A method to appraise the
functional cost of the complex
medical-managerial system
“hospital” in Greece
B. Spyropoulos
Medical Instrumentation Technology Department
Technological Education Institute of Athens, 12210 Athens, Greece
basile@teiath.gr
1. Introduction
The modern hospital consists of various functional units, which contribute to the
achievement of common goals, which are set for the entire structure, but they
maintain a relative autonomy and perform unique roles. Each unit adopts similar
methodological approaches, but it informs these approaches with the corresponding
concrete content peculiar to its scope. This wide variety of activities, express the
diverse levels of complexity, which should be encountered, in order to efficiently
develop, the appropriate information technology supported means to estimate the cost
distribution in each specific Hospital.
This is necessary, since all health-care Institutions seek to optimize their care
delivery processes, and their management needs are becoming more complicated, as
they try to keep pace with the rapid development and the diversity of Biomedical
Technology applied, as well as, the users cost by contemporary health-care. However,
within the Greek National Health System, there is not at the time, a standardized
approach, neither to the operational cost effectiveness and efficacy of the Hospitals,
nor a "case" classification algorithm, allowing for a rational remuneration of the
Institutions, by the various health-insurance organizations.
A first important step is to deal with the process of decision making in the specific
professional and scientific context of the hospital. The formulation of a general
decision making model, on the one hand, allows for the expedient problem solving
but, on the other, it imposes limitations which are due to the specific characteristics
and the nature of each problematic case in combination to the limitations of the
particular model used.
2. Decision making in the hospital departments
Hospital operation is impossible without the knowledge of a decision-making
mechanism while this operation necessitates the awareness of the need to apply this
mechanism on the basis of the uniqueness of the context and the individual case. It is
already mentioned that the modern hospital consists of various functional units, which
maintain a relative autonomy. Therefore, the most important issues concerning
decision making, in the hospital departments, are addressed in the following
paragraphs [Spyropoulos 1997].
2.1. The outpatient and the accident emergency departments
The outpatient and the accident emergency departments constitute a major interface
between society in general and the modern hospital. These departments act as input
gates to the structure and they receive patients under different constraints and with
different needs; hence they apply different approaches.
The main task of the outpatient department is to perform a diagnosis in
complicated cases and to provide treatment wherever this is feasible. A crucial
problem in this context is the mode of establishing criteria which are employed in the
selection of the data and the reasons for the potential omission of factors which might
have a bearing on both the diagnosis and the subsequent therapeutic intervention.
Decision making in the accident emergency department while it necessarily
implicates a narrower spectrum of parameters than the outpatient department, it
requires accelerated procedures in order to carry out the task in question, namely to
prevent death or the irreparable and irreversible damage to the patient.
The limited and specific nature of the data employed as well as the expedient
nature of the decision process in question, tend towards the adoption of standard
procedures both by the personnel and by the inference-supporting equipment involved
in the accident emergency departments.
2.2. In vitro diagnostic and medical imaging laboratories
The function of the in vitro diagnostic laboratories is on the one hand, to produce
data, which will be incorporated in the decision-making procedures of other
departments and, on the other, to chart courses of action on the basis of the
information they extract out of the specimens.
The main type of decision making in medical imaging concerns the discrimination
of forms and the ascription of specific characteristics on to them and the degree of
personal involvement of an expert in the final evaluation is still very strong because
of the great variation of the forms of biological structures.
2.3. The operating room
Decisions in the operating room result from the evaluation of the clinical features of
the patient, from the contacting of bio-signals and from the medical images evaluation
during the operation. Crucial decisions such as those concerning the operation itself,
the establishment of priorities among patients and the technique selection are
normally made long before entering the operating theater. An important parameter of
decisions related to surgery are the ethical questions, which pertain not only to the
patient himself, his needs and his wishes, but also involve the health care
professionals and the hospital as an institution.
2.4. The intensive-care units and the wards
A manifold of decisions is made at the intensive care unit. Among those is, first, the
selection of the patients to be admitted. Second, the development of treatment
strategies. Third, the evaluation of the data deriving from both the clinical
examinations and the bedside monitoring equipment and lastly concerning the
handling of critical situations. It should be noted that this manifold of decisions is
imbued with ethical dilemmas at every stage of the patient’s treatment.
Treatment is also provided at the inpatient wards where the major tasks to be
performed usually lack the urgency of those at the accident emergency department,
the operating theater or the intensive care units. However, the nature of duties at
these sections includes decisions pertaining to the administration of the wards. It
involves also decisions on the creation, the management and the evaluation of the
medical records of the patients.
2.5. Hospital administration
There are some very crucial decisions made in hospitals, which are of non-medical
nature but have a direct bearing on the operation of the whole structure. Such
decisions pertain to the overall policy of the hospital and in particular to matters of
allocation of human and material resources, the efficiency of its operation, the degree
of automation of various procedures, the level of decentralization of the various
departments necessitated by the rising cost of health care and last but not least, the
social role of the hospital, that is the position of the hospital in the health system
adopted by a given society.
3. Inferences leading to actions.
The hospital environment exhibits a variety of actions, administrative, medical,
technical, etc. and there are a variety of reasons employed in the various inferences,
which lead to corresponding actions. We claim that these reasons fall within three
categories. The first consists of reasons, which refer to biological functions, financial
and administrative indicators and technical procedures. These reasons contain
parameters, which are expressed in objective and quantifiable terms. Deterministic or
probabilistic models can adequately express reasoning schemata, which consist solely
of such reasons. The second category consists of reasons, which refer to the aspects,
which are unique to the problem in question. They express the specific parameters,
which allow for the instantiation of the specific problem. Such reasoning in its pure
form constitutes the casuistic approach. The case of a real patient, however,
exemplifies a form of reasoning, which exhibits both types of reasons, i.e. the
deterministically or probabilistically expressed parameters and the individuating
indicators. The third category of reasons that appear in inferences, which are carried
out in hospitals and lead to actions express, the ethical and axiological parameters of
the specific case in question. Such reasons are usually ignored in decision-making
models due to the fact that the information provided therein is usually not expressed
by typical quantification methods. Further, these reasons are considered to be of less
importance in the decision inferences since they are regarded as subjective and
depending on the individual patient’s preferences and the physician’s predisposition.
The optimization of the decision-making [Spyropoulos 1995] processes in the
hospital would require the development of tools so that the personnel involved would
acquaint itself effectively and efficiently with the subject matter. Such an effort would
also contribute to the promotion of interdisciplinary research, and, it would address in
an effective manner the problem on intra-hospital communication between the various
specialties.
4. A resources allocation and cost-capturing technique
As a first step, a resources allocation and cost-capturing technique has been
developed, allowing for the major Hospital Departments and Units, to monitor their
resources consumption, established on an appropriate combination of technical,
managerial and clinical of information. The technique is based on an appropriate
correlation of the major cost components, in each department of a specific hospital,
that is:
4.1. Estimation of the Medical Equipment median cost contribution
Equipment is classified according to ECRI. The actual Medical Equipment item
purchase price is transformed into a virtual daily leasing rate. The actual maintenance
cost is transformed into a virtual daily maintenance rate. An expected mean lifetime is
assigned. A mean daily throughput or load factor is calculated. An appropriate
approximate mean elementary operational cost is then calculated. For example, CT
Scanner Image, Stress test, Ophthalmic Unit examination, ICU-Monitoring Time etc.
4.2. Estimation of the Building average leasing and maintenance cost
The actual approximate building value is transformed into a virtual annual leasing
rate. The actual maintenance personnel salaries and outsourcing cost is transformed
into a virtual annual rate. The annual maintenance material costs are calculated. The
actual Maintenance Equipment item purchase price is transformed into a virtual
annual leasing rate. A daily maintenance cost / quadrate meter is calculated. Weight
factors for special Departments, as for example, ICU/CCU, Operating Rooms etc. are
also determined.
4.3. Approximate Reagents and other Material Cost
The starting point for the cost estimation is a three-level in vitro Laboratory test
Classification. A similar approach may be used in other in vivo procedures or any
other relevant medical activity that can be selected out of Medical Procedures
Classification, as for example ICD-10, and includes material cost. Data related to
reagents, consumables, disposable and other material components of annual
functional cost are related to the actual output, i.e. the number of the performed
examinations, tests etc. A mean approximate Reagents and other Material Cost is
assigned to each procedure.
4.4. Medication Cost
Medication Cost can easily be calculated since the Drugs are classified and priced by
a National Authority. The algorithm and its implementations allows for and assist the
discrimination between commonly used low-cost medication resulting in approximate
mean values for each Ward, and, individually prescribed medium and high cost
medication, such as antibiotics, cytostatics, antiviral cocktails etc. Both, personal and
statistical mean contribution of Medication Cost can be calculated.
4.5. Weighted Manpower Cost
The algorithm discriminates between medical and other scientific staff, nurses,
technical and administrative personnel, and auxiliary personnel. The internal structure
of each personnel category may also be taken into account. The overall salaries cost
of a Department or a Unit is distributed over the service outcome, resulting in an
elementary activity cost component. Distribution patterns can be defined and
modified, according to the relevance of the locally performed activities. Thus,
weighted manpower cost can virtually be associated to any medical activity that can
be selected out of a four-level Medical Procedures Classification and/or a three-level
in vitro Laboratory test Classification.
The proposed algorithm, implemented on specific department-customized
software-tools, allows for an efficient and realistic approximation of single-activity
cost, by combining the five main cost components, as described. Furthermore, this
technique allows for an estimation of the mean cost associated to specific medical,
laboratory and administrative activities.
5. The Patient Record as an appropriate tool for the allocation of
resources
Medical records usually include, first, clinical information obtained by the case
history, and by inspection, palpation, percussion, auscultation etc., second, signals
related to bodily functions which are collected from in vivo diagnostic procedures
(ECG, EEG, EMG etc.), third, data acquired through in vitro diagnostic tests, fourth,
images related to the morphology and the functions obtained from medical imaging
procedures, fifth, information related to various therapeutic interventions and, lastly,
data which are of administrative and of financial importance such as insurance, costs
of medical treatment, cost of hospitalization etc. Medical records are used in a variety
of ways and they serve a multiplicity of purposes. Beyond the explicit involvement of
records in the therapeutic process, there are several other discernible uses, such as in
research, in teaching, in the allocation of resources, and in the construction of the
patient’s personal history. These functions as well as the variant purposes served by
records, raise a number of important ethical issues.
The social role of patient’s records [Papagounos, 1998] is evident from the use
made of information -related to their contents- by non-medical institutions and
agencies. Insurance companies, various employers, state agencies, the pharmaceutical
industry, and others use data extracted or derived from these records in order to
formulate policies and chart courses of action, which affect the whole population.
These courses of action include both directives concerning the provision of health
care such as, e.g., drugs to be prescribed or guidelines to be followed in diagnostic
and treatment procedures, and more general practices such as the selection criteria of
hiring and firing of employees.
A major problem of contemporary health services is the scarcity of resources.
Patients’ records provide information concerning the expenses incurred and the
resources allocated for the treatment of the individual patient within the health care
system. This information is value laden since they express a society’s preoccupation
with costs -private or public- of services provided to its members. However, it is on
the basis of these preoccupations that the available data is evaluated and processed.
As a result, strategic decisions are made which concern the distribution of resources,
first, in terms of regional planning, second, among the components of the health
sector and, third, among individuals. Consequently, the structure, the contents and the
use of medical records, especially the electronic ones, as well as the state of the art
capabilities of information processing necessary for the evaluation of data, affect the
establishment of criteria for the allocation and distribution of resources. The decision
for the construction of, e.g., a specialized health care unit is based, among other
factors, on the epidemiological data derived from the medical archives of a given
region.
It should be kept in mind that the patient’s profile [Papagounos, 1999] which
emerges out of the records, lends itself to various uses and interpretations according
to the purpose they are called to serve. An insurance company, e.g., although it has no
official access to the full patient’s medical record, it can infer, by means of the
indicators necessary for the estimation of the compensation for particular medical
expenses incurred, an accurate health profile of the specific individual and it then
could then use this profile, to enforce an insurance policy on the individual concerned.
The insurance providers could further extrapolate, on the basis of a sufficient number
of such profiles, and create a policy which would apply to specific groups of people.
An extreme case of the use of inferences made by employing data derived from
medical records, could be the adoption of hiring and firing policies based on the
individuals’ health state and their predisposition to certain diseases, an event which
would violate the principle of equality.
Patient’s records, however, assume an additional dimension which is related to the
nature and function of technology. The development of patients’ records exhibits the
state-of-the-art biomedical technology, the dominant -at the time- medical knowledge
and practice and it reflects the existing value system of a given society. The
assimilation of new technologies is a complex phenomenon which is affected by the
wider socio-cultural, epistemic and axiological parameters at work within the specific
social context in question. In other words, the technologies employed presently in the
construction of medical records were developed to satisfy concrete social needs. The
availability of these technologies permitted new questions to be asked and new needs
to arise.
Out of these needs, a cardinal one is the documentation and the storage of
comprehensive financial data, relevant to health-care, in easily accessible forms.
The “vehicle” for our own approach, constitute, first, the combination of several
administrative and statistical hospital databases, second, the extraction of information
concerning the allocation of resources, out of the medical records, and finally, the
approximation, through procedures observations and simulations, of the manpower
distribution pattern, within the various departments, and among the specific treatment
cases. It should be kept in mind, that, unfortunately, only a small portion of the above
mentioned data are available in digital form, in the Greek National Health System
hospitals.
In the software level [Spyropoulos 2002] we are developing, first, departmentcustomized software-tools, that allow for a rational approximation of the effective
mean cost, for several elementary medical activities, coded according to ICD 10, over
different hospital departments. The tool can create, thus, a valuable statistical data set,
for the definition of an efficient financial compensation procedure of the involved
Units. In a second phase, the acquired data, in turn, will constitute the input data for
the formation of an appropriate medical record, that is intrinsically, however latent,
related to the costs caused and the expected reimbursement. The developed model
allows for every user, first, to select one or more out of the 37 discrete medical
specialties, second, to set up a custom-made menu of available diagnostic and
treatment activities out of the ones contained in the system, third, to individually price
them, according to the data acquired by the department-customized software-tools,
already described. Thus, the updating of the patient's relevant medical data ignites,
when relevant, the corresponding revision of an implicitly associated financial record
that allows for an approximation of the individually coded case-cost. The Australian
DRGs have served as a starting point for the case codification.
6. Concluding remarks
Technical details and partial results of the system have already been presented
elsewhere [Spyropoulos 2003]. After an employment of the system, for at least three
years, in pilot projects covering all types of Hospitals in Greece, that is in the two
major cities, in the rest of the mainland and on the islands, the accumulated
experience should allow for a rather realistic approach, to the real cost of the
treatment of the coded cases. Then, it would become possible to switch, from the fully
outdated daily flat-rate reimbursement of the Hospitals, towards a more efficient and
managed-care oriented system. The system, presently tested within a virtual hospital
environment, offers a simple and time efficient method, for the acquisition of realworld Hospital operational cost data, necessary for instance, for the introduction of a
Hellenic-DRG based reimbursement procedure.
References
Papagounos, G., & Spyropoulos, B., 1998, Professional ethics: A theoretical analysis of their
relative value and limitations in the era of information technologies, in “Ethics and
Information Technology”, Edited by G. Collste, New Academic Publisher, Delhi.
Papagounos, G., & Spyropoulos, B., 1999, The Multifarious Function of Medical Records:
Ethical Issues, Methods of Information in Medicine, 38, 317.
Spyropoulos, B., & Papagounos, G., 1995, A theoretical approach to Artificial Intelligence
Systems in Medicine, Artificial Intelligence in Medicine, 7, 455.
Spyropoulos, B., & Papagounos, G., 1997, Decision-Making in a Modern Hospital: Some
Theoretical and Practical Aspects, CD-ROM Proceedings of the 17th Annual International
Conference of the IEEE Engineering in Medicine and Biology Society, Chicago, IL.
Spyropoulos, B., 2002, Supporting the Acquaintance of Health-Care Personnel to OperationalCost Monitoring, CD-ROM Proceedings of TEHRE 2002, London, United Kingdom.
Spyropoulos, B., & Marinis, M., 2003, Supporting the acquisition of Hospital-Care Cost Data
to feed a refined DRG system for the Greek National Health System: Status report,
Proceedings of the 29th Annual IEEE Northeast Bioengineering Conference (NEBE),
Newark, NJ, 144.