RE-ENGINEERED WORKFLOW IN THE AP LABORATORY Costs and

RE-ENGINEERED WORKFLOW IN THE AP LABORATORY: Costs and Benefits Erin Grimm, MD Rodney Schmidt, MD, Ph. D University of Washington Medical Center Seattle, WA

Disclosures The UW-developed software (Power. Trax and e. Path. Image) is licensed through the University of Washington. The speakers have no relationship with IMPAC Medical Systems, owners of Power. Path, or any of the other mentioned companies.

Objectives 1. Review current workflow in Anatomic Pathology and the need for change 2. The UW Anatomic Pathology Automation Project • A detailed look at each step 3. Starting the automation process • Building a business case 4. Questions for the future

The scope of the problem Histology laboratory workflow has not changed in decades 1959 Yet • Volumes increase • Laboratories expand http: //history. library. ucsf. edu/imagelib/med_sci_b uilding_histology_lab_1959. gif

Problem #1 Inefficiencies exist that cause waste Waste increases expense • Labor costs • Poor resource utilization

Problem #2 Errors Happen • Patient ID errors occur in AP • • 4. 3 / 1000 surgical specimens 1 1. 9 / 1000 amended reports • 19. 2% of amended reports were due to patient ID errors 2 1. 2. Makary MA et al. Surgical specimen identification errors…. Surgery 2007 Apr; 141(4): 450 -5 Nakhleh RE, et al. Amended reports … Q-probes study of 1, 667, 547 accessioned cases. . . Arch Pathol Lab Med. 1998 Apr; 122(4): 303 -9.

Achievable Error Rates Error rates 1/100 1/1, 000 1/100, 000 Error Prevention Real world Examples Methods Clear process Errors filling out lab requisition Failure give results rom to 000 to patients o f Suboptimal specimen o g 1/10, T Clear process ion 0 error tomat 0 0 u Mislabeled specimens Systems for 1, and a / 1 identification s mitigationre ui req design + Specimen loss Advanced Reliance on education/vigilance Automation Error ID/ mitigation Computer interface errors Resar RK. Making noncatastrophic health care processes more reliable. . . Health Serv Res. 2006; 41: 1677 -1689.

1 6 4 % A 0 4 , 0 2 a i 0 4 r n 0 e a 0 p e p r r o e c o p r a n r s r e i u o e a u t r n s e g e p i o r c c u o Dilemma What if 1% of tests were errors and 6% of the errors led to inappropriate care?

Objectives 1. Review current workflow in Anatomic Pathology and the need for change 2. The UW Anatomic Pathology Automation Project • A detailed look at each step 3. Starting the automation process • Building a business case 4. Questions for the future

UWMC Pathology A complex academic environment with: • >36, 000 surgical pathology cases/year • 178 Faculty Members • 40 faculty with clinical duties • 29 Residents and Clinical Fellows • 35 Graduate Students • $32 million in NIH grants (2006)

UW Goal for Automation 1) Decrease mislabeling opportunities Resident/PA dictates gross description Resident/PA requests additional blocks Stickers with labels applied (Gross Room) (Histology) Case accessioned Signout Cassettes preprinted and placed with specimen Slides prelabeled by hand Pathologist calls up case to enters diagnosis (Gross Room) (Histology) (Offices) = Opportunity for transcription error

UW Goal for Automation 2) Streamline Workflow Save Labor (FTEs) • Automate manual processes • Ex. Histology order completion, specimen discard, image uploads • • Make location/progress of all assets (specimens, blocks, slides, and paperwork) visible and trackable in the AP-LIS Eliminate preprinting/prelabeling Initial phase Start with projects having • • ↑ Yield ↓ Developer hrs

Staged UW Automation Clinical Database (75 hrs) 2005 Slide tracking (1500 hrs) 2006 Document scanning with imaging suite (150 hrs) Gross room • Photography (80 hrs) • Specimen container disposal (50 hrs) 2007 New Clinical Database (400 hrs) 2008 Whole line automation Cassette barcoding (500 hrs? ) Approximate developer hours noted for each project

Technical info Custom software was written as a Windows application using Microsoft Visual Studio C#. Net and SQL Server UW Clients Power. Path Client PC Thin Client Power. Path Database UW Database SQL Server

Staged UW Automation Clinical Database (75 hrs) 2005 Slide tracking (1500 hrs) 2006 Document scanning with imaging suite (150 hrs) Gross room • Photography (80 hrs) • Specimen container disposal (50 hrs) 2007 New Clinical Database (400 hrs) 2008 Whole line automation Cassette barcoding (500 hrs? ) Approximate developer hours noted for each project

Document Scanning Goal • Develop an electronic document management system • All case-related paperwork is viewable from the case specific repository in our AP-LIS Workflow • Paperwork is barcoded when accessioned • Scanner reads paperwork barcode • Document is scanned, accepted by office staff, and automatically uploaded to the image tab of the AP-LIS

Scanning benefits Benefits: • 3. 8 hours/day saved for 26 pathologists and residents • Staff satisfaction: • 10. 0/10 • Saved 0. 25 min/case • Current usage: • 10, 614/month 1. Schmidt RA, et al. Integ. of scanned doc mangmt. . . Am J Clin Pathol. 2006 Nov; 126(5): 67883 2. Routbort M, Grimm E, Schmidt R. Optimized Document Management…. APIII 2006

Staged UW Automation Clinical Database (75 hrs) 2005 Slide barcoding (1500 hrs) 2006 Document scanning with imaging suite (150 hrs) Gross room • Photography (80 hrs) • Specimen container disposal (50 hrs) 2007 New Clinical Database (400 hrs) 2008 Whole line automation Cassette barcoding (500 hrs? ) Approximate developer hours noted for each project

Slide Tracking Goal: 1. Provide real-time status and location of slides Benefits include: • Providing real-time case progression information • Easier location of slides for conference/sendouts • Facilitates workflow analysis via time-stamps 2. Drives AP-LIS functionality • Automates histology order completion and other processes Name

Slide Tracking Workflow Histology Pathology Offices Sendouts File Room Histology work order completes with scanning Pull for conference Ship Faculty signout Resident review Deliver

Slide Tracking Benefits FTE Savings Histology +0. 5 FTE Reduced time hunting for mis-delivered slides +0. 5 FTE Office staff Auto completion of outstanding orders when slide is scanned +. 5 -1 FTE Reduced time for conference preparation +. 25 FTE Increased efficiency regarding send outs

Staged UW Automation Clinical Database (75 hrs) 2005 Slide tracking (1500 hrs) 2006 Document scanning with imaging suite (150 hrs) Gross room • Photography (80 hrs) • Specimen container disposal (50 hrs) 2007 New Clinical Database (400 hrs) 2008 Whole line automation Cassette barcoding (500 hrs? ) Approximate developer hours noted for each project

Gross Photography Gross photography • Photo is automatically imported into casespecific AP-LIS image tab Results • • • Improved Quality Quantity Increased Labor Savings Focus 50. 1% 77. 8% 310 photo/mo 503/mo • Resident/PA • Office Staff • IT help requests > 1 min/case by 1 FTE (bulk image upload) 1. 7/mo 0. 5/mo • Cost Savings • Eliminated cost of darkroom materials • Eliminated kodachrome storage

Specimen Discard Workflow • Device scans specimen barcode • Handheld device queries AP-LIS • If case signout occurred <2 wks prior • If case signout occurred >2 wks prior

Staged UW Automation Clinical Database (75 hrs) 2005 Slide tracking (1500 hrs) 2006 Document scanning with imaging suite (150 hrs) Gross room • Photography (80 hrs) • Specimen container disposal (50 hrs) 2007 New Clinical Database (400 hrs) 2008 Whole line automation Cassette barcoding (500 hrs? ) Approximate developer hours noted for each project

Cassette barcoding Goals: • Streamline workflow • Cassette barcode drives gross room and histology workflow • Eliminate cassette preprinting • Eliminates work for accessioners • Eliminates an error-prone step • Enable resident/PA to obtain cassettes without interruptions Photo Courtesy of General Data

Objectives 1. Review current workflow in Anatomic Pathology and the need for change 2. The UW Anatomic Pathology Automation Project • A detailed look at each step • • 3. Starting the automation process The business case The issues 4. Questions for the future

The Business Case 1. Efficiency • • More volume with same personnel $2. 50 - $3. 00/case (slides, specimens) 2. Patient safety • • Optimize patient care Prevent rare, catastrophic errors 3. Compliance • Custodial responsibility for patient materials (paperwork, slides, blocks, etc).

Buy vs. Build Decision • “Buy” is now possible • Some LIS vendors (IMPAC, Co. Path, et al) • Others (RA Lamb, Dako, Ventana, UW) • Others in development • Most are expensive (S/W and H/W) • No current product is comprehensive

Hardware • • Label printers – inexpensive Bar-code readers – inexpensive Cassette printers – expensive (most) Slide printers – expensive For distributed JIT workflow, we need “personal” cassette printers and slide printers that are as inexpensive, reliable, and ubiquitous as label printers.

Key Considerations 1) This is disruptive technology! • • • Use automation to change habits (prelabeling/preprinting) Don’t automate bad workflow Each user must benefit 2) Select carefully • • • Hardware compatibility Software compatibility Appropriate technology/solution

Questions Where are the boundaries for the APLIS? Who provides bar-coding solutions? Major automation providers are not AP-LIS vendors (Dako, Ventana, RA Lamb, UW) Implicit challenge to LIS vendor “lock-in” • Reporting/billing in one app • Lab/material handling in different app

Questions Where are the boundaries for the APLIS? What will be tracked? Traditional: Specimens, blocks, slides New derivatives: Cells, DNA, tissue banks, ancillary labs, biorepositories Pre-lab tracking: From OR, offices • Reduce ID (pre-analytic) errors

Questions How much of the financial benefits will labs be able to retain? • Hardware? • Implementation? • Software? • Purchase/support pricing model • Per-item metering

Conclusions Bar-coding automation • More than just tracking – disruptive technology! Workflow changes. • Allows processing of increased workloads with static FTE levels • Improves patient safety • Quantifiable gains can be made by upgrading the most inefficient/error prone processes in your laboratory

Thank you UW development team UW Program Operations Manager Dan Luff Erin Grimm, MD grimme@u. washington. edu Rodney Schmidt, MD, Ph. D schmidtr@u. washington. edu

Questions for the Future 2. What materials will be tracked? When does tracking start? • • • Traditional materials: specimen/blocks/slides More specimen derivatives arise: ancillary lab tests, tissue banking, biorepositories ? ? Will there be introduction of prelab tracking to reduce preanalytical errors No current product is comprehensive