Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
Predictive Analytics in Radiation Oncology
Michael Peters, MBA. R.T.(R)(T)
September 2014
USCANCERSPECIALIST,LLC
About USCANCERSPECIALIST
u President/CEO of USCANCERSPECIALIST,LLC u 20 total years of Oncology and Imaging Experience:
§ 4 years direct Clinical Imaging and Radiation Therapy Care § 11 in Management/Administrative Positions § 5 years of Oncology and Imaging Service Line Consulting
The State of Healthcare
Meeting the Changing Market Demands within the Healthcare Environment
Improve Quality and Safety
Op5mize Resource U5liza5on
Evidence-‐based Outcomes
Reduce Costs (Efficiency)
Maximize Revenue & ROI
Landscape changes +
Gaps +
New solution validation
Cost Effec*ve, High Quality Pa*ent-‐Centric Care
Healthcare Inefficiency in the US
• United States health care system is the most expensive in the world.
• U.S. underperforms relative to other countries on most dimensions of performance. (AUS, CAN, FRA, GER, NETH, NZ, NOR, SWE, SWIZ, UK)
• U.S. is last or near last on dimensions of access, efficiency, equity, quality and health outcomes.
• Affordable Care Act is increasing the number of Americans with coverage and access to care, but still lacks in Universal Health Insurance and accessibility of care.
AUS CAN FRA GER NETH NZ NOR SWE SWIZ UK US
OVERALL RANKING (2013) 4 10 9 5 5 7 7 3 2 1 11
Quality Care 2 9 8 7 5 4 11 10 3 1 5
Effective Care 4 7 9 6 5 2 11 10 8 1 3
Safe Care 3 10 2 6 7 9 11 5 4 1 7
Coordinated Care 4 8 9 10 5 2 7 11 3 1 6
Patient-Centered Care
5 8 10 7 3 6 11 9 2 1 4
Access 8 9 11 2 4 7 6 4 2 1 9
Cost-Related Problem 9 5 10 4 8 6 3 1 7 1 11
Timeliness of Care 6 11 10 4 2 7 8 9 1 3 5
Efficiency 4 10 8 9 7 3 4 2 6 1 11
Equity 5 9 7 4 8 10 6 1 2 2 11
Healthy Lives
4 8 1 7 5 9 6 2 3 10 11
Health Expenditures/Capita, 2011** $3,800 $4,522 $4,118 $4,495 $5,099 $3,182 $5,669 $3,925 $5,643 $3,405 $8,508
COUNTRY RANKINGS
Top 2*
Middle
Bottom 2*
EXHIBIT ES-1. OVERALL RANKING
Notes: * Includes ties. ** Expenditures shown in $US PPP (purchasing power parity); Australian $ data are from 2010.Source: Calculated by The Commonwealth Fund based on 2011 International Health Policy Survey of Sicker Adults; 2012 International Health Policy Survey of Primary Care Physicians; 2013 International Health Policy Survey; Commonwealth Fund National Scorecard 2011; World Health Organization; and Organization for Economic Cooperation and Development, OECD Health Data, 2013 (Paris: OECD, Nov. 2013).
A Personal Perspective-Radiation Oncology Challenges within Healthcare
Staffing Scheduling Technology Patient, Staff, Physician Satisfaction
• Temporary Staffing • Daily Patient
Volume fluctuation • Maternity/Paternity
Leave • Medical Leave
Mergers/Consolidation
• Seasonal Fluctuation,
• Linac Availability • Time from Referral
to Consult • Time from
Simulation to Treatment
• Physics QA
• New Technology Justification
• New Technology Adoption (SRS/SBRT, PET/CT, HDR-Mammosite©)
• Patient Wait Times • Staff Turnover • Physician Referrals
Which way do we ^ Move Forward?
Patient Flow How do I determine the right type of staff, the optimum schedule, and the correct staffing requirements for optimized patient flow? Workflow Process Improvement Initiatives Which one(s) should be implemented and how? Facility/Project Planning Do we have to expand or just re-design our existing facilities to meet demand and capacity issues? New Technology Evaluation and Implementation Do we have the right technology to meet the changing physician and patient demand for care?
What Value Added Services Provide Value
u Have access to high value Clinical and Professional Consulting services, specializing in Oncology specific Data Analytics.
u Have access to detailed customized
analytic reports that offer oncology-specific planning and management capability with the ability to offer a solution that supports capacity and scenario planning.
Value of Predictive Modeling Software
• Enables hospital administrators to: • Benchmark workflow, standardization
• Perform forward predictive planning of resources
• Budget for patient treatment capacity
All without affecting the operational service
• Allows for more informed decisions about cancer care for the benefit of patients – saving everyone involved valuable time and money
Managing the Challenges of
Predictive Analytics
Improving Performance and Patient Satisfaction through Predictive Analytics
u Effectively utilizing Computer Modeling to account for "what if" questions regarding various scenarios of Cost, Capacity, Demand, and Workflow.
u Effectively Interpreting Utilization and Efficiency data to collectively derive an operational plan to improve Patient Satisfaction. PATIENT
SATISFACTION
Cost
Efficiency Utilization
Shaping Efficiency and Utilization
Actual Patient Throughput* Efficiency = ---------------------------------------------- Effectiveness of Service Capacity Actual Patient Throughput* Utilization = --------------------------------------------- Designed Service Capacity
*Daily, Monthly, Annually
Revealing the Bottlenecks and Constraints
DEMAND/CAPACITY = THROUGHPUT
Providing Effective Capacity Planning
1. An Understanding of Current Services 1. How resources are utilized 2. Cost of resource utilization per patient
2. Experiment with alternatives 1. Model the effects of introducing a new Linac,
increase of patient volume, staffing level changes
3. Strategize for the Future 1. New satellite facility, consolidation of services,
Program Center of Excellence
Benefit of Predictive Analytics
Tech. and Equipment
Scheduling
Treatment Service
Staffing
• Improve patient flow • Track highly variable arrival patterns • Reducing long delays, door-to-provider
• Compare and evaluate similar technologies • Analyze implementation strategies • Develop technology integration plans
• Capacity analysis • Optimized room and resource utilization • Analyze movement of patients through the department
• Predicting effects of absenteeism • High level capacity demand modeling and analysis • Cost effective staff competency improvements
Defining the Process
18
The Process
1. Create a Model of the current environment (Baseline)
2. Understand/ Evaluate areas for opportunity (Review Baseline)
3. Simulate new scenarios (Simulation)
4. Plan for the future (Evaluation)
Breaking Down the Door on Big Data
Radical Palliative Diagnosis Overhead Salaries/Wages Overtime Wait times Tx Commencement Utilization
Creating the Baseline: Data Requirements
Creating the Baseline: Process Mapping
• Representation of Patient and Staff workflow
• Represent the delivery of RT (technique or protocol) to a patient.
• Built using components representing processes (planning, treatment delivery).
• Comprised of tasks (time based) performed by resources in the department.
• Establishes detailed level of operational efficiency.
Providing Value in Throughput Modeling:
Case Scenarios
Case 1: Acquisition/Consolidation
Problem Statement: Health System is considering an acquisition of a freestanding center. Should the health system maintain or consolidate services? Hypothesis: Overall sense within the health system is no need to keep acquired center open due to low volume and antiquated technology. Objective: Determine the costs and financial savings that consolidation would provide, while maintaining efficient throughput as to not impact patient wait times and staff/physician utilization.
Breaking Down the Feasibility
US Cancer Specialist 24
925
224
701
0 200 400 600 800 1000
Consolidation
Freestanding
Main
AvgPatientsPerYear
Consolidation
Freestanding
Main
2027632
870693
1994359
$0 $1,000,000 $2,000,000
Consolidation
Freestanding
Main
Overhead Costs per Year
Consolidation
Freestanding
Main
18 Total Staff Members (5 RTT’s)
• 701 New/Treated RO Patients
• 8am-6pm Operating Hours
• Financial Overhead $2M
• 7 Total Staff Members (2.5 RTT’s)
• 225 New/Treated RO Patients
• 8am-5pm Operating Hours
• Financial Overhead $870k
$0 $100 $200 $300 $400 $500
Administration-Utilised
Oncology-Utilised
Physics-Utilised
Radiography-Utilised
Staff-Utilised
Linac-Utilised
Pre-Treatment equipment-Utilised
Equipment-Utilised
Overheads-Total
Costs-Total
Thousands
Total Cost Category Cost
Freestanding $0 $100 $200 $300 $400 $500
Administration-Utilised
Oncology-Utilised
Physics-Utilised
Radiography-Utilised
Staff-Utilised
Linac-Utilised
Pre-Treatment equipment-Utilised
Equipment-Utilised
Overheads-Total
Costs-Total
Thousands
Total Cost Category Cost
Main Center
Consolidated Patients
$0 $100 $200 $300 $400 $500
Administration-Utilised Oncology-Utilised
Physics-Utilised Radiography-Utilised
Staff-Utilised Linac-Utilised
Pre-Treatment equipment-Utilised Equipment-Utilised
Overheads-Total Costs-Total
Thousands
Total Cost Category Cost
KEY POINTS: • Main Center Total Costs as part
of Consolidation increased only $33k • No Increase of Staff or Resources
to Treat Additional patients • Freestanding alone Total Costs for Same Pt. Load at $870k
Overall Utilization by $$$$
Reading Between the Lines
0 5000 10000 15000 20000
Consolidation
Freestanding
Main
TotalFractionsPerYear
0 10 20 30 40 50 60 70
Consolidation
Freestanding
Main
Average Utilization % per Linac
0 0.5 1 1.5 2 2.5
Consolidation
Freestanding
Main
AvgFractionsPerHour
0 20 40 60 80 100
Consolidation
Freestanding
Main
Average Utilization % of RTT's
Threshold
Monday Tuesday Wednesday Thursday Friday
Main L1 20.91836735 21.23076923 20.57142857 21.49056604 21.14
Consolidation 29.26530612 30.5 30.18367347 30.01886792 31.04
Main L2 24.14285714 25.86538462 25.57142857 25.24528302 26
Consolidation 29.24489796 30.40384615 30.12244898 30.26415094 30.82
0
5
10
15
20
25
30
35
# o
f P
atie
nts
per
Day
Patient Throughput per Linac
Demand on Resources (Linac)
Consolidation of Services Utilization Impact
Main Center Main + Pt. Consolidation
60.36% 64.04% 61.86% 61.20% 63.86%
18.67% 20.67% 22.70% 22.71% 20.87%
21% 15% 15% 16% 15%
2.53% 2.54% 2.24% 2.43% 2.41%
0.00%
20.00%
40.00%
60.00%
80.00%
100.00%
120.00%
OverTime
Modifiers
Unutilised
Utilised 68% 72% 71% 70% 73%
10% 11% 10% 13% 11%
21% 17% 19% 17% 16%
5% 6% 6% 6% 6%
0.00%
20.00%
40.00%
60.00%
80.00%
100.00%
120.00%
OverTime
Modifiers
Unutilised
Utilised
Radiation Therapist Utilization
0 50 100 150 200 250 300 350 400
<=3 Days
3-7 Days
8-14 Days
15-28 Days
>28 Days
Consolidation
Main Center
Freestanding
True Measure of Efficiency & Quality
Increase Operating Hours
Increase Resources (Linac
and Staff)
Decrease Days from Referral to
Tx
20
13 12
0
5
10
15
20
25
Consolidation Main Center Freestanding
AverageDaysToTreat
Conclusion:
• Limited financial impact to Main Center for Consolidation of Services
• Significant financial savings associated with closing freestanding center.
• Utilization of RTT’s above threshold and Linac Utilization change minimal.
• Increase of wait time (days)from Referral to Tx with Consolidation.
Case Example #2
Problem Statement: Center has experienced increased overhead costs and increased Patient wait times for Breast Cancer Patients, in comparison to other disease sites. Hypothesis: Non-conformity in the methods that physicians treat Breast Cancer Patients within our center has led to the increases in costs and wait times. (Seven Different Breast Care Pathways) Objective: Determine the cost difference, staff utilization, and patient wait times for breast Treatment based upon the different methods our physicians treat breast cancer patients within our center.
What is this Costing us?
Overhead cost per patient
• On a per patient basis, Prone Breast Tx accounts for the highest associated cost.
• Poses a clinical question of interest: Does Prone Breast offer a better Quality Outcome?
Cost breakdown per disease site (breast)
• Breast accounts for 25% of the total 701 New Patients seen in this center.
• Respectively it also accounts for 27% of the Total associated department Costs, specifically in Staff Utilization Costs.
$0 $1,000 $2,000 $3,000 $4,000 $5,000
Prone Breast Tx (11)
IMRT (13)
Breast 3D with E-Boost (72)
Accelerated 15FX (6)
HDR Breast (47)
Routine Pathway 20Fx (8)
SBRT (18) $0 $100 $200 $300 $400 $500 $600
Administration
Oncology
Physics
Radiography
Staff-Total
Linac
Pre-Treatment equipment
Equipment-Total
Overheads-Total
Costs-Total
Thousands
What is the Staff time Utilization?
0 100 200 300 400
Oncology
Immobilisation
Imaging
Pre-treatment Eqmt
Physics planning
Treatment
Linac
Administration
15fxs
IMRT
20fxs
SBRT
Prone
3D
HDR
Time in minutes
Total Time per Competency/Capability per Pathway
What is our Referral to Tx time?
5.33
15.78
16
13.61 16.45
17.66
19.03 HDR
3D
Prone
SBRT
20fx
IMRT
15fx
0
2
4
6
8
10
12
14
16
18
20
Booking Simulation / virtual
simulation
Physics planning
QA Plan Approval
Total Time: Referral to
Tx
HDR
3Dw/electron
Prone Tx
Breast SBRT
20fxs
IMRT
15fxs
Average time (days) per pathway component By Patient classification
Total Time (days) from Referral to Treatment
How does Breast delay rank?
• Breast has the greatest percentage of patients 24% waiting at least 1hr for a Task to start. Indicating lack of resources to complete associated tasks.
• Lung has the greatest % of patients waiting under a 1/2hr demonstrating better efficiency.
• Not a daily basis wait time, but the percent of patients that will expect to see a wait time of that length during at least once during their care cycle.
DISTRIBUTION OF DELAYS
FOR TASK COMMENCEMENT
0.00% 5.00% 10.00% 15.00% 20.00% 25.00% 30.00%
0
0.5
1
1.5
2
2.5
3
Breast (175)
Prostate (96)
Rectum/LGI (74) Lung (87)
hour
s
Conclusion:
• Prone Breast Tx Care Path is the least cost effective model of care.
• Breast Tx accounts for 27% of Total department costs and 25% of Total patients.
• Breast Tx takes an average of 16 days to start Tx, 4 days longer than any other disease Tx path.
• Breast Tx Care Paths have the highest percentage of patients waiting at least 1hr at any given point for a Task completion.
FINAL RECAP
A Recap of the Value of Predictive Analytics
u Baseline modeling of operations and financials.
u Reduction in business and financial risks using real life models and simulations.
u Quantitative data for budget justification for staffing, equipment and facilities.
u Ability to Benchmark Nationally with centers of same size and demographics.
u Increase businesses efficiencies, productivity and profitability.
u Standardization to deliver best practices amongst centers.
Contact Information: [email protected]