leveraging cloud computing for systematic performance

Procedia Computer Science 98 ( 2016 ) 316 – 323

1877-0509 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).Peer-review under responsibility of the Program Chairsdoi: 10.1016/j.procs.2016.09.048

ScienceDirectAvailable online at www.sciencedirect.com

The 6th International Conference on Current and Future Trends of Information and Communication Technologies in Healthcare (ICTH 2016)

Leveraging Cloud Computing for Systematic Performance Management of Quality of Care

Benjamin Ezea*, Craig Kuziemskyb, Rubina Lakhania, Liam Peytona aSchool of Electrical Engineering and Computer Science (SEECS), University of Ottawa, Ottawa, ON, K1N 6N5, Canada

bTelfer School of Management, University of Ottawa, Ottawa, ON, K1N 6N5, Canada

Abstract

Governments and healthcare providers are under increasing pressure to streamline their processes for efficiency, to reduce operational costs while improving service delivery and quality of care. Systematic performance management of healthcare processes is important in determining if quality of care goals are being met at all levels of the healthcare ecosystem. The challenge is that reaching these goals requires the aggregation and analysis of large amounts of data from various stakeholders in the healthcare industry. With the lack of interoperability between stakeholders in current healthcare compute and storage infrastructure as well as the volume of data involved, our ability to measure quality of care across the healthcare system is either limited or non-existent. Cloud computing is an emerging technology that can provide the needed interoperability and management of large volumes of data across the entire healthcare system. In this paper, we review various cloud computing applications for surveillance and performance management of quality of care in healthcare. We also introduce a framework for achieving systematic surveillance and performance management across the entire healthcare system. © 2016 The Authors. Published by Elsevier B.V. Peer-review under responsibility of the Conference Program Chairs.

Keywords: Healthcare; Cloud computing; Big Data; Interoperability; Personal Health Information; Performance Management, Surveillance

1. Introduction

Quality healthcare delivery requires coordination of numerous entities within and outside the circle of care. In the inner circle of care are hospitals, physicians, diagnostic imaging centers, laboratories and pharmacies. In addition,

* Corresponding author. Tel.: +1-613-562-5800 x2122; fax: +1-613-562-5664.

E-mail address: [email protected], [email protected]

© 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).Peer-review under responsibility of the Program Chairs

http://crossmark.crossref.org/dialog/?doi=10.1016/j.procs.2016.09.048&domain=pdf

317 Benjamin Eze et al. / Procedia Computer Science 98 ( 2016 ) 316 – 323

there are specialist clinics, public health clinics, ambulatory centers, long-term care homes, insurance companies, technology vendors, employers, and medical device manufacturers. Outside the circle of care are national and regional governments that devise and enforce regulations to ensure quality of care the promotion of public health. The overall goal of the healthcare system is to provide a cost-effective and high-quality integrated environment for efficient healthcare service delivery1. Other important objectives are to enable safe care delivery, streamline clinical and administrative tasks, and to safeguard patient data.

Continuous, aggregation and analysis of healthcare data is needed for performance management across the healthcare system to ensure quality of care. Most surveillance processes today are either manual or semi-automated and take excessive time for decision makers to receive the necessary data for critical decision-making across the various levels of care delivery. According to Sun and Reddy2, worldwide digital healthcare data is expected to reach 25,000 petabytes in 2020 while an average hospital would be required to manage up to 665 terabytes of patient data, 80% of which would be unstructured medical imaging data.

Internal infrastructure within most healthcare organizations cannot sustain this massive, and exponential growth of health data. Moreover, a complete analysis of quality of care requires integration of data across many different organizations (family doctors, specialists, hospitals, clinics, etc.). Timely analysis of this health data is impossible using traditional data analytics methods. Cloud computing could provide the infrastructure for big data analytics, helping reduce the cost of care delivery, streamline health care processes, and analyze health data to detect disease outbreaks. It would also pave the way for incorporating the growing number of Internet Connected Devices (ICD) and Internet of Things (IoT) in healthcare3,4.

In this paper, we survey applications of cloud computing and describe why it is an effective tool for surveillance and performance management of quality of care. We also introduce an architectural framework for applying this technology in a systematic manner across the healthcare system.

2. Cloud Computing and Healthcare Data

Systematic performance management of quality of care requires a scalable, highly available data aggregation infrastructure. With the rapid explosion of healthcare data2, many organizations are slowly hitting various thresholds in the amount of data they can handle using their internal IT infrastructure. The need for the right infrastructure with the capacity to aggregate and analyze such large volumes of data in real-time is essential to providing needed operational and governance related statistics as well as key performance indicators for healthcare performance management.

Cloud computing fills this need since it offers unlimited scalability of computing resources and capacity. Ochian et al.5 describe cloud computing as a mechanism for generating ubiquitous access to a pool of convenient, on-demand computing resources (compute, storage, platform, application and services) through a web interface with low administration overhead and the least intervention from a cloud service provider. Cloud-based infrastructure and applications offer a consolidated view of patient-relevant data to healthcare providers6 while providing better and more effective quality of care and facilitating their ability to share information, improve collaboration and reduce expenditure on infrastructure3.

Popular cloud providers like Amazon, Google, and Microsoft offer Software as a Service (SaaS), Platform as a Service (PaaS) and Infrastructure as a Service (IaaS) cloud packages to end users7,8. Some of the significant benefits9 of cloud computing include device and location independence, 24x7 support, lower total cost of ownership (TCO), reliability, scalability, sustainability, agile deployment, lower capital expenditure and a single infrastructure to fulfill all computing, networking and storage needs for various applications.

Cloud computing is a potential platform for national and regional healthcare systems interoperability. Torre-Diez et al.9 explored interoperability options for the Spanish Public Health National System and came to the conclusion that cloud infrastructure should be used for data sharing and service orchestration. Similarly, Biswas et al.10 propose an "eHealth Cloud" platform for the government of Bangladesh. The eHealth platform would connect physicians, patients, hospitals, government departments, insurance companies, and pharmaceutical companies to the same


platform. This platform would also give patients’ access to their EMR through federated credentials across all available healthcare services.

Notwithstanding these strengths, the healthcare system presents unique challenges to using cloud computing. Moumtzoglou11 characterizes the healthcare industry as a high-risk, highly regulated, process-heavy industry with multiple stakeholders that are slow in adopting new technologies, and have a higher tendency to maintain long-term relationships with technology vendors. The author points out that the shortcomings of clouding computing11 and the use of cloud services usually do not go well with the structure and operations of the healthcare industry. For example, cloud computing requires data owners to relinquish control over compute and data storage resources while healthcare regulations and policies require such controls and oversight. It is also difficult to ensure full accountability with cloud environments regarding patient privacy and confidentiality. Some of other notable issues, identified by Moumtzoglou, include functionality creep with cloud data; monopoly and vendor lock-ins; as well as questionable privacy protection since patient data could be stored in locations beyond the jurisdictional boundaries of care providers.

2.1. Relevance of Big Data Analytics

Big data refers to datasets that are so large, and complex that they cannot be processed using traditional data analytics and processing systems12,13. According to Gartner14, “Big data is high-volume, high-velocity, or high-variety information assets that demand cost-effective, innovative forms of information processing for enhanced insight, decision making, and process automation”. One can specify 5 Vs that define big data, Volume is influenced not only by the amount of data but also by the dimensionality. Variety looks at data both in terms of heterogeneity of data formats - structured databases, hierarchical data, unstructured and semi-structured data. Velocity refers to static vs. dynamic for streaming and versioned data15. Veracity manages inconsistencies and ambiguities in data especially with unstructured clinical notes, and finally, Value measures the added-value that data brings towards the creation of knowledge.

Cloud-based big data analytics is being used for the development of elastic services for integrating, storing, and processing large scale disparate data16. According to Mathew and Pillai17, most health data is unstructured and exists in silos within different applications such as imaging systems, hospital EHR systems, physician EMR systems, billing systems, etc. Cloud-based big data solutions combine these disparate data sources in real-time, giving healthcare providers access to a 360-degree view of each patient health intervention and outcome. For example, The BRAIN initiative18 is a US National Institutes of Health (NIH) project that is expected to change our understanding of the human brain and therefore, accelerate the development of new treatments for many neurological diseases. This is a typical big data problem because brain activity recordings (EEG); cardiac measurements (ECG) and other physiological measures are high-volume and high-velocity data. According to this study18, a five-day epilepsy evaluation would generate about 1.6GB of data for a single patient, a single 3D MRI would generate about 150MB of data; a single x-ray would generate about 30MB of data while a 3D CT scan would generate up to 1GB of data.

Evidence-based medicine17 provides treatments that are validated through scientific evidence to supplement to a physician’s skills and knowledge. Big data tools provide analytics results and reports in support of evidence-based medicine (Fig. 1). These tools include Hadoop File System (for the storage of large files), Hadoop MapReduce (for scalable job execution), Apache Hadoop YARN (for resource management), HBase (non-relational data management solution), and Hive (for providing SQL-like functionalities to Hadoop clusters). These technologies are making analytics possible on massive volumes of data running into the terabytes and petabytes.

Some of the sources of big data in healthcare include: i. Electronic Health Records: This include both structured and unstructured data that is generated from Patient

Management (PM) and Electronic Medical Records (EMR) systems. Structured and semi-structured data requires various competing standards like HL7, SNOMED19, LOINC20, ICD-9 and ICD-1021 and DICOM 22. EHR include unstructured data like clinical notes, medical imagery like x-rays, CT-Scan, as well as ECG and EEG scans.


ii. Public Health Data: This is data that comes from monitoring population health. This data helps public health authorities measure incidents and spread of infectious diseases like the common flu, West-Nile virus, as well as sexually transmitted diseases.

iii. Sensor Data: These are data that come from various devices used in monitoring patient vital signs and organ functions, for example, sensors that track heart rate or blood pressure.

iv. Genomic Data: Sequencing the human genome is providing researchers with information on effects of our genetic makeup to our predisposition to various diseases. Analysis of this data is the basis for precision medicine.

v. Behavioral Data: In today’s world of social media, this is data from Facebook, Twitter, LinkedIn, Health blogs, and various Web 2.0 tools. Behavioral data from these interactions is becoming an important source of big data.

Fig. 1. Evidence-based Medicine and Big Data

2.2. Cloud Computing and Internet of Things (IoT)

The Internet of Things (IoT) is the concept of interconnecting physical objects to each other or to the Internet to create domain-specific intelligence through seamless pervasive sensing, data analytics and information visualization with cloud computing23. Sensor technologies and networks are fast becoming the core drivers of IoT24. IoT applications like asset tracking and Body Area Sensor Networks (BASN) are built for ubiquitous mobile health surveillance. Most current smartphones now come with some embedded sensors like microphone, camera, gyroscope, accelerometer, compass, proximity sensors, GPS, and ambient light detector3 that can be leveraged for various health applications. The large volume of data being generated by these devices requires a scalable and highly available big data environment for aggregation and analytics.

Wearable IoT can be defined as a technological infrastructure that interconnects wearable sensors to enable monitoring of health, wellness, behaviors and other data useful in enhancing individuals’ everyday quality of life. Individuals are seamlessly tracked by wearable sensors for personalized health and wellness information; body vital parameters, physical activity, behaviors, and other critical parameters impacting the quality of daily life25.

BASN can perform classified learning and analysis in real time to provide early medical warnings26. Data streams from devices are analyzed in real-time to help ecologists, public health and city officials in their day-to-day activities. Some of the analytics include determining correlation and association amongst variables; performing time and location partitioning to investigate how variables evolve over time in different geographical regions, and the use of k-means algorithms to generate Voronoi-like partitions to provide geographic context.

The uHealth system described in the work of Ganchev et al.26 is an IoT project that integrates doctors, patients, other caregivers as well medical devices and sensors on the cloud. The cloud-based X73 infrastructure is used for medical data collection and as a fault-tolerant real-time computing system for data storage, mining, and analysis. An


interactive web-based visualization tool is used to explore these relationships through time-sliced correlation maps and to navigate the geographic clusters generated through k-means clustering.

The challenge with sensor data is how to effectively manage the data being generated by these devices. Wiki-Health 3, is a big data service platform designed to handle the problem of explosive growth in healthcare data. Real-time analytics allows Wiki-Health to meet performance and scalability goals with big data through the Wiki-Health Analysis Framework. The Analysis Tasks Allocation Scheme (ATAS) allows analysts to add data analysis tasks dynamically with access to specific data streams while cloud resources are automatically provisioned to support each new task as needed.

A cloud computing infrastructure for big data can facilitate the management of wearable data and can support advanced functionalities of data mining, machine learning, and medical big data analytics. Cloud-assisted BAS (CaBAS) is emerging as a promising technology that provides integration of Mobile cloud computing and Wearable body area sensors to facilitate the growth of scalable, data-driven pervasive healthcare 25.

2.3. Maintaining Privacy and Confidentiality with Cloud Computing

Perhaps the biggest challenges to cloud computing are privacy and confidentiality27. While encryption helps secure data exchanges and storage, it does not protect the data from authenticated users that have access to the decrypted data. It also sacrifices the analytical utility of cloud data. Privacy measures, on the other hand, ensure confidentiality of patients and other members of the circle of care while still allowing data to be available for analytical processing17,28.

Many government legislations do not yet permit the storage of health data on the cloud because of privacy regulations11,29. According to Donnelly et al.30, one approach is to store data on the cloud but maintain all identifying data locally. The PACE30 healthcare architecture uses a combination of cloud and peer-to-peer technologies to model healthcare units or clinics where off-cloud data is identifying data while cloud data is completely anonymized. The model was demonstrated in collaboration with some dementia researchers in Ireland.

There is the need to deal with security and privacy related issues with cloud implementations9, and this requires cloud providers to be HIPAA compliant31. Integration solutions must be HIPAA certified through proper authentication and authorization as well as access control practices, data integrity, accountability and audit trail measures31.

3. Proposed Framework

The ability to support continuous performance management of quality care delivery requires an infrastructure that can aggregate data from multiple stakeholders in the healthcare industry. In the architectural framework depicted in Fig. 2, we take the best practices, techniques and technologies from the papers we reviewed in this study and integrate them into an architectural framework that supports systematic surveillance and performance management of quality of care. The target is to ensure that our framework supports heterogeneous data sources – data streams, batched data and connectors for various forms of data and process integration. The framework also addresses the challenges of supporting various healthcare data interoperability standards, as well as the high-dimensional and longitudinal nature of personal health information (PHI), and the need for privacy and confidentiality protection.

Data processing is done in 3 phases – Aggregation, Transformation, and Analytics. For the Aggregation phase, we assume that the data sources are heterogeneous, streaming and high variety – as is typical of big data. Data at this stage contains a large amount of PHI and requires storage-level encryption if analysis cannot be done in real-time. To address semantic interoperability issues, and eliminate ambiguities in data elements from various sources, semantic matching of data attributes must be done to translate data elements to a common data ontology. Subsequently, profile matching is done to match data streams to the relevant actors in the system. Data analysts would be responsible for updating semantic metadata registries, defining and uploading ontology and semantic mapping rules.


For the Transformation phase, data is prepared for storage and retrieval through data conditioning, anonymization and other needed transformations for big data analytics. At this point, personally identifying information (PII) in the aggregated data is either removed or anonymized before analytics models are applied to the data. A major bulk of these transformations would be triggered through fact and context-based rules engines that trigger various transformation processes based on the type, nature or content patterns of an incoming data stream.

Fig. 2. Cloud-based Big Data Framework for Systematic Performance Management of Quality of Care (QoC)

In the Analytics phase, continuous data mining for associations and correlation is run to create patient clusters or to detect strange or new behaviors or patterns in the data. In this architecture, data analysts can define elements of the data feed that are interesting for analysis using a set of big data analytics templates. For example, the stream of heart attacks to hospital emergency rooms (ER), stream of stroke admission to the ER, etc.

Healthcare providers could be mandated to provide individual data streams for different analytics buckets. In our model, each analytics streams would have a separate storage for data aggregation and continuous mining. Key performance management indicators can be set up to trigger alerts when interesting patterns come up based on the parameters of the various analytics feed.

Finally, it is important that data analytics reports and indicators are fed back to healthcare stakeholder to influence decision making and healthcare governance. These alerts can be in the form of emails, analytics reports, data push to remote services, and updates to dashboards for various health providers. Operational systems can subscribe to advisory alerts that help guide physicians, nurses and other care providers in their day-to-day decisions on patient care, therefore improving the quality of care.

4. Future Works

Our framework depicts the building blocks for implementing a surveillance and performance management framework for continuous data aggregation of disparate data, semantic attribute matching, patient profiling, and


various data transformation tasks. Extracted and conditioned data is then pushed to databases for subsequent data analytics processes like clustering, classification, and determination of various associations between health care data attributes in support for prescriptive and predictive analysis needs. Implementation will depend on the choice of the cloud provider and big data analytics platforms supported by such providers. Web services will play a critical role in orchestrating the various components of this system.

Our framework also addresses some of the challenges to implementing big data analytics applications in healthcare such as semantic and process interoperability, support for legacy systems, maintaining privacy and confidentiality of the patient through anonymization and bridging skillset gaps by providing pre-packaged analytics modules that can be activated dynamically by business users. We will expand upon the components of the framework and provide further implementation details in subsequent publications.

5. Conclusions

Healthcare system surveillance and performance management require collaboration amongst healthcare stakeholders and an infrastructure that can scale in response to data provided and the analytics needs of the healthcare system. Cloud computing provides the infrastructure to support dynamic resource allocation, and provision on a massive scale while our framework then provides an architectural framework to support wide-scale surveillance.

We do recognize that legislation plays a critical role in the adoption of cloud-based applications in healthcare. It is, therefore, important that privacy laws like HIPAA32 for the US and PIPEDA33 for Canada, include provisions to support cloud-based data aggregation and analytics. Big data applications will need to support some level of anonymization capability to guard against privacy breaches and vulnerabilities.

Finally, systematic performance management requires regional and national health authorities to mandate the healthcare stakeholders to provide data streams to an aggregation system. They also need to consume and incorporate analytics feeds into their business processes.

References

1. Sabooniha N, Toohey D, Lee K. An evaluation of hospital information systems integration approaches. In: Proceedings of the International Conference on Advances in Computing, Communications and Informatics - ICACCI ’12. New York, New York, USA: ACM Press; 2012:498.

2. Sun J, Reddy C. Big data analytics for healthcare. In: 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM, 2013. ACM Digital Library; 2013:1525-1525.

3. Li Y, Guo Y. Wiki-Health: From Quantified Self to Self-Understanding. Futur Gener Comput Syst. August 2015. 4. Raptopoulos A, Xydis S, Soudris D. Acceleration of Big Bio-Data: The AEGLE Approach. In: Sano K, Soudris D, Hübner M, Diniz PC, eds.

Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol 9040. Lecture Notes in Computer Science. Cham: Springer International Publishing; 2015:531-541.

5. Ochian A, Suciu G, Fratu O, Voicu C, Suciu V. An overview of cloud middleware services for interconnection of healthcare platforms. In: 2014 10th International Conference on Communications (COMM). IEEE; 2014:1-4.

6. Bhaskaran S, Suryanarayana G, Basu A, Joseph R. Cloud-Enabled Search for Disparate Healthcare Data: A Case Study. In: 2013 IEEE International Conference on Cloud Computing in Emerging Markets (CCEM). IEEE; 2013:1-8.

7. Amato F, Mazzeo A, Moscato V, Picariello A. A Framework for Semantic Interoperability over the Cloud. In: 2013 27th International Conference on Advanced Information Networking and Applications Workshops. IEEE; 2013:1259-1264.

8. Furht B, Escalante A. A Handbook of Cloud Computing. In: A Handbook of Cloud Computing. ; 2010:3-19. 9. de la Torre-Díez I, González S, López-Coronado M. EHR Systems in the Spanish Public Health National System: The Lack of Interoperability

between Primary and Specialty Care. J Med Syst. 2013;37(1):9914. 10. Biswas S, Akhter T, Kaiser MS, Mamun SA. Cloud based healthcare application architecture and electronic medical record mining: An

integrated approach to improve healthcare system. In: 2014 17th International Conference on Computer and Information Technology (ICCIT). IEEE; 2014:286-291.

11. Moumtzoglou A. Cloud Computing Applications for Quality Health Care Delivery. In: Moumtzoglou A, Kastania AN, eds. Cloud Computing Applications for Quality Health Care Delivery. IGI Global; 2014:284-301.

12. Wang L, Alexander CA. Big Data in Medical Applications and Health Care. Am Med J. 2015;6(1):1.


13. Srivathsan M, Arjun KY. Health Monitoring System by Prognotive Computing Using Big Data Analytics. Procedia Comput Sci. 2015;50:602-609.

14. Big Data. Gartner. http://www.gartner.com/it-glossary/big-data. Published 2015. Accessed December 12, 2015. 15.Orit L (Microsoft). Big Data Ecosystem Reference Architecture.; 2013. http://bigdatawg.nist.gov/_uploadfiles/M0015_v1_1596737703.docx. 16. Bohlouli M, Merges F, Fathi M. Knowledge integration of distributed enterprises using cloud based big data analytics. In: IEEE International

Conference on Electro/Information Technology. IEEE; 2014:612-617. 17. Mathew PS, Pillai AS. Big Data solutions in Healthcare: Problems and perspectives. Innov Information, Embed Commun Syst (ICIIECS),

2015 Int Conf. 2015:1-6. 18. Jayapandian C, Chen CH, Dabir A, Lhatoo S, Zhang GQ, Sahoo SS. Domain ontology as conceptual model for big data management:

Application in biomedical informatics. In: Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol 8824. Springer Verlag; 2014:144-157.

19. SNOMED CT: The Global Language of Healthcare. http://www.ihtsdo.org/snomed-ct. Published 2015. Accessed December 12, 2015. 20. LOINC: A universal code system for tests, measurements, and observations. https://loinc.org/. Accessed December 12, 2015. 21. International Classification for Diseases. WHO. http://www.who.int/classifications/icd/en/. Published 2015. 22. DICOM: Digital Imaging and Communications in Medicine. http://dicom.nema.org/. Published 2015. Accessed December 12, 2015. 23. Gubbi J, Buyya R, Marusic S, Palaniswami M. Internet of Things (IoT): A vision, architectural elements, and future directions. Futur Gener

Comput Syst. 2013;29(7):1645-1660. 24. Hromic H, Le Phuoc D, Serrano M, et al. Real time analysis of sensor data for the Internet of Things by means of clustering and event

processing. Commun (ICC), 2015 IEEE Int Conf. 2015:685-691. 25. Hiremath S, Yang G, Mankodiya K. Wearable Internet of Things: Concept, architectural components and promises for person-centered

healthcare. In: Wireless Mobile Communication and Healthcare (Mobihealth), 2014 EAI 4th International Conference on. ; 2014:304-307. 26. Ji Z, Ganchev I, O’Droma M, Zhang X, Zhang X. A cloud-based X73 ubiquitous mobile healthcare system: design and implementation.

ScientificWorldJournal. 2014;2014:145803. 27. AbuKhousa E, Mohamed N, Al-Jaroodi J. e-Health Cloud: Opportunities and Challenges. Futur Internet. 2012;4(3):621-645. 28. Spruijt-Metz D, Hekler E, Saranummi N, et al. Building new computational models to support health behavior change and maintenance: new

opportunities in behavioral research. Transl Behav Med. 2015;5(3):335-346. 29. Andry F, Ridolfo R, Huffman J. Migrating healthcare applications to the cloud through containerization and service brokering. In:

HEALTHINF 2015 - 8th International Conference on Health Informatics, Proceedings; Part of 8th International Joint Conference on Biomedical Engineering Systems and Technologies, BIOSTEC 2015. SciTePress; 2015:164-171.

30. Donnelly N, Irving K, Roantree M. Cooperation across Multiple Healthcare Clinics on the Cloud. Vol 8460. (Magoutis K, Pietzuch P, eds.). Berlin, Heidelberg: Springer Berlin Heidelberg; 2014.

31. Mendelson DS, Erickson BJ, Choy G. Image Sharing: Evolving Solutions in the Age of Interoperability. J Am Coll Radiol. 2014;11(12):1260-1269.

32. Summary of the HIPAA Privacy Rule. http://www.hhs.gov/ocr/privacy/hipaa/understanding/summary/index.html. Published 2015. Accessed December 12, 2015.

33. PIPEDA: The Personal Information Protection and Electronic Documents Act. Office of the Privacy Commissioner of Canada. https://www.priv.gc.ca/leg_c/leg_c_p_e.asp. Accessed December 12, 2015.