Download to read offline
More Related Content
Lost Orders Project_System Operations Presentation July 2013
- 1. Based in GreatFalls, Montana, Benefis serves a 15-county region with a population of approximately 230,000 people. With 2,600 employees, it is the largest non- governmental employer in the Great Falls area. The organization is comprised of two campuses with a total of 516 beds and partners with more than 250 area physicians.
- 2. Why This Project Asignificant number of charges are going to be written off secondary to missing orders or invalid orders exists.
- 3. Process Name andPurpose: Lost Physician Orders There are a significant number of medical record charts that are going to being written off, versus billed, secondary to missing physician orders. In trying to find missing orders, there is waste in search, duplication, and over processing, that is producing caregiver and provider frustration. There is a need to identify process root causes to ensure efficient and effective solutions are identified. Problem Statement: 1.Multiple requests for re-work/finding lost orders to ancillary services and physician offices 2.There are currently 648 invalid physician orders ($184,809) 3.There are currently 1216 missing physician orders ($635,274) Process Sponsor: Steve Ballock,CFO Process Boundaries: Starting Point: Patient presents for Procedure at Benefis Health System after outpatient physician care referral. Stopping Point: Order is scanned and available in medical record to code and drop bill timely (within 4 days) Process Owner: Julie Wall Team Lead: Laura Merchant Team Members: Joe LoDuca, Judy Rosales, Patty Harris, Sarah Hall, Greg Hilpert, Marci Huntsinger, Ingrid Dieudonne, Vicki LeBrun, Amy Linder, Shellie Curtis, Eric Peterson, Nathan Hough, Kristen Rowen, Hasim Turhan, Laura Merchant, Peter Gray, Julie Wall Team Leaders: Laura Merchant Key Customers: Patients, Physicians, Internally – Accounting, PBS, HIMSS Dept, Ancillary Services Departments, Quality Improvement Project Start Date: 10 December 2012 Proposed End Date: 15 June 2013 Project Vision Statement: All charts, following outpatient procedures will have a valid order 4 days post DOS, available for coding and billing on the electronic medical record. Project Deliverables: 1. A reliable and capable process. 2.Reduction of charts with missing orders by 50% by June 2013 3.Reduction of charts with missing orders by 80% by Sept. 2013. Resource Representatives: Coding, Billing, Radiology, Laboratory, BMG offices, Quality Improvement -SCIP Coordinator Project Charter 2012:133 Missing Outpatient Physician Orders
- 4. The Vision • Allcharts, following outpatient procedures will have a valid order available for coding and billing on the electronic medical record within 4 days of service • CTQ- Order on chart within 4 days from date of service on right patient D number
- 7. Volume of LostOrders by Dept
- 8. Charges on theseaccounts
- 10. 2012 Performance Level– Locations Impacted • 2012 Total # opportunities= 107,711 • Number Defects: 3290 • DPMO= 30,545 *** 1.9 Baseline Sigma Level ***
- 11. Current Process – Outof Control
- 12.
- 13. 5 Why Analysis •WHY #1:How service departments manage orders and paperwork – unorganized, multiple paper charts are collected in baskets WHY #2: The process for getting charts to medical records varies from daily pick up by medical records staff to departments delivering when big enough batch (can be 2 weeks from DOS before gets to medical records) • WHY #3 Silos - departments and functions. – Coders are only looking for orders in Meditech 4 days after service, faster than medical records gets charts processed and scanned into Meditech – Service area do not know the process of how orders get scanned into Meditech -- they have to be scanned to patient’s D number so that coders can find order (they are off site) – Charts with non-valid orders are sent back to review without any communication of what is missing.
- 14. 5 Why Analysis(Continued) WHY #4 Training and access to systems is lacking. •Not all services being coded "no orders" require an order (a physician is providing the service) •Not all orders for services are in the Meditech documentation system (many providers are in Next Gen). •Coders do not have access to systems they need - only Meditech WHY #5 Lack of standard process for capturing orders where needed and education HOW: TEMPORARY COUNTERMEASURE 1. Training - what is a valid order, what services have orders in what system, 2. Provide coders access to systems needed FINAL COUNTERMEASURE- A standard process with reduced handoffs and batching, capturing the power of technology
- 15. Scatter Plot andRegression Analysis
- 16. Importance Cost toFeasibility Cost Leverage Total Project to Customer Implement (Likelihood Reduction (Positive Impact Project of Success) On Other Processes) Priority Rate 1 to 5 Rate 1 to 5 Rate 1 to 5 Rate 1 to 5 Rate 1 to 5 High = 5 High = 1 High = 5 High = 5 High = 5 Low = 1 Low = 5 Low = 1 Low = 1 Low = 1 POS Scanner 3 X 5 x 5 x 3 x 4 = 900 Central Repository 5 x 4 x 5 x 3 x 3 = 900 Coder Training 5 x 3 x 3 x 3 x 4 = 540 Eliminate Time Stamping 1 x 5 x 5 x 2 x 3 = 150 Eliminate Triplicate order sheett 1 x 2 x 3 x 5 x 3 = 90 Coder Access to Next Gen 4 x 2 x 5 x 4 x 5 = 800 Coder Access to Manager of clinic 4 x 5 x 5 x 1 x 3 = 300 Barcode on 90% of documents 1 x 1 x 4 x 1 x 3 = 12 Project Prioritizer
- 17. FMEA FMEA Process Name: CentralRepository and POS Scanning Process Number: SBT 445 Date: 1/1/2001 Revision Level: 1.3 FAILURE MODE A) SEVERITY B) OCCURRE NCE Probability C) DETECTIO N Probability RISK PRIORITY NUMBER ACTION TO IMPROVE REVISED VALUESRate 1-10 Rate 1-10 Rate 1-10 RPN 10=Most Severe 10=Highest Probability 10=Lowest Probability AxBxC A B C RPN 1) Long list to search though to find order -increase staff time 3 4 1 12 Auto delete established 3 1 1 3 2) Select right patient but wrong order 9 2 8 144 When scanning have to enter test ordered 9 1 4 36 3) Scan order to wrong D number 10 5 6 300 Have to match date of service during entry 10 2 3 60 4) Order scanned into wrong category into Meditech 3 3 5 45 Made only access is to order category 3 1 5 15 5) Scan an order that is not "valid" 10 6 10 600 Training, written reference document provided at desktop. MR random audit 10 4 8 320
- 18. Trial Results **Continue tohave 0 Missing Orders through May 31, 2013**
- 19. Go Live • 31Departments • Built access for 441 in Repository • Set Up 8 Kodak Scanners and 30 All in One Scanners and access for 200 • 33 G drive folders built • Set up 190 with new Meditech access • Employees Trained in all 31 departments
- 20. Control Phase Sustaining andContinuing Improvement
- 21. CONTROL CHARTS Before andAfter Comparisons
- 22. 71645750433629221581 0.30 0.25 0.20 0.15 0.10 0.05 0.00 Sample Proportion _ P=0.0429 _ P=0.0075 UCL=0.0716 UCL=0.0334 LCL=0.0141 LCL=0 1 58 1 1 1 1 1 1 1 1 11 1 1 1111 P Chartof SCI Missing by Week Number Tests performed with unequal sample sizes
- 23. 1715131197531 0.20 0.15 0.10 0.05 0.00 Sample Proportion _ P=0.0075 UCL=0.0334 LCL=0 1 P Chart ofSCI Missing Tests performed with unequal sample sizes
- 24. 71645750433629221581 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.00 Sample Proportion _ P=0.02954 UCL=0.04007 UCL=0.02373 LCL=0.01901 LCL=0.00401 1 67 _ P=0.01387 1 1 1 11 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1111 11 1 1 1 1 1 1 1 1 P Chartof Missing Orders by Week Number Tests performed with unequal sample sizes
- 25. 6457504336292215811 0.08 0.06 0.04 0.02 0.00 7473727170696867 0.08 0.06 0.04 0.02 0.00 Sample Proportion _ P=0.02954 UCL=0.04007 LCL=0.01901 1 _ P=0.01387 UCL=0.02373 LCL=0.00401 67 111 1 1 1 1 1 1 1 11 1 1 1 1 11 111111111 1 1 1 11 1 1 P Chart ofMissing Orders by Week Number Tests performed with unequal sample sizes
- 26. Go Live withNew Process Date April 9, 2013 Go Live with Pilot (SCI and S.Tower Radiology)
- 27. Average Days toBill All Orders 2013 YTD Jan 12.72 Feb 11.15 Mar 12.68 April 9.19 May 10.63 2012 Days to Bill Jan 20.22 Feb 18.12 Mar 16.67 Apr 17.25 May 23.34 Jun 24.82
- 28. Financial Impact • Codershave to final code to be able to bill for services • Prevented Write off of $635,274 (at 40% reimbursement rate = $254,109.60 revenue ) Days to Final Abstract Status by Coding: 2012 = 23.97 Days 2013 YTD=11.25 Days Increased cash flow in one time collections by $705,177 Total Missing Orders ( areas impacted by project) Ave Days to Bill these Encounters Sum of Charges 2012 3290 95.45 DAYS $1,845,984.00
- 29. Lessons Learned • TestPilot Invaluable • Struggled with being able to ID go live date • 4/15/13 IM 318 orders email- attaching before deleting (no attached order document in CR) • Communication Opportunities • Standardization – How search; enter provider names, how year entered • Remembering where training resources are • How to and Valid Order • Once Live – need to still hear about issues - continue to improve • DOB • Education Opportunities- ongoing
- 30. Project Closure • TeamCelebration – Recognition with Sponsor • Project Owner in place with accountabilities • Project Sponsor- seeing the results • Departments impacted love the new process
Editor's Notes
- #12 The process currently is out of control. To get to a process in control, we determined the significant inputs that would have the greatest impact on achieving control in the process and improvements in capability. We identified a trial solution to see if we could achieve an improved process - the trial solution focuses on eliminating variation in valid order understanding, and handoffs and travel of orders. The trial solution involved POS Scanning of order from electronic holding tank or upon receipt from patient walking in- the order at time of service is scanned to patient ID, real time. The hypothesized result would be improvement – an order in the EMR within customer required time. Because it is a process that is discrete data, capability analysis beyond the control chart is not appropriate. ( in addition to the fact it is not in control)
- #16 In our attempt to quantify the significant inputs and relationships between the input of how much variability there is in our current process of getting orders to Medical records with the orders, we completed a scatter plot and regression analysis comparing the days to get to medical records (variation on how departments get charts with orders to medical records) and to the number of charts missing orders. This demonstrates a high correlation to the longer it takes to get the charts to medical records, the higher chance of missing orders. This demonstrates correlation but not causation. We completed our trial process with standardization of process, and our hypothesis test (slide 32) discussed our results.
- #19 The trial solution achieved 0 defects, no missing orders. All orders by the SCI providers to the radiology areas (CT, PET, MRI, RAD, RADMOB) were present in the system. The above graph demonstrates over the 4 week period, we had a total of 103 orders with 100% achieving valid orders in place for coders to code. (the coders were the individuals who measured/determined if valid order was present or not) During this trial on five separate dates, this individual went to the radiology department and randomly selected 20 patients names that received services on that date, at the end of day shift. Of the randomly selected charts, all had valid orders in place. This was another method to test the capability of the process – valid order for every test in the chart. ( A significant process step, prior to going to Medical Records)
- #23 SCI control chart – pre and post new process in pilot (Week 1 - corresponds to the first week in January 2012. Week 58 corresponds the first week in February 2013 when the new process (pilot location) went live) The p value went from 0.0429 to 0.0075.
- #24 Control chart of pilot area since go live date 2/1/2013
- #25 All orders to missing orders since the first week in January 2012 through June 1, 2013.
- #26 This is all orders to missing orders by week. Week 1 corresponds to the first week in January 2012. Week 67 is the first week we went live with the process to all impacted departments. The trend at week 58 is when we started the pilot in our SCI area (this was an area with a large volume of missing orders).
- #27 We have reduced our missing orders rate to 0.74% of all orders since our April 9, 2013 go live date. This is compared to the 6.3% performance rate in 2012. Our baseline sigma level performance was 1.9, our new process since go live on April 9, 2013 has a sigma level performance at 2.4. (This is not quite 2 months into the new process – we learned CVS area training gap, MOBLAB challenges with variation in patient processes and not all providers are able to use Central Repository (only employed or internal areas) at this time.
- #28 Due to a greater percentage of patient encounters having the orders in the chart the first time coders enter the chart, average days to bill is improving. Capturing 41% of total charges on encounters that initially were indicated as missing orders, secondary to the actions in place to dig into reason missing/find and get coded. In 2012, we only captured 11% reimbursement rate of the total charges on all encounters with missing orders.
- #29 Increase in one time collections ($561 * 1257.10 = 705,177.00).
- #31 External provider offices are asking to be able to be a part of this process. We are looking into trying to make this possible.