Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

LIFE: International Journal of Health and Life-Sciences

ISSN 2454-5872

Al-Badarneh et al.

Regular Issue Vol. 2 Issue 3, pp. 26-37


Date of Publication: 04th January, 2017
DOI-https://dx.doi.org/10.20319/lijhls.2016.23.2637

AN ADAPTIVE ROLE-BASED ACCESS CONTROL APPROACH


FOR CLOUD E-HEALTH SYSTEMS
Amer Al-Badarneh
Jordan University of Science and Technology, Irbid 22110, Jordan
[email protected]
Hassan Najadat
Jordan University of Science and Technology, Irbid 22110, Jordan
[email protected]
Enas 'Hassan Abu Yabes'
Jordan University of Science and Technology, Irbid 22110, Jordan
[email protected]

Abstract
Securing and protecting electronic medical records (EMR) stored in a cloud is one of the most
critical issues in e-health systems. Many approaches with different security objectives have been
developed to adapt this important issue.This paper proposes a new approach for securing and
protecting electronic health records against unauthenticated access with allowing different
hospitals, health centres and pharmacies access the system, by implementing role-based access
control approach that could be applied smoothly in cloud e-health systems.
Keywords
EMR, Security, Privacy, Role-Based Access Control, Cloud Computing.

1. Introduction
2017 The author and GRDS Publishing. All rights reserved.
Available Online at: http://grdspublishing.org/

26

LIFE: International Journal of Health and Life-Sciences


ISSN 2454-5872

Many healthcare providers recently have assumed some design of electronic health
record systems, in spite of the fact that most of them keep medical records in concentrated
databases in the framework of electronic records, As a rule, a patient may have numerous health
care providers, including medical attendants, doctors, experts, dental specialists, and others..
Likewise a patient may utilize a few medicinal services insurance agencies for various types of
protections.
Currently, every provider has its own database for electronic health records. Exchanging
data between medicinal services institutions is limited and managed by administrative tasks,
sharing information between electronic health systems through various health centers is slow.
Costly and inefficient usability are the biggest difficulties to adopt health systems (Carlos Roli et
al., 2010).

Cloud computing supplies an efficient information technology design to reduce the cost
of e-health systems related to both ownership and repair efforts for various medical operations. It
is generally realized that cloud computing and open standards are imperative techniques to

arrange and manage healthcare whether it is for saving health records, supervision of patients,
administrating diseases or analysis of data. Managing health care systems with clouds will make
revolutionary change in the way of health care operations (Carlos Roli et al., 2010).
In traditional health systems, protecting the privacy of patients information is the
responsibility of the independent Institutions like hospitals, clinics and medical centers. By using
highly information technologies like cloud computing where the system data is distributed in
different storage areas. Patient information can be accessed with patients, physicians and any
other related institutions.
This will force health care industry towards a more open medical information
infrastructure and to upgrade the quality of health systems. In addition, health care industry also
gives a number of new issues in insuring patient medical information to keep it in confidentiality
and availability. An open health care requires complementary security management approaches
and policies to prevent the existence of serious potential violations of security and privacy ( Rui
Zhang, and Ling Liu ,2010).

2017 The author and GRDS Publishing. All rights reserved.


Available Online at: http://grdspublishing.org/

27

LIFE: International Journal of Health and Life-Sciences


ISSN 2454-5872

The main requirements for securing electronic health records are first, effective access
control model that satisfies expressiveness and flexibility needed by privacy policies, second, the
system should be operated smoothly in various environments and the system should contain and
manage rich meta information (Ming Li, et al 2010).
Operative administration of electronic health records is a very complicated and critical
problem. Patient protection issues, at the same time with dangers that could uncover medical data
insures the importance for privacy and security mechanisms combined with medical systems and
could be implemented toward a different collection of dissimilar systems and networks, a shared
electronic health record simply a complicated combination of critical information, containing
patient particulars, medical histories, checking results, etc. There is also an important necessity
for protection models that respond to legitimate and administrative approaches, while at the same
time guaranteeing that entrance to delicate data is restricted and committed just to those beings
that have a permissible need to know approvals allowed by patients. For example, a patient may
choose to obviously conceal his medical information from general medical information sharing
session unless a certain treatment choice is implied. Its important to look for safe and sharable
policy that permits patients to rapidly and smoothly grant a collection of medical sub
organizations to access their critical information partially or in whole.
Role-Based Access Control (RBAC) is one of the access control mechanisms to classify
and determine many of the requirements of security administration in distributed information
systems. So this paper concentrates on how to implement role based access control in cloud
electronic health systems. There are some essential benefits in electing RBAC for ensuring
patient privacy and security. First, roles, those are an important indirection between users and
authorizations, immediately arranges healthcare institutional positions, like doctors, physicians,
and nurses.
The access rules should support important procedures to achieve distinct jobs tasks. For
example, the system should confine read, duplicate, and print work on basic data to just the
imperative stuff for a specific period.
Access control need to recognize who has entry to the information, which type of access
is permitted, what works that are provided, under what circumstances, and for what period This
2017 The author and GRDS Publishing. All rights reserved.
Available Online at: http://grdspublishing.org/

28

LIFE: International Journal of Health and Life-Sciences


ISSN 2454-5872

paper presents a role based access control approach, that takes into account the function in hand
and the turn of the user and writing analysis on the patients record on the other hand.
The paper is organized as follows. Section 2 illustrates the motivation of the research
.Section 3 reviews related work. Section 4 introduces the proposed approach. Section 5 gives
more discussions about the proposed approach. Section 6 concludes the paper.

2. Motivation
In spite of the fact that forcing privacy by RBAC could be easily deployed in systems
already adapting RBAC. The embracing of role based access control in cloud e-Health systems
requires some essential additional requirements. Healthcare is a complicated environment which
naturally includes multiple domains. Also, not only the content of electronic medical records
needed to be protected and secured but also some meta information about electronic medical
records, such as owners of records, integrity constraints, privileges granted to users, default
column values. So role based access control mechanism should be customized in an efficient
manner to guarantee a burst secure and protected environment.

3. Related Work
A number of solutions have been proposed to address security and access control issues
associated with e-health systems. (Lhr, Sadeghi & Winandy, 2010) proposed a model where
electronic medical records are managed by health specialists only. In most countries this needs
various allowable prerequisites and an obvious differentiation between personal health records
and electronic health records, so infrastructures that contain electronic health records are more
complicated than simple e-health cloud model which clarified previously. The common needs
that proposed are still the functional and meaningful of the information saved in the electronic
health records. The electronic health records are designed, directed, and administrated by health
care providers, and can be shared across the central server in the cloud with other health
specialists. Also the health care providers have the ability to access billing services that manage
their accounting with the health insurances of the patients.

2017 The author and GRDS Publishing. All rights reserved.


Available Online at: http://grdspublishing.org/

29

LIFE: International Journal of Health and Life-Sciences


ISSN 2454-5872

According to the proposed approach by (Josh Benaloh et al. 2009) the patient may
determine to give his dental specialist access to both the dental records feature and the
fundamental medical information feature. This will permit to the dentist to read all data related to
the dental clinic visits, dental x-rays, surgeries and medications. The dentist will not be able to
decrypt any of the health records of the patient, or his personal data, for instance, the server that
saves the health information will not be able to get the secret key, or any of the sub keys given to
the doctor, so he will not have the ability to decrypt the data.
The hierarchical model that is proposed by (Josh Benaloh et al. 2009) is easily
expandable; the patient and possibly other clients to whom the patients give the appropriate
authentications can append additional sub features during any existing group. So within the
medications group, the patients dentist might append a new feature for doping or visiting on a
certain date. Once the patient gives the specialist authorizations to his drugs, if his dentist adds a
new feature, all records in that subcategory will be spontaneously accessible to the doctor.
According to this, the patient can easily give access to a group, without the need to know all the
types of files that might be involved in it.
The same thing is for doctors; they can add sub features with random names, without
support from the patient. This will be certainly valuable if we can't foretell the names of all
probable sub features, for example if a specialist needs to append a feature for a new kind of test,
or if features are assorted by visit dates. The drawback of this model is that there is only one
method in which can split the records, and if we need to grant access privileges according to
something else, for example, record sort or receptiveness of information, we should take a
gander at all the low-level characteristics included, and give a different decryption key for each.
A specific supposition is taken about the configuration of the patients' record; the patient's record
is stored as an accumulation of passage, where every passageway contains the name of the
document, additionally the name of the minimal feature including that record which the patient
can use to get to the record, and the encrypted copy of the file (Josh Benaloh et al. 2009).
(Jacques Wainer et al., 2006) addressed the Integrity confidentiality and control issues
with the following key features:

2017 The author and GRDS Publishing. All rights reserved.


Available Online at: http://grdspublishing.org/

30

LIFE: International Journal of Health and Life-Sciences


ISSN 2454-5872

Availability: the electronic health records should be available when the health specialists
require it, so all attention in making the system firm and reliable is important.

Up-to-datedness: the electronic health records should include all of the recent related
information with taking into account the patients health; so there should be no crucial
postponement from when data is entered into the record and when it becomes obtainable
to different specialists. If the health specialist decides some medication to the patient, that
information should be contained in the electronic health record directly, so if the patient
asks another health specialist for another treatment, then the information must be
obtainable.

Usability: In spite of the fact that usability is not considered as an integrity issue, it is also
important to perfect utilization of the electronic health records, the health specialist
should not have to read through all of the patients records to determine the usability
conditions.

No access rights to the patient: the patient has no privileges to access or modify the
electronic health record, and the patient can only depute access privileges to his own
records to health specialists.

Emergency Access: There are believable cases in where health specialists can access a
patients record without his predetermined delegation can happen obviously in
emergency cases, if the patient comes to an emergency clinic then the health specialist
must be able to access to his electronic health records.

Implicit acceptance of health organization structure: by granting the patients electronic


health record to the health specialist, the patient according to this may accept any deputy
needed for health specialist.

Limited read access for public health: legitimate and professional humans may have a
restricted and unknown read access to the electronic health records without taking into
account the patients agreement.

2017 The author and GRDS Publishing. All rights reserved.


Available Online at: http://grdspublishing.org/

31

LIFE: International Journal of Health and Life-Sciences


ISSN 2454-5872

The authors (Ammar Alkassar et al., 2011) improved a usable and secure end-client sample
that can save basic medicinal information from being gotten to or adjusted by illegal clients, they
called it MediTrust. The following objectives are defined to be achieved:

Preserving medical data that are manipulated on the same developed model with other
functions.

Building security architecture that securely divides the data of different operations.

Supplying an efficient and smooth model that does not require any load in the normal
process of health specialists.
The scenario that is applied here is that the doctor utilizes a computer system to prepare

electronic health records of patients that are registered on a centric server. A similar PC
framework is utilized to send information to a human services accounting and billing server, the
PC is likewise used to connection to alternate services , for example web sites on the internet.
Privacy domains are built for medical data as a technical magnitude to help in the execution of
privacy and data protection mechanisms.
The client model, like desktop or notebook computers, should divide execution
situations for applications into discrete scopes that are distinguished from each other. The
information inside a security space and the area foundation ensures that only the legitimate
customers can join this area. In addition, data infiltration from the domain is prohibited by the
security model and the domain infrastructure. A similar framework ought to be able to be used
for different operations that are completely isolated. So, the importance of MediTrust is on the
execution of secure customer structures that can be used not just to access basic health records
and joined accounting data securely, but also working frameworks and programming that are
completely isolated from the basic information and the medicinal services customer application
(Ammar Alkassar et al., 2011).
A set of roles are proposed (Hema Narayanan& Mehmet HadiGne , 2011) starting by
implicit access control and tasks to stand by explicit access control. A lessee in the cloud system
has various users. Every user is given a role; roles are given to a workflow or non-workflow
functions, and functions are given to permissions. Users with an assigned role are able to run
different tasks through workflow and non-workflow functions given to their role. Authorizations
2017 The author and GRDS Publishing. All rights reserved.
Available Online at: http://grdspublishing.org/

32

LIFE: International Journal of Health and Life-Sciences


ISSN 2454-5872

are given to roles according to their functions and assigned authorizations and changes
dynamically according to the current function.
Authorization specifies who can do which functions with what role under what
circumstance. It is determined by the states or rows (U, R, T, P, and C) where U is the set of
usersui, R is the set of roles ri, T is the set of tasks ti, P is the set of permissions pi, and C is the
set of constraints ci. For instance, the tuple (Jack, doctor, read patient information, read, daytime
and office location) determines the way that Jack as a doctor reads patient information from
office during office hours.
The factors important to healthcare cloud systems are tenant: is customer like clinic
hospital or pharmacy in a health care system, user, task, information resource, business rule,
permission, session: is a map of different rules for certain user.

The factors that are important to healthcare cloud systems are (Hema Narayanan
& Mehmet Hadi Gne, 2011): Work Flow, Business Rules: is a set of regular activities of users
that the organization follows, Least Privilege: the essential privileges assigned to each user,
Least Separation of Duty separating the obligations regarding assignments in a work process
between various members and ensure against trick activities ,Tenant_User Assignment : A cloud
system has many tenants each with different users, User-Role Assignment: A user can be
assigned one or more roles, And the same role can be given to different users.
Task-Role Assignment: A role can have multiple tasks and a task can have multiple
roles. Permission-Task Assignment: Tasks are given permission to be executed, Task-Workflow
Assignment: the duties that belong to approval classes are assigned to a particular workflow,
Least Privilege: the essential privileges that is assigned to the user.
The significant imperatives in getting to control are Least Privilege: the most essential
privileges that every user have in the cloud to see a certain information, Dynamic and Static
Separation of Duty: dynamic separation of duty prevent Simultaneous execution of at least two
sole assignments by a similar role, Delegation: delegation means assigning a role by another
user, Spatiotemporal Constraints; Users location and time is considered for giving access to a

2017 The author and GRDS Publishing. All rights reserved.


Available Online at: http://grdspublishing.org/

33

LIFE: International Journal of Health and Life-Sciences


ISSN 2454-5872

task. When a tenant registers in the system, its office and clinic locations are registered for more
temporal checking (Hema Narayanan& Mehmet HadiGne , 2011).

4. Proposed Approach
The proposed approach supplies access control decisions based on tasks associated with
authorized evaluators. For instance, a doctor at a certain hospital has features including his title,
appointments, specialty; every preserved record contains implicitly a complicated access control
approach to specify which users can access the record. This approach helps clients who do not
have recognized functions, like medical scientists. For instance an individual may be given
access only to records created within certain time period, with a particular record kind.
The proposed approach includes the following entities: Users, Roles, Permissions,
Preconditions, Data, and Metadata.

The User: is the human being like doctor, physician, nurse and more.

The Role: is a specific job or a specific task in a job.

Users are granted roles based on their credentials and responsibilities in the organization.

Roles are granted to users by many to many relationships: The role structure is hierarchal
where every role has 0 to 1 parent role and 0 to N leaf roles; whenever the user granted
certain role, the parent role will be implicitly granted to the user but not the leaf roles.

Permissions are the types of privileges granted to Role: query, modification, insert,
administrative roles, executing functions, permissions are granted to roles; a many to
many relationships applied between the permission and roles. An implicit grant of
permission is executed whenever a certain function that is granted to certain role contains
query of modification permissions.

Preconditions are predetermined validations that are used to decide if a certain user can
access a role or not; this property can be used in an efficient way if for example a certain
user have the authorization to grant particular roles to another users, according to the
preconditions the system will determine if the user needs the roles he granted or not
according to the user features and tasks.

2017 The author and GRDS Publishing. All rights reserved.


Available Online at: http://grdspublishing.org/

34

LIFE: International Journal of Health and Life-Sciences


ISSN 2454-5872

Data are the objects that are stored as resources in the cloud such as tables, images,
columns, rows, raw files and X-ray photographs.

Meta Data are the information about the data such as definition, allocated and utilized
capacity estimate for information objects, default column values, integrity constraints,
names of and privileges granted to users, auditing information and more.
In the proposed approach the meta data can be combined with roles to make assigning

roles more efficient and secure by assigning roles to users according to the data description of
objects, for example, the pediatrician can have access to information related to patients with ages
less than 12 years old, and gynecologist can access records related to female patients only.

5. Further Discussion
So far this paper addressed a role based access approach that can be implemented easily
by relying on the structure of the health system as various roles will interact with the
information. Access roles to resources must be given to users according to their tasks. For
instance, a specialist ought to be allowed access to restorative history of a patient.
Cloud e-Health systems should allow access to resources only when it is required and
protect users from unintended problems. Furthermore, access policies should support core
operations to achieve job tasks; the system should manage read, copy, and print operation on
critical information to only the specialized users according to their tasks. Access control should
determine who has access to information, which type of accesses are permitted, what duties are
given and under what circumstances.

6. Conclusion and Future Work


Patient privacy and data security can be achieved using role based access control
techniques by determining the main entities for the model, and assigning each user one or more
tasks after explicit and implicit grants of the predetermined roles for the legal users, the
advantage of applying implicit grant is to simplify the complicated structure of the role based
access control hierarchy.

2017 The author and GRDS Publishing. All rights reserved.


Available Online at: http://grdspublishing.org/

35

LIFE: International Journal of Health and Life-Sciences


ISSN 2454-5872

The future work is to improve the role based model for adapting access to electronic
health records in the situations of emergency where access typically occurs in an ad-hoc and
spontaneous way.

REFERENCES
Rui Zhang, and Ling Liu. (2010). Security Models and Requirements for Healthcare Application
Clouds, IEEE, and 3rd International Conference on Cloud Computing, 268-275. doi:
10.1109/CLOUD.2010.62.
Josh Benaloh, Melissa Chase, Eric Horvitz, and Kristin Lauter (2009). Patient Controlled
Encryption: Ensuring Privacy of Electronic Medical Records, in the ACM workshop on
Cloud computing security, 103114. doi: 10.1145/1655008.1655024.
Hans Lhr, Ahmad-Reza Sadeghi, Marcel Winandy (2010). Securing the E-Health Cloud, in the
ACM

International

Health

Informatics

Symposium,

220-229.

doi.

10.1145/1882992.1883024.
Ammar Alkassar, Biljana Cubaleska, Hans Lhr, Ahmad-Reza Sadeghi, Christian Stble, Marce
Winandy. Medi Trust (2011). Secure Client Systems for Healthcare IT to Protect Sensitive
Data of Patients, The International eHealth, Telemedicine and Health ICT Forum,
Luxemburg, 4, 385-389.
Jing Jin, Gail-Joon Ahn, Hongxin Hu, Michael J. Covington, Xinwen Zhang (2010). PatientCentric Authorization Framework for Electronic Healthcare Services, 30,116-127.
doi:10.1016/j.cose.2010.09.001.
Hema And Al Jayaprakash Narayanan, Mehmet Hadi Gne (2011).Ensuring Access Control in
Cloud Provisioned Healthcare Systems ,In the Consumer Communications and Networking
Conference (CCNC), IEEE.doi:10.1109/CCNC.2011.5766466.

2017 The author and GRDS Publishing. All rights reserved.


Available Online at: http://grdspublishing.org/

36

LIFE: International Journal of Health and Life-Sciences


ISSN 2454-5872

Ming Li, Shucheng Yu, KuiRen, and Wenjing Lou (2010). Securing Personal Health Records in
Cloud Computing: Patient-Centric and Fine-Grained Data Access Control in Multi-owner
Settings. doi:10.1007/978-3-642-16161-2_6
Jacques Wainer, Carlos Jose Reis de Campos, Daniel Sigulem (2006).Security requirements for a
lifelong electronic health record system based on non-standard ethical principles,
Department of Health Informatics. 2, 160-165. doi: 10.2174/1874431100802010160.
Carlos Oberdan Rolim, Fernando Luiz Koch, Carlos Becker West Phall, Jorge Werner, Armando
Fracalossi, Giovanni Schmitt Salvador (2010). A Cloud Computing Solution for Patients
Data Collection in Health Care Institutions. 95-99. doi: 10.1109/eTELEMED.2010.19.

2017 The author and GRDS Publishing. All rights reserved.


Available Online at: http://grdspublishing.org/

37

You might also like