Table of Contents
A System Dynamics-based Study on
Elderly Non-A cute Service in Norway
Yueping Chen
Contents
1. ABSTRACT
2. INTRODUCTION
2.1 Health condition and elderly caring capacity in Norway
2.2 Responsibility and type of elderly caring service in Norway
2.3 Problem Identification
2.4 Hypothesis
3 LITERATURE REVIEW
4 METHOD AND RESEARCH DESIGN
5 EXPECTED RESULTS
REFERENCES
uo PWN DN
10
1, Abstract
Modem non-acute service sector for old people is a highly dynamic and complex
system. Currently in Norway, it includes nursing home service, community house
service and home-based service. Patient flow between different service sectors is
influenced by various factors like population composition, waiting time and relative
satisfaction. Traditional analyses and comparisms between these services ignore the
dynamics and complexity underlying the patient flow and fail to properly represent
the structure of nonacute service sector for old people. Other researches on the flow
in public health service system, although from a dynamic view, usually take the
non-acute care system as a supplement to acute care system. Thus they do not
differentiate the nonacute services and make comparisons for the investment
decision-support purpose in non-acute sector.
This research uses the system dynamics methodology to map the dynamic patient
flow in elderly nonacute caring system in Norway. The model will differentiate the
characteristics of nonacute service sector and thus can increase understanding of the
complexity and dynamics caused by influencing elements in the system. The ultimate
purpose of this research is to deliver a tool to all local communities in Norway for
their long-term budget planning in non-acute service sector for old people.
2. Introduction
2.1 Health condition and elderly caring capacity in Norway
There has been considerable improvement regarding health status of Norwegian
people. One manifestation of this is the fact that life expectancy at birth is six to seven
years longer than it was 50 years ago. In 1998, the average life expectancy for males
was 75.5 years and 81 years for women. However, if compared with other countries in
Westem Europe, life expectancy in Norway changed from ranking highest during the
period 1950- 1970 to average levels at the end of the 1990s. (Health Care Systems in
Transition (Norway) 2000) This implies that the increase in life expectancy during the
last decades has been far more moderate than the average within the European Union.
And behind that, Norway had over 10 nursing home beds per 1,000 population in
the mid of 1990's, which is the highest among developed countries for which data
were available. It also has exhibited much higher expenditure on nursing home care at
over US$250 per capita and that amount accounts for over 15% of its total health
expenditure.(Health services and resources 2000) See Figure 1
Norway 1 1 Norway 4
Denmark 2 ——_— in
mn? ———
usag == Denmark 4
Irolond 5 Canada 6
Australia 6 Netherlands 6
Germany 7 Aunivala 7
Netherlands 8 Germany &
UKs New Zealand 9 [7]
‘Switzerland 10 ‘Spain 10 [7
France 11 [77] France 11
Jepan 12 Japan 12 []
span 13 5) Greece
Sweden 14 1 Hong Kong
Greace 15 |] Ireland
Hong Kong lerael
lara aly
ttaly ‘Singapore
New Zealand ‘Sweden
‘Singapore UK
0 2 4 6 B 0 2 0 100 ©=6200 300A
Number of beds PPP $US
Figure 1: Nursing home beds per 1000 population (left)
Expenditure per capita on nursing homes (right), mid-1990s
But with such high absolute capacity in Nursing home service, there are still
some communities in Norway reporting that they are suffering from having to set up a
waiting list for those who want to move into nursing homes. Sometimes the waiting
list is really long that patients currently in the nursing homes could only get limited
service amount and then just be transferred to other lower-level service compartment,
like community houses or home-based service. (Eldreomsory 2001) And that may
partly explain why the increase in life expectancy in Norway has been slowed down
during the last decades.
So it is the time for the govemment and communities in Norway to think about
whether there are some structural problems causing the delivery of elderly non-acute
service less efficiently.
2.2 Responsibility and type of elderly caring service in Norway
Norway has a small population and is sparsely settled. The responsibility for the
Norwegian health service is therefore decentralized. The country has five health
regions, 19 counties, and 435 municipalities/communities. Each is responsible for its
part of the health service. (Bolstad 2000)
The Local Authority Health Care Act was passed in 1984 and made local
communities responsible for all primary health care. In 1988, the task of managing
nursing homes was shifted from counties to communities. So right now local
commumities in Norway should take the responsibility for all the non-acute caring
service for the old people, as shown below:
Social welfare schemes (Health Care Systems in Transition (Norway) 2000)
Public social security
Pension schemes } administered by the central government
Acute-care in hospitals
Home-based service
Community-based care administered at the local community level
Nursing Home Service
The health care for old people can be divided into two groups: acute and
non-acute service care. A cute care refers to the surgery operation in professional
hospitals, which normally lasts for a short period of time. Non-acute care refers to the
longer-term service provided to those who cannot take care of themselves due to
physical or mental disabilities.
The non-acute health care service in Norway mainly consists of: (a) nursing home
care; (b) community house care; (c) home based care. Nursing homes provide
specialized medical, nursing and social services in an institutional setting. Ina
community house, old people live together in a specialized house where there are
some self-serviced medical facilities, easy for old people to use. Nurses will come by
regularly and they will only come with a higher frequency if individual patient needs
them. For home-based care, people just stay at their own homes and nurses will only
come by if they call for ask.
The main difference between these three services is the service level providing
and service capacity flexibility. For providing nursing home service, communities
need to establish a fixed staff and some physical facilities, like houses, beds and
medical equipments. For providing community houses, communities just need to set
up some physical facilities, but do not need to have a fixed staff, so the capacity in
community houses enjoys a more flexible attribute than in nursing homes. And for
giving home-based service, communities neither set up any physical facilities nor a
fixed staff, and so it enjoys the highest capacity flexibility but it can only deiverlow
level of caring service.
2.3 Problem Identification
Sometimes, people do not really need much service but still stay in some
high-level service compartments, like a nursing home or a community house. Figure 2
describes the service need by patients in a community house in Norway. Over half of
all the patients need less than 200 minutes’ service every week, which means they can
be just sent back to their own home and get the home-based service. Only 25% of the
patients need more than 600 minutes’ service per week.
Figure 2 Service need by patients in a community house in Norway
Therefore, to make the delivery of elderly caring service more efficiently, we
must firstly find out what is the real reason for some communities to be unable to
service the people just in time even though their capacity is not low, compared with
other countries. This is the first part of the problem can be called as the capacity
problem in the non-acute elderly service system in some Norwegian communities.
After the reason is found, then comes the second question needs to be answered:
What the local community should do to prevent this problematic situation from
happening? And what is the cost of implementing different policies?
Last, since the local communities’ goal in providing non-acute health care service
can be generally stated as: to gain a high level of public satisfaction at a relatively low
cost, this research also needs to find out how should local communities build different
non-acute service capacity to satisfy the demand for different types of caring from the
old people in a relatively long run.
The time horizon chosen in this research is 10 year, which is long enough because
the performance of local communities will be assessed by the central govemment
every 45 years. And the boundary should first of all include all three non-acute
service type compartments as endogenous structures, which are of most interest to the
local communities. And then take population, acute hospital and funding system as
some exogenous inputs.
2.4 Hypothesis
It is normal to have a waiting list in the health service system, if people want to
get the service but cannot get it due to the capacity limit, they will be just put onto a
waiting list. The waiting list will be expanded by new arrival of patients, either from
people newly having health problems or people transferring from other level of
service compartment. And it can be depleted by the admission to nursing homes,
which depends on service capacity limits and the number of current patients living in
the nursing homes. The average time people need to wait before they can get into the
nursing homes is termed as Waiting Time.
There exist some psychological effects of Waiting Time on people, as shown in
Figure 3. Firstly, when the waiting time is long, people will be reluctant to leave even
if they feel they are ok. Because they know it will take them quite long time to get in
the nursing home again if they need the service again. (Wait long, Stay long) Secondly,
if people perceive a long waiting time, they will tend to apply for the service earlier.
(Wait long, Arrive earlier) And this will make the waiting list artificially long because
not everyone on the waiting list really wants the service right now.
These two effects actually block the flow of patients in nursing home system,
forming two reinforcing loops that will lead to even more expanded waiting list and
more serious shortage of service capacity with the time passing.
pe Average Stay
)
Waiting Time Time
Gap
Wait long,
Wait long, + - Stay long
Arrive early
U Current £
Waiting Lift A——)_ Patient >)
Arrival Admission Serviced Leaving
+ Death
Checking Caring
Capacity Capacity
(More flexible) (Less flexible)
Figure 3 Patient flow in Nursing Homes
One policy to break the vicious cycles is that to set up some service capacity for
checking function. This checking capacity is used to check if the people really want
the service when they arrive and after they have stayed some time in the nursing
homes. Of course this policy need communities to increase some capacity first, like
extra nurses and some equipment need to be used to check those people health
condition. But these checking capacity are usually flexible, which means they can be
discarded easily and never cost any more once they are not needed. And if checking
can help reduce the waiting list and waiting time successfully, and the system can
benefit from these reinforce loops, then communities may not need to have so much
inflexible caring capacity. So in the long run, they can reduce the service capacity in
nursing home service, which can save the maintain cost of capacity for them.
The shortcoming of the policy lies in that in the public service sector, it is
impossible to enforce people not to stay longer and arrive earlier if they really want to
do so. And if the policy pushes people very much to do otherwise, it will tum down
the public satisfaction towards local communities, which brings other problems.
So some communities in Norway have adopted the idea of “Old people at right
place” (Hestnes GE 2000). The idea mainly refers to that caring service should be
provided at different levels with different service intensity and facility occupancy,
which can thus satisfy the different demands from different old people. Actually this
idea focuses on the flow between different service level compartments within elderly
caring sector. It is intended to realize that the service is only provided to those who
need it most. People in a higher-level service compartment who do not need so much
service can be transferred to a lower-level service and on the other hand, people can
first get some lower-level caring when they are waiting for higher level service. And
by doing so, the patient flow within sectors can be controlled without loss of public
satisfaction.
Figure 4 describes the patient flow between different services and some decision
rules of people when they choose different services. (Intermediate Service in the
figure aggregates the home-base service and community house service)
i 2
> Waiting List
Nursing Home Admission
+ +
ry Death
Life expectancy from NH
Fi in ee,
+
Public Satisfaction
Leaving
inNH _
Elderly Populatjon| F
(potential patiqnt)| “Samat ite
Re- Life expectancy f
IS Taker
ie
Fractional Rate of Public Satisfaction
Taking IS forIS +
=__
ae
New IS Taker
Death of Intermedii
Service Taker
Figure 4 Patient flow between different services
3 Literature Review
Different types of elderly caring services have long been discussed. Before
1980's, the prevailing philosophy on welfare emphasises what was then termed
"indoor relief" in institutional settings(Trattner 1974). During a long-time period,
non-acute care and relief for the elderly was only provided in kind of indoor settings,
like Nursing Homes, which provide nursing care, assistance with personal care
activities, and room and board. However, from the later 1980s, more and more
countries had a significant reduction in the proportion of elderly people in institutional
care together with a strong emphasis on expanding home care services— “outdoor
relief’, which is intended to maintain as many elderly people as possible in their own
homes for the reason of limited capacity of nursing homes.(Caring for Frail Elderly
People Policies in Evolution 1996)
The difference between institutional care and home care service has been
compared and various indicators conceming to the satisfaction level under both
service conditions have been evaluated. (Ruusland 2001) Some surveys show that
older disabled persons and their families prefer to receive services in their own homes
and communities, rather than in institutional settings, because many seniors enjoy a
higher quality of life in their own homes and communities. And some argue that more
money should be put on increasing home- and community-based care because the
average annual cost of nursing home care is twice as much as home- and
community-based care and its capacity is less flexible. (Indiana's Need for More
Home- and Community-Based Options and Fewer Institutions 2002)
These comparisons and analyses are very helpful to differentiate types of elderly
caring service and model people's decision rules when choosing different services in
this research. But they usually ignore the dynamics and complexity within the system
and so cannot do not properly represent the structure of non-acute service sector for
old people.
There is another group of researches related to the topic in this research, which is
more specified on the patient flow in public health service system. Modelling the flow
rate of Geriatric patients in health service system has been gradually developing
during the last decade. In early 1990’s, long stay patient and short stay patient were
firstly differentiated in modeling the flow in health service system. A number of
works have been done to create and solve a so-called two-stage dynamics model of
flow through geriatric departments. (Harrison GW 1991) In a two-stage model,
patients are initially admitted to the first stage, from which they may depart from the
system, by death or discharge or move to the second stage, from which they
eventually depart by death or discharge. (Millard 1989) Later studies showed that
factors outside the service relating to the availability of altemative long term care
placements has a major influence on the outcome of clinical care (Millard 1993). In
1996 a three. compartment model of flow was published which extends a
two-compartment model by considering patients released from geriatric departments
and their subsequent length of stay in the community. (Taylor G 1996)
Such models are all used to estimate the average numbers and lengths of stay for
short-term and long-term patients and the average number and length of stay in the
community for released patients, which allows for a significant improvement in the
forecasting of future bed requirements to aid the planning of geriatric departments.
Thereafter, some researches argue that the average length of stay in the acute
compartment will be artificially high if some who would be long-term patients are
kept waiting in the short stay acute compartment until beds become available in
long-stay compartments, like community houses and nursing homes. (El-Darzi E 1998)
And policy to reduce the waiting time for acute service was suggested to set up more
long-term non-acute services, like nursing home, which can serve as a secondary
caring institution after old people receive a surgery operation. This policy was tested
and demonstrated to have a much more profound effect on total patient wait times
than more obvious wait time solutions, such as increasing acute hospital bed capacity
directly. (Wolstenholme 1999)
In all, this group of models and researches, although from a system dynamic view,
usually take the non-acute care system as a supplement to acute care (like surgery
operation) in hospitals, which is more medical oriented and can last for a shorter
period of time. Because they do not look into and do not care the non-acute service
sector very much, they never differentiate the services within this sector and make
comparisons of them. So their results are of very little meaning to the decision-makers
in non-acute service sector (like the local community leaders in Norway) and their
suggestions to reduce the waiting time in acute service hospitals will be very likely to
increase the waiting time in non-acute nursing home services and thus only shift the
burden of problem to next taker.
In this paper, the three-compartment system dynamics models on the patient flow
in acute service can be applied in non-acute sector with some modifications. E.g. the
patient flow from hospital acute compartment to non- acute nursing home
compartment should be conversely considered because the focus in this research is on
non-acute rather than acute service sector. And attention also needs to be paid on
other characters, which differentiate either type of service, like different flexibility of
capacity, different waiting time attributes, etc.
4 Method and Research Design
This research may become a part of a bigger project, namely “Norwegian
Community Project’. The whole project deals with all responsibilities of local
communities in Norway, which include education, infrastructure, industry and social
welfare system. So the purpose of this research is to deliver a sub-model structure in a
policy-testing tool applicable to all Norwegian communities for their long term
planning in funds distribution.
Bearing this purpose, the research is designed as follows:
First of all, system dynamics should be implemented as a main methodology
because it can promote two things at the same time, -system thinking and system
dynamics modeling.
System thinking philosophy breaks the open loop and static thinking pattem.
Therefore, in the light of system thinking, one type of service system is no longer seen
as an isolated and distinct system but the one associated with other service systems
dynamically. As we understand from the illustration of the hypothesis, the capacity
problems in a high-level service compartment are highly interactive with other
lower-level service compartments available within the same sector. So if the problem
is not considered systematically, high leveraging policies can never be found.
System dynamic modeling enhances our ability to understand the capacity
problems better by capturing and illustrating the iterative complexities. The stock and
flow network is a strong language in this case to represent various material and
information flows and service process stages in the whole nor-acute elderly caring
system, so that it can successfully capture the associated accumulation, multiple
feedbacks, delays and non-linearity whose implications are usually too difficult to
understand if people only use their mental capacity. Moreover, a simulation model
allows laboratory experiment under altemative scenarios and visualizes the outcomes,
so that it can be used to exhibit the problems to the decision-makers in the local
communities and to test and explain the effects of various policies developed to deal
with the problem.
After a rudimental model has been set up, it should be reexamined in a group
modeling workshop where all the experts in different domains, e.g. medical
professionals, managers of service institutions, financial decision-makers in the
community, etc. will come together to give their ideas and recommendations on the
structure improvement. The workshop will be established several times during
process of the whole project, and this research can benefit from getting continuous
feedback and assessment from practitioners during different stage of modeling. And
this will obviously help increase the validity of the model.
At the same time, data also need to be collected for some parameters. In this case,
the data conceming the population distribution can be easily found, but the estimation
of population evolution can be very different from different information sources since
they use different methods to estimate. Second group of data of interest is related to
the attributes for different types of service, which include cost, average construction
time and loss rate. They can be easily found in the accounting records in the
communities. Third group includes the factor indexes influencing patient satisfaction
and their preference in choosing the service under different conditions. These data
could be found in some reports related, but it will be ideal if some experiments or
interviews are specifically established.
Finally, since the purpose is decision support, various policies of how much
money should be invested on each service from the local communities need to be
tested. For this purpose, an Interactive Leaming Environment can be set up to
facilitate the understanding of system structure and trial of their decisions without any
teal loss.
General stages of research are as follows:
1. Setup an SD-model based on the hypothesis
2. Let the model examined in a group modeling workshop
3. Interview the elderly and collect data for parameter estimation
4. Policy testing in ILE environment
5 Expected Results
First of all, the confirmation of the hypothesis conceming reasons of capacity
problems in some Norwegian communities is expected. Then the results by using the
Checking Capacity to solve the problem are also expected. Its advantage on the
flexibility and its possible negative effect on the public satisfaction should be shown
to the decision-makers in local communities.
When it comes to problem of how much money should be invested on each
service in the elderly non-acute sector, there may not be so-called optimal policies.
Every different community may have different weights on cost, satisfaction and other
elements. But the function of the model in this research is just to let the
decision-makers try their trade-offs and see the forth-coming results in short-run and
longer-run (up till 10 years) brought about by their different decisions.
Through trying different policies in the simulation model established by this
research, the decision-makers in local communities can get a general understanding of
why the capacity problems appear, how much capacity should be really invested for
either type of service and which type of service should be focused on. This
understanding will better their long-term budget planning in non-acute service sector
for old people, to help make further progress in the health status of old population on
one hand, and to keep the cost under control on the other.
References
Bolstad, Maalfrid. 2000. Norway's social security and health service. In Nytt fra
Norge: Ministry of Foreign A ffairs.
Caring for Frail Elderly People Policies in Evolution. 1996. OECD.
El Darzi E, Vasilakis C, ChaussaletT, Millard PH. 1998. A simulation modeling
approach to evaluating length of stay, occupancy, emptiness and bed blocking in a
hospital geriatric department. Health Care Management Science 1:143-9.
Eldreomsorg. 2001. Omsorysboliger i pleie- og omsorystjenesten. Tidsskriftet
Sykepleien.
Hanison GW, Millard PH. 1991. Balancing acute and long-term care: the
mathematics of throughput in departments of geriatric medicine. Meth Inform Med
30:221-8.
Health Care Systems in Transition (Norway). 2000. European Observatory on Health
Care Systems, WHO Regional Office for Europe, Govemment of Norway.
Health services and resources. 2000. In Intemational health-How Austral ia
conpares: Australian Institute of Health and Welfare.
Hestnes GE, Otterstad HK. 2000. Eldre paa rett plass. Tidsskriftet Sykepleien
11:52-55.
Indiana's Need for More Home- and Community-Based Options and Fewer
Institutions. 2002. Indiara Family & Social Services Administration.
Millard, P.H. 1993. Flow rate modeling: a method of comparing performance in
departments of geriatric medicine. Doctor of Philosophy in Medicine thesis,
University of London.
Millard, P.H., Higgs, P.F., Rochon, P.A. 1989. Occupancy plotting: a method of
estimating bed usage in departments of geriatric medicine., Division of Geriatric
Medicine, St. George's Hospital Medical School, London.
Ruusland, Anneline. 2001. Needs, experience of self-determination and satisfaction
with the nursing home, Section of Nursing Sciences, University of Bergen.
Taylor G, McClean S, Millard P. 1996. Geriatric-patient flow -rate modeling. IMA
Journal of Mathematics Applied in Medicine & Biology 13(4): 297-307.
Tratiner, W.I. 1974. From Poor Law to Welfare State: A History of Social Welfare in
America. Free Press.
Wolstenholme, Eric. 1999. A patient flow perspective of U.K. health services:
exploring the case for new intermediate care initiatives. System Dynamics Review 15
(3): 253-271.
Back to the Top
