1. The document summarizes a presentation on using blockchain in healthcare. It discusses using blockchain for securely sharing patient information across different healthcare providers and systems to improve interoperability. 2. Blockchain can provide traceability across clinical supply chains and ensure accuracy of clinical trials and research data. Its transparency and immutability properties make it suitable for tracking medical devices and medications. 3. Artificial intelligence and analytics of patient data on the blockchain can help provide personalized preventative care through early risk prediction and detection. This could improve healthcare outcomes and quality.

1. GCCCF Roundtable, Paris, May 22nd
Blockchain in Healthcare
May 22nd, 2019
Quæfacta
Lea Dias, CEO & Co-founder
David Andrianavalontsalama, CPO & Co-founder

3. Blockchain
1. Decentralised, not owned by a single entity;
2. Data is cryptographically stored inside a block;
3. Immutable, tamper resistance;
4. Transparent, so data can be tracked.
!3

4. Participant
A
Participant
B
Participant
C
Participant
…
Blockchain
!4
……
Data is immutable & transparent
Participants can join / leave
A distributed consensus protocol is shared by all participants
block
1834
block
1835
block
1836
block
1837

5. Use cases in healthcare
Sharing of information – between tertiary health care facilities,
community healthcare and telemedicine including; pathology and
radiology results, discharge information, medicines information;
Clinical trials traceability – for clinical research and development of
medications;
Genomics research – precision, tailored-made medicine for individual
genetic makeups;
Medication supply chain – tackling issues such as counterfeit
medications, opioid misuse and vaccination distribution;
Interoperability – IoT, medical devices, robotics, wearable devices,
sensors, applications;
Partial views – mandatory reporting for governments, health
departments and health institutions.
!5

6. !6
Improve quality of healthcare:
1. Intelligent workflow management
2. Smart data
Let’s start with a case study…

7. !7
Patient

8. !8
Case study
Patient:
• Mr J. Doe
• 57yo male, morbidly obese, smoker
• PMHx: hypertension, hypercholesterolaemia, NIDDM
• Presents: Stroke like symptoms, dizziness, confusion,
weakness in limbs, speech difficulty, facial drooping
• Medications: perindopril 4mg om, frusemide 40mg om, pravastatin 20mg on, metformin 500mg bd,
gliclazide 80mg od
• Risk factors: hypertension, diabetes, smoking, obesity, lack of exercise

9. !9
1. Examinations 2. Diagnosis 3. Emergency
treatment with
medications
4. Procedure 5. Treatment
team may
include:
6. Medications
post stroke
Ischaemic stroke
J. Doe
Intravenous injection
of tissue
plasminogen
activator (tPA)
Angioplasty and
carotid artery stent
• Doctor trained in
brain conditions
(neurologist)
• Rehabilitation
doctor (physiatrist)
• Nurse
• Dietitian
• Physical therapist
• Occupational
therapist
• Recreational
therapist
• Speech
pathologist
• Social worker
• Case manager
• Psychologist or
psychiatrist
High blood pressure
medication
Blood thinners
Anti - coagulants
Diuretics
Cholesterol medication
Anti-fibrillation drug
Diabetic medication
Antidepressants
Physical examination
Blood tests
CT scan
MRI scan
Carotid ultrasound
Cerebral angiogram
Echocardiogram

10. !10
ADMISSION ICU INPATIENT DISCHARGE
• Patient brought to ED by
ambulance
• Patient is non responsive
• Admitted to hospital recorded
on PAS
• Pathology, radiology
investigations performed
• Call for medical records
• Close neurologic and
hemodynamic monitoring
provided in the ICU to minimize
the risk of secondary injury
• Monitor ventilation
• Commence IV saline and
mannitol 20%
• Access to pathology and
radiology results with ICU
systems
• Managed care on ward
• Rehabilitation begins
• Allied health and pharmacy
follow up
• Discharge summary prepared
on inhouse software system
• Communication with GP via
phone, fax, mail
• Follow up outpatient
appointment booked manual
NO INTEGRATION (NI)
MANUAL PROCESSES (M)
• GP medical history, medications
or allergies (NI)
• Paper record of ambulance
information (M)
• Medical information record (M)
• PAS with inpatient system (NI,
M)
• Allergies recorded on PAS and
paper chart (NI, M)
• Pathology, radiology systems
(NI)
• Smart pumps and ICU system
(NI)
• Medical devices and ICU
systems (NI)
• ICU and theatres booking
system (NI)
• ICU and inpatient paper
recorded (NI, M)
• ICU system and paper record
(NI, M)
• Pathology and radiology
systems (NI)
• Allied health information (NI, M)
• Medication reconciliation (M)
• Medication reconciliation (M)
• Pathology, radiology input (M)
• Allied health information (NI, M)

11. Global
EMR adoption
Current State of Healthcare (1/4)
Patient information
is siloed
Incomplete
information
Vulnerability
& Exposure
BCMA
(Barcode Medication Administration)
EMR Adoption Model
In US, 2016, 97% of
hospitals unit dosing,
96% CPOE adoption,
94% BCMA and
40% paperless hospitals
(200-400 beds)
OpenEHR Standards
for customisable, flexible, open source
platforms facilitating interoperability
!11

12. Patient information
is siloed
Global
EMR adoption
Current State of Healthcare (2/4)
Incomplete
information
Vulnerability
& Exposure
There is fragmentation
and gaps in the transfer
of information between
hospital care and
community care
Patient
Hospital
Providers
Hospital HealthcareCommunity Healthcare
Outpatient
clinicsGP Clinic /
Community Health
Home Health
Pharmacy
Wearable
devices
Laboratory
Rehabilitation
Screening
& diagnosis
Ambulatory
care

13. Global
EMR adoption
Current State of Healthcare (3/4)
Patient information
is siloed
Incomplete
information
Vulnerability
& Exposure
Medicines information, inpatient records, admission and discharge information are often missing or
poorly communicated by health professionals within hospitals and to community health providers.
This may lead to:
‣ hospital readmissions;
‣ adverse drugs events;
‣ compromised patient care;
‣ serious or fatal outcomes;
‣ litigation.
!13

14. Global
EMR adoption
Current State of Healthcare (4/4)
Patient information
is siloed
Incomplete
information
Vulnerability
& Exposure
Patients and health providers are left feeling vulnerable and exposed.
!14

15. 1. Observable gaps in the transfer of
information
2. Lack of interoperability — Many
devices and practitioners interact and
do not share the full data
3. Procedures that should be implemented
are not, or not followed, or incomplete
!15

16. !16
1. Intelligent workflow management to improve quality of healthcare

17. !17
Patient
Role of blockchain
Securely sharing
information
Interoperability
Traceability
Accountability
Fraud detection
Incentives
Data privacy
Analytics & AI
Digital Identity
Matching

18. !18
Digital Identity
Matching
Patient
“Matching the correct individual to his or her
health data is critical to their medical care.”
“Statistics show that up to one in five patient
records are not accurately matched even within
the same health care system. As many as half of
the patient records are mismatched when data is
transferred between healthcare systems.”
— Shaun Grannis, Director of Center for Biomedical Informatics (CBMI)

19. Multi-vendor + smart contracts
Vendor A
Vendor B
Vendor C
Auditing
system
hash
data
data
data
hash
hash
The data and
results are accurate
certification!
command
+ hash
data
Anchoring system using blockchain + smart contracts
!19
Interoperability

20. Patient
Hospital
Providers
Hospital HealthcareCommunity Healthcare
Outpatient
clinicsGP Clinic /
Community Health
Home Health
Pharmacy
Wearable
devices
Laboratory
Rehabilitation
Screening
& diagnosis
Ambulatory
care
!20
Securely sharing
information

21. !21
Pharma Pharma
Pharma /
Med device
Product
Development
Innovation
Active
Pharmaceutical
Ingredient
Manufacturing
Secondary
Manufacturing
ERP ERP ERP
Logistics Logistics
Distribution Supplier
ERP ERP
Logistics Pharmacist Customer
Wholesaler
Reseller
(Pharmacist
dispenses Rx
2D scan - WF1)
Patient
ERP
ERP
PIS
Smart device
Direct
to patient
Retail
Pharmacy
(Pharmacist
dispenses Rx
2D scan - WF1)
Hospital
Pharmacy
(Pharmacist
dispenses Rx
2D scan - WF1)
Supply chain logistics workflowTraceability

22. !22
TR TR TR
TR
TR
TR
TR
TR
TR
TR
TR
TR
Acquiring traces

23. !23
Traceability content: Who? What? Where? When? Why?
Traceability actor: Any known user + key
Acquisition tools
Anchoring: Any known blockchain
Metrics: How many traces per device? How often? How long?
What a trace holds

24. !24
How we acquire a trace
Tool suite:
• API
• Mobile & desktop apps
• Dashboards

25. !25
TR TR TR
TR-1 TR-2 TR-3

26. !26
TR TR TR
TR-1 TR-2 TR-3
TR
TR-4smart contract

27. !27
TR
TR-XXX = 2000 lines

28. !28
TR
TR-XXX = 2000 lines
smart contract
TR TR-XXX-1
TR TR-XXX-2
TR TR-XXX-2000
TR TR-XXX-3
…

29. !29
Chain reconstitution
Whole chain
Partial view #1
etc.data
analysis

30. !30
Chain anchoring
Whole chain
Partial view #1
etc.
+ supervisor
validation
+ supervisor
validation
smart
contract
smart
contract
TR
TR-x
TR
TR-y

31. !31
Trace composition
Supply chain model
Diversity of actors
Incremental level of trust
What a chain holds

32. Data
Lake
!32
TR
TR
TR
BLOCKCHAINS Analysis
TR-SPEC TR-SPEC
Trace
reports Traceability
chains
Traces
validated
data
Data
retrieval
Analytics
dashboard
BUSINESS RULES
event
anchoring

33. + 1 level
of trust
Data
Lake
!33
TR
BLOCKCHAINS Analysis
+ supervisor
validation
BUSINESS RULES
TR
TR-SPEC TR-SPEC

34. !34
Analytics & AI
Weighting the trust
Feedback loops
Pattern matching from theoretical chains
Statistical inference from actual chains

35. !35
2. Smart data to improve quality of healthcare

36. !36
The 4 P’s of Personalised healthcare
Identification of individual risks of
developing certain diseases based
on the person’s genetic profile and
other personal information
Predictive
Methods and treatments to avoid,
reduce and monitor the risk of
developing certain diseases
Preventive
Clinical interventions based on the
unique genetic, medical and
environmental characteristics of
each patient-citizen, and genomic
profile of his/her diseases
Personalised
Citizens are fully engaged in
personal health management
Participatory

37. !37
Case study
Patient:
• Mr J. Doe
• 57yo male, morbidly obese, smoker
• PMHx: hypertension, hypercholesterolaemia, NIDDM
• Presents: Stroke like symptoms, dizziness, confusion,
weakness in limbs, speech difficulty, facial drooping
• Medications: perindopril 4mg om, frusemide 40mg om, pravastatin 20mg on, metformin 500mg bd,
gliclazide 80mg od
• Risk factors: hypertension, diabetes, smoking, obesity, lack of exercise

38. !38
Analytics & AI
Google AI team:
• Analyse retinal images, extract personal health risks, and make
predictions based on the knowledge received.
• Identifying risk factors critical for CV and stroke,
• body mass index (BMI)
• hemoglobin A1c (HbA1c)
• systolic and diastolic blood pressure
• smoking status.
Smart data to diagnose ischaemic stroke?
Researchers reported their algorithms succeeded in predicting the chances of particular patients
developing stroke or heart attack in a five-year period with a 70 percent accuracy.

39. !39
Analytics & AI FDA Approved, Viz.AI Contact 2018
AI Algorithm
Clinical decision support for triage
Analyse CT scans and detect stroke signs in medical images
Detects slightest deviations on CT and MRI scans
ML algorithms can distinguish ischaemic from haemorrhagic stroke
System suspects stroke, alerts neurovascular specialist via smartphone
Specialist’s attention refocused to the acute cases
Radiologist proceeds with review of less urgent scans
AI-enabled process optimization ensures timely care for patients

40. !40
Viz.AI Contact 2018

41. !41
Analytics & AI
• Support health specialists and provide actionable insights to
accelerate diagnosis.
• Ensure accurate medication and intervention decisions in the
shortest possible time.
• Reduce the risk of developing conditions, elicit subtle warning
patterns and alert clinicians to upcoming crisis.
Artificial Intelligence

42. !42
Incentives
• Insurance companies may incentivise patient’s (data) for good
behaviour via a reward mechanism.
e.g. tokens for following a care plan or staying healthy.
• Pharma companies/medical institutions may incentivise patients
who provide data for research and clinical trials.

43. !43
4,567 steps
2 coins
cash in

44. !44
Fraud detection
Pharma companies
• Detection of counterfeit medications.
Governments/healthcare
• Detection of opioid/medication misuse, abuse and theft;
• Detection of inappropriate use of medications (including high cost
medication).
Insurance companies
• false claims/information by patients and providers to receive
payable benefits.

45. !45
“Blockchain is not meant for storage of large data sets.
Blockchain is not an analytics platform.
Blockchain has very slow transactional performance.
However, as a tamperproof public ledger, blockchain
is ideal for proof of work.
Blockchain is highly resilient”.
— John Halamka, CIO of Beth Israel Deaconess Medical Center in Boston

46. Quæfacta
!46
We do:
• Blockchain traceability solutions in
healthcare
• AI, data acquisition and analytics
https://quaefacta.com
contact@quaefacta.com
Thank you!
May 22nd, 2019