Cloud Computing in Health

A Global View to
Infrastructure, Platform and
Software as a Service
on Health

Joaquín Blaya
Juan Bautista Bru

Cloud Computing
Presentation: Objective and Targets

 Objective
 To provide a global view to Cloud Computing
in the Health domain

 Targets:
 Health staff involved in management at any
level
 ICT staff working in the health domain

Cloud Computing
Index
 The basics of the model with some definitions

 Benefits and problems focused on health domain

 Related organizations, standards and certifications

 Real use cases

 Conclusions

The Cloud Computing Model
What it is?

 Cloud Computing is a model
 Cloud Computing Model links:
 Customers (on the Health domain in this case)
 Providers of ICT resources

 ICT Resources

What it is not?

 Cloud Computing is not any ICT
resource as hardware, software,
technology or ICT service itself
(despite it is supported by these ICT
resources)

 Example: Cloud computing is not Virtualization but
Virtualization supports Cloud Computing

Definition by NIST

 “Cloud Computing is a model for enabling
convenient, on-demand network access to
a shared pool of configurable computing
resources”
*NIST is the US National Institute of Standards and Technology

Definición
 Es un modelo de prestación de servicios de
tecnologías de la información, en el que:

 Uno o mas clientes acceden bajo demanda a un
conjunto de recursos ICT proporcionados por y
pertenecientes a un proveedor de servicios

 El acceso a los recursos se consigue mediante una
red de telecomunicaciones (no necesariamente
Internet)

Question
 It has to be remarked that Cloud Computing model
references a provider and a customer.

 In the case that a Cloud belongs to an organization A
which provides services to the same organization A, it
makes no sense to talk about Cloud Computing, despite
of resources and management can be similar

 If so, we just give a new name to something being done
since 50 years ago: datacenters exists from the sixties

Components

 Service Models
Define what level of architecture is delivered
as service for the customer

 Deployment Models
Define who is served by the cloud

Service Models
 Infrastructure as a Service (IaaS)

 Platform as a Service (PaaS)

 Software as a Service (SaaS)

*Combinations can exist

Infrastructure as a Service (IaaS)

 The capability provided to the consumer is
to provision processing, storage,
networks, and other fundamental
computing resources where the consumer
is able to deploy and run arbitrary
software.

Platform as a Service (PaaS)

 The capability provided to the consumer is
to deploy onto the cloud infrastructure
consumer-created or acquired applications

So the cloud provides
 Operating Systems
 Databases
 Application Servers
 Other…

Software as a Service (SaaS)
 The capability provided to the consumer is to
use the provider’s applications running on a
cloud infrastructure

So the cloud provides
 EHR
 CPOE
 Medical Imaging
 Interoperability nodes
 Administration systems: CRM, HR,…
 Others…

Deployment Models
 Public
Services are available for general public or for
concrete market in case of specialized services

 Private
Services are available for only one organization

 Community
The cloud is shared by two or more organizations

*Combinations can exist

Deployment Models for Health

 Medium to big size Health organization
would use private clouds or reserve a
restricted area within a community cloud,
that is, a community cloud and combine
IaaS, PaaS and SaaS

 Small Health organizations can use on
public clouds combined with SaaS

Roles

 Provider
It is the owner of the cloud and provides the services to the users

 Customer
It is the one paying for and receiving the services provided

These are the typical two roles which appear in definitions

 Other Roles
Manager
It is the one operating the cloud. It will be usually the Provider, but it could not.
User
Is that receiving the end services. It will be usually the Customer, but it could not.

Roles: example
Let’s think on a Regional Health Agency provides to
patients an Appointment service for Primary Care via a
web site:

 Provider
 Company A providing IaaS

 Customer
 The Regional Health Agency
 Platforms (operating systems, databases an server applications) for
Appointment Software, integrated with an EHR Software, and owned by
the customer are deployed on the insfrastructure

Roles: example
 User
 Clinicians belonging to the health Agency (customer)

 The Patients (differs from customer)

 Managers
 A company B providing managed services, managing the
IaaS, but directly outsourced by the customer

 A company C managing and supporting the software
(usually the company which developed the software),
directly outsourced by the customer

Cloud Computing
Main Characteristics
 On-Demand Service Provisioning
Resources are provided (committed and released) at a customer’s request

 Quality of Service (QoS) ICT indicators
(Performance, availability, capacity, others)

 Autonomy of the Cloud
Functioning of the systems on the Cloud has to be managed by the provider in a
completely transparent mode for customers

(makes sense when cloud computing is accepted as a Customer/Provider)

 Scalability and Flexibility
Resources have to be easily tailored to customer’s capacity and functionality
requirements

Cloud Computing
Regarding QoS and Autonomy of the Cloud

 QoS and Authonomy of the Cloud should be taken
carefully when talking on Health domain

 Let’s think in automotive industry
Outsourcing is a reality since 60 years ago
When the thing works, Providers are a complete abstraction
for the Customer
But when things go wrong, agreements allow Customer take
control of Provider production systems

 So medium to big size Health organizations thinking in
moving to cloud core systems have to be plan this

Cloud Computing
Regarding day-to-day operation

 Day-to-day operation of a big organization using clouds needs very
good process definition and operation between all Customer,
Provider and Managers

 Make to work these processes in a proper way is very complicated:
in automotive, everybody knows about process management at
required level, and at any management level, most of staff has
process management certifications as Six Sigma

 This can become a handicap to move big projects to Cloud
Computing because Health organization are not so used to work
under a process paradigm as industries as Automotive

Cloud Computing
Main Benefits
 Allows organizations focusing on core business
(A Health organization with an ICT department should
not become an ICT organization with a Health
department, but some times it looks like)

 Lower cost of ownership

 Capital costs transferred to operational costs

 Easy access to cutting edge technology

Cloud Computing
A main issue in Health

Affecting to EHR raw databases :

 Loss of Control

 Security

 Legal

Cloud Computing
Organizations
 Eurocloud (http://www.eurocloud.org/)

 Cloud Standards Customer Council (http://www.cloud-council.org/)

 Distributed Management Task Force (http://www.dmtf.org/)

 Storage Networking Industry Association (http://snia.org/)

 Open Cloud Consortium (http://opencloudconsortium.org)

 Cloud Computing Interoperability Forum (http://occi-wg.org/)

 Cloud Security Alliance (https://cloudsecurityalliance.org/)

 Open Grid Forum (http://www.gridforum.org/)

* Some of them have strong dependence on ICT companies

Cloud Computing
Standards
 ISO/IEC JTC 1 SC 38 WG3
 ISO/IEC JTC 1 is the standards development environment within ISO on ICT standards for business and consumer applications
 The Working Group 3 on Cloud Computing focuses on:
 Developing a standard which provides common terminology for Cloud Computing
 Developing Cloud Computing Reference Architecture

 An initial document was expected by the end of 2011

 Open Virtualization Format (OVF) and Virtualization Management (VMAN)
 The OVF specification is intended for packaging and describing how to deploy virtual machines and applications on heterogeneous virtualization
platforms
 VMAN's Profiles standardize many aspects of the operational management of a heterogeneous virtualized environment
(supported by DMTF)

 OpenNebula
 Is an open-source industry standard for data center virtualization offering the most feature-rich, flexible solution for the comprehensive,
complete management of virtualized data centers to enable on-premise IaaS clouds in existing infrastructures
(Supported by a community but also a commercial versión exists supported by a company)

 ISO focused on the model
 OVF-VMAN and Open Nebula reference mainly technical
aspects of Cloud

Cloud Computing
Standards
 Pure Cloud Computing Standards are at a very early stage

 Standards focus in the technical questions of technology and
products supporting cloud, as virtualization and security

 Focus in eHealth standards as HL7, ISO13606, SDTM, ADaM, BRIDG
Model, CDASH LAB, LOINC, DICOM

 Commercial, data access and legal aspects of Cloud Computing are
a very important question when trying to apply Cloud Computing to
Health Domain

Cloud Computing
Certifications

 Provided by companies: focused in common
technology supporting Cloud, system security
and management, not in health

 For the Health domain, it is interesting to get
knowledge in the Legal scope covering data
protection and data access which deal with EHR
data

Cloud Computing
What are the companies offering for Health?

 EHR/CPOE/RIS under SaaS, mainly in US

 IaaS, PaaS specialized for Health, assuring data
access and security in agree with EHR legal
aspects

 EHR Storage systems


Use Cases in Health
 Tipical software applications in Health organizations include:

 Electronic Health Record systems (EHR)

 Clinical Prescribing Order Entry (CPOE)

 Medical Imaging (MI)


 Laboratory

 Clinical Data Management

 Enterprise Applications (HR, ERP,…)

 Classifications (ICD, ATC, SNOMED…)

Cloud Computing
SaaS: ICD, ATC

 Two cases of accepted classifications used
daily from the cloud:

 Classification of Diseases (ICD) http://
www.who.int/classifications/icd/en/

 Anatomical Therapeutic Chemical (ATC)
classification systems plus the Defined Daily Dose
(DDD) http://www.whocc.no/atc_ddd_index/

Cloud Computing
Two cases in Health

 Spain: Current situation in a large regional
government health agency

 Chile: Improving Diabetes programs by using
Cloud

Cosmos project in Chile
The Business Problem: Diabetes

 Let’s start from the business problem that is, from the Clinical
problem

 It is well known that Diabetes is becoming a pandemic for the XXI
Century with a 15% of prevalence

 In Chile, the public health care system is required by law to confirm
or rule-out Diabetes Mellitus Type 2 (DM2) within 45 days of a
patient’s high blood glucose test

 Despite this:
 40% of patients with poor glucose control are not diagnosed

 15% of those diagnosed do not initiate treatment

A pilot project in 2005

 In 2005, a team from the School of Nursing at Pontificia
Universidad Católica de Chile implemented a practice
linked telehealth model of self- management support

 The target was to promote lifestyle changes and reduce
hemoglobin A1c in low income patients with DM2 treated
in public primary care centers

 Qualitative evaluation following the initial
implementation trial suggested that:
 many of the tele counseling session could be replaced by short
message service (SMS) text messages or automated calls
delivered using an interactive voice response (IVR) service.

A final solution in 2010
 In 2010, with funding from the Inter-American Development Bank
(IADB), the team developed a mHealth model based on tele
counseling, SMS, and IVR for DM2

 The goals were:

 Activate individuals with high blood sugar levels to confirm or rule-out
the diagnosis of DM2 within 45 days

 Improve the efficacy of the diagnostic confirmation process

 Improve adherence to pharmacological treatment and lifestyle changes

 COSMOS was deployed in two health centers and evaluated during
an 11-month period.

Architecture
 COSMOS is built on:

 OpenMRS, a web-based, open-source
Electronic Medical Record (EMR)

 Open Data Kit Voice (ODK-Voice) to make IVR
calls

The Cloud Model: IaaS, PaaS, Hibrid Deployment

 A US company based in Texas provides the service
provides the ICT infrastructure where the software is
deployed

 Telecom is provided another Chilean Telecom company

 These leads to a mix of IaaS, PaaS and managed
services

 Software is owned by the customer but mantained by a
Chilean Company

Conclusions

 Cloud Computing can push health
programs

 Cloud computing fits to implement pilot
projects and small to medium size projects

 Cloud computing can be applied in
developing countries

Use Cases in Health:
An approach to Comunidad Valenciana
Public Health System

 Population: approximately 5.5

 Health centers: +1000

 CS 2011 budget: +5.500.000.000 €

ICT Picture
 Primary Care EHR/CPOE: Abucasis/Gaia

 Hospital EHR/CPOE: Orion Clinic and others

 +100 business applications

 Network: Arterias

 Many National and Multinationa ICT blue chip contractors:

 1 Corporative Datacenter

 1 corporative spare system under PaaS on a datacenter owned by a company

 datacenters in each hospital

 And much more…

ICT: ¿What about cloud?

 Spare system is under Cloud model

 Small projects under PaaS

 Some parts of the system are under Managed
Services model

 Cloud adoption is starting up

Use Cases in Health
Corporative Datacenter
 Located in old Hospital La Fe de Valencia

 Building, hw and sw resources belongs to CS/AVS

 Managed by outsourcing companies leaded by
CS/AVS managers: 100 staff

 Applications hosted
 Abucasis/Gaia (servicing 14.000 clinicians every day)
 +100 Software Applications

Hot Spare Datacenter
 Located in an ICT provider builging

 Providing IaaS, PaaS services

 Access to services through a private network (Arterias),
not Internet

 Managed by both provider and outsourcing companies
leaded by CS/AVS managers

 Applications hosted
 Abucasis/Gaia

 A part of the Administration Applications

Hospital Datacenters

 Each hospital has its own small datacenter hosting:
 EHR
 CPOE
 RIS
 Some other Clinical specialized software
 Business software

 Some hospitals using Orion Clinic, expected to be deployed in all
hospitals in the regional health system

 This datacenters are managed by staff:
 Located in the hospitals
 Located in the corporative datacenter

Business Intelligence under PaaS
 A BI tool is loaded monthly

 The BI tool is deployed on Wintel platform provided by a
provider, so PaaS model is applied

 The provider is the same providing spare system, within
private Arterias Network and under strong controls

 The BI tools is used by Dirección General de Farmacia
staff to provide information at different levels following a
very strong legal protocol

 This is a small project which provides high speed
outcome delivery in some process

Cloud Computing options
What could be done?

 Short Term
 Move all systems which do not include EHR information to PaaS
models
 Move EHR hospital software from Hospital Datacenters to
Corporative Datacenter under a Managed Service model

 Medium term
 Move all not EHR related sofware to SaaS
 Move EHR software to IaaS

* a main handicap for these actions is the fact of the lack of process management
adoption

Why to keep an strong control on EHR/
CPOE/RIS raw data?

 Stakeholders have strong interest in large
databases of EHR raw data

 Wait some time to see what happens…

Conclusions
 Technology and companies are ready to provide cloud computing
services to Health domain

 Cloud computing is a very interesting option for all that no-clinic
application as Human Resources, Purchasing, Accounting and
similar

 Cloud computing model can boost eHealth adoption in Health
organizations by applying the model to:
 Proof of concept and pilot projects
 Small to medium size projects

 Cloud computing can be used in both technologically advanced
developed countries and developing countries

 For big EHR projects, it would be better to wait due to the legal
aspects, and adopt soft cloud computing options as IaaS

Cloud Computing in Health

More Related Content

What's hot (16)

Viewers also liked (14)

Similar to Cloud Computing in Health (20)

Cloud Computing in Health