BUSINESS INTELLIGENCE: MINING OUR DATA TO IMPROVE OUR OUTCOMES - PowerPoint PPT Presentation

BUSINESS INTELLIGENCE: MINING OUR DATA TO IMPROVE OUR OUTCOMES Michael P. Recht Louis Marx Professor and Chair NYU Langone Medical Center

USING DATA TO CREATE VALUE  To succeed, Imaging departments need to change and adapt to new environment and external pressures  The only way to successfully do that is to collect, aggregate and analyze data  And then to Change and https://www.pinterest.com/pin Adapt to what the data tells you

DATA  Needs to be :  Accurate  Customizable

DATA  Needs to be : “The issue is that companies don't create value by owning the latest  Accurate technology, or by managing data  Customizable access rights. …Value is created by putting data in the hands of business  Accessible people (“radiologists”). Successful CIOs are moving to an Agile approach  Transparent/ easily where technologists, data scientists understood and business people work in close  Real time collaboration to focus the technology investments and analytical activities on the areas with the greatest potential to derive real, meaningful insights for the business.” Berez, Bain

DATA  Needs to be :  Accurate  Customizable  Accessible  Real time  Transparent/ easily understood  Actionable http://smallbiztrends.com/2016/01

DATA “ 90% of the data you collect will never be actionable or even helpful.” Albert Einstein http://smallbiztrends.com/2016/01

DATA  Data should not be punitive  Group/section oriented rather than individual  Take advantage of the Hawthorne effect  Not a “report card” but a tool to allow everyone to be involved and improve

FIVE STAGE RESPONSE  Denial  Anger  Bargaining  Depression  Acceptance

DASHBOARDS  Customize for users  Leadership  Section heads/section members  Admin  Chief techs  Billers/ Pre-cert 

DASHBOARDS  Volume  Scheduled exams  Performed exams  Exams per day  Historical trends

DASHBOARDS  Quality  On time starts  Turnaround time  Unread cases  Radiation dose  Enhanced imaging report  Structured reporting  Retake

DASHBOARDS  Finance/admin  Insurance payments by payer by modality  Denial information  AR  Pre-certification data

DASHBOARDS  Research/Education  Grants  NIH  Non-NIH  Mentors  Tenure timelines  Resident case logs

MR REENGINEERING  Backlog  Inpatients  Pediatric Anesthesia  On Time Performance  Scan duration  Exam volume

RESULTS  Large number of patients not scanned on scheduled magnet – led to significant downstream delays  Developed “electronic greaseboard”  Tracked every step of patient experience  Available on ipads and large screens  Drag and drop

Results  Outpatient anesthesia  Moved slots to 7-11 every day and all day on Saturday  Increased slots from 17 to 25  Changed hour of operations – more weekends and evenings  Lack of standardization across sites/scanners  Quality manager  Enforced standard protocols by all radiologists  Appropriately “timed” exams  Technologists key to success –Best in Practice

RESULTS  Increased # of techs per magnet  Compared times with either 1 or 2 techs per magnet  Average time saved/transition with 2 techs = 5 min 8 sec  At 16 cases/day/magnet = 1 hr 22 min saved – allowed 2 more cases per day  Regular follow up meetings  Meet with MR “leadership” team biweekly

RESULTS 59% growth from 2013 to 2017 13% growth first quarter 2018

NEXT STEPS  Increase duty cycle of scanners  Decrease turnaround times between patients  Dockable tables and dedicated prep rooms  RTLS  Machine “up Time”  Service costs  Scheduling Efficiency  No shows/ Cancellations  Same day add ons

SCANNER UTILIZATION

INTRASCANWORKFLOW

NEW MR CENTER  Dockable tables  Dedicated prep room  2 doors in and out of each MR scan room  1 directly connected to prep room  Scanner angled to provide straight line to prep room  Coils in prep room  Key coils duplicated  3 techs per scanner

BENEFIT OF SPEED  If save 5 minutes per exam # of # of scanners avg additional  At 16 cases/day/magnet and an average of 40 in reimbursement avg days scans per increased revenue system per scan scanning day per year minutes per exam = 2 extra cases per day  At $500 per case = $260,000 per magnet per year $ 45 500.00 260 1 $ 22,500.00 $ 5,850,000.00 $ 45 500.00 260 2 $ 45,000.00 $ 11,700,000.00 $ 45 500.00 260 3 $ 67,500.00 $17,550,000.00 $ 45 500.00 260 4 $90,000.00 $23,400,000.00

ULTRASOUND  Typically 1 tech per room  Exam slots 30 min Registration Changing US Room 21 min 18 min 20 min 22 min 19 min 21 min

ULTRASOUND  Pilot  2 techs per room  20 minute room time  Enables you to eliminate 1 room out of 3 or increase volume by 33% with same number of rooms  Pilot demonstrated  improved on-time performance  Increased tech satisfaction

QUALITY  Need to learn form our mistakes (“failures”)  Book contrasts airline industry with radiology (mammography)  How do we get feedback (data) about our performance

QUALITY- RAD-PATH CORRELATION

QUALITY- RAD-PATH CORRELATION

RAD-PATH CORRELATION  Approximately 5000 per month  Now expanded to OR notes  Goal is to discover systematic trends rather than anecdotal or individual's errors  Peer Learning vs Peer Review  Approx 150 learning opportunities per month  Each section working on research project based on results  Now extended to pathologists

SUMMARY  Data needs to be  Accurate  Transparent  Non-threatening  Actionable “Without a goal, analytics is aimless and worthless ” Michael Porter