Technology Evolution in Pathology: The University Health Network Experience Across Ontario

Technology Evolution in Pathology:Technology Evolution in Pathology:The University Health Network Experience The University Health Network Experience

Across OntarioAcross Ontario

Sylvia L. Asa, MD, PhDPathologist-in-Chief

Medical Director, Laboratory Medicine Program

2

ObjectivesObjectives

• The nature of pathology practice in Ontario• The reason for a centralized laboratory program• The IT requirements for success of centralized

pathology• The reason for using digital imaging

Participants should have an understanding of:

AssumptionsAssumptions

• A single payer, publically funded health care system

• A large geographic area with population concentration in 5 large centers

• A shortage of Pathologists

3

Initial StatusInitial Status• Multiple hospitals of variable size scattered

throughout the province– Toronto (GTA) has 7 major teaching hospitals and 35

other hospitals

– 5 medical schools in various cities with 1-5 affiliated hospitals

– Other large cities with large, full-service hospitals

– Many small towns with hospitals of varying size

• Each hospital is operated as an independent entity with funding from the Ontario Ministry of Health and Long-term Care

4

Historical IssuesHistorical Issues

• 1990s Ontario determined that – Health care costs were too high– Pathology was a dying field – There would be no need for Pathologists in the

next century– Training programs in Pathology were slashed

Outcome: major shortages of Pathologists emerged in late 1990s-2000

5

Healthcare Reform 1990sHealthcare Reform 1990s

• Regional planning for healthcare (LHINs)

• Consolidation of hospitals6

The University Health NetworkThe University Health Network

• A consolidation of three U f T affiliated teaching hospitals

• Programmatic restructuring– TGH cardiac care; transplantation; advanced

medicine and surgery– PMH cancer care– TWH neurosciences, musculoskeletal care,

community health

• Laboratory consolidation7

The University Health NetworkThe University Health Network

8

~ 1 mile

The Challenge: Lab ConsolidationThe Challenge: Lab Consolidation

• 3 physical sites

• 3 cultures

• 3 missions of the academic institution:– Complex patient care– Education– Research

9

Proposed SolutionProposed Solution

• A single core department

• Electronic support for specimen tracking and handling at 3 sites

• Highly subspecialized expertise– Biochemistry - Microbiology– Hematology, Transfusion & Hematopathology– Subspecialty Anatomical Pathology– HLA - Molecular/Genetics

10

Solution: Step 1Solution: Step 1

LIS implementation goals:

• Best-of-breed approach to support high volume complex testing

• Integration in e-chart with e-orders

• Specimen tracking and management

• Integration of lab data from all disciplines into a consolidated report

11

Solution: Step 1Solution: Step 1

LIS implementations:

• Core Lab automation and middleware

• CoPath solution for Pathology

• Transfusion Medicine LIS

• HLA Histotrack

• Upgrade existing Shire for molecular lab and interface with CoPath

12

Solution: Step 2Solution: Step 2• Analyze workflow

and clinical needs

• Build core labs and satellites– State-of-the-art

space and equipment– Tubes where possible– Rapid response labs where required– On-site accessioning and grossing for surgical

pathology with enhanced PA support

13

14

Informatics: Voice RecognitionInformatics: Voice Recognition

• Dragon-speech integrated with LIS

means instant reporting without

the need for dictatyping

Solution: Step 3Solution: Step 3

• Recruit appropriate medical and technical expertise

• Create teams of experts who integrate with clinical staff in priority programs: The Pathologist as Medical Consultant

“As is your pathology, so goes your clinical care.”

Sir William Osler15

16

Subspecialty PathologySubspecialty Pathology

• All cases reported by a pathologist with expertise in the specific subspecialty required

• Benefits:– Better quality and faster patient care– Fiscal responsibility: 1 pathologist per case– Pathologist satisfaction – enhanced academic

excellence

• Challenges:– Requirement for appropriate staffing in all areas

and redundancy

Solution: Step 4Solution: Step 4

• Implement telepathology for intraoperative consultations and frozen sections at non-core sites– Phase 1: Robotic microscopy– Phase 2: Digital WSI

17

Historical Data: TelepathologyHistorical Data: Telepathology

• 1973: Washington DC diagnosis of leukemia via satellite from Brazil

• 1986: Dr. Ronald Weinstein coins name

• 1990s: Norway implements robotic microscopy to support frozen sections in remote hospitals

• 2003 ? Why Not UHN

18

19

Barriers to TelepathologyBarriers to Telepathology

• Cost – cheaper than another pathologist!• FDA approval – not applicable in Canada• Billing/CPT codes – not applicable• Turnaround time - overcome• Pathologist issues

– learning curve/accuracy– “images are good, but not ready for prime time”

Th Philosophical ResponseTh Philosophical Response

20

In a time of drastic change it is the learners who inherit the future. The learned usually find

themselves equipped to live in a world that no

longer exists.Eric Hoffer

21

Due Diligence Before Going LiveDue Diligence Before Going Live• Medical Malpractice Insurance Provider

– Canadian Medical Protective Association (CMPA)– telepathology will not affect coverage

• UHN Medical Advisory Committee– SOP presented for approval

• Health Canada – Therapeutic Products Program– telepathology does not involve “medical devices” (no direct contact between

instrument and patient) – no federal approval required

• Surgeon Education– demonstrating the robotic microscope/slide scanner

• essential to get surgeon buy-in!

22

The Robotic System:The Robotic System: November 2004-October 2006 November 2004-October 2006

23

Toronto GeneralTelepathology Work Station

Toronto WesternSurgical Pathology

The Robotic System:The Robotic System: November 2004-October 2006 November 2004-October 2006

24

Whole-Slide Imaging:Whole-Slide Imaging: October 2006-Present October 2006-Present

25

Whole-Slide Imaging:Whole-Slide Imaging: System Parameters System Parameters

26

UHN Telepathology ProtocolUHN Telepathology Protocol• System test each morning • Pathologist reviews daily O.R. list and communicates

game plan for the day to histotechnologist • Surgeon defines tissue of interest• Histotechnologist contacts Pathologist

- specimen description, processing specimen • Histotechnologist at TWH scans the slide and calls the

Pathologist • Pathologist speaks with the surgeon by telephone• QA the next day

27

1003 Frozen Sections from 802 1003 Frozen Sections from 802 Patients (Nov 2004-Sept 2008)Patients (Nov 2004-Sept 2008)

0

200

400

600

800

1000

1200

Robotic Total

# Frozen Sections

653

1003

350

28

Performance: 1003 Cases/4 Years Performance: 1003 Cases/4 Years • Accuracy

– 98% concordance with final pathology– Not a function of technology

• Deferral rates– Identical to on-site rates– NOT a function of technology

• Sometimes you just don’t know for sure• Sampling issues in the frozen section biopsy

• Turnaround times– Well within 20 minutes required

29

TAT Single Block Frozen SectionsTAT Single Block Frozen Sections

02468

101214161820

Robotic WSIFrozenSection

WSIFrozen +

Smear

Total TAT

**

* p < 0.0001

Receipt of tissue to report of diagnosis

30

Pathologist Interpretation TimePathologist Interpretation Time

0123456789

10

Robotic WSIFrozen

WSIFrozen +

Smear

Time/slide (min)

4-fold

Pathologists tended to go to TWH site for multi-block cases when using the robotic microscope – not so for whole-slide imaging.

*

*

* p < 0.00001

Receipt of image toReport of diagnosis

31

WSI Pathologist Interpretation TimeWSI Pathologist Interpretation Time

0

5

10

15

20

25

30

35

40

< 1 1-2 > 2

% of Cases

Minutes/slide

32%

38%

30%

* 70% of cases reported in < 2 minutes after scan is received

Failure Mode AnalysisFailure Mode Analysis

32

• PRE-CASE:– Network failure

– Moving the scanner within the surgical pathology lab• static vs dynamic IP addresses

• discovered on morning test run.

• MID-CASE: – Minute/pale pieces of tissue that the scanner would not

“recognize”

– Excess mounting media causing the cover slip to stick to the scanner objective

33

Subspecialty Support for FS Subspecialty Support for FS

Subspecialty Model Subspecialty Model

• How do we get the two liver pathologists to read transplant biopsies and attend all academic meetings?

Telepathology solution– USCAP 2008 all rush biopsies read on laptops

at the meeting

34

Subspecialty Model Subspecialty Model

• How do we get the subspecialty support for weekend coverage?

Telepathology solution– Summer 2008 all weekend cases read on

laptops at the home/cottage etc.

35

Subspecialty Model Subspecialty Model

• How do we get the pituitary expert to read a tough biopsies when she is in Istanbul?

Telepathology Blackberry solution

36

Ontario-Wide ImplementationOntario-Wide Implementation

• Timmins and District Hospital forms an alliance with 9 other hospitals in North East Ontario

• Seeks Laboratory Medical Directorship

• UHN provides a suitable proposal– Team of subspecialists to support all clinical

needs from core in Toronto

• Initiation of a new model

37

Google Maps 2008

422 miles

LHIN # 13

***

******

38

LHIN # 7

Ontario NE Cluster ImplementationOntario NE Cluster Implementation

Configurations in NE OntarioConfigurations in NE Ontario

• Small hospitals going to POCT only

• Medium hospitals on-site labs with POCT

• Largest hospital with full lab and surgical pathology accessioning, grossing by PA with webcam support– All smaller hospitals send AP specimens to

core in Timmins – Complex testing referred to UHN

39

Subspecialty ModelSubspecialty Model

• Requires sign-out of all cases by subspecialist– Slides shipped to Toronto by overnight courier

• FS review by subspecialist must be available

• Ultimately no pathologist on siteTelepathology solution

40

The Ultimate SolutionThe Ultimate Solution

• $3M grant from government to implement high resolution digital imaging at all sites– All abnormal blood smears, malaria,

microbiology gram stains, CSFs, etc

• Plan to expand FS service to hospitals that have not had this available

• CoPath integration of digital imaging in future will alleviate need for any slide transportation

41

Pros and Cons of LIS IntegrationPros and Cons of LIS Integration

Pros

• Fast

• E-filed into right location

• Integration of gross, micro, EM, molecular

• Remote access and who has (need) access

Cons

• Images “trapped” and need for export for other purposes

42

Google Maps 2008

422 miles

LHIN # 13

***

******

43

LHIN # 7

Addition of New ClientsAddition of New Clients

44

The Future of Pathology?The Future of Pathology?

+

45

The Future of PathologyThe Future of Pathology

+

46

The Future of PathologyThe Future of PathologyThe best way to predict the future is to invent it

Alan Kay

What About Academia?What About Academia?

• Digital education

• Digital documentation of the biobank– The “Biobank” is the current phraseology for

the “Department of Pathology”

• Scanning and automated analysis of TMAs

• Scientific Advances– Laboratories must evaluate,

develop, and apply the genotypic and phenotypic analyses of specimens

47

48

AcknowledgementsAcknowledgements• Pathologists

– Andrew Evans– Runjan Chetty– Blaise Clarke– Sidney Croul– Bayardo Perez-Ordonez– Rasmus Kiehl

• Surgeons– Mark Bernstein– Abhijit Guha– Fred Gentili– Chris Wallace– Michael Fehlings– Mojgan Hodaie– Jaime Escallon

• Histotechnologists– Suganthi Ilaalagan– Sofia Aguierre– Alfreda Antonio– Carsen Chan– Gordon Chin– Norman Hew-Shue– Pam McCartin– Aparna Pant– Ann Marie Scott– Henry Wu

• IT Support– Greg Lewis– Karen Jaquardt

Vendor SupportLeica MicrosystemsQuorum Technology/Aperio