=Paper=
{{Paper
|id=Vol-2623/paper12
|storemode=property
|title=Medical Information System Development: Practical Aspects
|pdfUrl=https://ceur-ws.org/Vol-2623/paper12.pdf
|volume=Vol-2623
|authors=Kateryna Yalova,Kseniia Yashyna,Nataliia Savina
|dblpUrl=https://dblp.org/rec/conf/intelitsis/YalovaYS20
}}
==Medical Information System Development: Practical Aspects==
https://ceur-ws.org/Vol-2623/paper12.pdf
Medical Information System Development: Practical
Aspects
Yalova Kateryna1[0000-0002-2687-5863], Yashyna Kseniia1[0000-0002-8817-8609]
and Savina Nataliia2[0000-0001-8339-1219]
1Dniprovsk State Technical University, Kamyanske, Ukraine
2 National University of Water and Environmental Engineering, Rivne, Ukraine
yalovakateryna@gmail.com, yashinaksenia85@gmail.com,
n.b.savina@nuwm.edu.ua
Abstract. This paper deals with the development of the medical information
system for automation blood transfusion station operations such as: accounting
and availability monitoring of qualified blood in the storage, accounting of do-
nors, monitoring of temporary and permanent blood-donating termination. The
functional and software requirements are listed in the paper. Data domain mod-
eling and system design were obtained on the base of the Kamyanske Blood
Transfusion station (Dnipropetrovsk region) data and input requirements. The
target audience of the information system are blood transfusion station employ-
ees. Object-oriented data domain analysis allows describing the functional fea-
tures of a specific data domain in terms of classes, class objects, and their inter-
actions. The results of data domain analysis are presented in the mathematical
form. The scientific significance consists in developing an approach to utiliza-
tion of system analysis, object-oriented and functional modelling of the data
domain for the purpose of architectural design of the specific medical infor-
mation system. Practical relevance of the proposed solutions results in creation
of a multiple-use software product that is a source of and a tool for further de-
velopment. Interface design has been focused on determining a set of form con-
trols providing minimum user’s physical effort towards required result and
maximum level of protection against user’s mistakes. The example of standard-
ized and unified interface prototype is also presented.
Keywords: medical information system, blood transfusion station, object-
oriented data domain analysis
Introduction
Rapid development of information technologies results in their introduction into vari-
ous areas including information processes automation in health-care. The main pur-
pose of information systems is acceleration and optimization of information processes
involving data acquisition, storage, and processing. Information systems intended for
automation of health care institutions, medical personnel workplaces, and business
processes conducted during various phases of medical care are called medical infor-
Copyright © 2020 for this paper by its authors. Use permitted under Creative Commons
License Attribution 4.0 International (CC BY 4.0). IntelITSIS-2020
mation systems (MIS) [1]. The MIS is a hardware-and-software integrated complex as
a set of software tools and databases, the most important functions of this system are
improvement of medical care and optimization of management decision-making in
health care system [2].
Research topic of this article is the process of information system for blood trans-
fusion station development as a special type of MIS. Proposed decisions describe the
process of donors registering and accounting, availability of blood and its compo-
nents, and donation prohibition at blood transfusion stations. The main phases of au-
tomation of blood transfusion station operation are as follows:
• data domain analysis required to determine functional features of interactions be-
tween actors, characteristics of static and dynamic objects, document structures and
document flow schemes;
• design of database structure, system architecture, and user interface prototype ac-
counting for user profiles and data access rights;
• software implementation of design solutions;
• software debugging and testing;
• information system commissioning.
The scientific significance consists in developing an approach to utilization of system
analysis, object-oriented and functional modelling of the data domain for the purpose
of architectural design of a specific MIS.
1.1 Related Works
Introduction of web-oriented and LAN-based MIS’s made it possible to automate
medicine accounting processes in drug stores, create digital references and ontologies
for medicines and diseases, develop digital patient records, create robotized drug
stores, implement remote interaction between doctors and patients by means of tele-
medicine tools, etc. Such researchers as O. Chaban, O. Boyko, K. Kopnyak,
V. Stepanov analyze in their papers [3-5] the current state of MIS introduction in
Ukraine and other countries. They address the issues of health care digitalization, as
well as the problems linked to creation of a uniform medical-informational space. The
papers by researchers S.V. Tymchyk, S.M. Zlepko [6], S.V. Kostishyn,
T.I. Ovcharuk, A.A. Ovcharuk compare such MIS’s as: C-hospital, Digital Hospital,
TherDep, and Medialog against a number of criteria. One of the most prospective
problems nowadays is development of expertise and advisory diagnostics systems for
supporting treatment decisions, this problem is addressed by the following Ukrainian
researchers [7-11]: T. Hryhorova, M Iepik, N.I. Melnikova, O.V. Kravchenko, K.V.
Steblina and others. Artificial intelligence and neural network technologies are widely
used in development of automated medical diagnostics systems. In addition to MIS
added to the Ukrainian nationwide health care automation projects E-health [12] such
as: Helsi, Medics, Medstar, nHealth, SimplexMed, Health24, Askep.net, Skarb,
Ukrmedsoft, MedInfoService, Medikit, MedAir, information technologies can also be
utilized in personalized web-oriented services or mobile applications providing user-
patients with useful information.
Considering a hard political situation in the country, blood services have become one
the most important strategic health care services, digitalization of these services is one
of priorities for efficient management of blood assets and donors accounting in general.
Automated information systems provide means for integrated digitalization of blood
collection, examination, storage, and distribution, significantly improving production
safety and efficiency, introducing additional quality control stages, and allowing fast
informing the management. Paper [13] describes digitalization and blood bar-codding at
Zaporizhzhya Regional Blood Transfusion Station. The authors of paper [14] have de-
veloped a blood donor companion system in the form of Android app used for blood
donation planning accounting for individual nuances and health state, as well as for
getting information regarding actions before and after donation.
Despite the large number of research papers devoted to the development of MIS,
automation of processes related to the provision of medical services remains an actual
scientific and practical problem especially important for Ukrainian medical system.
Proposed Data Domain Model
2.1 Functional and Software Requirements
The target audience of the MIS are blood transfusion station employees. For imple-
menting the distributed data access mechanisms, the system can be divided into the
following functional modules:
• record-keeping module dealing with processing the blood-donors records;
• doctoral activities module intended for keeping the results of examinations and
prescriptions;
• laboratory module responsible for registering the results of donors’ analyses;
• module for blood accounting, passporting, and quarantine issues.
The functional modules correspond to the organizational structure of the blood trans-
fusion station, allowing to automate the following information processes:
• accounting and availability monitoring of qualified blood in the storage, which
includes accounting of all blood donations, considering the dates of these dona-
tions, as well as donor blood type and rhesus. Blood and its components are ac-
counted per blood type and rhesus using absolute and relative values;
• accounting of donors including recording the data regarding all their visits, keeping
the electronic records of donors and reserve donors;
• monitoring of blood-donating termination, both temporary and permanent, consid-
ering the time intervals of such pauses;
• report generation regarding donors with absolute contraindications, which are not
allowed to donate blood or its components.
Input functional requirements were formed on the base of the Kamyanske Blood
Transfusion station (Dnipropetrovsk region) data. The generalized main functional
and software requirements for the application package of the MIS under development
are as follows:
• all data, accompanying documents, and reports should be developed based on the
information stored in the database of the MIS;
• the user interface should be implemented as a Window – Image – Menu – Pointer
(WINP) interface in the form of hierarchical menu, with input or search fields,
functional buttons triggering specific actions when pressed, and tabular representa-
tion of the data from the database;
• the user dialogues should allow efficient users training regarding general principles
of system interaction, and be available in utilization. The software product should
be experienced user-oriented;
• the MIS should provide the means for adding, modification, erasing, and searching
for reference and input data, as well as for output data generation;
• depending on data access level, the MIS should provide the medical personnel with
the following information: blood amount available in the storage; information on
active and healthy donors; information on donors who are not allowed for blood
donation and the reasons for such temporary/permanent exclusion; the system
should allow producing printed forms of medical documentation and certificates.
The MIS should process the data in multi-user mode utilizing the client-server tech-
nology, as well as in interactive mode, i.e. generate the output information by execut-
ing dynamic inquiries to the database. A database designed using the object model of
the data domain should be an integral part of the MIS.
2.2 Object-Oriented Data Domain Analysis
According to the software engineering standards the life cycle of the MIS, like for any
software product, consists of five phases: data domain analysis, design, software im-
plementation, testing and operation. An object-oriented analysis was used for analyz-
ing the data domain. Object-oriented analysis of the data domain allows describing
the functional features of a specific data domain in terms of classes, class objects, and
their interactions as a model of information, labor, financial, and material flows [15].
Data domain data were analyzed and presented by the data domain model, which
mathematical form, can be represented as follows:
D=〈K, R, BP〉, (1)
where D is a data domain limited by the requirements of the task to be automated;
K={k1,…,kN} is a multitude of classes of the data domain; R={r1, r2, r3} is a multitude
of relations type between classes: «one-to-one», «one-to-many» and «many-to-many»
connections, rξ= {1 – 1:1; 2 – 1:N; 3 – N:N}; BP={bp1,…,bp3} is a multitude of data
domain business processes. Each data domain class is defined as:
K=〈O, P, TK, TP〉, (2)
where О = { o1,…,oN }is a multitude of instances of the ki class; P = { p1,…,pN } is a
multitude of properties of the ki class, TK = { tk1, tk2, tk3 } is a multitude of the ki
class types: static class type 1, which is implemented in the MIS as simple data do-
main directories; static class type 2, which is implemented in the MIS as multi-table
data domain directories; dynamic class type 3, which is implemented in the MIS as
data domain input documents. TР = {tp1, tp2, tp3} is a multitude of the ki class pi
property types: constant-by-convention tp1 → const; dynamic tp2; calculation-related
ones tp3. Based on (1)-(2), it is possible to specify predicates that describe the struc-
ture of data domain classes and business-processes. The predicate, named
class_type(K, TK), determines a simple correspondence between the ki class and its tkξ
type. A multitude defining this predicate can be described as follows:
Mct= K ×TK ⊃ { (ki, tkξ)}| ∀ ki∈K, ∃! tkξ∈TK. (3)
Based on (3), there is a single tkξ class type for each ki class. This allows to create
classes multitudes of similar type:
KS={ki |class_type(ki,1)}; KM={ki |class_type(ki,2)}; KD={ki |class_type(ki,3)}, (4)
where KS – is a multitude of simple directories; KM – is a multitude of multi-table
directories; KD – is a multitude of input documents. The predicate, named sin-
gle_directories(K, R), sets KS classes. A multitude determining the predicate sin-
gle_directories(K, R) is defined as follows:
Msd= K ×R ⊃ { (ki, r1)}| ki∈KS. (5)
The predicate, named multi_directories(K,K, R), sets KM classes. A multitude de-
termining the predicate is defined as follows:
Mmd= K ×K, × R ⊃ { (ki,kt, rj)}| ki∈KS,kt ∈K, kt ∈KS∪KM, rj∈R, (6)
whereK – is a multitude of the child classes, which is joined with parent class by the
relation type r2 or r3.
In order to fill in dynamic type classes data, it is necessary to use date from simple
and multi-table directories. The predicate, named input_doc(K, R), sets KD classes. A
multitude determining the predicate is defined as follows:
Mid= K × R ⊃ { (ki, rj)}| ki∈ KS∪KM, rj∈R. (7)
KS, KM and KD multitudes joining sets a multitude of the data domain classes МK:
МK=KS∪KM∪KD. (8)
Static type classes of the data domain include those whose property type is constant
by convention. The predicate, named property_type(P, TP), determines a simple cor-
respondence between the pi property and its tрξ type. A multitude determining the
predicate property_type(P, TP), can be represented as follows:
Mpt= P ×TP ⊃ { (pi, tpξ)}| ∀ pi∈p, ∃! tpξ∈TP. (9)
Based on (9), there is the only one type tpξ for each property pi.
The main purpose of static type classes existence in the MIS is decreasing the
number of manual data input operations, replacing manual input with more efficient
data fetching from the database via the elements of the graphical forms. Multiple use
of static type classes during creation of dynamic type instances and their unambigu-
ous representation in the database increases data reliability, as well as data input
speed and correctness. The main static type classes are: the types of donors, health-
care institutions, medical personnel, territorial subordination hierarchy, blood types
and rhesus, temporary donation prohibition, absolute contraindications, etc.
The data of dynamic type classes is the main information source for the database.
The frequency of creating dynamic type instances is unknown and depends on the
environment conditions. The main dynamic type classes are donors, reserve donor
record, blood donation assignment, examination certificate, donor benefit medical
note, etc. The selection of the classes and their properties depends on the task to be
automated. Also, the analysis approach, as well as the end-users of the information
services provided by the software product should be considered. As for the described
data domain, the following roles have been determined that are given different data
access privileges, for the data stored in the database or generated by the MIS: donors,
medical registrars, laboratory doctors, station doctors.
Object-oriented analysis has also allowed decomposition of data domain business-
processes. It made it possible to determine the set of mandatory actions and event
triggering conditions, where their execution is influenced by the environment. The
predicate, named b_process(BP, K), describes interactions between classes joined by
the bpi business-process. A multitude determining the predicate b_process(BP, K),
can be represented as follows:
Mbp= BP ×K ⊃ { (bpi, kj)}| bpi∈BP, kj∈K. (10)
The object-oriented analysis of the data domain and data normalization rules allow
transition to relational database design according to the following flow: data domain
entity → class → database table. Database functions for creation, deletion, storage,
and data search are to be implemented for all entities of the data domain.
2.3 User Interaction Algorithm
Considering the nuances of information flows and the list of functional requirements
set for the MIS, the following algorithm for user interaction has been made up:
1. A potential donor applies to the registry desk of the blood transfusion station where
he/she confirms his/her identity with a valid document.
2. The medical registrar searches for applying person’s record in the database.
3. In case the database contains no such record, the registrar enters the person’s data
into it, namely – information on person’s ID, occupation/study, registered and ac-
tual residential addresses. Otherwise, the MIS returns donor’s record containing the
exhausting information about the donor, e.g. whether he/she has donated in the past
2 months, and whether there are contraindications against donation. After that, the
registrar creates reserve donor record and prints donor form. The fields in the do-
nor record and donor form are automatically filled with the donor information.
4. The Reserve Donor Record is given to the donor, then he/she is directed to medical
examination. Before examination, the donor should fill the form at the backside of
the Reserve Donor Record, the answers should be confirmed with the personal sig-
nature of the donor.
5. Determination of blood donation qualification. The laboratory doctor conducts the
required tests, makes corresponding entries into the Reserve Donor Record, and di-
rects the donor to the doctor who analyzes the donor form and enters information
on examination results. Based on medical test and examination results, the doctor
determines whether the donor is qualified for blood donation. In this case, a medi-
cal examination note is issued that is automatically filled with the information on
examination date, the reason of donation prohibition, prohibition term, examination
results, etc. In case the donor is well-qualified for donation, the doctor enters the
amount of blood to be donated and issues blood donation assignment and medical
note about donation.
6. Blood donation. Having a blood donation assignment, the donor donates blood, the
results should be written to the donor record.
7. Passporting and quarantine storage of the donated blood. Based on the data on
completed donations and complying with the regulations, the donated blood is
quarantine-stored for a specific period, after that the blood well-qualified for utili-
zation in medical institutions is to be pasportised.
The developed data flows diagram, which also describes the algorithm of user interac-
tion with MIS and its components, is shown in Figure 1 within the Business Process
Management Notation (BPMN). BPMN allows to describe graphical notation for
mapping business processes as data flows diagram.
Fig. 1. Data flows diagram (Source: own work)
The pools in the figure 1 display the available actions for each group of data domain
actors. Users accesses to the database for performing data input operations and obtain-
ing the queries results are presented as data flows.
Results
In the course of system implementation three-level Model – View – Controller
(MVC) architecture has been utilized, thus the MIS is composed of data storage, busi-
ness-logic, and visualization levels. Data storage level has been implemented using
relational database mechanisms, with the database designed accounting for data nor-
malization rules. The business-logic level has been implemented as stored procedures
and views, both in the database and at the application level. The visualization level
has been implemented as a set of user forms intended for interaction with the MIS, the
forms being designed according to Window – Image – Menu – Pointer (WIMP)
graphics standard. All user forms have been design using a single style and account-
ing for standardization and unification requirements.
On the base of data domain functional and object models the main components of
the proposed MIS were developed. They are as follow: reference sources (simple and
multi-table directories), input (digital documents) and output (data fetched from the
database) information. The interaction of data domain actors is documented according
to approved forms, such documents are legal evidence of donation. The output infor-
mation is generated based on input parameters of database queries; this information
contains the data fetching results regarding:
• the amount of blood available at the storage per blood types, rhesus, etc.;
• the number of people having absolute contraindications against blood or its com-
ponents donation for a specific period of time;
• the number of people under temporary prohibition of blood or its components do-
nation, as well as the time intervals of such prohibition;
• the total number of donor’s visits to the blood transfusion station regardless of
whether blood donation has been completed;
• the total number of blood donations for each donor.
Usability is one of the main quality criteria for the user interface developed, the crite-
ria it is assessed against are as follows [16]: easy learnability, ergonomics, number of
user mistakes, subjective satisfaction with operation, memorability level. Ergonomic
level of the developed interface has been assessed as a total degree of convenience
that depends on intellectual effort and decision-making speed. The efficient imple-
mentation of interface usability mechanisms influences competitiveness of the appli-
cation compared to existing ones, its marketing attractiveness as a commercial soft-
ware product, and popularity. Taking into account the analysis of existing MIS having
similar functionality, interface design has been focused on determining a set of form
controls providing minimum user’s physical effort towards required result and maxi-
mum level of protection against user’s mistakes. WIMP standard has been used for
user interface implementation. One of characteristic features of this standard is that
user-computer dialog is conducted using windows, graphic menus, cursor, and other
elements [17]: all operations with applications, files, and documents are to be per-
formed in windows; applications, files, documents, devices, and other items are repre-
sented as icons that transform into windows when opened; all operations with the
objects are to be performed using menus that is the main control; the mouse serves as
the main controller. The graphic form for input data document Reserve Donor Record
is shown in Figure 2 as an example of proposed standard interface of the MIS.
Fig. 2. Ukrainian language user interface prototype (Source: own work)
To meet the interface usability criteria, all graphic forms of the digital documents
have been divided into four areas:
1 – search input area for accelerating selection of required values form the database;
2 – MIS reference area for decreasing incorrect inputs and keyboard use;
3 – visualization area for digital document display;
4 – functional panel area where buttons are located, the buttons are used for running
user dialog scenarios.
To satisfy the user with the application, all graphic forms have been designed using
reasonable color scheme and consistent font, the overall appearance decreases user’s
effort as to processing the data displayed on the form. Other features improving inter-
face usability include blocking wrong user actions, automatic input validation, and
use of signal colors.
Practical relevance of the proposed solutions results in creation of a multiple-use
software product that is a source of and a tool for further development. Suggested
practical usefulness of proposed design solutions includes the following:
• Acceleration of information processing during registering and servicing the donors due
to generation and database-storage of reference information regarding data domain;
• Ensuring reliability and display-ability of database data for each donor at any mo-
ment for the purpose of reporting to the management or monitoring authorities;
• Reduction of inefficient manual data processing operations and the number of in-
correct inputs due to automatic input validation;
• Optimization of information processes and improvement of managing decision-
making in general.
Unlike MIS described in [13,14] proposed system provides multiuser distributed data
access and automates information processes from the donor arrival for donation to
determining the suitability of blood for use.
Conclusions
This paper deals with the development of the MIS for blood transfusion stations
based on functional requirements of the Kamyanske Blood Transfusion station
(Dnipropetrovsk region). Accuracy and coordination in medical staff activities in
blood transfusion institutions are of primary importance, since they directly influence
the health, safety, and anonymity of donors and recipients; reducing losses due to
monitoring blood expiration dates; optimization of the number of donations and blood
amounts, as well as efficient distribution of blood among medical institutions.
The actuality of the MIS development is an improvement of donors and recipients’
services, as well as of medical staff and blood services working conditions due to
automation of information processes. The proposed solutions are integral components
of the process of creating a uniform information space in Ukrainian health care for the
purpose of strengthening the public health and ensuring civil rights in health care. In
addition, they conform to the National Blood System Development Strategy approved
by the Cabinet of Ministers of Ukraine No. 120-p of Feb 20, 2019. The strategy em-
phasizes the need in creation of a unified national donors’ registry and digital MIS in
health care institutions for the purpose of monitoring blood assets circulation.
The results of the object-oriented and functional data domain analysis are present-
ed by the means of the mathematical model. Suggested mathematical model defines
classes, its objects, properties, types and relation between them. In the order to repre-
sent data domain futures the predicates logic have been used.
The developed data domain models are generalized and universal and can be ap-
plied to the MIS development not only for the Kamyanske Blood Transfusion station
but also for others similar medical institutions. Introduction of the MIS will result in
the following:
• improvement of donors accounting process;
• toughening supervision over medical prohibitions (exclusion from donations);
• minimization of human factor;
• maximum acceleration of operations along with protection from mistakes;
• simplification and optimization of keeping all accompanying documentation;
• acceleration of donor search per blood type, rhesus, phenotype, etc.
Development and introduction of the information system will accelerate information
processing and ensure accuracy, reliability, operation speed and convenience, elimi-
nate inefficient manual data processing, increase staff management efficiency by co-
ordination of operations and data storage in a single database.
References
1. Yukhymets, V.O.: Medical information system as a component of electronic health care
system E-health. Ukranian Pulmonology Journal 2, 1–8 (2018).
2. Yukhymets, V.O., Kuts, V.V, Melnyk, О.L., Lisnevych О.S.: Avtomated medical infor-
mation system: technical requirements and features of implementation in the clinic. Ukra-
nian Pulmonology Journal 1, 1–10 (2016).
3. Chaban, O., Boiko, O.: Overview of world practice on the implementation of medical in-
formational systems and the problems of creating a uniform medical information environ-
ment. The bulletin of Lviv Polytechnic National University «Computer sciences and in-
formation technologies» 771, 365–370 (2013).
4. Kopniak, К.: Evaluating the effectiveness implementation of medical information systems.
Economics and organization of management 2(26), 109–119 (2017).
5. Stepanov, V.Yu.: The information space of health care in Ukraine. Theory and practice of
public administration 3(54), 1–5 (2016).
6. Tymchyk, S.V., Zlepko, S.M., Kostishyn S.V.: Classification of medical information sys-
tems and technologies for combined integrated criterion. Information processing systems
3(140), 194–197 (2016).
7. Hryhorova, T.: Architectural solution for the information system of rapid diagnosis of hu-
man immune. Transactions of Kremenchuk Mykhailo Ostrohradsky National University
4(81), 59–63 (2013).
8. Iepik, M.O.: The development concept for the disease diagnostics intellectual system. In-
formation and control systems at railway transport 2, 25–30 (2019).
9. Melnikova, N.I.: Features of medical information processing for support medical decisions
making systems. The bulletin of Lviv Polytechnic National University «Information sys-
tems and networks» 832, 190-204 (2015).
10. Development of medical diagnostic decision making systems and their cost-effectiveness,
http://journals.uran.ua/tarp/article/download/128455/143734, last accessed 2019/09/10.
11. Melnikova, N.I., Steblina, K.V.: Features of designing treatment decision support systems.
Mathematical machines and systems 1, 92–100 (2014).
12. E-health, https://ehealth.gov.ua/pidklyuchenni-do-ehealth-mis/, last accessed 2019/09/20.
13. Computerization and barcoding in the blood service, http://cfm.com.ua/Crystal_files/
KYSK.pdf, last accessed 2019/09/20.
14. Shpinkovski, O.A., Tsybulko, А.V.: Information systems of resource support of blood do-
nor activities. In: III All-Ukrainian Scientific Conference on Problems of energy saving in
the industrial region, pp. 159–160. Mariupol, Ukraine (2017).
15. Shvedenko, V.N., Veselova, N.S.: Enterprise information resources modeling for process
organization of enterprise management system. Software and Systems 4, 260–264 (2014).
16. Yen, P.-Y., Bakken, S.: Review of health information technology usability study method-
ologies. Journal of the American Medical Informatics Association 19(3), 413–422 (2012).
17. Iannessi, A., Marcy, P., Clatz, O. et al.: A review of existing and potential computer user
interfaces for modern radiology. Insights Imaging 9, 599–609 (2018).