Successful Deployment of Cloud-hosted Services and Performance

Management for Community Care

Benjamin Eze

1,3

, Craig Kuziemsky

, Jamie Stevens

, Paul Boissonneault

and Liam Peyton

School of Electrical Engineering and Computer Science, University of Ottawa, 800 King Edward Ave, Ottawa, Canada

Telfer School of Management, University of Ottawa, 55 Laurier Ave E, Ottawa, Canada

Champlain Local Health Integration Network (LHIN), 4200 Labelle St, Ottawa, Canada

Keywords: Healthcare Interoperability, Data Aggregation, Surveillance, Performance Management, Community Care,

Cloud Computing, Regional Health Authority.

Abstract: Achieving systematic performance management of care processes across a health region requires an

architecture that balances interoperability and data standardization with data governance and privacy

compliance. This paper presents a case study of a successful pilot of cloud-hosted performance management

for community care by a Regional Health Authority mandated with coordinating home care amongst 54

Community Support Services agencies. Cloud-hosted data services enabled data integration to a common

data model. Formal data sharing agreements and privacy definition documents controlled aggregation and

data masking to protect privacy while enabling accurate and comprehensive performance management

services for all agencies.

1 INTRODUCTION

Community healthcare is diverse with multiple

settings, actors, data sources and communication

channels (Eze et al., 2017a). Regional health

authorities (RHA) charged with ensuring quality of

care and population health would like to measure, on

a continuous basis, performance management across

the entire healthcare ecosystem. This is challenging

because of the aforementioned distributed nature of

healthcare delivery (Foldy et al., 2014).

To date, hospitals have provided much of the

healthcare outcome data through data sources such as

discharge summaries, referrals, emergency room

visits, wait times and procedures. However, this data

only tracks isolated hospital-based events and not

patient outcomes at the health systems level (Veillard

et al., 2010). Healthcare delivery to support complex

patient care such as chronic illness takes place in the

community and is an ongoing process rather than an

isolated event. Monitoring system performance for

community care is challenging because of the need to

track and measure care delivery across settings

(Durovich and Roberts, 2018; Maruthappu et al.,

2015; Roughead et al., 2011).

Health system management is often coordinated

at an RHA level (Molinari, 2014) that has the goal of

delivering a cost-effective and high-quality

collaborative environment (Sabooniha et al., 2012).

Yet quality management and system accountability

are challenging because of the diverse settings in

which healthcare delivery is provided (Denis, 2014).

Performance management requires a systematic

framework that enables continuous data integration

and monitoring of care processes (Lemieux-Charles

and Greengarten, 2014). Cloud computing is one

potential infrastructure for developing interoperable

healthcare solutions (Andry et al., 2015; Bhaskaran et

al., 2013; Li and Guo, 2015). There is a need for

studies that describe actual implementation of cloud

computing systems and how issues such as privacy

and data sharing were managed (Griebel et al., 2015).

This paper presents a case study of a successful

pilot of cloud-hosted performance management for

community care at by a RHA mandated coordinating

home care amongst 54 Community Support Services

(CSS) agencies. A cloud-hosted system owned and

managed by the health authority solves

interoperability and security issues when you want 54

community care organizations to share data. Cloud-

hosted data services enabled data integration to a

common data model. Formal data sharing agreements

248

Eze, B., Kuziemsky, C., Stevens, J., Boissonneault, P. and Peyton, L.

Successful Deployment of Cloud-hosted Services and Performance Management for Community Care.

DOI: 10.5220/0007365902480255

In Proceedings of the 12th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2019), pages 248-255

ISBN: 978-989-758-353-7

and privacy definition documents controlled

aggregation and data masking to protect privacy

while enabling accurate and comprehensive

performance management services for all agencies.

2 BACKGROUND

Community healthcare is the delivery of healthcare

services outside hospitals to manage chronic illness,

and provide services like rehabilitation support,

nursing, physiotherapy, and end of life care to

patients (CIHR, 2017). Community healthcare is

delivered to clients of all ages whose illness or

condition necessitates long-term health care delivery

at home. The purpose of healthcare performance

management as it relates to community care is to

measure the extent that quality of care goals are

achieved through the delivery of healthcare processes

(Vanhaecht et al., 2007).

Patient needs are met through service-level

planning and coordinated care delivery provided by

health care providers in community settings such as

public health units and community services agencies,

(CIHR, 2017). Continuous efforts are being made to

provide good quality and cost-effective care with

support from government and non-profit

organizations (Boissonnealth and Lafreniere, 2014).

As patient needs may need to be coordinated across

different settings and providers, it is critical to build a

strong collaboration platform for all aspects of

community healthcare to coordinate service delivery

to patients while also monitoring outcomes such as

cost-effectiveness (Berler and Apostolakis, 2014).

2.1 Regional Health Authorities

Community care is typically governed by an RHA. In

Ontario, Canada for example, the Regional Health

Authority is the Local Health Integration Network

(LHIN, 2018), with the mandate to plan, integrate and

fund local healthcare in a region. Ontario has 14

Local Health Integration Networks that target each

sub-region within the province.

Norway employs a similar model with 4 main

RHAs, each with many subsidiaries (Ringard et al.,

2013). For both Ontario and Norway, the RHA is

responsible for patient treatment, medical staff,

planning, research, and development, as well as

support and training for patients and their caregivers.

RHAs are associated with improved healthcare

outcomes, healthcare equality, increased life

expectancy, improved coordination and reduced cost

to healthcare services (Vida et al., 2012).

2.2 Complex Patients

Patient complexity can be due to factors such as

comorbid health conditions that make patient care

management very challenging (Grant et al., 2011).

Providing care for complex patients is putting

increased strain on healthcare budgets and service

delivery performance goals such as wait times

(Sheikh et al., 2015). As a consequence, there is a

growing need to transform the health care system and

the services it delivers to more efficiently provide

care for complex patients (Sabooniha et al., 2012).

Government and healthcare organizations want

better accountability for money spent on healthcare

delivery (Bohmer, 2016), which requires

performance management of care processes across all

stakeholders in the healthcare ecosystem (Berwick et

al., 2008). Achieving this requires coordination and

integration of data across disparate healthcare

information systems (Sabooniha et al., 2012).

Complex patient management is challenging and

expensive as it requires care and service delivery

from a variety of providers (Mcgregor et al., 2016).

Further, complex patients may be managed using

multiple clinical practice guidelines which may have

conflicting recommendations about medications or

treatments (Wilk et al., 2017).

2.3 Performance Management

Performance management provides a mechanism for

translating strategic objectives and business goal to

operational processes (Kemper et al., 2013).

Performance management involves planning, setting

expectations, continuous monitoring of performance,

developing the capacity to perform, as well as rating

and rewarding of performance (OPM.GOV, 2017).

Continuous monitoring of KPIs is key to effective

monitoring and management of strategic goals

(Sanchez and Robert, 2010). However, each strategic

goal also needs to be linked to these KPIs to measure

the extent that the performance of the organization is

achieving goals (Kuziemsky et al., 2010).

Heterogeneous data silos and inconsistent patient

identity approaches, coupled with patient privacy

regulations, limit our ability to correlate data for

complex patients as part of performance management

(Eze et al., 2016). This results in the inability of

stakeholders to coordinate care delivery across

multiple healthcare domains (Adler-Milstein and Jha,

2012). Attempts to address these factors often lead to

unintended consequences (e.g., social, legal and

workflow consequences) that arise from technology-

mediated connectivity (Kuziemsky et al., 2016).

Successful Deployment of Cloud-hosted Services and Performance Management for Community Care

249

2.4 Cloud Computing

Cloud computing is a distributed, configurable

approach for generating ubiquitous access to a pool

of convenient, on-demand computing resources

(compute, storage, platform, application and services)

through a web interface. Cloud computing has been

shown to provide for regional, national and

international data aggregation using a broad range of

topologies that could integrate various devices, data

sources and services very quickly in a scalable and

cost-effective manner (Andry et al., 2015). In

addition, Cloud infrastructure provides an infinitely

scalable storage for very data-intensive applications

(Bhaskaran et al., 2013; Ochian et al., 2014).

Cloud deployment models are differentiated by

the location of the infrastructure, the user of the

infrastructure and the entity that manages the

infrastructure (Furht and Escalante, 2010). There are

three deployment models of cloud computing –

Private, Public, and Hybrid Clouds. A private cloud

is operated by a single organization, which has full

control over the infrastructure, data, security, and

quality of service (QoS). The public cloud is operated

by a 3

party and can be used by with applications

mixed together on cloud servers, storage systems, and

networks (Eze et al., 2016). Finally, a hybrid cloud is

a mix of public and private clouds. In the hybrid

cloud, data and applications are distributed across

both public and private clouds using secure data

bridges (Ma et al., 2014).

A private cloud provides the owners full control

over everything – compute, storage, networking, as

well as the quality of service. While having full

control increases the complexity associated with the

development and deployment of a cloud application

and services, it provides better security and

confidentiality with user data. Unfortunately,

compared to other cloud deployment models, a

private cloud is more expensive (Ma et al., 2014).

The public cloud provides the lowest Total Cost

of Ownership (TCO) of the cloud types but also

provides the least control. Also, data security cannot

be guaranteed since cloud resources are shared by

many organizations. Public clouds are also prone to

resource contention issues, SLA breaches, and

service disruptions. For healthcare organizations with

high volumes of highly sensitive data, this would not

be acceptable since it violates data privacy laws in

many countries (Furht and Escalante, 2010;

Gazzarata et al., 2015).

3 PILOT PROJECT

The Champlain Local Health Integration Network

(LHIN) provides at-home care services for patients in

a metropolitan area with a population of over 1.2

million. About 60,000 annual active patients are

receiving over two dozen community care services

from the LHIN and its 54 Community Support

Services (CSS) agencies.

Figure 1: CSS Agencies and Spectrum of Services.

Fig. 1 shows examples of some of these services.

Services provided by these CSS agencies are intended

to improve the quality of life of patients with complex

and chronic health conditions.

3.1 Lack of Performance Management

The lack of performance management of community

care in the Champlain LHIN before the pilot project

is depicted in fig. 2. CSS agencies target specific

populations with niche community services. Usually,

these agencies are small organizations with limited

budgets and their own small ad-hoc IT systems.

Figure 2: Lack of Performance Management.

As a result, there is minimal interoperability and

limited performance management. Data collection is

limited to ad-hoc invoices (typically in MS Excel

HEALTHINF 2019 - 12th International Conference on Health Informatics

250

Figure 3: Cloud-based Performance Management Architecture for Community Care.

format) for services rendered by each agency and

usually submitted by email to the RHA financial

officers to process payments. The RHA depends on

ad-hoc reports, mostly prepared for funding and

budgeting needs.

Each CSS agency has their own data silo resulting

in service duplication and limited coordination of care

delivery.

3.2 Architecture

The multi-tenanted private cloud architecture shown

in figure 3, hosts, for each of the 54 CSS agencies,

their patient management applications and

operational databases, while providing data

integration and performance management services.

This preserves the autonomy of each CSS agency,

allowing for quick adoption by each organization.

As depicted in figure 3, the LHIN CSS agencies

that want to participate in the performance

management architecture must sign a Data Sharing

Agreement (DSA) and provide patient consent.

These are formally defined and enforced using

Privacy Definition Documents (PDD). The three

aspects of data integration, privacy compliance and

performance management are described in detail in

the sections that follow.

3.2.1 Data Integration to a Common Data

Model

A Systematic Data Hosting Service ensures that the

data from each CSS agency is in a secure, controlled

environment. This provides the trust needed to allow

systematic data integration into a Common Data

model to support Performance Management services.

As shown in Figure 3, each organization’s patient

management application is hosted in a load balanced

cloud-hosted virtual machine (VM) using a cloud-

hosted Microsoft SQL Server database cluster. Users

from each organization access their respective

application instances through a remote VPN service

using an SSL VPN Client.

A Systematic Data Collection Service connects to

the MS SQL Server cluster and systematically

collects data from each database instance. This

service supports heterogeneous data sources through

data integration (Platform-as-a-Service) PaaS

containers customized for the RHA and 54 CSS

agencies. Data from a particular CSS is only collected

if there is a signed DSA in place as described in

section 3.2.2. The major challenge with this

architecture is that CSS agencies and the Champlain

LHIN do not share a common patient identifier such

as government-issued health card number (HCN). In

addition, identity attributes of the patient like first and

last names, date of birth, gender, phone numbers, and

addresses are not collected in a consistent manner

across the agencies. It is therefore difficult to match

data from different databases for the same patient.

Imposing a common region-wide patient identifier

would be a costly long-term initiative and was

therefore not a viable short-term option. Instead, a

patient-identity matching service was provided.

After evaluating the data sets across the agencies,

we came to three conclusions about identity

matching. First, each agency had an identifier specific

to their database for the patient. Second, agency data

could have data entry errors. Third, patient identity

matching could leverage attributes like current and

historical addresses, phone numbers, to fine-tune and

verify matches.

Successful Deployment of Cloud-hosted Services and Performance Management for Community Care

251

A probabilistic matching algorithm was

implemented to address identity matching for the

infrastructure. This algorithm derives from existing

work in probabilistic record linkage domain like the

Expectation-Maximization (EM) algorithm

(Dempster et al., 1977), as well as the theories of

record linkage (Fellegi and Sunter, 1969) This

algorithm is described in details in a previous

publication (Eze et al., 2017).

It addresses the record

linkage problem by dividing a data set into blocks to

minimize comparisons to only records within the

same block. A block is a combination of one or more

identifier attributes with an associated weight. The

matching process accumulates matches across all

block passes for each patient identity attributes to

determine the matches that are full, partial or

ambiguous or non-matches.

Patient Identity Matching is carried out on all

patient-level records across all the incoming data

streams based on published matching rules. At the

end of the process, a global identifier is issued to each

cluster of profiles belonging to the same patient

across the collaborating organizations. This identifier

is then used to map the rest of their data into the

Common Data Model (CDM).

Figure 4: Common Data Model Depicted.

The CDM uses a hierarchical model (fig. 4). Data

elements found in each source database must be

organized into this hierarchical model. The 1

level is

patient-centric data with patient identifiers,

demographic data, contact details – phone numbers

and addresses, personal and emergency contacts,

caregivers, family physicians. The 2

level captures

episodes of care and data that map to each episode of

care like the patient population (a categorization of

patients into population groups with similar care

needs), referrals, care plans, and medication history.

The 3

level is longitudinal events and maps to data

on the care episodes defined at the 2

level. These

include service visits for home care, diagnoses and

medications.

3.2.2 Data Sharing Agreements and Privacy

Definition Documents

An essential need for community care data integration

and performance management services is adherence

to privacy legislation. Since the CSS agencies are

independent organizations, they are not under any

mandates to provide data to the LHIN apart from

billing and support reports. Binding agreements that

protect patient privacy and confidentiality need to be

signed between these organizations to allow their data

to be shared.

Privacy compliance is addressed through the

creation of formal Data Sharing Agreements (DSA)

for organizations and a Privacy Compliance

Definition Document (PCDD) that captures both

organizational and patient consent to configure and

regulate the processing of data.

DSAs consents

regulate each incoming data stream from a CSS

agency. The PCDD applies globally (across all data

streams), locally (specific to each organization data

stream), and partially (controlling access to specific

data entities and attributes). Enforcing these consents

can result in complete removal of patient data from

the common data model. In some cases, there could

be full or partial anonymization (data masking,

generalization, suppression) of patient data to meet

set risk thresholds for the infrastructure.

For this pilot project, an all-or-nothing type of

privacy compliance was implemented. The PCDD

provides meta-data on data elements that describe

their semantic types, determining how the privacy of

each attribute should be addressed. It also specifies

fields that needs be nulled, removed, or even masked

based on DSAs to ensure that restricted data never

makes it to the CDM.

All stakeholders that signed the DSA, including

the RHA, have access to all aggregate reports from

the CDM. They and the RHA are also able to see

patient-level report across all agencies except for

those patients that did not consent.

While the all or nothing approach provides good

privacy protection, it does impair the accuracy and

completeness of performance management reporting.

An anonymization service that improves accuracy

HEALTHINF 2019 - 12th International Conference on Health Informatics

252

and completeness is part of proposed future

enhancements to the architecture.

3.2.3 Performance Management Services

Our architecture supports four performance

management services – anonymization, analytics,

reporting, and subscription services. For this pilot

project, three performance management services

were leveraged – anonymization, reporting and

subscription services.

A simplified anonymization model - where all

patients that opted out of data sharing are

automatically excluded from performance

management by the data collection service, was

leveraged. Agencies only see aggregate service data

for patients in their circle of care. Patient-level data

from patients that belong to other providers are

excluded. All performance management reports are

made available to only those participating agencies

that have signed the DSA.

The reporting service use two hosted MS SQL

Server Reporting Services (SSRS) that mirror each

other. Each analytics report created is published on

both servers. Personnel from the LHIN and the

agencies use the reporting portal to access the

standard reports made available to them. Since SSRS

has Web services support, these reports can also be

streamed dynamically, using the subscription service.

Shared Services Subscription service is the

component of the framework that closes the loop

regarding pushing information from the shared data

model to the stakeholders and decision makers at all

levels across all collaborating organizations and their

partners. It is also the component of the performance

management framework that supports process

interoperability described in (Benson, 2012;

Kuziemsky, 2013). The subscription service

leverages data from the CDM and LDAP to provide

knowledge, collaborative, and operational data

needed for performance management of community

care processes.

The key features of the subscription service that

make it adaptable to a cloud computing environment

include 1) Support for dynamic data-driven

subscriptions with declarative and SQL executable

definitions for subscribers and report parameters. All

settings for subscriptions and reports are in

configurable XML definitions hosted in a source

repository. 2) The ability to dynamically package and

deliver multiple reports to users in these different

formats – HTML, MS Word, MS Excel, and Acrobat

PDF. 3) Support for multiple delivery modes – email,

file system, and calendar appointments. 4) Dynamic

scheduling – daily, day periods, weekdays, weekly,

monthly, and quarterly, specific days of the week,

month, quarter. 5) Supports rich failure notifications

for administrators.

For each report developed and published, a

custom report definition file that allows the

subscription service to dynamically stream the report

in many formats is also published to the resource

library. They are created by Business Intelligence

officers with more in-depth knowledge of the model

and the framework. The key utility of the subscription

service is its ability to dynamically package and

deliver multiple reports in various formats through

multiple delivery mechanisms to data recipients.

4 PRELIMINARY EVALUATION

The pilot project is considered a success within the

LHIN. 48 of the 54 the CSS agencies with over

150,000 patients are currently cloud-hosted. 17

agencies with about 30,000 patients have signed the

DSA and currently participate in the performance

management infrastructure. There are nightly data

collection and aggregation of the data across the CSS

operational databases to the Common Data Model.

There are nightly patient identity matching and

progressive clustering of patient profiles. Identity

matching results show that at least 25,000 patients

have matches from another agency or the LHIN with

about 3,000 ambiguous matches that would require

manual approval. However, research has been done to

suggest improvements to the identity management

framework (Eze et al., 2017b).

There are up to 8 active report subscriptions set

up for the LHIN and CSS Agency contacts that

publish and emails various reports. The subscription

service has been adopted by the LHIN and is used to

package and send via email and file transfer, hundreds

of reports to LHIN employees multiple times a day.

Data quality feedback is provided to CSS agencies on

various data quality issues with patient profiles that

need addressing.

The current implementation uses an all-or-

nothing approach to address privacy consent (Eze et

al., 2018). Ongoing research is looking at extending

this through a privacy compliance framework that

leverages anonymization to provide more complete

and accurate reports.

Successful Deployment of Cloud-hosted Services and Performance Management for Community Care

253

5 CONCLUSION

Modern healthcare delivery is about connected

healthcare delivery and patient care management

across providers and settings. Achieving these require

modern approaches to supporting patient care.

Though there are many architectures for patient

management, implementation has remained the last

mile problem. But this is not a technological problem

per se but rather an issue that encompasses multiple

areas such as software design, patient needs, technical

interoperability, privacy considerations, patient data

interoperability (e.g. standards), and governance

agreements across settings.

This paper describes an initial deployment of a

cloud-based performance management system. A

multi-tenanted private cloud infrastructure with

cloud-hosted data services provide a trusted

environment which can enable secure, well-regulated

systematic data integration to a common data model

(CDM) to facilitate comprehensive performance

management for community care. Formal data

sharing agreements (DSA) and a privacy compliance

definition document (PCDD) provide a robust

mechanism for controlled aggregation and data

masking to protect privacy while enabling accurate

reporting.

ACKNOWLEDGEMENTS

This work was supported by the National Science

and Engineering Research Council of Canada

REFERENCES

Adler-Milstein, J., Jha, A.K., 2012. Sharing clinical data

electronically: A critical challenge for fixing the health

care system. JAMA - J. Am. Med. Assoc.

Andry, F., Ridolfo, R., Huffman, J., 2015. Migrating

healthcare applications to the cloud through

containerization and service brokering, in: 8th

International Conference on Health Informatics,

Proceedings;. SciTePress, 164–171.

Benson, T., 2012. Why Interoperability is Hard. In

Principles of Health Interoperability HL7 and

SNOMED SE - 2, 21–32. Springer London.

Berler, A., Apostolakis, I., 2014. Research Perspectives on

the Role of Informatics, Health Policy and

Management, El Morr, C. (Ed.), IGI Global, 168–207.

Berwick, D.M., Nolan, T.W., Whittington, J., 2008. The

triple aim: care, health, and cost. Health Aff. 27, 759–

769.

Bhaskaran, S., Suryanarayana, G., Basu, A., Joseph, R.,

2013. Cloud-Enabled Search for Disparate Healthcare

Data,2013 IEEE International Conference on Cloud

Computing in Emerging Markets IEEE, pp. 1–8. h

Bohmer, R.M.J., 2016. The hard work of health care

transformation. N. Engl. J. Med. 375, 709–711.

Boissonnealth, P., Lafreniere, N., 2014. Deploying

Information Systems throughout the Community Care

Sector of the Champlain Region, in: E-Health 2014

Conference, Vancouver, BC. British Columbia.

CIHR, 2017. Community-Based Primary Health Care, Can.

Institutes Heal. Res., accessed 10.15.17 at

http://www.cihr-irsc.gc.ca/e/43626.html

Dempster, A.P., Laird, N.M., Rubin, D.B., 1977. Maximum

likelihood from incomplete data via the EM algorithm.

J. R. Stat. Soc. Ser. B 1–38.

Denis, J.-L., 2014. Accountability in healthcare

organizations and systems. Healthc. Policy 10, 8–9.

Durovich, C., Roberts, P., 2018. Designing a Community-

Based Population Health. Popul. Health Manag. 21.

Eze, B., Kuziemsky, C., Lakhani, R., Peyton, L., 2016.

Leveraging Cloud Computing for Systematic

Performance Management of Quality of Care. Procedia

Comput. Sci. 98, 316–323.

Eze, B., Kuziemsky, C., Peyton, L., 2017a. Cloud-based

performance management of community care services.

J. Softw. Evol. Process.

Eze, B., Kuziemsky, C., Peyton, L., 2017b. A Patient

Identity Matching Service for Cloud-based

Performance Management of Community Healthcare,

in: Procedia Computer Science.

Eze, B., Kuziemsky, C., Peyton, L., 2018. Operationalizing

Privacy Compliance for Cloud-Hosted Sharing of

Healthcare Data, in Software Engineering in

Healthcare Systems (SEHS). pp. 18–25.

Fellegi, I.P., Sunter, A.B., 1969. A theory for record

linkage. J. Am. Stat. Assoc. 64, 1183–1210.

Foldy, S., Grannis, S., Ross, D., Smith, T., 2014. A Ride in

the Time Machine:

Information Management

Capabilities Health Departments Will Need. Am. J.

Public Health 104, 1592–1600.

Furht, B., Escalante, A., 2010. A Handbook of Cloud

Computing, in: A Handbook of Cloud Computing.

Springer US, pp. 3–19.

Gazzarata, G., Gazzarata, R., Giacomini, M., 2015. A

Standardized SOA Based Solution to Guarantee the

Secure Access to EHR. Procedia Comput. Sci. 64,

1124–1129.

Grant, R.W., Ashburner, J.M., Hong, C.C., Chang, Y.,

Barry, M.J., Atlas, S.J., 2011. Defining patient

complexity from the primary care physician’s

perspective: a cohort study. Ann. Intern. Med. 155.

Griebel, L., Prokosch, H.U., Köpcke, F., Toddenroth, D.,

Christoph, J., Leb, I., Engel, I., Sedlmayr, M., 2015. A

scoping review of cloud computing in healthcare. BMC

Med. Inform. Decis. Mak.

Kemper, H.G., Rausch, P., Baars, H., 2013. Business

Intelligence and Performance Management:

Introduction. Bus. Intell. Perform. Manag. 3–10.

HEALTHINF 2019 - 12th International Conference on Health Informatics

254

Kuziemsky, C., Liu, X., Peyton, L., 2010. Leveraging goal

models and performance indicators to assess health care

information systems, Quality of Information and

Communications Technology Conference, 222–227.

Kuziemsky, C.E., Randell, R., Borycki, E.M., others, 2016.

Understanding Unintended Consequences and Health

Information Technology. IMIA Yearb. 53–60.

Lemieux-Charles, L., Greengarten, M., 2014. Performance

management systems and the reality of Canadian

healthcare organizations from industrial to learning

models. Heal. Q 17, 5–7.

LHIN, 2018. Local Health Integration Network accessed

15.10l.18 at http://www.lhins.on.ca/

Li, Y., Guo, Y., 2015. Wiki-Health: From Quantified Self

to Self-Understanding. Futur. Gener. Comput. Syst. 8

Ma, J., Peng, C., Chen, Q., 2014. Health Information

Exchange for Home-Based Chronic Disease Self-

Management -- A Hybrid Cloud Approach, 4 5th

International Conference on Digital Home. 246–251.

Maruthappu, M., Hasan, A., Zeltner, T., 2015. Enablers and

Barriers in Implementing Integrated Care. Heal. Syst.

Reform.

Mcgregor, J., Mercer, S.W., Harris, F.M., 2016. Health

benefits of primary care social work for adults with

complex health and social needs: A systematic review.

Heal. Soc. Care Community.

Molinari, C., 2014. Does the Accountable Care Act aim to

promote quality, health, and control costs or has it

missed the mark? Comment on “Health system reform

in the United States”. Int. J. Heal. policy Manag.

Ochian, A., Suciu, G., Fratu, O., Voicu, C., Suciu, V., 2014.

An overview of cloud middleware services for

interconnection of healthcare platforms,

Communications Conference, 1–4.

OPM.GOV, 2017. Performance Management Overview

[WWW Document]. URL https://www.opm.gov/

policy-data-oversight/performance-

management/overview-history/ (accessed 20.10.18).

Ringard, Å., Sagan, A., Sperre Saunes, I., Lindahl, A.K.,

2013. Norway: health system review. Health Syst.

Transit. 15, 1–162.

Roughead, E.E., Kalisch, L.M., Ramsay, E.N., Ryan, P.,

Gilbert, A.L., 2011. Continuity of care: When do

patients visit community healthcare providers after

leaving hospital? Intern. Med. J. 41, 662–667.

Sabooniha, N., Toohey, D., Lee, K., 2012. An evaluation of

hospital information systems integration approaches,

Advances in Computing, Communications and

Informatics. ACM Press, New York, New York, USA,

p. 498.

Sanchez, H., Robert, B., 2010. Measuring Portfolio

Strategic Performance Using Key Performance

Indicators. Proj. Manag. J. 41, 64–73.

Sheikh, A., Sood, H.S., Bates, D.W., 2015. Leveraging

health information technology to achieve the “triple

aim” of healthcare reform.

J. Am. Med. Inform. Assoc.

22, 849–856.

Vanhaecht, K., De Witte, K., Depreitere, R., Van Zelm, R.,

De Bleser, L., Proost, K., Sermeus, W., 2007.

Development and validation of a care process self-

evaluation tool. Heal. Serv. Manag. Res. 20, 189–202.

Veillard, J., Huynh, T., Ardal, S., Kadandale, S., Klazinga,

N., Brown, A., 2010. Making Health System

Performance Measurement Useful to Policy Makers:

Aligning Strategies, Measurement and Local Health

System Accountability in Ontario. Healthc. Policy.

Vida, M., Lupse, O., Stoicu-Tivadar, L., 2012. Improving

the interoperability of healthcare information systems

through HL7 CDA and CCD standards, 7th

International Symposium on Applied Computational

Intelligence and Informatics, 157–161.

Wilk, S., Michalowski, M., Michalowski, W., Rosu, D.,

Carrier, M., Kezadri-Hamiaz, M., 2017. Comprehensi-

ve mitigation framework for concurrent application of

multiple clinical practice guidelines. J. Biomed. Inform.

66.

Successful Deployment of Cloud-hosted Services and Performance Management for Community Care

255