Modeling a Policy for Managing Polio Vaccine in J apan:
Scenario Planning based on System Dynamics
Yosuke NAKA] IMA*, Toshiyuki YASUF*, Yoshiaki OHKAMP**, Naohiko
KOTAKE***
* Graduate School of System Design and Management, Keio University
**: Keio Advanced Research Center, Keio University
er; Advanced Research Center for System Design and Management,
Keio University
er: Graduate School of System Design and Management, Keio University
SDM, Collaboration Complex, 4-1-1 Hiyoshi, Kohoku-ku,
Yokohama-City, 223-8526, J apan
* +81 (0)90 8599 8198
* y.n@z5.keio.jp
Abstract
This research is to provide a methodology for making policy scenarios
based on the system dynamics. The authors deem this new methodology
would be a useful tool for policymakers to make policy scenarios. As for the
case study, this research deals with the policy scenarios for managing
polioviruses in Japan as an example. This methodology includes both the
simulation part of using System Dynamics and the conversation part
related to Scenario Planning. Through using this methodology, we had
structural understanding of the problem with the visible simulation results
and conversation with member which was focusing on the parameters that
would be a part of suggested scenarios. This methodology is expected to
improve the public deliberations for making policy scenario based on data.
1 Introduction
This research is to propose a methodology using system dynamics with scenario
planning in a case of making policy scenarios for managing polioviruses in Japan. And
we validated the usefulness of the methodology.
Polio is known as an acute infection which arises when the poliovirus is infected to
central-nerves tissue. Currently, the only Oral Polio Vaccine can be used as the
polioviruses prevention depends on the law in Japan. It is known that Oral Polio
Vaccine rarely causes symptoms of paralysis and secondary infection. In 2011, the
parents who have infants worried about the side effects of the Oral Polio Vaccine, the
movement which kept from polio vaccination have occurred in Japan.
According to an investigation for polio vaccination rate, which is conducted by the
Government of Japan, the number of Oral Polio Vaccine inoculation rate from A pril to
June in 2011 was decreased by 17.5% as the national average comparing with the
corresponding-period-of-last-year in Japan (MHLW, 2011).
There is an Inactivated Polio Vaccine as a substitute of Oral Polio Vaccine. It is known
that it has very little possibility of appearance of symptoms of the paralysis which is the
side effects that Oral Polio Vaccine has. However, the clinical trials for introducing
Inactivated Polio Vaccine are currently on-going in Japan.
Besides, the wild polio viruses are decreasing in the world, but they continued to
circulate in some countries (Nigeria, India, Pakistan, Afghanistan, etc.) in 2010. And
there are much traffic between the countries and Japan. Therefore, the possibility of
infection to Japan from the area where wild polio viruses circulate cannot be denied.
And since the curative drug to polio does not exist, the prevention of polioviruses using
a vaccine is still an important policy countermeasure.
2 Literature Review
System dynamics is known as an effective systematic method of making a policy
scenario. And there are some researches which discussed System Dynamics and
scenarios making related to the theme of the infection disease policy (Pruyt, 2010;
2007). However, the main interest is the model comparison for the simulation, and the
framework which derives policy scenarios was not discussed as the main subject.
There is a research which discussed System Dynamics and Scenarios Planning related
to the theme of making strategy (Andrea, 2004). However, the main interest is the
differences of the approaches between the standard manager analysis and the
methodology proposed, and the application of using the methodology was limited to
making company strategies.
There are research which treated epidemic processes using simulations of
mathematical models (e.g., Kermack & McKendrick, 1927; Anderson and May, 1991;
Nishimura, Kakehashi & Inaba, 2009), and a research which treated decision-making of
polio vaccine policy in 1950s using SD model of system dynamics (Thompson KM, et
al, 2006). However, we could little find a research which described the policy scenario
in the group which keeping high immunization rate like present Japanese situation.
As a research of the behavior of epidemic process, there is a research which
mentioned the usefulness of simulation regarding the efficacy of the preventions (Yang,
2009). However, that did not show the conclusion as a policy scenario which was
derived from simulation results.
Regarding the importance to make a prediction about the future of the environment
changes using mathematical methods is descried, but we rarely find the way of finding
the controllable parameter from the mathematical method and relation between the
parameter and the scenario which is lead as a solution (Kees, 2004).
This research also pursues the new way to create a public policy by combining the
system dynamics and the scenario planning. One the one hand, the conventional
approach of the system dynamics to create a public policy stresses the significance to
make the complex relations of casualties simple and visible (e.g., Stave, 2002; Homer
and Hirsch, 2006). However, this approach often refers to less method to make a policy
maker visionary for the future policy courses. On the other hand, the conventional
approach of scenario planning refers to the importance of assumptions for the future
policy courses, but not to causal backgrounds why such assumptions can be rational
(Fleisher and Bensoussan, 2002). Thus by bridging the causal backgrounds drawn by
the system dynamics and visionary outcome enriched by the scenario planning, this
research provides a consistent and cause-based platform of policy creation.
3 Design of SSP Methodology
3.1 Outline of the methodology
This research called the methodology System Dynamics Driven Scenario Planning
“SSP” (Figure 1.). The hypothesis of this research is that the methodology which this
research proposed is effective for policy scenario making. And we empirically validated
the effectiveness of SSP in a case of policy scenarios for current Japanese polio issue.
To confirm the effectiveness of SSP, we derived policy scenarios using this
methodology. The design of the methodology is shown below.
System Dynamics Driven Scenario Planning (SSP)
>» Solution
@: Information transaction based on simulation results/graphs
Figure 1. Design of SSP methodology
The SSP is a methodology for creating policy options based upon the system dynamic
and scenario planning in the cycling process. Once an issue is input in it, then scenarios
are output as a solution. When the issue inputs and the solution outputs, conversations
will be held in accordance with the method of Scenario Planning.
Those conversations are an approach in the process of scenario making by
considering the environmental structure which surrounds the issue or causal relationship
with the stakeholders.
And in this methodology, the simulations using SD model are performed repeatedly.
Then, this research gets a better scenario as a result. In this process, the simulation
results play the role to tie System Dynamics and Scenario Planning. So, the simulation
result is used as a vehicle of information to connect both methods.
The methodology finally generated the policy scenarios against the issue and we
proposed them.
3.2 Basic Causal Loop Diagram (CLD) And Stock-Flow Diagram (SFD) for
Infection Disease
We are able to easily perform the simulation using any tools or programming with
those basic models which are the Causal Loop Diagrams (hereinafter referred to as the
“CLD") (Figure 3.) and the Stock-Flow Diagram (hereinafter referred to as the “SFD")
(Figure 4.) for infection disease.
Those diagrams are known as CLD and SFD to show an infection disease expansion
process (Sterman, 2000; Pruyt, 2010). To be consistent with the mathematical model,
the parameter of “Contact rate” was excluded from those diagrams.
Recovery
time
Susceptible Infection Infected
loop + loop ee
Susceptible (~ A iakecioa ( Recovered
Infections ( Recovery —»
ae. oe population
Na Infection
rate
Figure 3. Causal Loop Diagram (CLD): Basic Infection Model
Total population
Recoverytime
Susceptible > Infected \ > Recovered
population 4 population population
Infections Recovery
Infection rate
Figure 4. Stock-Flow Diagram (SFD): Basic Infection Model
SD model strongly relates to CLD and SFD. Once the CLD and SFD are generated,
the analyst could realize the model and get simulation results using them. The CLD and
SFD become basic tools whenever making a change to the model. Then after generating
the model for the simulation, it is easy to modify them depends on the diagram changes.
Then, this approach is easy and very effective to keep the accuracy of the simulation
results for not only policymaker but also everyone.
3.3 Why does a problem-owner need to have conversations to generate scenarios?
Conversation is an approach in the process of scenario planning by considering the
environmental structure which surrounds the issue or causal relationship with the
stakeholders. Scenario Planning is known as an effective way to cope with uncertainty
by considering multiple, equally plausible futures as well as traditional and conventional
way of thinking (Kees, 2004).
And this research defines the meaning of scenario is below:
Thinking the issues associated with the future, two or more possibilities are made
as stories while thinking about the movement of the environment in the surrounding
and other relating subjects. Then the outputs from this thinking are called scenarios.
The scenario is a kind of hypothesis. If it is validated, it becomes an established
theory. Therefore, scenario is a speculative idea until being validated.
Merits and demerit of conversation approach
There are both merits and demerits to have conversation in order to generate
scenarios. But the authors considered there are more merits than demerits to improve the
accuracy of the generated scenarios as a solution through the conversation. The merits
and demerits are below:
Merits:
There are merits to have conversations to generate scenarios.
Understand and share ideas among its members and implement organizational
leaming
Examine an essential solution to the issue by multi aspects
Reduce missing some critical things
Get rid of the self-righteous of the person in charge of the analysis
Construct network between stakeholders
Secure social trust by examination in a group
Demerits:
The two biases are thought as the demerits when we have conversations to generate
scenarios.
Selection bias is:
- Bias that happens when the group is not correctly representing proper opinion.
For instance, only noncooperation persons or aggressive cooperator.
Information bias is:
- Bias that happens because information obtained is not correct when it is
observed. For instance, very limited information or depending on their
memories.
And the conversation is a kind of human-interface whose objective is to confirm the
important decision among the member.
3.4 Process of SSP methodology
This section describes the process of methodology (Figure 6.).
Firstly, this process defined the issue having a conversation with the stakeholders as
an approach of Scenario Planning. Then we made the model of the issue with CLD and
SFD of System Dynamics. A fter that, this research conducted a simulation to know the
behavior of polio epidemic process. Once the simulation was implemented, the authors
were able to make hypothesis and find target parameter through having conversation.
And it re-generated the model depends on the simulation results and reflected the target
parameter which is discussed among the member. Finally, this research confirmed the
policy scenarios based on the simulation results by having conversation.
@ start
{
| Define Issue l
J mj Conversation (Scenario Planning)
Doyot confirm
No theissue definition?
Generate CLD, SFD Yes r Current status
pee / Analy’
(| \[ systembynamies “Tt
modeling
Simulation \
\
\
hypothesis \L_ lessonslearned
/- fromsimulations
Conversation (Scenario Planning) jaan j /
\ Doyou finitt [Generally Known |
= —_____—
No- target parameter ? | Information |
” 2 Xe /
a _ i /
passin —___| Svstembynamies {SD model)
ete \[L_Be-modeling ]
\ —ft
\ Simulation |-—'
\ Canto
\
‘ Do you confirm
the scenario’?
ves |
Conversation (Scenario Planning) [uma Oe
Figure 6. Process of SSP methodology
4 Example Application
In order to recognize the current status of Japanese polio issue, the authors performed
the simulations to know an epidemic process. This research proposes the use of SD
model for the simulation to derive a policy scenario. Though it is known that a
simulation is useful for a discussion regarding infectious diseases countermeasures, the
discussion process which is based on a simulation result is tied to a policy scenario is
rarely described in previous researches. This research realized that this is a point which
was not covered by using SD model for making scenarios. And the authors deemed that
we are able to resolve this issue by using SD model with Scenario Planning method.
In this section, firstly, this research described the importance of the Herd Immunity
with vaccine in the area of epidemiological research. Next, we mentioned the usefulness
of the simulation of System Dynamics in epidemic process research. In addition, it
showed the results of simulation related to the polio epidemic process by using SD
model. Finally, it performed the Scenario Planning based on the results of the
simulations.
4.1 Importance of the Herd Immunity with vaccine
There are two important concepts in epidemiological research.
Q@H: It is believed that immunizing some members of a community against a
disease would protect the entire community and that this called herd
immunity could be used to control infection outbreaks. The immunized
rate in a group is called Herd Immunity (hereinafter referred to as the “H").
@RO: The number shows the strength of reproductions that the infectious disease
spreads from the human to the human. This is called Basic Reproduction
Number (hereinafter referred to as the“RO"). This is an average number of
second infected persons who were infected from an infected person.
The followings are the relational expression of these concepts (1.1).
(1.1) H=(1-1/R0) x100
For instance, RO 5-7 and H 80-86% is known as Poliovirus. RO 5-7 means that a
person who is infected poliovirus will infect 5-7 persons on the average. H 80-86% is a
concept that the epidemic of the infectious disease doesn't happen if a group has
immunization rate of 80-86% or more. In other words, the value of H becomes the
boundary whether the epidemic happens or doesn't happen. It is called the Herd
Immunity Threshold.
Also, RO and H of the infectious diseases other than poliovirus are known below
(Table 1.).
Table 1. Basic Reproduction Number and Herd Immunity thresholds
Infection RO: Basic Reproduction H: Herd Immunity
Number thresholds
Mumps 4-7 75-86
Polio 5-7 80-86
Measles 12-18 83-94
Malaria 5-100 80-99
H=(1-1/R0)*100 Fine PEM: Epidemiologic Reiews 15; 265, 1993 (modified)
4.2 Purpose of simulation for the epidemic process research using System
Dynamics
First of all, the use of SD model for the research of epidemic process is intended
NOT to find the time point and the route where an expansion of epidemic happen. But,
the purpose of using simulation is rather to confirm how the behavior of the epidemic
process is changing by setting different values of parameters (Meadows and Robinson,
1985). In addition, by changing values of the parameters, or by changing the model
according to analyzed environment and object, we could leam the behavior of the
epidemic process from the simulation result.
4.3 Simulations: Polio epidemic process using SD model
This research confirmed the behavior of the epidemic process of polioviruses using
simulations. The purpose of those simulations is to make policy scenarios for the polio
issue in Japan.
Firstly, this research performed the simulations to know the possibility of occurrence
of the polio epidemic for two years among a group of one million people that reflected
the Japanese status which is decreasing Oral Polio Vaccine vaccination in 2011. Next, it
performed the simulations to know the possibility of epidemic in a group of 50 infants
which is including the infants who did not have the polio vaccination in 2011. Then, it
discussed a parameter which could be a factor of solution. It was a contact rate. The
authors thought the contact rate could be a parameter which has possibility to control
the behavior of the epidemic process. Finally, this research conducted the simulation to
know the effectiveness with the virtual group of 50 infants using contact rate as a
parameter.
Simulation 1: Confirmation of the process of polio epidemic in a group of one million
person including adults
From the results of simulation 1, the authors realized that the epidemic would not
happen in a group of one million people including adults even if its Herd Immunity is
decreasing by 17.5% per year for over 2 years from 2011.
The number of births in 2010 is estimated that 1.57 million people and birthrates per
1000 population are 8.4 in Japan (MHLW, 2011). In 2011, The decreasing of the Oral
Polio Vaccine vaccination having been seen in the infants who were recommended to
dose in the age of from six months to one year and half years old. The decreasing rate in
2011 was 17.5% compared with the simultaneous in period 2010.
Based on these Japanese situations, the authors conducted the simulation using a
virtual group of one million people which is the same population ratio of the adults and
infants of Japan. Before the simulation, they generated the SFD which reflected the
Japanese situation decreasing the polio vaccination of infants for over two years by
17.5% annually (Figure 7.). This model is almost the same as the basic SFD of infection
disease (Figure 4.), but the model is constructed by two parts. One is the part of adults
and another is infants. And we put the “contact rate” in the both parts to clearly specify
it as a parameter. In addition, they put “Immunized decreasing rate” in the part of infant.
The “immunized decreasing rate” means the rate of infants who kept from polio
vaccination in 2011.
Immune Contact rate
rate A * Infection rate gee
V Susceptible AVA infected V7 Recovered
lationdA ————— |
a 7 population A \ population A TAN > copulation A
Infections ] Recovery
Total population
a:
~—
y Susceptible Infected Recovered
populationB -———<——=>| | ——— S|
a population B a population B aN population B
Infections Recovery
Immune ik
pre immunized Infection rate
Contact rate!
decreasingrate
Figure 7. Modified Stock-Flow Diagram (SFD): Infection Model Including
Immunized Decreasing Rate
The simulation result using the model is below (Figure 8.).
soft SD model Tesult, 2010'to 2011, B: SD model result, 2011 to 2012.
Condition: Disease=Polio, RO=6.0, Recovery time=28 (day), Population=1M, Default
immunization rate=90% & Current
Figure 8. Polio epidemic process in a group of one million person including adult
The authors learnt that there were no changes from the simulation results. This means,
even if the influence of keeping from infant vaccination in 2011 is considered or not,
that the epidemic does not occur in the group of one million people including adults and
infants. The reason why the epidemic does not occur is that the population of the infant
is very small. Therefore, the epidemic of polio will not be caused in the group of one
million persons even if the group includes the infants who kept from vaccination
decreasing at the rate of 17.5% for over two years in the future.
Simulation 2: Possibility of polio epidemic in a place where infant gathers a lot
The authors learnt that there was a possibility of the polio epidemic in a group which
gathers infants after 2011. Possibly, the places where the infants are gathered lot are a
child care place or hospital, etc. The infants who including the infants kept from the
polio vaccination are gathered those places.
In this simulation, they would like to know the possibility of occurrence of polio
epidemic in a group of 50 infants including the infants who kept from the vaccination.
As the condition in this simulation, we set the immunized rate in 2010 is 90%, in 2011
is 74.25% and in 2012 is 61.25% which reflected Japanese situation (Figure 9.).
A: Year 2010 (Immune B: Year 2011(Immune C: Year 2012 (Immune
90%) ADH) 1.25%)
Condition: Disease=Polio, RO=6, Recovery time=28 (day), Population=50, Default
immunization rate=60, 70 and 80%
Figure 9. Influence of immunization rate in an infant group a lot in 2012, 2011 and
2010
In the simulation, it was learnt that the polio epidemic would occur after 2011. This
meant, if a polio virus holder entered the group, an epidemic of polio was caused in the
group. In this case about 10 infected persons by the accumulation in 2011 are generated.
In addition, the result is understood that there is a possibility that about 20 infected
persons in the group are caused by the accumulation in 2012 when the decreasing rate
of kept from vaccination was 17.5% in the annual rate continues after 2011.
Simulation 3: Behavior of polio epidemic process by change of the parameter “ contact
rate” that could be controlled in an infant group
The authors learnt that it caused the delay of infection expansion by controlling the
contact rate through the simulation.
Before the simulation, based on the results of Simulation 2, this research discussed
the parameter to control the infection expansion. Then it focused on contact rate as a
parameter which could be controlled by any regulations or restrictions.
When it focused on the contact rate, we had known that some parameters which
effect changes to the behavior of epidemic process were seen in the SFD, Figure 6. For
instance, they are the parameters like group immunization rate (Immune rate), infection
rate (Infection rate), recovery period (Recovery time) and etc. However, these
parameters are characterized by each infectious disease. Therefore, it is difficult to use
10
them as controllable parameters. On the other hand, the contact rate (Contact rate) may
be possible to control it as a parameter. Then we listed those parameters below (Table
2.) to show those parameters are whether controllable or not.
Table 2. Listing of parameters which can be possibly controlled or not (in case of
Polio)
Possibility of control Descriptions Value
No Basic Reproduction Number (RO) 5~7
No Recovery time (infectious period) 3~5 weeks
No Herd Immunity (H) 90% (-17.5% per yr for
infant)
Yes Contact rate Greater than or equal to 0
Yes Frequency of hand washing Greater than or equal to 0
It was understood that the parameter which has effectiveness against the poliovirus
expansion is contact rate and a frequency of hand washing from the listing, Table 3.
However, the research considered a quantitative effect of the frequency of hand washing
might be not clear. Then this research excluded it from the candidate of parameter
which would derive scenarios.
Then it performed simulations that the contact rate is set as a parameter. The values of
contact rates were set to 0.5, 1, and 2 in this simulation (Figure 10.). In this case, the
authors set the average normal contact rate in the group is Contact (rate) =1. And RO=6
which is known by the epidemiologic research for polioviruses. The group which is
used for the simulation is a virtual 50 infants group including the infants who kept from
polio vaccination which reflected present Japanese situation in 2011.
__A:Contactrate=0.5 BB: Contactrate=1_ = C: Contactrate=2
Condition: Disease=Polio, RO=6, Recovery time=28 (day), Population=50, Default
immunization rate=74.25%
Figure 10. Influence of contact rate in an infant group including infants who kept from
polio vaccination
When it set to Contact=1, as the peak of the infection, about 10 cumulative infected
persons were caused around the day 80. And when the contact rate is set to 0.5, the
epidemic did not happen. In contrast, when contact rate is set to 2, about 15 cumulative
infected persons were caused around the day 40 as the peak.
11
4.4 Simulation results summary
This research exercised performed three simulations above. And the results lead the
scenarios for Japanese polio issue. In this process, it focused on the effectiveness of
contact rate which was a parameter for changing the behavior of the epidemic process in
the simulation. In particular, it implemented Simulation 3 followed by Simulation 2.
This sequence is very important to find and confirm the parameter that leads the
scenario.
The following were the results which were clarified by above simulations.
+ Polioviruses would not occur epidemic over the next two years in the group of one
million person including the adults and infants.
+ There was a possibility of the polio epidemic in the group of 50 infants including the
infants who kept from polio vaccination.
- A parameter "contact rate" caused the delay of expansion of polioviruses in the
group.
This research completed two scenarios from policy aspects with conversations based
on the results of simulation. The conversations are an approach in the process of
Scenario Planning. One of the scenarios is for a polio prevention scenario in usual status
and another is for an epidemic scenario in a large group including adults and infants.
Two scenarios from policy aspects
- Prevention in usual status
- Delay expansion in epidemic status by decreasing contact rate
4.5 Flow of leading scenarios through this methodology
The flow of the Scenario Planning using System Dynamics for epidemic issue is
shown below.
Define and confirm the issue through conversation
Recognize current status using simulation results
Make a hypothesis to solve the problem
Figure out the parameters which change the behavior of epidemic process using
CLD, SFD, simulation results and generally known information. And confirm it.
5. Lear the effectiveness of the parameter which influence the epidemic process
using re-modeling SFD and simulation
6. Verify the feasibility of the parameter using generally known information and
generate scenario.
7. Confirm the scenarios through conversation
PwNre
Some of those procedures would be repeatedly performed for making scenarios with
CLD, SFD and the simulation.
12
4.6 Policy scenarios for J apanese polio issue
These authors conducted the following scenarios by using this methodology.
Scenario 1: Vaccination policy using Oral Polio Vaccine in usual status
The research developed a scenario that the vaccination policy using Oral Polio
Vaccine in usual status from a policy aspect by taking into accounts the result of
Simulation 1. This scenario means that we thought it was an important policy to
continue the vaccination with Oral Polio Vaccine to current Japanese polio issue.
Because, the polio epidemic is observed in some foreign countries, and the spread to
Japan is also undeniable. Therefore, it is necessary to keep the high immunized rate for
the prevention of polioviruses in Japan.
And the following are generally known information as prevention ways of keeping
from infectious disease,
Generally known information: Prevention of infection disease
Population-level Herd Immunity by vaccination
Water and sewerage
It was learnt that the current Herd Immunity did not cause polio epidemic over next
two years from the simulation results. And, it is known that the necessity of keeping
Herd Immunity level by vaccination to prevent infection disease in a population level.
As described above, the authors judged that keeping Herd Immunity in a group is the
most important thing from a policy aspect. Besides, the confirmation of the risk of
Inactivated Polio Vaccine in Japan was still ongoing. So, the necessity for the usage of
Inactivated Polio Vaccine with unknown risks and spending additional public expense is
not high priority. Therefore, in this research, it proposed as a prevention scenario in a
usual status was the necessary of continuing the vaccination using Oral Polio Vaccine.
However, it also would suggest that the Inactivated Polio Vaccine should be promptly
taken the place of Oral Polio Vaccine after evaluating its efficacy and safety in Japan.
Scenario 2: Delay expansion of infection disease in the epidemic status by decreasing
contact rate
From the results of Simulation 2 and 3, this research developed a scenario from a
policy aspect that is by controlling the contact rate in epidemic status, it would delay the
expansion of infection disease.
Details process of the scenario making is as follows. It had been understood that Herd
Immunity in Japan had been decreasing because of keeping from the vaccination of Oral
Polio Vaccine. And it had been learnt that the possibility of epidemic in an infant group
in 2011 was undeniable from the result of the Simulation 2. At the same time, by the
Simulation 2, it could be understood structurally how the infection was expanded.
Additionally, the authors researched population level countermeasures which were
generally known in an epidemic status.
The following are generally known as population level countermeasures against
13
epidemic.
Generally known information: Countermeasures against epidemic
Population-level Temporary closing of classes, School
closure
Canceling of meeting
This research selected the contact rate as a parameter that influenced the epidemic
process based on the result of the Simulation 2 and generally known countermeasures
for a population level in epidemic status. And, it performed the Simulation 3 to confirm
the effectiveness of contact rate in the simulation. As the result, it confirmed that
decreasing contact rate caused delay of expansion of infection. It identified this was a
factor which leads scenario from a policy aspect.
5 Discussion
This research proposed two scenarios from a policy aspect against the Japanese polio
issue
1) Inusual status
Continuation of the vaccination with Oral Polio Vaccine.
From the safety concern, recommendation of early introduction of the
Inactivated Polio Vaccine.
2) Inepidemic status
By decreasing contact rate, it is expected that controlling expansion of
infection in epidemic status.
These scenarios supported the generally known countermeasures from epidemiologic
researches. However, the behavior of the polio epidemic process could be recognized
quantitatively by the simulation results using SD model and Scenario Planning method
in this research.
By being clear the reasons why these scenarios were selected against the Japanese
polio issue, additional policy discussion can be expected to be activated. For instance,
they might be discussed below. Those are additional topics which lead by the
conversation with stakeholders. This is the secondary efficacy of using this
methodology.
+ Improvement of the environment of watching infection disease and information
transmission in domestic and foreign.
+ Operation procedure in case of epidemic happens and the way of keeping
everyone informed it.
+ Confirmation of time to secure the vaccine in emergency
+ Backup of forcibly isolate from the infection, traffic interception, work restriction
and etc. which are expected when the infection disease expanded.
In addition, the authors expect other applications of SSP to other topics in addition to
14
the case of Japanese polio issue. For instance, it may be an unemployment policy issue.
The model that workers move from current industry to other industry is similar to the
model of infection disease. Then the simulation results would show the transition
process visibly. And the results could lead a scenario against the unemployed issue by
finding the factor that controls the behavior of transition. After that, the scenario might
cause decreasing the insurance cost which is a finance pressure for the government.
Therefore, it is expected that the possibility of making scenarios based on data not
only the Japanese polio issue but also the unemployed issue and others by using
simulation results of SD and discussion.
6 Conclusion
This research was to propose a methodology for making scenarios called SSP. And it
tried to validate whether SSP lead the policy scenarios in the case of Japanese polio
issue. And it validated that SPP could lead some policy scenarios.
The SSP is a methodology composed of understanding of the structural clarification
of the problem with visible results from performing simulations based on System
Dynamics and scenario making through conversations which are an approach in the
process of Scenario Planning. This methodology generated the scenarios as the solution
derived from the simulation results for Japanese polio issue.
The flow of this research is that, firstly, it defined the issue with conversions. And it
performed simulations to know the epidemic process in a case of current Japan polio
issue. By performing the simulations, it found a parameter which would change the
behavior of the epidemic process in the simulation. Then it executed another SFD model
and simulation using the parameter that is contact rate. From the simulation results, it
visually and quantitatively recognized the parameter impacted the behavior of the
epidemic process. Therefore, the authors were able to lead the scenarios against the
Japanese polio issue from a policy aspect using the parameter.
The authors expected this methodology which we proposed in this research will
contribute to activate the discussion of making policy scenarios based on data.
7 Further research
As a further research, there might be other aspects and parameters that should be
discussed in this research. In This research, we generated two scenarios against the
vaccination of Oral Polio Vaccine in Japan from a policy aspect. But, for instance, In
addition to the policy aspect, there may be some aspects from health care practitioners
such as medical doctors and the nurse who related to the infectious disease treatment, or
an aspect from parents of infants. Moreover, although this research specified the contact
rate for the parameter, other parameters should be discussed using different models
which cause change of the behavior in the epidemic process. The authors would like to
reflect them into the models in the further research.
References
Andrea M. Bassi (2004) Strategic Analysis Evolution: scenario planning and
simulation based on the methodology of System Dynamics, In Proceedings of the
22th International Conference of the System Dynamics Society, Oxford, England,
UK. System Dynamics Society.
Fine PEM. (1993) Herd immunity: history, theory, practice, Epidemiologic Reviews
15(2), 265-302.
Fleisher, C. and Bensoussan, B. (2002) Strategic and Competitive Analysis: Methods
and Techniques for Analyzing Business Competition, Upper Saddle River, NJ,
Prentice Hall, 284-297.
Homer, J. and Hirsch, S. (2006) System Dynamics Modeling for Public Health:
Background and Opportunities, American Joumal of Public Health, 2006 March;
96 (3), 452-458.
John D. Sterman. (2000) Business Dynamics: Systems thinking and modeling for a
complex world, New York, Irwin/McGraw-Hill.
Kees van der Heijden (1996; 2004 2nd edition) Conversation Scenarios: The Art of
Strategic Conversation, Chichester & New York; John Wiley & Sons
Kermack,W.O. and McKendrick,A.G. (1927) A contribution to the mathematival
theory of epidemics, Proceedings of Royal Society A, 115, 700-721.
Krystyna A. Stave. (2002) Using System Dynamics to Improve Public Participation in
Environmental Decisions, System Dynamics Review Vol.18 No.2 (Summer 2002),
139-167.
Meadows, Donella H. and Jenny M. Robinson (1985) The Electronic Oracle:
Computer Models and Social Decisions, Chichester.
Ministry of Health, Labour and Welfare (2011) Annual estimation of vital statistics in
2011 [in Japanese]. See
<http://www.mhlw.go.jp/toukei/saikin/hw/jinkou/suikei11/dl/honbun.pdf>
Ministry of Health, Labour and Welfare (2011) Transition of the number of inoculation
persons of poliomyelitis oral vaccine [In Japanese], The 2nd investigative
commission for introduction of an _ inactivation polio vaccine. See
<http://www.mhlw.go.jp/stf/shingi/2r9852000001nbw-att/2r9852000001rlfm.pdf>
Nishimura, H. Kakehashi, M. and Inaba, H. (2009) Two critical issues in quantitative
16
modeling of communicable diseases: Inference of unobservables and dependent
happening Chapter 3. In: Mathematical and Statistical Estimation Approaches in
Epidemiology, Chowell G, Hyman JM, Bettencourt LMA, Castillo-Chavez C Eds.,
Springer, 53-87.
Pruyt, E. (2007) Dealing with Uncertainties? Combining System Dynamics with
Multiple Criteria Decision Analysis or with Exploratory Modeling. In Proceedings
of the 25th International Conference of the System Dynamics Society, Boston,
MA, USA. System Dynamics Society.
Pruyt, E. and C. Hamarat (2010) The Influenza A(H1N1)v Pandemic: An Exploratory
System Dynamics A pproach, In Proceedings of the 28th International Conference
of the System Dynamics Society, Seoul, Korea.
Thompson KM, el al. (2006) Development and consideration of global policies for
managing the future risks of poliovirus outbreaks: insights and lessons learned
through modeling. Risk Anal., 26(6), 1571-80.
Yang, Y. el al. (2009) The Transmissibility and Control of Pandemic Influenza A
(H1N1) Virus, Science, Published Online doi: 10.1126/science.1177373.
17
