Conference PaperPDF Available

Abstract and Figures

p>Singapore is undergoing an epidemiological shift and has to provide services for an aging population with a higher burden of chronic disease. In order to address this challenge, enhancing the provision of primary care by improving the ability of more primary care providers to offer care to more complex patients over the continuum of needs is seen as a promising solution. Developing capabilities and capacities of primary care services is far from straightforward and requires careful analysis of how increasing the number of primary care providers with enhanced capabilities influences the multiple objectives of the health care system. The paper demonstrates how group model building can be used to facilitate this planning process, and provides potentially valuable initial insights regarding the tradeoffs engendered by policies aimed at meeting the health care needs of a more complex population.</p
Content may be subject to copyright.
Proceedings of the 2016 Winter Simulation Conference
T. M. K. Roeder, P. I. Frazier, R. Szechtman, E. Zhou, T. Huschka, and S. E. Chick, eds.
David B. Matchar
John. P. Ansah
Peter Hovmand
Steffen Bayer
Department of Health Services and Systems
Social System Design Lab, Brown School
Washington University
Duke-NUS Medical School
8 College Road
St. Louis, MO 63130, USA
Singapore is undergoing an epidemiological shift and has to provide services for an aging population with
a higher burden of chronic disease. In order to address this challenge, enhancing the provision of primary
care by improving the ability of more primary care providers to offer care to more complex patients over
the continuum of needs is seen as a promising solution. Developing capabilities and capacities of primary
care services is far from straightforward and requires careful analysis of how increasing the number of
primary care providers with enhanced capabilities influences the multiple objectives of the health care
system. The paper demonstrates how group model building can be used to facilitate this planning process,
and provides potentially valuable initial insights regarding the tradeoffs engendered by policies aimed at
meeting the health care needs of a more complex population.
Singapore, like many other countries, is aging rapidly. Less than a decade ago, the proportion of the
resident population aged 65 years and older was less than 10%; this is expected to nearly triple by 2030
(Singapore Department of Statistics 2015; Singapore Department of Statistics 2016; National Population
and Talent Division 2012). Accompanying this demographic shift is a growing number of people with
illnesses requiring chronic care, such as diabetes and hypertension (Ministry of Health 2015a). These
conditions can progress to become complex as complications, such as heart disease and stroke, become
manifest. A consequence is that the acute hospitals and specialist services that have been the focus of a
health system designed for a younger population have been showing signs of systemic stress. This is
evident in the high bed occupancy rates in acute hospital (Lam 2014; Ministry of Health 2016a),
overburdened specialty outpatient clinics (SOCs) (Ministry of Health 2001), increasing emergency
department (ED) utilization (Anantharaman 2008), long waiting times for admission to acute hospital
wards (Ministry of Health 2016c), as well as workforce shortages (Ministry of Health 2012; Kang and
Leong 2012). Hospital admissions due to diabetes, for instance, are much higher in Singapore (432 per
100,000 population) than in countries such as the United Kingdom (64 per 100,000) or the United States
(149 per 100,000) (OECD 2016).
To adapt to the evolving needs of a population with chronic conditions, many countries have been
reassessing their healthcare system to focus on chronic care management by the “front line” providers,
namely primary care physicians and teams (World Health Organization 2002). The various models of
healthcare delivery that have been adopted include changes in the role and incentives of the private sector
978-1-5090-4486-3/16/$31.00 ©2016 IEEE 2123
Matchar, Ansah, Bayer, and Hovmand
and general practitioners (GPs), changes in subsidies and the use of empanelment to assign individual
patients to designated primary care providers or care teams (Department of Health 2014; Grumbach and
Olayiwola 2015; Ham 2009). These are among the dominant policy alternatives under consideration in
The challenge of meeting the needs of a rapidly changing population is complicated by the many
moving parts with complex interactions (e.g., hospital services, outpatient services, private and public
providers, and so on), and the many stakeholders whose interests must be considered if a major policy
change is to be successful.
In this paper, we describe our strategy for approaching the challenge of meeting the growing need for
chronic care through enhanced primary care, as a dynamic problem. We focus on initial work completed,
in which we have worked in partnership with community stakeholders to begin a facilitated discussion on
how investments in primary care services and regulatory changes will potentially affect overall health
system success. A summary model is presented which illustrates potentially important insights that can
inform the policy discussion.
Singapore is an island city-state with a total population of 5.5 million people, of which 3.9 million are
citizens or permanent residents. To serve their needs, the healthcare system in Singapore is a hybrid of
public and private elements that deliver primary, acute and step-down care in various institutions island
wide. An estimated 80% of acute hospital care is in the government hospitals; in contrast, 80% of primary
outpatient services are provided by private practitioners (Ministry of Health 2016b). Eighteen public
polyclinics and about 1,500 private GP clinics largely operating as solo or small group practices provide
the primary care services and people are free to choose venue of care (Ministry of Health 2015b).
Polyclinics act as “one-stop” healthcare centers, offering comprehensive range of services, including
outpatient medical care, health screening, education and vaccinations, and x-ray and laboratory services
(Ministry of Health 2015b); however, their capacity to address a full range of needs of an individual with
multiple health problems can be limited by high volume and relatively short consultation times. Private
GP clinics are generally run by solo practitioners and usually do not possess a full range of services. On
the other hand, patients in a private clinic usually see the same physician whereas in polyclinics patients
are in most cases assigned any available doctor each time they visit (Khoo, Lim, and Vrijhoef 2014). Not
infrequently, patients seek care at multiple sites.
Not only are more hospitalizations occurring in the public sector, public polyclinics also get a
disproportionate share of patients requiring complex, chronic care. While 20% of primary care attendees
are seen by public polyclinics, nearly half of patients with conditions requiring chronic care are seen in
the public sector (Sng 2010). This may be attributed to healthcare financing in Singapore, which
encourages transition to public services as needs become more complex. Healthcare financing in
Singapore is a combination of out-of-pocket payments, payments from a personal medical savings
account (Medisave) and subsidies. Public polyclinics are highly subsidized by the government. Singapore
citizens referred to specialty care from a polyclinic will also continue to receive subsidized treatment and
medication, up to 75 percent of the bill (Ministry of Health 2014). While private GP services have
historically not been subsidized, there is a move to make chronic care in the private sector more
affordable. In the past 10 years, individuals have been able to use their Medisave accounts to pay for
specific chronic conditions, subject to the GPs’ participation in the Chronic Disease Management
Programme introduced by the Ministry of Health (MOH). More recently, Singaporeans from lower-to
middle-income households eligible for the Community Health Assist Scheme (CHAS) receive subsidies
for medical care at participating GP clinics (CHAS 2016). The concern remains that patients with greater
needs continue to migrate to the public sector due to the larger range of services available (e.g., a patient
with previously mildly symptomatic heart failure will tend to go to the ED with the first onset of severe
Matchar, Ansah, Bayer, and Hovmand
shortness of breath as all X-rays and lab tests are available and highly subsidized; thereafter they are
likely to remain within the public polyclinics and/or SOCs).
3.1 The Objective: What is a Successful Health System?
As a first step in this exercise to project the impact of various policies intended to enhance primary care,
we established the key outcomes for measuring success of a health care system. We chose to apply the
Quadruple Aimframework developed by the Institute for Healthcare Improvement (Institute for
Healthcare Improvement 2016). That is, policy options are assessed in terms of the degree to which
alternatives achieve an optimal mix of service effectiveness (to improve population health), patient
satisfaction, service efficiency (to minimize need for expensive services, reducing per person costs), and
provider satisfaction (to minimize staff burnout and turnover).
3.2 Preliminary Work: Group Model Building
We applied group model building (GMB) to gain insights into the dynamic forces that promote or inhibit
the development and uptake of enhanced primary care, which in turn could be useful in guiding actions
that would make enhanced primary care effective and sustainable. GMB is a participatory method for
involving stakeholders in the process of understanding and changing systems using the methodology of
system dynamics (Vennix 1992; Scott, Cavana, and Cameron 2016; Forrester 1961; Homer and Hirsch
2006). System dynamics is a simulation modeling method used to represent the structure of complex
systems in an analytic framework to facilitate understanding systems behavior over time. The
involvement of various stakeholders with vast knowledge in the continuum of care, particularly in
primary care, in the development of a system dynamics model increases the relevance and usefulness of
the model. This provides a strong framework for analytic deliberation and testing of various hypotheses
(Scott, Cavana, and Cameron 2016; Sterman 2000). Examples of its application in healthcare include the
work of Vennix and Gubbels (1992), and Homa et al. (2015).
To better understand the dynamics of chronic disease management in a primary care setting, a 2-day
roundtable workshop was organized around a GMB exercise. The workshop was comprised of 50
stakeholders and experts providing chronic care to patients or with an interest in chronic care
management, including private GPs, polyclinic doctors, hospitals administrators, and government policy
makers, as well as academic health services researchers. Patient groups were not involved in this stage,
but will be engaged in the following stages of model development.
Day 1 of the workshop began with 11 interactive presentations covering a wide range of experience
on enhanced primary and chronic disease care in the United States, United Kingdom and Singapore. After
this, a panel discussion consisting of three local GPs was conducted. The presenters and stakeholder
participants generally agreed that primary care capacity and capability is not adequate to serve the local
population over the medium- to long-term and that the essential features of enhanced primary care would
include: (1) providing the first point of contact for patients, (2) offering a broad range of services for
patients with chronic conditions, and (3) coordinating patient care across venues.
On day 2, participants were divided into breakout sessions consisting of 6-8 individuals with a mix of
backgrounds. Each breakout group had a facilitator, modeler, recorder, and reflector. Their task was to
identify key causal relationships that impacted the ability of enhanced primary care to affect health, cost,
and satisfaction of patients and providers. They were encouraged to use causal loop diagrams to represent
their hypotheses, and these diagrams were then presented to all the participants and examined in an often
highly animated plenary discussion. A summary model capturing the various insights gained from the
GMB exercise was later developed by the research team. For purposes of the current presentation, we
describe a simplified version of the summary model, which captures major issues that might promote or
Matchar, Ansah, Bayer, and Hovmand
inhibit the successful development of an enhanced primary care sector (i.e., one that achieves the declared
aims of a modern healthcare system).
3.3 Model Structure
In this section, we describe the simplified summary model illustrating key insights from the workshop.
Specifically, it reflects the hypothesized causal relationship between the provision of services and the
quadruple aims” of a health care system, noted above (Figure 1). Since the two types of primary care
services (normal and enhanced) are represented in identical structures, for clarity only the single common
structure is shown; technically the separation is accounted for using variable subscripting. Table 1 shows
the parameter values used to initialize the model.
Figure 1: Causal diagram. Boxes indicate accumulations (stocks), black arrows are flows into or out of
stocks. Blue arrows indicate causal relationships. The quantities in red indicate the “quadruple aim” of the
healthcare system: population health (here represented by the proportion of the population with complex
conditions), patient satisfaction, service efficiency (cost per person in the total population), and provider
satisfaction (here represented by the doctor-patient relationship).
heal thy, at r isk
stable chronic
chronic condition
bir ths
incidence rate
rate progression
mortali ty
heal thy
mortali ty rate
heal thy
mortali ty
stabl e
mortali ty rate
stabl e
mortali ty
mortali ty rate
clinic volume
patient satisfaction
change in
attractiveness of
out of pocket
relative patient
initial patient
uptake rate of
service gap
relative service
time t o change
patient satisfaction
waiting time
waiting time
unit c ost of
number of
lengt h of
cons ulta tio n
max time per
doctor patient
relationship building
relative length
of consultation
frequency of seeing
same doctor
time to adj ust
relative doctor
patient relationship
initial doctor
patient relationship
relative frequency of
seeing same doctor
service gap
regression rate
total c ost of
service provisi on
total c ost of
proportion of
popula tion wit h
complex condition
cost per
total labor
cost per
rate healthy hospitalization rate
stable chronic
hospitalization rate
cost of
average patient
average doctor
pati ent re lati onship
cos t per
pers on
Matchar, Ansah, Bayer, and Hovmand
3.3.1 Population Health
In considering how primary care relates to population health, the participants agreed on the notion that
individuals can be considered as occupying health states of increasing levels of severity (Homa et al.
2015). This is reflected in Figure 1 by the three stocks: (a) healthy, at risk; (b) stable chronic condition;
(c) complicated chronic condition. Each stock corresponds to the nature (type and intensity) of medical
service needs that would be expected to inhibit progression into or regression from more symptomatic or
disabling health states (represented by black flows). Chronic conditions are taken to be an aggregate of
all types of chronic diseases. This “needs-based” (as opposed to “disease-based”) perspective reflects the
fact that many health conditions are associated with similar type and level of needs (e.g., care of patients
with chronic obstructive pulmonary disease (COPD) and heart failure both entail close longitudinal
follow up by a physician, chronic medications, and the potential for sudden, life-threatening exacerbations
requiring emergency care). Moreover, the majority of individuals with chronic conditions have several
conditions and these occur in a myriad of combinations, many of which can be treated in groupings (e.g.,
asymptomatic diabetes, hypertension, and mild chronic kidney disease often occur together, and require
similar levels of management skills and intensity (how often seek medical care).
Progression refers to the transition into health states with higher needs, with greater instability,
untoward impact on quality of life, and higher costs (particularly hospitalization), while regression refers
to the reverse The rates of flow are represented byvalves(here by double triangles) which are
influenced by the degree to which service needs are met. Specifically, service gaps lead to higher rates of
progression and lower rates of regression.
3.3.2 Patient Satisfaction
In the lower right section of the causal diagram indicates the relationships influencing patient satisfaction.
Patient satisfaction is represented as a stock that can increase or diminish. Stakeholders deemed that the
main determinant of satisfaction is out-of-pocket-cost; this is consistent with published literature (Derose
and Petitti 2003; Himmel, Dieterich, and Kochen 2000; Kulu-Glasgow, Delnoij, and de Bakker 1998).
Additional factors influencing change in patient satisfaction with care is waiting time for care (Sloan and
Kasper 2008), quality of the doctor-patient relationship (Anderson, Camacho, and Balkrishnan 2007), and
perceived quality of care by patients (Williams, Weinman, and Dale 1998). For simplicity, this last factor
is assumed to be directly related to meeting medical needs (i.e., as service gaps diminish, change in
patient satisfaction becomes more positive. This section of the diagram also includes a feedback loop: all
things being equal, patients who are satisfied with care are more likely to find the venue attractive which
leads to higher patient volume, and, in turn, longer waiting times and lower patient satisfaction. This is a
balancing loop, inducing a constraining force on clinic growth.
Patient satisfaction was deemed to be related, in part, to the quality of the doctor-patient relationship.
This relationship is importantly related to the number of doctors relative to demand for services
(Alrubaiee and Alkaaida 2011) since relationships are established through longer consultation time
(Murray 2007), and the consistency of the relationship over time (i.e., ability to see the same doctor)
(Howie 1991). Stakeholders felt that current consultation time are too short and thus consistent doctor-
patient relationships are not possible in local public polyclinics. Note that this establishes another
feedback loop that balances clinic growth: if a clinic becomes more attractive by increasing consultation
time and access to the same doctor, resulting increased clinic volume forces consultation times to be
reduced and decreases the chance that a person’s doctor will be available when needed and so will need to
see other doctors, reduces patient satisfaction and clinic attractiveness.
3.3.3 Service Efficiency
Overall costs of chronic care services include two major components. The first is the cost of providing
Matchar, Ansah, Bayer, and Hovmand
clinic services, a cost that is dominantly driven by numbers of providers. The second component of cost is
services that may be reduced by high-quality care. As noted above, filling service gaps reduces
progression of chronic conditions. This progression not only has a profound effect on quality of life, it is
also expensive. More medically complex individuals are more likely to use emergency services, be
hospitalized, and require expensive specialty care.
Evidence suggests that as length of consultation decreases, fewer of the needs of patients with chronic
conditions can be addressed during a visit, which impacts quality of care (Fan 2005). In addition,
integration of chronic care services among various providers, and capability of care providers influence
quality of care (Murray 2007; Campbell 2001). While addressing these problems with chronic disease
management has immediate costs, the question is whether it will ultimately reduce net costs, or at least
lead to extra costs deemed worth the health benefits.
3.3.4 Provider Satisfaction
Here the level of the doctor-patient relationship is used as an indicator of provider satisfaction. This stock
represents crucial characteristics our stakeholder physicians valued in their day-to-day work life: the
ability to have an adequate amount of time to see patients, and the opportunity to follow the same patient
over time.
3.4 Model Inputs
3.4.1 Model Parameters
Steady state:
Because the model is intended for policy exploration, stylized numbers are used to initialize the model.
To make it easier for policy analysis, the model is initialized in a steady state (i.e. a hypothetical situation
in which population, service gap, cost of care, patient satisfaction and doctor-patient-relationship are
equal and constant across the two care venues normal and enhanced primary care). To initialize the
model into steady state the following assumptions were made: (1) the population in the three health state
are assumed to remain constant; (2) there is no difference in the effectiveness of care (a proxy for service
gap) provided across the two venues; (3) both care venues have the same number of providers (medical
staff) to care for patients; and (4) the out-of-pocket cost for patients across the care venues are the same. It
is important to note that the initial steady state is a dynamic equilibrium and is numerically sensitive to
model parameters, but typical consequent behavior is not. The parameters used in the model are shown in
Table 1.
3.4.2 Policy Experimentation
To illustrate the potential impact of policy changes, the simulation was run under four scenarios in which
the steady state was perturbed by changing three main model parameters (i.e. service gap, out-of-pocket
cost, and number of doctors) stepwise under the following scenarios:
I. Equilibrium: All key policy variables were kept constant. Under this scenario, effectiveness of
normal and enhanced primary care was assumed to be equal and remain unchanged over the simulation
time. Likewise, out-of-pocket cost, and number of doctors were assumed to be equal across the two care
venues and remain constant over the simulation. All outputs were expected to remain constant at their
steady state values and provide a reference trajectory.
II. Effective enhanced primary care: The effectiveness of receiving care at an enhanced primary
care venue was assumed to be significantly higher to that of normal primary care. This was implemented
by reducing service gap from 0.5 to 0.1 at time 5 while the service gap for normal primary care remained
at 0.5 over the simulation time.
Matchar, Ansah, Bayer, and Hovmand
III. Reduced out-of-pocket costs for primary care: In addition to introducing enhanced primary
care that could favorably affect clinical course, in this strategy the out-of-pocket costs for such care was
assumed to be half of normal care, in order to increase the attractiveness of seeking care in enhanced care
IV. Proactive increase of enhanced primary care providers: This scenario was the same as
scenario III except that number of providers of care at the enhanced primary care (here indicated by
number of doctors) is proactively increased as demand for enhanced primary care increases.
Table 1: Model Parameters.
Model Parameter
 [] = 0.5
 []= 0.5 + (0.4,5)
  ℎ = 0.1
   ℎ = 0.3
   = 1
  [] = 10
  [ℎ]=   ( = 1:  : 5,  [],10)
   =15
   =100
ℎ =350
  ℎ  =2500
   [] = 1
 ℎ,  =10000
  = 0.05
   [] = 50
   [ ]=100
   = 0.025
    =22500
   = 0.07
  ℎ = 0.01
   = 0.015
   [] = 50
   [ ]=50 +(25,5)
  = 0.12
  ℎ  =5000
    = 5
  ℎ   = 5
    =12
   ℎ = 3
    ℎ = 6
Figure 2 shows the impact of the four scenarios on the quadruple aim of the health care system
population health (proportion with complex conditions), per person cost, patient satisfaction, and provider
satisfaction (doctor-patient-relationship).
As expected, in scenario I, “equilibrium”, given the assumptions, all the outcome variables remained
constant over the simulation period.
Under scenario II, “effective enhanced primary care”, where enhanced primary care was assumed to
be 40% more effective than normal primary care, the proportion of the population with complex
conditions decrease by 88% over the simulation time. Initially, per person cost of care increased
following the sudden increase in effectiveness for enhanced primary care and decreased gradually over
time. While average patient satisfaction increased with improved quality of care, the average doctor-
patient-relationship actually decreased.
Under scenario III, “reduced out-of-pocket costs for primary care” is the same as scenario II with the
elimination of out of pocket costs to patients (thereby enhancing the attractiveness of enhanced primary
Matchar, Ansah, Bayer, and Hovmand
care to patients), the behavior of the outcome variables are similar to that of the effectiveness of care
scenario. Again, average patient satisfaction increased, now enhanced by a somewhat higher proportion of
the population receiving higher quality services and with this increase is seen a diminished doctor-patient
relationship as enhanced primary care venues become more crowded.
Lastly, under scenario IV, “proactive increase of enhanced primary care providers”, also includes a
provision for the increase in providers to meet growing service demand. Of all the alternative scenarios,
this leads to the best population health (i.e., lowest proportion of the population with complex conditions)
and highest costs. Notably it maintains high patient satisfaction as well as average doctor-patient-
Figure 2: Impact of the four scenarios on the quadruple aim of the health care system.
In this exercise, we demonstrated the application of simulation modeling incorporating GMB to address a
complex social issue involving potentially substantial change to the organization of crucial services. In
addition to engaging stakeholders, the resulting model provided some preliminary insights into the
dynamic impact of introducing a new model of primary care services intended to serve the needs of a
broader range of patients than typically provided in the current system. It reinforced the general intuition
that enhanced primary care services could improve outcomes, and that costs may, at least partially, be
offset by the reduced utilization induced by better population health. Reducing the out-of-pocket costs to
1 6 11 16 21 26 31 36 41 46 51 56 61
Proportion of population with complex
Equilibrium Effect ive enhanced PC
Low out-of-pocket cost Proactive incre ase in providers
1 6 11 16 21 26 31 36 41 46 51 56 61
Average patient satisfaction
Equilibrium Effect ive enhanced PC
Low out-of-pocket cost Proactive incre ase in providers
1 6 11 16 21 26 31 36 41 46 51 56 61
Cost per person
Equilibrium Effect ive enhanced PC
Low out-of-pocket cost Proactive incre ase in providers
1 6 11 16 21 26 31 36 41 46 51 56 61
Average doctor patient relationship
Equilibrium Effect ive enhanced PC
Low out-of-pocket cost Proactive incre ase in providers
Matchar, Ansah, Bayer, and Hovmand
patients had minor effects on population health or per person costs (not unexpected as costs were merely
shifted from the patient to the government); however, since patients are quite price sensitive, this
reduction did improve patient satisfaction by promoting their use of higher quality services. Moreover,
changing the model of care without increasing the number of providers in this illustration had the
undesirable consequence of decreasing doctor satisfaction as new venues become overwhelmed with
patients seeking enhanced services, reducing continuity of providers and consultation time, and thus
detracting from the doctor-patient relationship.
There are several limitations to the simple summary model. In particular, it represents only two key
players: the normal primary care providers and enhanced primary care providers. In reality, chronic care
services can be provided in multiple venues, including private GP offices, enhanced private sector clinics,
public sector polyclinics, and specialty clinics. Each sector will offer a different mix of services and thus
have different capacities to reduce service gaps, as well as different levels of attractiveness to patients.
Thus, a practical set of policy options will not merely include the expansion of existing services (in this
case, public polyclinic services) but also the encouragement through funding and regulation to promote
enhanced primary care in the private sector. Thus, a realistic causal model must account for the tendency
for individuals to vote with their feet: for doctors to participate in alternative models of chronic care
service. Further, particularly in a system involving a mix of public and private sectors providers, quality
of care will depend on their ability to integrate (i.e., share the care of the same patient) which requires a
level of trust and communication.
A second limitation is the dearth of available data on current patterns of health service use, and the
factors likely to influence both doctors and patients to participate in alternative care venues. The
conclusions based on more complete evidence may very well be quite different. Indeed, we see this work
as a first step in a coherent sequence of activities to promote informed decision-making regarding the
development of enhanced primary care in Singapore. Based on the current exercise, we identified two
areas where reliable information would be crucial to plausible projections of the impact of policy change
and for which existing data are not available: (1) epidemiological data on the proportion of the population
in various health needs segments and the rate of transition between segments, as a function of their
patterns of care; and (2) the desirability of venues of care for patients and primary care providers based on
characteristics of those venues which emerge from policy options. Thus, we are proceeding with
collecting these data via a population-based survey and a discrete choice experiment with a broadly
representative group of primary care physicians and individuals in various health needs segments. Since
patient groups were not represented in the first GMB exercise, we will seek input from patient groups in
the refinement of the model. In addition, we have engaged multiple representatives of the Regional Health
Systems (RHSs); the RHS serve as the units responsible for planning the public response to expanding
population health needs. The RHSs have agreed to partner with the research team to tailor the model to
unique conditions in each region.
Alrubaiee, L., and F. Alkaa'ida. 2011. "The Mediating Effect of Patient Satisfaction in the Patients'
Perceptions of Healthcare Quality-Patient Trust Relationship." International Journal of Marketing
Studies 3 (1):103.
Anantharaman, V. 2008. "Impact of Health Care System Interventions on Emergency Department
Utilization and Overcrowding in Singapore." International Journal of Emergency Medicine 1 (1):11-
Anderson, R. T., F. T. Camacho, and R. Balkrishnan. 2007. "Willing to Wait?: The Influence of Patient
Wait Time on Satisfaction with Primary Care." BMC Health Services Research 7 (1):1.
Campbell, J. L., J. Ramsay, and J. Green. 2001. "Practice Size: Impact on Consultation Length, Workload,
and Patient Assessment of Care." British Journal of General Practice 51 (469):644-650.
Matchar, Ansah, Bayer, and Hovmand
CHAS (Community Health Assist Scheme). "What Is CHAS (Community Health Assist Scheme)?".
Accessed 30 March 2016.
Department of Health. 2014. "Chronic Disease Management (Formerly Enhanced Primary Care or
EPC) GP (General Practitioner) Services." Last modified 2 April 2014. Accessed 22 March 2016.
Derose, S. F., and D. B. Petitti. 2003. "Measuring Quality of Care and Performance from a Population
Health Care Perspective." Annual Review of Public Health 24 (1):363-384.
Fan, V. S., M. Burman, M. B. McDonell, and S. D. Fihn. 2005. "Continuity of Care and Other
Determinants of Patient Satisfaction with Primary Care." Journal of General Internal Medicine 20
Forrester, J. W. 1961. Industrial Dynamics. Cambridge, Massachusetts: Massachussetts Institute of
Technology Press.
Grumbach, K., and J. N. Olayiwola. 2015. "Patient Empanelment: The Importance of Understanding Who
Is at Home in the Medical Home." The Journal of the American Board of Family Medicine 28
Ham, C. 2009. "Chronic Care in the English National Health Service: Progress and Challenges." Health
Affairs 28 (1):190-201.
Himmel, W., A. Dieterich, and M. M. Kochen. 2000. "Will German Patients Accept Their Family
Physician as a Gatekeeper?" Journal of General Internal Medicine 15 (7):496-502.
Hjortdahl, P., and E. Laerum. 1992. "Continuity of Care in General Practice: Effect on Patient
Satisfaction." British Medical Journal 304 (6837):1287-1290.
Homa, L., J. Rose, P. S. Hovmand, S. T. Cherng, R. L. Riolo, A. Kraus, A. Biswas, K. Burgess, H.
Aungst, and K. C. Stange. 2015. "A Participatory Model of the Paradox of Primary Care." The Annals
of Family Medicine 13 (5):456-465.
Homer, J. B., and G. B. Hirsch. 2006. "System Dynamics Modeling for Public Health: Background and
Opportunities." American Journal of Public Health 96 (3):452-458.
Howie, J., A. Porter, D. Heaney, and J. Hopton. 1991. "Long to Short Consultation Ratio: A Proxy
Measure of Quality of Care for General Practice." British Journal of General Practice 41 (343):48-54.
Institute for Healthcare Improvement. 2016. "IHI (Institute for Healthcare Improvement) Triple Aim
" Accessed 22 March 2016.
Jackson, C., J. Tsai, C. Brown, D. Askew, and A. Russell. 2010. "GPs with Special Interests: Impacting
on Complex Diabetes Care." Australian Family Physician 39 (12):972.
Kang, S. H., and C.-H. Leong. 2012. Singapore Perspectives 2012: Singapore Inclusive: Bridging
Divides: World Scientific
Khoo, H. S., Y. W. Lim, and H. J. Vrijhoef. 2014. "Primary Healthcare System and Practice
Characteristics in Singapore." Asia Pacific Family Medicine 13 (1):1.
Kiran, T., A. Kopp, R. Moineddin, and R. H. Glazier. 2015. "Longitudinal Evaluation of Physician
Payment Reform and Team-Based Care for Chronic Disease Management and Prevention." Canadian
Medical Association Journal:cmaj. 150579.
Kulu-Glasgow, I., D. Delnoij, and D. de Bakker. 1998. "Self-Referral in a Gatekeeping System: Patients’
Reasons for Skipping the General-Practitioner." Health Policy 45 (3):221-238.
Lam, P. M. 2014. "Bed Crunch." Last modified 20 January 2014. Accessed 23 March 2016.
Ministry of Health. 2001. "Waiting Times in Specialist Outpatient Clinics of Restructured Hospitals and
Specialty Centres." Last modified 20 March 2015. Accessed 23 March 2016.
Matchar, Ansah, Bayer, and Hovmand
Ministry of Health. 2012. "MOH (Ministry of Health) 2012 Committee of Supply Speech Healthcare
2020: Improving Accessibility, Quality and Affordability for Tomorrow's Challenges (Part 1 of 2)."
Last modified 6 March 2016. Accessed 22 March 2016.
Ministry of Health. 2014. "Higher Subsidies and Pioneer Generation Benefits for Subsidised Treatment at
Specialist Outpatient Clinics in Public Hospitals from 1 September 2014." Last modified 22 August
2014. Accessed 26 February 2016.
Ministry of Health. 2015a. "Disease Burden." Last modified 30 July 2015. Accessed 22 February 2016.
Ministry of Health. 2015b. "Primary Healthcare Services." Last modified 2 January 2015. Accessed 23
March 2016.
Ministry of Health. 2016a. "Beds Occupany Rate (BOR)." Accessed 23 March 2016.
Ministry of Health. 2016b. "Introduction to Healthcare Institution Statistics." Accessed 23 March 2016.
Ministry of Health. 2016c. "Waiting Time for Admission to Ward." Accessed 23 March 2016.
Murray, M. 2007. "Panel Size: How Many Patients Can One Doctor Manage?" Family Practice
Management 14 (4):44.
National Population and Talent Division. 2012. "Issues Paper 2012 - Our Demographic Challenges and
What These Means to Us." Accessed 23 March 2016.
OECD. 2016. "Health Care Quality Indicators: Primary Care." Accessed 23 March 2016.
Scott, R. J., R. Y. Cavana, and D. Cameron. 2016. "Recent Evidence on the Effectiveness of Group
Model Building." European Journal of Operational Research 249 (3):908-918.
Singapore Department of Statistics. 2015. "Population Trends 2015." Singapore.
Singapore Department of Statistics. 2016. "Latest Data." Last modified 17 March 2016. Accessed 23
March 2016.
Sloan, F. A., and H. Kasper. 2008. Incentives and Choice in Health Care. Boston, Massachusetts:
Massachussetts Institute of Technology Press
Sng, Q. S. 2011. "Primary Care Survey 2010 Profile of Primary Care Patients." Ministry of Health
Sterman, J. 2000. Business Dynamics: Systems Thinking and Modeling for a Complex World. Boston:
Thorsen, O., M. Hartveit, and A. Baerheim. 2012. "General Practitioners’ Reflections on Referring: An
Asymmetric or Non-Dialogical Process?" Scandinavian Journal of Primary Health Care 30 (4):241-
Matchar, Ansah, Bayer, and Hovmand
Vennix, J. A. M. 1996. Group Model Building: Facilitating Team Learning Using System Dynamics.
New York;Chichester, England;: John Wiley.
Williams, S., J. Weinman, and J. Dale. 1998. "Doctor-Patient Communication and Patient Satisfaction: A
Review." Family Practice 15 (5):480-492.
DAVID B. MATCHAR is Professor of Medicine at Duke University and Director of the Program in
Health Services and Systems Research at Duke-NUS Medical School. He holds a MD from University of
Maryland. He focuses his research on the evaluation of clinical practice based on “best evidence,” and the
implementation and evaluation of innovative strategies to promote practice change. His email address is
JOHN P. ANSAH is an Assistant Professor in the Program in Health Services and Systems Research at
Duke-NUS Medical School. He holds a Ph.D. in the System Science methodology of System Dynamics
from University of Bergen. His research interests lie broadly in developing healthcare strategy and
planning simulation models that are rigorous, evidence-based and customized to the optimal usefulness of
stakeholders to inform policy. His email address is
STEFFEN BAYER received his Ph.D. from the University of Sussex, and currently is Assistant
Professor in the Program in Health Services and Systems Research at Duke-NUS Medical School. His
research interests include planning of health services, in particular the use of simulation modeling. His
email address is
PETER HOVMAND is the founding director of the Brown School’s Social System Design Lab. He
received his Ph.D. from Michigan State University. His research focuses on using participatory group
model building methods to involve communities and other stakeholders in the process of understanding
systems and designing solutions using system dynamics models and computer simulations with specific
emphasis on promoting social justice. His email address is
... Other references, such as [31,47], investigate macrosystem reforms and feature entire primary care systems. Still, the agent-based model [31] differs from SiM-Care in its objective: It investigates the external effects of treatments in primary care on the entire health care system, whereas SiM-Care focuses on the processes within primary care systems. ...
... Hence, [31] does not model internal processes, such as appointment scheduling. Model [47] implements the system dynamics paradigm and thus focuses on a higher level system representation than SiM-Care. While system dynamics models do not consider the level of microdetail offered by agent-based simulations, they require less computational effort to run simulation experiments. ...
... The modeling team regularly consulted with health care practitioners including primary care physicians, health insurance representatives, as well as representatives from industry associations and administrative authorities. Generally, we find that explaining the simulation model through the agent-based paradigm and presenting results from related studies allows Cooperation between academics and patients, caregivers, and clinicians Model, software, and worksheets available for download and discussion [47] SD Entire primary care sector ...
Full-text available
Primary care systems are a cornerstone of universally accessible health care. The planning, analysis, and adaptation of primary care systems is a highly non-trivial problem due to the systems’ inherent complexity, unforeseen future events, and scarcity of data. To support the search for solutions, this paper introduces the hybrid agent-based simulation model SiM-Care. SiM-Care models and tracks the micro-interactions of patients and primary care physicians on an individual level. At the same time, it models the progression of time via the discrete-event paradigm. Thereby, it enables modelers to analyze multiple key indicators such as patient waiting times and physician utilization to assess and compare primary care systems. Moreover, SiM-Care can evaluate changes in the infrastructure, patient behavior, and service design. To showcase SiM-Care and its validation through expert input and empirical data, we present a case study for a primary care system in Germany. Specifically, we study the immanent implications of demographic change on rural primary care and investigate the effects of an aging population and a decrease in the number of physicians, as well as their combined effects.
... Yet, our participants also felt that a short consultation time appeared to limit trust and rapport building with patients. Such time constraints were identified in earlier studies locally (49,50) and globally (23,51), with an average consultation time to be less than 5 minutes across different healthcare settings (52). Literature suggests that short consultation times are associated with poorer health outcomes for patients such as polypharmacy and antibiotic overuse (53,54) whereas longer consultation length is likely to result in reduced hospital admission (52) and better disease control in patients with diabetes (55). ...
Full-text available
Introduction The perspectives of healthcare professionals (HCPs) are pivotal to co-development of self-management strategies for patients with diabetes. However, literature has been largely limited to perspectives of patients within the context of a Western healthcare setting. This study aims to explore factors influencing diabetes self-management in adult patients with diabetes from the perspectives of HCPs and their views of the value of mHealth application for diabetes self-management. Materials and Methods We conducted focus group discussions (FGD) with purposively selected HCPs in Singapore. All FGDs were audio-recorded and transcribed verbatim. Thematic analysis was conducted using NVivo 12. Results A total of 56 HCPs participated in the study. Barriers to self-management included limited patient commitment to lifestyle changes, suboptimal adherence to medication and treatment, patient resistance to insulin initiation and insufficient rapport between patients and HCPs. Patients’ perceived susceptibility to complications, social support from family and community, multidisciplinary team care and patient’s understanding of the benefits of self-care were viewed as facilitating self-management. HCPs saw mHealth apps as a vital opportunity to engage patients in the self-management of conditions and empower them to foster behavior changes. Yet, there were concerns regarding patient’s limited digital literacy, lack of integration into routine electronic system and reluctance. Discussion We identified a set of factors influencing self-management in adult patients with diabetes and useful app features that can empower patients to manage their conditions. Findings will inform the development of a mHealth application, and its features designed to improve self-care.
... Some studies provided descriptions about how they facilitated this input, which ranged from structured and active methods where stakeholders were asked specific questions (31) or engaged in purposeful storytelling exercises (32), to unstructured and passive methods where stakeholders provided feedback about or annotated an existing model (30,40). More structured methods of facilitation were used in early stages when studies were engaging stakeholders in designing the model from scratch (24,31,32,36,37,44,53), and more passive methods were used when stakeholders were engaged at a later stage and a draft model had already been designed (16,30,35,38,40,48). Further details about the Modes of Engagement & Facilitation are found in Appendix C. ...
Full-text available
Objective: To identify processes to engage stakeholders in healthcare Simulation Modeling (SM), and the impacts of this engagement on model design, model implementation, and stakeholder participants. To investigate how engagement process may lead to specific impacts. Data Sources: English-language articles on health SM engaging stakeholders in the MEDLINE, EMBASE, Scopus, Web of Science and Business Source Complete databases published from inception to February 2020. Study Design: A systematic review of the literature based on a priori protocol and reported according to PRISMA guidelines. Extraction Methods: Eligible articles were SM studies with a health outcome which engaged stakeholders in model design. Data were extracted using a data extraction form adapted to be specific for stakeholder engagement in SM studies. Data were analyzed using summary statistics, deductive and inductive content analysis, and narrative synthesis. Principal Findings: Thirty-two articles met inclusion criteria. Processes used to engage stakeholders in healthcare SM are heterogenous and often based on intuition rather than clear methodological frameworks. These processes most commonly involve stakeholders across multiple SM stages via discussion/dialogue, interviews, workshops and meetings. Key reported impacts of stakeholder engagement included improved model quality/accuracy, implementation, and stakeholder decision-making. However, for all but four studies, these reports represented author perceptions rather than formal evaluations incorporating stakeholder perspectives. Possible process enablers of impact included the use of models as “boundary objects” and structured facilitation via storytelling to promote effective communication and mutual understanding between stakeholders and modelers. Conclusions: There is a large gap in the current literature of formal evaluation of SM stakeholder engagement, and a lack of consensus about the processes required for effective SM stakeholder engagement. The adoption and clear reporting of structured engagement and process evaluation methodologies/frameworks are required to advance the field and produce evidence of impact.
... Simulation models aimed at investigating macrosystem reforms of primary care systems mostly include an entire primary care system, however they are usually much more high level. Matchar et al. [48] use the methodology of system dynamics to develop a simulation model to aid primary care planning in Singapore. The model captures the causal relationships between the stakeholders' aims and the provision of services in an analytical framework. ...
Full-text available
Demand for health care is constantly increasing due to the ongoing demographic change, while at the same time health service providers face difficulties in finding skilled personnel. This creates pressure on health care systems around the world, such that the efficient, nationwide provision of primary health care has become one of society's greatest challenges. Due to the complexity of health care systems, unforeseen future events, and a frequent lack of data, analyzing and optimizing the performance of health care systems means tackling a wicked problem. To support this task for primary care, this paper introduces the hybrid agent-based simulation model SiM-Care. SiM-Care models the interactions of patients and primary care physicians on an individual level. By tracking agent interactions, it enables modelers to assess multiple key indicators such as patient waiting times and physician utilization. Based on these indicators, primary care systems can be assessed and compared. Moreover, changes in the infrastructure, patient behavior, and service design can be directly evaluated. To showcase the opportunities offered by SiM-Care and aid model validation, we present a case study for a primary care system in Germany. Specifically, we investigate the effects of an aging population, a decrease in the number of physicians, as well as the combined effects.
Full-text available
An ageing population, with increasing prevalence of diabetes and hypertension, is expected to increase the number of people with chronic kidney disease (CKD) and end-stage renal disease (ESRD) needing dialysis. This paper explores the impact of upstream and downstream interventions on the future number of CKD, ESRD patients needing dialysis, and the cost of dialysis. A system dynamics model was developed based on Singapore national data. Results indicate that under the base case scenario the number of people with CKD is projected to increase from 437,338 in 2020 to 489,049 by 2040. As a result, the number of patients requiring dialysis is projected to increase from 7669 in 2020 to 10,516 by 2040. The cost of dialysis care, under the base case, is projected to increase from S$417.08 million in 2020 to S$907.01 million by 2040. The policy experiments show that a combined policy will cumulatively save S$1.042 billion from 2020 to 2040.
Full-text available
Purpose: The paradox of primary care is the observation that primary care is associated with apparently low levels of evidence-based care for individual diseases, but systems based on primary care have healthier populations, use fewer resources, and have less health inequality. The purpose of this article is to explore, from a complex systems perspective, mechanisms that might account for the effects of primary care beyond disease-specific care. Methods: In an 8-session, participatory group model-building process, patient, caregiver, and primary care clinician community stakeholders worked with academic investigators to develop and refine an agent-based computer simulation model to test hypotheses about mechanisms by which features of primary care could affect health and health equity. Results: In the resulting model, patients are at risk for acute illness, acute life-changing illness, chronic illness, and mental illness. Patients have changeable health behaviors and care-seeking tendencies that relate to their living in advantaged or disadvantaged neighborhoods. There are 2 types of care available to patients: primary and specialty. Primary care in the model is less effective than specialty care in treating single diseases, but it has the ability to treat multiple diseases at once. Primary care also can provide disease prevention visits, help patients improve their health behaviors, refer to specialty care, and develop relationships with patients that cause them to lower their threshold for seeking care. In a model run with primary care features turned off, primary care patients have poorer health. In a model run with all primary care features turned on, their conjoint effect leads to better population health for patients who seek primary care, with the primary care effect being particularly pronounced for patients who are disadvantaged and patients with multiple chronic conditions. Primary care leads to more total health care visits that are due to more disease prevention visits, but there are reduced illness visits among people in disadvantaged neighborhoods. Supplemental appendices provide a working version of the model and worksheets that allow readers to run their own experiments that vary model parameters. Conclusion: This simulation model provides insights into possible mechanisms for the paradox of primary care and shows how participatory group model building can be used to evaluate hypotheses about the behavior of such complex systems as primary health care and population health.
Full-text available
It is crucial to adapt and improve the (primary) health care systems of countries to prepare for future patient profiles and their related needs. The main aim of this study was to acquire a comprehensive overview of the perceptions of primary care experts in Singapore about the state of primary care in Singapore, and to compare this with the state of primary care in other countries. Notwithstanding ranked 2nd in terms of efficiency of health care, Singapore is facing significant health care challenges. Emails were sent to 85 experts, where they were asked to rate Singapore’s primary care system based on nine internationally adopted health system characteristics and six practice characteristics (response rate = 29%). The primary care system in Singapore received an average of 10.9 out of 30 possible points. Lowest ratings were given to: earnings of primary care physicians compared to specialists, requirement for 24 hr accessibility of primary care services, standard of family medicine in academic departments, reflection of community served by practices in patient lists, and the access to specialists without needing to be referred by primary care physicians. Singapore was categorized as a ‘low’ primary care country according to the experts.
Population health care is health information and clinical services provided to individuals of a defined population. From a population health care perspective, quality of care involves the health status of the entire population, and thus issues of access, cost of care, and efficiency matter. In this paper, we describe the definitions of quality health care and the framework for measuring quality, with emphasis on the performance of organizations involved in the delivery and assurance of population health care. We describe quality measurement sets and systems, criteria for the choice of measures, data sources, and how quality measurements are used to improve health care and outcomes from a population health care perspective.
Variations in practice list size are known to be associated with changes in a number of markers of primary care. Few studies have addressed the issue of how single-handed and smaller practices compare with larger group practices and what might be the optimal size of a general practice. AIM: To examine variations in markers of the nature of the care being provided by practices of various size. DESIGN OF STUDY: Practice profile questionnaire survey. SETTING: A randomised sample of general practitioners (GPs) and practices from two inner-London areas, stratified according to practice size and patients attending the practice over a two-week period. METHOD: Average consultation length was calculated over 200 consecutive consultations. A patient survey using the General Practice Assessment Survey instrument was undertaken in each practice. A practice workload survey was carried out over a two-week period. These outcome measures were examined in relation to five measures of practice size based on total list size and the number of doctors providing care. RESULTS: Out of 202 pratices approached, 54 provided analysable datasets. The patient survey response rate was 7247/11,000 (66%). Smaller practices had shorter average consultation lengths and reduced practice performance scores compared with larger practices. The number of patients corrected for the number of doctors providing care was an important predictor of consultation length in group practices. Responders from smaller practices reported improved accessibility of care and receptionist performance, better continuity of care compared with larger practices, and no disadvantage in relation to 10 other dimensions of care. Practices with smaller numbers of patients per doctor had longer average consultation lengths than those with larger numbers of patients per doctor. CONCLUSION: Defining the optimal size of practice is a complex decision in which the views of doctors, patients, and health service managers may be at variance. Some markers of practice performance are related to the total number of patients cared for, but the practice size corrected for the number of available doctors gives a different perspective on the issue. An oversimplistic approach that fails to account for the views of patients as well as health professionals is likely to be disadvantageous to service planning.
Background: We evaluated a large-scale transition of primary care physicians to blended capitation models and team-based care in Ontario, Canada, to understand the effect of each type of reform on the management and prevention of chronic disease. Methods: We used population-based administrative data to assess monitoring of diabetes mellitus and screening for cervical, breast and colorectal cancer among patients belonging to team-based capitation, non-team-based capitation or enhanced fee-for-service medical homes as of Mar. 31, 2011 (n = 10 675 480). We used Poisson regression models to examine these associations for 2011. We then used a fitted nonlinear model to compare changes in outcomes between 2001 and 2011 by type of medical home. Results: In 2011, patients in a team-based capitation setting were more likely than those in an enhanced fee-for-service setting to receive diabetes monitoring (39.7% v. 31.6%, adjusted relative risk [RR] 1.22, 95% confidence interval [CI] 1.18 to 1.25), mammography (76.6% v. 71.5%, adjusted RR 1.06, 95% CI 1.06 to 1.07) and colorectal cancer screening (63.0% v. 60.9%, adjusted RR 1.03, 95% CI 1.02 to 1.04). Over time, patients in medical homes with teambased capitation experienced the greatest improvement in diabetes monitoring (absolute difference in improvement 10.6% [95% CI 7.9% to 13.2%] compared with enhanced fee for service; 6.4% [95% CI 3.8% to 9.1%] compared with non-team-based capitation) and cervical cancer screening (absolute difference in improvement 7.0% [95% CI 5.5% to 8.5%] compared with enhanced fee for service; 5.3% [95% CI 3.8% to 6.8%] compared with non-teambased capitation). For breast and colorectal cancer screening, there were no significant differences in change over time between different types of medical homes. Interpretation: The shift to capitation payment and the addition of team-based care in Ontario were associated with moderate improvements in processes related to diabetes care, but the effects on cancer screening were less clear.
Group model building (GMB) is a participatory approach to using system dynamics in group decision-making and problem structuring. This paper considers the published quantitative evidence base for GMB since the earlier literature review by Rouwette et al. (2002), to consider the level of understanding on three basic questions: what does it achieve, when should it be applied, and how should it be applied or improved? There have now been at least 45 such studies since 1987, utilising controlled experiments, field experiments, pretest/posttest, and observational research designs. There is evidence of GMB achieving a range of outcomes, particularly with regard to the behaviour of participants and their learning through the process. There is some evidence that GMB is more effective at supporting communication and consensus than traditional facilitation, however GMB has not been compared to other problem structuring methods. GMB has been successfully applied in a range of contexts, but there is little evidence on which to select between different GMB tools, or to understand when certain tools may be more appropriate. There is improving evidence on how GMB works, but this has not yet been translated into changing practice. Overall the evidence base for GMB has continued to improve, supporting its use for improving communication and agreement between participants in group decision processes. This paper argues that future research in group model building would benefit from three main shifts: from single cases to multiple cases; from controlled settings to applied settings; and by augmenting survey results with more objective measures.