3 13.7 75.3 $ Annual Net Project Benefit (4.6) (54.5) (70.5) (39.7) 18.8 21.8 29.1 30.0 30.9 31.8 (6.9) Cumulative Net Project Benefit (4.6) (59.1) (129.5) (169.2) (150.4) (128.6) (99.6) (69.6) (38.7) (6.9) Break Even in Years 10.7 10-year Project ROI
$ in Millions *Additional costs may be incurred by operational departments to cover the time for resources who spend <20 hours per week on the project including but not limited to time spent in end user training, advisory council/stakeholder meetings, and intitial requirements workshops.
43
If UW Medicine remains in current state, the following challenges will persist : > Enterprise-wide inefficiencies resulting from having to work in and maintain multiple EHR systems
— Workflow inefficiencies caused by having to navigate and find information across two systems and by re-entering information contained from one system to the other — Duplicative costs in IT for software licensing, ongoing maintenance and staff, and system interfaces — The large knowledge burden to know both systems in depth and to be an efficient, expert and safe user, adding cognitive overhead for clinicians
> High probability for errors in patient care, reduced quality, and continued challenges in the management of patients and populations across the inpatient-
> Misalignment with other Epic customers, limiting opportunities for collaboration via Care Everywhere > The need remains to replace the aging EHR platform at NWH to avoid the severe system risk as it will no longer be supported by the vendor
44
> Improved Clinical Quality & Patient Safety – Clinicians are able to view active
> Increased opportunity to interact with patient and improved perception of care provided > Increased Patient Satisfaction – Currently patients must visit multiple clinical patient portals with unclear separation of similar information; this translates into a perceived “lack of a health system” and low patient satisfaction. Enterprise EHR will provide a single clinical patient portal and a more cohesive appearance to the UW Medicine health system. > Enables UW Medicine to adapt to changing reimbursement where measuring and enhancing quality across the care continuum are imperative towards survival > Improved communications among clinical staff through a single source of and repository for clinical documentation; reduced risk
45
> Customization is key in many areas:
— CORES report: Custom clinical rounding report — Voice Recognition for Anesthesia: UW's anesthesia system uses highly customized voice recognition enhancements. — Enhancements in Epic: Concern regarding Epic -- innovation and improvements may be harder with Epic as they build to the masses.
> SCCA Power plans: The SCCA power plans were developed with organizational subject matter experts and there is a concern that a duplicative effort will be required to transfer this information into Epic. > Epic cost: The Epic solution has a high cost in time and money. There will be a time
Reviewing the intent of moving to an Enterprise EHR, concerns do exist: Despite concerns, Clinicians and Operational staff in attendance showed their enthusiastic support for moving to an Enterprise EHR.
> We’re currently setup for failure: The cognitive overhead and manual processes we deploy to ensure quality and safety for our patients are not sustainable; boils down to communication and safe transfers and our systems do not allow us to do it effectively. > We need this for the long vision: Enterprise EHR enables UW Medicine to serve the patient as a Health System, seamlessly, efficiently and effectively.
46
If UW is to realize benefits with an initiative of this magnitude, several key attributes, structures and processes need to be in place Common goals and shared vision of one enterprise record, including:
Executive sponsorship outside of IT, including:
Empowered governance structure for the enterprise including:
Engagement at all levels across all departments/disciplines, including:
47
48
49
Kelli Trosvig Vice President, UW-IT and Chief Information Officer Lead Executive Sponsor, HR/Payroll Modernization Program
50
> Revised Budget per approval of the Board of Regents, May 2016 > Go-live estimated June 2017 > Does not include stabilization costs post go-live
51
> Completed intensive future state business process design reviews with unanimous business owner endorsement of all 14 end-to-end critical processes > Completed Benefits future state business process design reviews, with Health Care Authority and Benefits Office sign-off on all eight key benefits business processes > Led Campus Foundations – providing orientation on the foundational elements of Workday, how the system works, and how we are designing it for UW to 350 campus leaders and administrators > Completed the fifth of nine data conversions (extracting employee data from the legacy system and converting it into Workday) > Completed 2nd cycle of parallel payroll testing; reached goal of 90% data accuracy within the identified comparison thresholds > Finalized the design for the Integrated Service Center > Conducted Future State Process sessions for more than 100 campus leaders to review key campus facing business process designs and demonstrate configuration
52
> Round 1 of Health Care Authority testing taking place now > Second set of Future State Process sessions with campus leaders to be held in October > All configuration and unit testing to be completed by October 31 in preparation for Regression and Integration Testing (RIT), which begins in November > Finalize Parallel Payroll Testing cycle 3 > Build and validate 6th system conversion in October > Complete step two of the seven-step change management process:
— Educating campus leaders about the changes coming with Workday
> Launch step three of the change management process:
— Helping campus identify change impacts and create readiness plans — These plans will detail how units must prepare for the change to come (e.g., update an internal website or form, communicate Workday timelines, provide training)
53
54
55
Kelli Trosvig Vice President, UW-IT and Chief Information Officer Lead Executive Sponsor, HR/Payroll Modernization Program
Enhanced Customer Service
> Single point of contact, eliminate unnecessary hand-offs:
— Does not preclude local support where available, but eliminates multiple points
> Unified approach to support, training, communication and ongoing updates > Development of comprehensive knowledge base > End-to-end process insight across the HR, Academic HR, Benefit, and Payroll domains
Optimized Operations & Data Integrity
> Ensure the right people/areas are engaged in right steps to improve operations and maintain compliance
Continuous Improvement
> Develop competencies to address existing and emerging requirements
Consistent Use of Technology
> Leverage UW Connect (ServiceNow) service and knowledge management capabilities
56
—Offer a single point of contact and subject matter expertise for payroll, HR, and benefits — Reduce customer handoffs and keep them personal and high-touch — Simplify the process of getting help, provide application support for Workday — Enhance collaboration between functions now operating separately
— Provide guardrails for data integrity so that we can validate, audit, and verify employment data more accurately — Empower more strategic employment decision-making for the University
57
— Increase workflow efficiencies by providing alternate routing option for higher volume processes (Example: within Academic HR) — Improved reporting capabilities for academic units
58
Work-stream Jan Feb Mar Apr May Jun Jul Aug Sept Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sept Overall Operating Model Technology (UW Connect, WordPress website, Call Center telephony) Staffing & Facilities Org Change/Comm's, Readiness, Training Deploy/ Implement
Project Timeline
2016 2017
ISC Implementation Project Milestones HRPM Program Workday Go-Live Go/No-Go Decisions Establish Service Deliver Framework Establish Service Deliver Framework Service Delivery Processes & Procedures Phone Menu, UW Connect ISC Website Space approved Wave 1 Staffed Wave 2 Staffed Space build Wave 3 Staffed ISC TAP Initiative Kickoff ISC TAP Initiative Completed ISC Training Operational Readiness Testing ISC Soft Go-Live HRPM Decision Workday Go/No-Go
59
60
61
62
63
64
65