Next Article in Journal
Gender Differences in Kinematic Parameters of Topspin Forehand and Backhand in Table Tennis
Previous Article in Journal
Media Coverage of Pedophilia: Benefits and Risks from Healthcare Practitioners’ Point of View
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
IJERPH
Volume 17
Issue 16
10.3390/ijerph17165741
ijerph-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

A System Dynamics Simulation Applied to Healthcare: A Systematic Review

by
Mohammad Reza Davahli
1,*,
Waldemar Karwowski
1 and
Redha Taiar
2
1
Department of Industrial Engineering and Management Systems, University of Central Florida, Orlando, FL 32816, USA
2
Université de Reims Champagne-Ardenne, 51100 Reims, France
*
Author to whom correspondence should be addressed.
Int. J. Environ. Res. Public Health 2020, 17(16), 5741; https://doi.org/10.3390/ijerph17165741
Submission received: 21 June 2020 / Revised: 4 August 2020 / Accepted: 6 August 2020 / Published: 8 August 2020
Browse Figures
Review Reports Versions Notes

Abstract

:
In recent years, there has been significant interest in developing system dynamics simulation models to analyze complex healthcare problems. However, there is a lack of studies seeking to summarize the available papers in healthcare and present evidence on the effectiveness of system dynamics simulation in this area. The present paper draws on a systematic selection of published literature from 2000 to 2019, in order to form a comprehensive view of current applications of system dynamics methodology that address complex healthcare issues. The results indicate that the application of system dynamics has attracted significant attention from healthcare researchers since 2013. To date, articles on system dynamics have focused on a variety of healthcare topics. The most popular research areas among the reviewed papers included the topics of patient flow, obesity, workforce demand, and HIV/AIDS. Finally, the quality of the included papers was assessed based on a proposed ranking system, and ways to improve the system dynamics models’ quality were discussed.

1. Introduction

Complex and intractable healthcare problems, as well as limited progress in discovering effective solutions, continue to challenge the healthcare industry [1,2]. In complex healthcare systems, the mutual effects of different variables are subtle, while the impact of interventions over time is not immediately identifiable [3,4,5,6,7]. Furthermore, many internal and external factors that influence healthcare outcomes are often nonlinear [3,4,5,6,7]. Under these conditions, simulation models can help to elucidate the counterintuitive behavior of complex healthcare problems [6]. Simulation modeling is considered to be a wise option because the complexities of problems are far beyond our capacity to solve them manually [6]. According to Sterman [6], it is challenging to develop accurate healthcare simulation models because there is no reliable and well-understood principle in this area. However, the substantial benefits of developing and applying simulation models are learning about complex problems and testing different interventions [6].
Operations research (OR) approaches, especially simulation methods, have been used to specify complex problems in healthcare since the 1960s [8]. Many healthcare aspects, such as extended care, specialty care, public health, hospital operation, rehabilitation, and long-term care, have been subject to such computer simulations in the past [9]. Simulation methods can be categorized into four main groups: Monte Carlo (MC), discrete-event simulation (DES), system dynamics (SD), and agent-based simulation (ABS) [10]. Each type of simulation has its advantages concerning specific aspects of healthcare [10]. For example, DES has been commonly used in medical emergency system applications, while SD was found useful in epidemic modeling and disease prevention efforts [10,11]. Other factors may also affect the applicability of the simulation method, including the goals, significance of the feedback loop, simulation time, amount and quality of the input data, and type of healthcare units [10,12]. Generally speaking, matching the most suitable simulation method with a specific aspect of healthcare is not an easy task [10].
System dynamics (SD) simulation emerged in late 1950 as a result of focusing on the behavior of complex systems over a specific period [11]. The main features of SD simulation are a feedback loop, stocks and flow function, and time delay [11]. These features have been used to model the emerging nonlinearity of systems behavior [11]. In SD, dynamic objects move though flows and accumulate in the stocks [13]. At each specified time, stocks calculate and report the quantitative status of the objects [13]. Feedback loops and delays in SD can create a dynamic system [13]. Therefore, if the change occurs anywhere in the system, it will result in a chain reaction throughout the system [13]. Furthermore, SD studies can be carried out qualitatively or quantitatively [10]. For example, if the objective of the study is to explain the problem in a better way or reliable quantitative information is not available, the study can be conducted qualitatively [10].
In general, the SD model can be developed through specific steps, including a clear explanation of the problem, generating a qualitative diagram of the system structure, converting the qualitative hypothesis to a quantitated simulation model, testing the model, and informing policy decisions about the model’s implications [14]. The main advantages of SD simulation are discovering the emergent properties and characteristics of a system, creating quantitative analysis for qualitative problems, identifying the most important system parameters, predicting the long-term effects of decisions, and helping stakeholders learn about the nature of their problems [15]. The main disadvantages of SD simulation are the inability to model low-level interactions and overreliance on qualitative validation [15].
The main objective of the present study is to review recent papers that utilize the SD method to analyze a variety of healthcare problems. We attempt to provide an overview of SD’s applications in healthcare by discussing different modeling approaches, procedures, and quality. This review is structured as follows: The first part presents the main outcomes of published review papers in healthcare. The Methodology section discusses the inclusion and exclusion criteria and the risk of bias. The Results section provides the outputs of the literature search, while the Discussion section explains healthcare aspects and research areas of studies. Finally, the quality of the reviewed papers is evaluated in the Quality Assessment section.

2. Evaluation of Review Papers

To date, there have been several literature reviews regarding the applications of simulation methods in healthcare. Some of the highly cited literature reviews published between 2000 and 2019 are shown in Table 1.
Two methodologies were used to summarize the published papers of interest: (a) the literature review, and (b) the systematic literature review. Of the 19 papers considered here, six papers used the literature review method, while 13 papers used the systematic literature review method. Of the systematic review papers, only He and associates [21] used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for reporting a systematic review. The majority of the papers review multiple simulation methods, and almost half of the papers concern the overall healthcare area. The average and median numbers of included papers among these reviews were 270 and 182, respectively. More specifically, the maximum number of papers used in the published reviews was 1441 [17], while the minimum number of papers was six [23].
Web of Science was the most popular database among the selected review papers, followed by PubMed, Science Direct, and Google Scholar. The outcomes of the review papers can be classified into five distinct groups, including (a) the categorization of the included papers based on country, publication source, and year; (b) grouping by areas of research; (c) comparing simulation techniques; (d) discussing trends of topics or trends of simulation techniques; and (e) describing specific areas of research in detail.
Two major sections in the selected review papers are suggestions for research and study limitations. In the suggestions portions of the papers, many of the review papers offer interesting topics for future research, while others attempt to describe their upcoming studies. Additionally, a few review papers suggest novel methodologies for conducting a literature review. The limitations sections in the review papers can be categorized into different groups. The first limitation is identifying and selecting papers published in a specific time frame. The second limitation is the inability to identify individual relevant records resulting from the use of a restricted number of search keywords. The third limitation is not covering all search databases, while the fourth limitation is restricting the search to specific regions or countries.

3. Methodology

The present systematic literature review followed PRISMA guidelines [30] and contains two main features: developing research questions and determining the search strategy. The following research questions were formulated for this review:
RQ1.
What are the main problems in the healthcare sector that have been studied in the past using the SD approach?
RQ2.
How was the SD approach utilized to model and address complex healthcare problems?
RQ3.
What can be learned from the past SD simulation research in the healthcare area that will support high-quality research in the future?
To answer the above research questions, a search strategy was developed to identify and review the relevant scientific papers. The search strategy included (a) defining keywords and identifying the relevant materials, (b) filtering the articles written by two authors, and (c) addressing the risk of bias [30]. As mentioned earlier, one of the main concerns in a systematic review is developing comprehensive and accurate keywords. The main objective of the present systematic review was to target all aspects of healthcare, such as medicine, management, costs, systems, technology, safety, and system operations. Since this is a broad and multi-faceted topic, it is important to generate unique keywords that cover all aspects of healthcare, including health, medical care, healthcare, and the healthcare system. Since classical medical research focuses on disease, health is defined as the absence of disease [31]. Referring to the definition and metrics of health described by the World Health Organization (WHO) and the global burden of disease (GBD), the main health-related words are diseases, healthy, life expectancy, morbidity, aging, and illness [31].
Health care, healthcare, or health-care is the improvement or maintenance of health via the diagnosis, prevention, and treatment of illness, disease, injury, and other mental and physical impairments in individuals [32]. Physicians, health professions, primary care, secondary care, tertiary care, and public health are part of healthcare [32]. Medical care is defined as the professional treatment of disease, illness, or injury [33]. A health care system, health system, or healthcare system is an institution or organization of individuals and resources that deliver health-related services to respond to the needs of target populations [34]. In conclusion, we developed the first set of keywords based on the presented definitions, as shown in Table 2.
Web of Science and Google Scholar were used as database search tools in the present review. The search conducted in this review was divided into two steps: (a) the first set of keywords was used to discover relevant articles (1162 articles have been identified and added to the central database), and (b) a new set of keywords was developed based on the identified papers, which was used to discover additional articles. The second set of keywords is shown in Table 3. The second set of keywords was developed to identify papers that did not mention specific keywords—including health, medical, disease, or illness—but whose contents are related to our review. Finally, 1231 articles with relevant content were identified.
After the development of the main database and the identification of all relevant papers, a formal screening process was applied to the database based on the exclusion and inclusion criteria. The inclusion criteria were as follows:
Articles with SD models or an SD procedure/framework;
  • Articles related to healthcare;
  • Articles published from 2000 to 2019;
  • Articles related to research questions;
  • Articles written in English.
The exclusion criteria were as follows:
  • Papers written in other languages;
  • Chapters of books;
  • Review papers;
  • Articles from secondary sources that were not free or open access;
  • Letters, newspaper articles, viewpoints, presentations, anecdotes, duplicated studies, short papers, and posters.
In this systematic review, bias could occur by (a) applying inclusion/exclusion criteria and (b) specifying the dimensions of the included papers. To address the first type of bias, two researchers (authors) independently reviewed the titles, abstracts, and conclusions of the identified records to choose articles for full-text review. Subsequently, the two authors compared their selected articles to create a unified list. After reading the full texts of the selected articles, the authors decided whether or not to include the articles. If there was an agreement between the two researchers, the article was considered and included. Disagreements between the two authors were resolved in sessions with the third author. To address the second type of bias, two authors independently specified aspects of healthcare and modeling in the included papers then compared the results, resolving disagreements by consulting with the third author.

4. Results

A chart of the selection strategy according to PRISMA guidelines is displayed in Figure 1. As a result, the present review included a total of 253 papers written by 940 authors. On average, a single paper was written by four authors. As shown in Figure 2, the trend of papers included increased over time.
Interestingly, only 15% of the included papers were published between 2000 and 2009, and there was a significant increase in the number of included papers after 2013. To confirm these results, we first compared the trend of included papers with that of unscreened records, as shown in Figure 3.
Secondly, we compared our results with those of other literature reviews. Kunc and associates [16] indicate that there was a significant and increasing trend in SD publications after 2010 in the areas of health improvement, health services, health epidemics, and health policy-making, although the trends were significantly more dramatic than those observed in our results. Torres [17] presented the number of papers cited in the System Dynamics Review Journal and found that only 37% of the citations were from 2000 to 2009.
Of the 253 included papers, 207 were published in journals and 46 were presented as conference papers. The most popular sources of publication among the included papers, in order, were the Journal of the Operational Research Society, System Dynamics Review, Winter Simulation Conference, PLoS One, Systems Research and Behavioral Science, and Health Care Management Science.

5. Discussion

The papers included in the present review are categorized into different aspects of healthcare, as shown in Table 4. The most popular categories among the included papers are healthcare operations, communicable diseases, non-communicable diseases, and healthcare systems. The most popular subcategories among the included papers are patient flows, HIV/AIDS, obesity, and workforce demand.

5.1. Patient Flow

The most popular research area among the reviewed papers is patient flow. Since healthcare facilities are under pressure to decrease delays and reduce costs, an investigation into streamlining patient flow is beneficial [35]. The objectives of the included papers in this area can be classified into analyzing the length of stay, studying the shift, delayed discharge, and developing SD models. Patient flow was analyzed in the context of admission, emergency departments, cardiac catheterization services, elderly patients, and hospitals in general. For SD model simulation, information held by hospital databases and staff was used for model calibration. The most common simulation time for patient flow modeling was one day to 2 weeks, and the most common simulation scenarios were changes in the number of beds, the value of demand, and the number of hospital staff. Almost half of the reviewed papers in this area used case studies, and the majority of the papers were quantitative studies.
It has been reported that more than 75% of instances of delayed discharge were a result of delays in making long-term care arrangements for elderly patients [36,37]. Several studies evaluated the impact of hospital resources on the length of stay [35,38,39], while others focused on elderly individuals, chronically ill patients, and their impact on bed occupancy, the average length of stay, and resource consumption [40,41]. The main conflicting results are the impact of changing bed capacity on performance. A study by Kumar [42] indicates that optimizing bed capacity could minimize patient waiting time in an urban country hospital. Lane and associates [39] conclude that reducing the number of beds does not increase the waiting time for emergency admissions. At the same time, Taylor and Dangerfield [43] mention that increasing bed capacity is not a practical solution for improving the average waiting time in cardiac catheterization services. Moreover, Rashwan and associates [37] report that increasing bed capacity in post-acute care services has a temporary impact and is insignificant in the long term. A study by German and associates [44] evaluated healthcare services in the Philippines and indicates that hospital bed requirements could be satisfied by 2021. Several studies analyzed decision-making in healthcare and blamed (a) applying informal policies, (b) focusing on short-term results, and (c) using a single-performance measure for problems in the process of decision-making [43,45].
The key variables of the simulation models include the patients awaiting discharge, inpatients waiting for rehabilitation, patients in care, length of stay, admission rate, bed capacity and demand, and waiting lists. The main output variables are the average daily occupancy, patient waiting time, number of patients, number of delayed discharges, and post-acute accessibility. The core part of the stock-flow diagram among the included papers for the patient flow model is illustrated in Figure 4.
In order to develop an expanded SD model, auxiliary variables such as the average daily arrivals, referrals, and walk-in admissions can be linked to the patient admission/arrival rate flow. Different auxiliary variables can be linked to patients in hospital or patients on hospital wards such as the numbers of occupied beds, patients, and empty beds. A variety of auxiliary variables can be connected to, and have an impact on, the patient treatment rate such as the treatment time, treatment conditions, and size of hospital staff. The patient in assessment for discharge can be linked to different auxiliary variables, including the inpatient discharge, the patient needs for other treatment, the patient needs convalescence, and the patient needs rehabilitation. The core part of the causal loop diagram for the reviewed papers is shown in Figure 5.

5.2. Obesity

Another popular research area in healthcare is obesity. Increasing rates of obesity and the incidence of being overweight are putting pressure on the economy and health services [46]. Many diseases and conditions, such as type 2 diabetes mellitus, coronary heart disease, cancer, hypertension, and musculoskeletal disorders, are associated with obesity [46]. The included papers investigate obesity among children [47,48,49,50], women of reproductive age [51], individuals in low-income urban areas [52], and lower-socioeconomic status (SES) groups [53,54]. The objectives of the selected papers can be categorized into two main groups: (a) reducing obesity (reactive responses) and (b) presenting interventions for obesity prevention (proactive responses).
Papers report different results for reducing and preventing obesity. For example, Abdel-Hamid [46] indicates that combining physical activity and diet composition is critical for effective obesity treatment; Abidin and associates [55] emphasized that efficient eating behavior to prevent obesity is a combination of decreasing food portions, controlling meal frequency, and limiting the intake of high-calorie foods. Furthermore, Roberts and associates [50] suggested implementing a comprehensive combination of different interventions to decrease obesity, including increasing the availability of healthy food, obesity prevention programs, bans on junk food advertising, and subsidies for healthy food. Allender and associates [56] discussed the causes of obesity in four domains: fast food and junk food, social influences, physical activity, and participation in sports. Although a study places more emphasis on building outdoor activity spaces for children [48], it was also reported that physical education is more effective than physical activity in reducing obesity among children [49]. Interestingly, Chen and associates [54] concluded that enhancing an individuals’ status from lower- to middle-income levels would be effective in controlling obesity and being overweight.
The most popular scenarios among the reviewed papers are changing the sensitivity of weight to diet, healthy food availability, weight loss rates, the availability of educational programs, and preventing fast food advertising. The key variables among the reviewed papers include the average daily energy intake, the expected weight of a user, change in expected weight, energy expenditure, energy metabolism, body composition, the effect of diet on weight, the sensitivity of weight to exercise, overweight adults, obese adults, normal-weight children, overweight children, obese children, physical activity, changes in average weight per year, participation in sports, junk food consumption, levels of physical activity, healthy weight, healthy food choices, and mental health. The main output variables of the reviewed papers in this area include weight, change in weight, carbohydrate intake and oxidation, fat intake and oxidation, change in physical activity, the number of overweight and obese individuals, and the prevalence of being overweight and obesity. The results indicate that qualitative studies, building causal loop diagrams, and integrating SD with other simulation methods are not common in obesity papers. The core part of the stock-flow diagram among the included papers for the obesity model is shown in Figure 6.
To expand the core model, different auxiliary variables can be linked to the weight loss rate and weight gain rate, including changes in average weight per day/month, engaging in weight loss/gain behaviors, daily energy balance, and healthy weight.

5.3. HIV/AIDS and Tuberculosis

Studying the application of SD in HIV/AIDS and tuberculosis (TB) is popular among researchers. In many countries, AIDS is the main cause of death in the adult population and a significant source of excessive demands on existing treatment and medical resources [57]. With the high cost of HIV/AIDS and the absence of medical solutions, behavioral approaches and prevention programs are the main solutions for preventing the spread of the epidemic and decreasing individual exposure to HIV [57]. The objectives of these studies can be categorized into different groups: (a) analyzing the impacts of different therapies and programs on diseases, (b) studying HIV care services, and (c) studying effective interventions.
Some of the main programs are mentioned in the following paragraph. Several studies have investigated the consequences of highly active antiretroviral therapy (HAART) [58,59,60]. HAART is the most suitable treatment for advanced HIV disease in developed countries and works by preventing the virus from replicating [58]. Other studies have investigated multidrug-resistant tuberculosis (MDRTB) control programs, HIV harm-reduction programs, and evidence-based HIV intervention programs [57,61].
Interestingly, almost all the papers mention model calibration, and different types of data are used for this purpose (e.g., European Network for HIV/AIDS Surveillance (ECDC) data [58], and data from literature reviews, the Estonian Ministry of Social Affairs, and the National Institute for Health Development [61]). The average simulation time among the papers is 10.5 years. The scenarios of the selected papers include changing therapy cure rates, populations sharing needles (PSH), the drug injection frequency (DIF), the preventive-program population coverage, the percentage of the target population with easy access to the program, the proportions of motivated and unmotivated members, the length of the programs, the proportions of non-drug users and active drug users, implementing new needle-exchange programs, financial hardship, program budgets, health education, and psychological counseling.
While Dangerfield and associates [58] indicate that implementing HAART could decrease new HIV infections, a study by Lebcir and associates [60] states that the decrease in HIV- and TB-related deaths would be small for HAART coverage of up to 50% of individuals, and that coverage of 70% or greater would be more effective. Lounsbury and associates [62] highlight that access to antiretroviral therapy (ART) would improve outcomes for HIV patients, while other studies report that upon the implementation of effective harm-reduction programs, the number of HIV-associated deaths would decrease by 30% [61]. Furthermore, Atun and associates [61] stated that effective harm-reduction programs would more effectively decrease the cumulative deaths from TB than effective multidrug-resistant tuberculosis (MDRTB) control programs. Lebcir and associates [60] did not mention effective harm-reduction programs, concluding that a high MDRTB cure rate has decreased the deaths from TB. Miller and associates [57] noted that shortening the evidence-based HIV prevention program would increase the number of individuals finishing the program. Zou and associates [63] indicate that the most efficient scenarios that decreased HIV infection involved condom promotion, needle-exchange programs, and psychological counseling. Lastly, a study by Shariatpanahi and associates [64] on disease awareness distribution indicated that disease awareness distribution for AIDS and tuberculosis is much faster than for breast cancer and autism.
The key variables in the developed models include the program participants, undiagnosed HIV-positive individuals with AIDS, family awareness, the support of family and friends, undiagnosed HIV, and diagnosed HIV. The main output variables include the annual HIV incidence, cumulative TB deaths, cumulative deaths from AIDS, cumulative MDRTB deaths, cumulative deaths from AIDS in non-TB individuals, and HIV infections and incidences.

5.4. Workforce Demand

One of the main research areas of healthcare is workforce demand. The objectives of the selected papers on workforce demand are planning and predicting the number of workforces, including medical specialists [65,66], the pediatric workforce [67], the cardiac surgery workforce [68], dentists [69], the ophthalmologist workforce [70], the radiology specialization workforce [71], physicians [72,73], registered nurses [74], and physical therapists [75]. Interestingly, the average simulation time among the selected papers in this area is 24.5 years, which is higher than that for previous research areas of healthcare. The main scenarios among these papers are changing the level of population growth, the medical school graduation rate, the national exam pass rate, medical demand, an aging population profile, the physician-to-population ratio, the level of student intake, and the age of retirement.
A study by Barber and López-Valcárcel [65] indicates that the shortage of medical specialists in Spain would increase from 2% in 2010 to 14.3% in 2025 with moderate population growth. Wu and associates [67] mention that the predicted decline in the supply of Taiwanese pediatricians could be delayed by interventions, while Vanderby and associates [68] model the future Canadian cardiac surgery workforce with different scenarios and a predict short-term excess of cardiac surgeons by 2030. Samah and associates [69] state that, according to the simulation, the total available Malaysian clinical dentists in the year 2030 will be fewer than the required dentists, while a study conducted about Singapore [70] reports that in all scenarios, the number of Singaporeans with eye conditions will double and thus there will be a significant increase in eyecare demand and a great need for ophthalmologists. Taba and associates [71] indicate that the Australian system will never meet the demand for radiologists in the simulation time. Ishikawa et al. [73] state that by 2020, the physician shortage in the Hokkaido Prefecture region of Japan will be resolved, and De Silva and associates [72] highlight that the shortage of doctors in Sri Lanka will be resolved by 2025. Lastly, Morii and associates [75] project that the number of physical therapists in Japan will be 1.74 times greater in 2025 and 2.54 times greater in 2040 than in 2014.
The critical variables among the models include immigration, emigration, the targeted workforce, the general population, births, deaths, graduated students and medical specialization schools, demand based on the population, and retirement. The main output variables among these papers include the number of specialists, population prediction, workforce demand and supply prediction, care supply, and demand. The core part of the stock-flow diagram for the included papers for workforce demand is shown in Figure 7.
Auxiliary variables can be linked to flows, including the application rate, immigrant workforce, national exam pass rate, graduation rate, national exam application rate, graduate school enrolment rate, rate of choosing a hospital as a workplace, and employment rate.

5.5. Chronic Diseases

The main area of non-communicable diseases is chronic diseases, defined as conditions that last a year or more and require ongoing medical attention and/or limit activities of daily living [76]. It has been reported that the use of the term “chronic disease” has large variation [77]. For example, the Centers for Disease Control (CDC) uses the term “chronic diseases” for cancer, type 2 diabetes, coronary heart disease, stroke, obesity, and arthritis; however, the Centers for Medicare and Medicaid Services include further conditions such as Alzheimer’s disease and depression [77]. It is reported that the same kind of variation in terminology exists in academic papers [77]. The authors of the reviewed papers take different paths to address chronic diseases, where some only focus on one area of chronic diseases—such as diabetes, cancer, or cardiovascular disease—while others address all aspects of chronic diseases. The objectives of this group of papers can be divided into different sets (1) projecting growth and the number of individuals with chronic diseases, (2) developing a simulation model of chronic disease, and (3) analyzing interventions, therapies, treatments, and programs to improve chronic care.
The average simulation time among the selected papers is 32.4 years, although a study by de Andrade and associates [78] used 1440 min as the simulation time in an SD model. The main scenarios of the included papers are changes in the intervals and intensity of treatments and therapies, stage-specific survival rates, patients and population size, adopting different levels of programs and interventions, and the existence of other conditions such as obesity.
A study by Tejada and associates [79] indicates that the best screening policy for breast cancer prevention is annual breast cancer screening for all women aged 65–80 years old, while Liew [80] emphasizes that either clinical breast examination (CBE) or mammographic screening has the potential to avoid some deaths from breast cancer. A study by Palma and associates [81] highlights that prostate-specific antigen (PSA) screening would be beneficial in decreasing prostate cancer mortality, while Hallberg and associates [82] mention important points for developing an SD model for patients with growing tumors, which were (a) categorizing patients based on a physiological point of view, and (b) considering a non-constant rate for tumor growth.
Kenealy and associates [83] concluded that implementing anti-smoking interventions has the potential to decrease cardiovascular events, while Lich and associates [84] identified a broad hypertension control effort to be the most effective intervention in increasing the quality and length of life (quality-adjusted life-years). De Andrade and associates [78] discussed the cause of delays in the treatment of ST-Elevation Myocardial Infarction (STEMI) patients, the main factors of which can be categorized as professional, equipment, and transportation logistics. A study by Recio and associates [85] investigates the impact of exposure to urban noise on the growth of cardiovascular mortality. Additionally, the improvement of chronic diseases has been investigated by implementing different interventions, such as increasing access to primary care [86], evaluating the behavior of cardiovascular drugs on the market [87], increasing income and employment, enhancing social cohesion and neighborhood attractiveness [88], and improving physician education and care coordination [89]. Mehrjerdi [90] analyzed the interconnections between health problems and weight, and found that teaching individuals about their health will have a considerable impact on the number of deaths from heart attacks.
Several studies indicate that the rates of diagnosed diabetes and prediabetes can be decreased by different interventions such as improving the clinical management of diabetes, reducing obesity prevalence, and innovative healthcare delivery, while other studies project that the population with diabetes will increase in the U.S. and Japan [91,92].
The key variables among the reviewed papers are access to treatment/therapies, age, adherence to treatment/therapies, the rate of chronic disease diagnoses, chronic disease mortality, moderate and severe stages of chronic disease, the population in different age groups, population growth, and the population with disease. The main output variables among the reviewed papers include the treatment/therapy cost, number of disease incidents and diagnoses, number of deaths, and fraction of the population with the disease or those at-risk. The core part of the stock-flow diagram for chronic diseases is displayed in Figure 8.
Some studies implement the entire core part of the stock-flow diagram [84,89,91,93,94,95], while other studies implemented only the diagnosed part (blue box), in which there is an additional flow from the population to diagnosed patients [82,96,97,98]. Furthermore, several of the selected studies only implemented the undiagnosed-to-diagnosed part of the stock-flow diagram (orange box) without expanding on different stages of chronic diseases [79,80,81,83,85,87,92,99]. Depending on the chronic disease, different auxiliary variables can be linked to the onset and progression flows, including the smoking prevalence, the use of primary care, stress, poor diet, physical activity, the transition rate between the different stages of chronic disease, the management of chronic disease, the elderly portion, and the obese portion. Screening, testing, and detection variables can be linked to the diagnosis flow.

5.6. Hybrid Modeling

One of the main research areas in healthcare is hybrid modeling and the comparison of simulation methods. The main objective of the selected papers is to discuss the integration or comparison of SD with other simulation methods in a healthcare context. Interestingly, all papers are qualitative studies, with the majority being conference papers published at the Winter Simulation Conference. The main results of the reviewed papers include (a) the advantages of using hybrid models [100,101,102], (b) selecting an appropriate simulation technique for different healthcare problems [103,104,105], and (c) presenting a framework for integrating simulation methods [106,107,108,109,110,111,112].

5.7. Chlamydia

Chlamydia—one of the research areas of communicable diseases—was studied by five papers [113,114,115,116,117]. The main objectives of the selected papers were to analyze interventions and screening programs. Interestingly, the simulation time in four of the five papers was 1 year, and all were quantitative studies. Several studies analyzed the potential cost–benefit of chlamydia screening programs [113,114,115,117]. The key stock variables include the different stages of the disease, susceptibility to sequelae, transmission incidents, and recovery. The main output variables among the papers are the cases of infertility prevented per year, cases of pelvic inflammatory disease (PID) prevented per year, number of sequelae vs. the screening rate, infection prevalence, screening rate, and per-capita cumulative cost. The core part of the stock-flow diagram for the reviewed papers for chlamydia is shown in Figure 9. It is possible to expand the core stock-flow diagram to cover multiple stages of the disease and low- and high-risk groups.

5.8. Other Research Areas

A handful of papers present SD models as the main results and outcome, including for a methadone maintenance treatment [118], water regulation in the human body [119], depression [120,121], fatigue [122], alcohol misuse [123,124], urban community health services [125], construction workers’ health behavior [126], Hispanic immigrant worker health [127], patient satisfaction [128,129], urban health policy-making [130], healthy eating [131], oral health [132], the total costs associated with hip joint endoprosthesis [133], mobile stroke units [134], occupational health and safety (OHS) performance [135,136], and electronic health information exchange (HIE) [137,138].

6. Quality Assessment

To build guidelines for SD applications in healthcare, we targeted (a) the categorization and analysis of relevant papers and (b) an assessment of the quality of SD models. The assessment criteria were adopted from a previous study since they provide an analytical and quantitative approach to analyzing the reviewed papers [3]. However, since those criteria were developed for quantitative studies with simulation models, the present criteria were adjusted to cover (a) qualitative and quantitative studies and (b) papers with and without SD models. The list of the quality criteria is shown in Table 5.
A good quality paper is one that has clear objectives and results, provides detailed information about developing the SD model, presents model variables, and, in the case of quantitative studies, describes scenarios, model validation, outcome variables, and the source of the input data. The quality of the papers is categorized into four groups: “good”, “medium”, “low”, and “very low”. The results of this quality assessment are shown in Figure 10.
It should be noted that more than 50% of the reviewed papers belong to the good- and medium-quality categories. Furthermore, the observed trend in quality has not increased over the 20 years. In fact, in the last five years, there has been a considerable number of papers classified as “bad” and “very bad” quality. Generally, papers belonging to the aspects of communicable diseases, non-communicable diseases, and disorders and stress exhibited higher quality than those belonging to other healthcare aspects, which may indicate evidence-based practice in these areas. It is important to note that we only assessed the information given in the papers and their additional/attached files. Moreover, we found that some highly cited papers did not present their models, and therefore, we adjusted the criteria to assess the papers based on model description or model presentation and not only based on model presentation.
The clarity of the study objectives is vital to modeling efforts and is related to the type of problems under consideration [6]. The reviewed papers’ objectives can be divided into different groups, such as developing models, analyzing elements/problems in a system, improving a system with different factors, and projecting elements in a system.
The quality of data was assessed based on two parameters: (a) the engagement of experts and stakeholders in terms of the survey, interview, discussion, community-based model development, and group model building, and (b) the presentation of a source of data for model calibration. A total of 65 out of the 253 papers engaged experts and stakeholders in their studies or used interviews or surveys in their research.
The reviewed papers have been categorized into different groups: presenting stock-flow diagrams, presenting causal loop diagrams, presenting SD frameworks, and describing SD model development. A total of 26 out of the 253 papers present no SD model, 11 of these 26 papers describe different frameworks for hybrid modeling, and the remaining 15 papers describe model development. The reviewed papers present 139 causal loop diagrams and 249 stock-flow diagrams. A total of 32 out of the 253 papers only present the causal loop diagram, 74 present both the casual loop diagram and the stock-flow diagram, and 121 only present the stock-flow diagram.
The time horizon is one of the main aspects of stock-flow diagram development and should go back far enough in history to show the emergence of problems and their consequences and symptoms. The time horizon must also extend far enough into the future to elucidate the effects of different policies and delays [6]. The choice of time horizon strongly affects the perception and assessment of problems and the evaluation of policies [6]. The included papers use different time horizons in their research, from a half of a day [139] to 120 years (1920–2040) [85], as shown in Figure 11.
The model validation can be a problematic task and should, therefore, be performed by engaging stakeholders and experts during the model-building process, where the stakeholders should understand and trust the behavior and internal structure of the model [40]. Relying too much on qualitative validation is one of the weaknesses of SD [140]. Hence, we divided model validation among the selected papers into three groups:
  • Direct structural tests, including empirical and theoretical tests;
  • Structure-oriented behavioral tests, such as extreme condition tests and sensitivity tests;
  • Behavioral pattern tests, considered to be model calibration [140].
It should be noted that out of the 235 papers selected for this review, 187 studies were carried out quantitatively. A total of 115 out of these 187 papers discussed calibration tests by comparing the model results with different sources, including (a) actual/historical/experimental data, and/or (b) data extracted from the literature reviews, and/or (c) data from surveys. A total of 57 reported direct structural tests, including empirical tests and theoretical tests. Finally, 104 studies performed sensitivity tests by evaluating the response of the model to fluctuation in the values of certain parameters.
A total of 187 quantitative papers contain a total of 707 and an average of four outcome variables in the forms of charts, tables, and graphs. The cost, number, and percentage of specific parameters over time are the most common forms of the outcome variable. The most popular outcome variables of the reviewed papers are the number of patients, individuals, or doctors; the cost of the treatment, disease, or program; and the mortality or diagnosis rate.
The included papers presented a total of 410 scenarios. The main types of scenarios were changing the value of established variables, adding new feedback loops or variables to the model, and removing specific variables. Common scenarios among the reviewed papers include program, legislation, or guideline adaptation; running the model with and without a specific system or feedback loop; changing the percentage or size of the population or population growth; changing the level of health services; changing hospital performance indicators; changing the rate of birth, death, or disease; and changing the treatment procedure or cost. The included papers used a total of 8627 variables in their models and an average of 22 variables per model. Moreover, papers included 2835 equations in their models and an average of 11 equations per stock-flow diagram.
Typically, SD simulation modeling has been used to study and explain high-level problems in complex healthcare systems, mainly related to the impact of strategy and policy decisions on health outcomes, and to design future healthcare systems [6]. The main advantages of SD applications in the healthcare sector include the relatively low cost of collecting input data, transforming complex problems into simple structures, and quickly constructing suitable models [6,141]. However, one of the important limitations of the SD modeling approach is that it cannot provide a deep description of the micro-behaviors of healthcare systems. Another disadvantage is the difficulty of validating complex SD models with a large number of inter-related variables [6,141].
Finally, to generate a better picture of the included papers, the titles and abstracts were also analyzed. Figure 12 and Figure 13 show a map of co-occurrence in two forms: a bubble chart and a heatmap. In Figure 12, the nodes correspond with specific terms, and their sizes indicate the frequency of term occurrence. Links between two nodes represent the co-occurrence of the terms in publications. Frequently co-occurring terms form clusters and emerge closer to each other with the same color. A first glance at Figure 12 reveals the central cluster (red color) with terms including system dynamics, approach, study, strategy, framework, technique, insight, and hybrid simulation. The next cluster (green color) identifies critical issues such as community, organization, public health, implementation, policy intervention, collaboration, and prevention. In Figure 13, darker colors represent a higher frequency of the analyzed terms.

7. Conclusions

The current paper offers a comprehensive review of the critical healthcare issues under consideration by the SD application community. A carefully chosen set of keywords was used for discovering relevant articles. Following PRISMA guidelines, we explored three main research questions related to fundamental problems in the healthcare sector that have been studied in the past using the SD approach. These questions focused on how the SD methodology has been utilized in the past and what could be learned from the application of the SD simulation approach. Finally, the specific criteria for inclusion were defined and applied to the subsequent analysis focusing on developing useful SD applications in the healthcare sector. In general, reviewing and assessing contributions related to SD applications in the healthcare sector is somewhat tricky because there is no unique or standardized way of developing and presenting such papers. However, the main criterion applied here was the overall organization of the manuscript and the quality of the technical content from which the reader could learn about the current trends in applying the SD approach to healthcare problems.
The research questions were answered as follows:
RQ1.
What are the main problems in the healthcare sector that have been studied in the past using the SD approach? To answer this question, the included records have been reviewed, and the main problems were extracted and are represented in Table 4.
RQ2.
How was the SD approach utilized to model and address complex healthcare problems? To answer this question, essential aspects of the SD approach, including “core part of SD diagrams”, “key variables”, and “output variables” are represented in the Discussion section.
RQ3.
What can be learned from the past SD simulation research in the healthcare area that will support high-quality research in the future? In order to answer this question, the main aspects of high-quality SD research have been assessed in the quality assessment section, such as the clarity of the objectives and results, the description of the scenarios, the source of the input data, and model validation.
This paper contributes to the body of knowledge related to developing system dynamics and simulation models that have been used in the past to analyze complex healthcare systems. We draw on a systematic literature selection and present a comprehensive view of system dynamics’ applications to the healthcare sector. The study includes papers from 2000 to 2019 and demonstrates the broad range of interests in the research community and the variety of healthcare topics. We demonstrate that the most popular research topics include patient flow, obesity, workforce demand, and HIV/AIDS. Moreover, the quality of the reviewed papers has been assessed using a proposed ranking system. Finally, the need for improving the quality of the system dynamics models was discussed.

Author Contributions

M.R.D.: methodology and writing—original draft and revisions; W.K.: conceptualization, funding acquisition, writing—review and revisions, editing, and supervision; R.T.: review and revisions, editing, and methodology. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Brailsford, S.C.; Harper, P.R.; Patel, B.; Pitt, M. An analysis of the academic literature on simulation and modelling in health care. J. Simul. 2009, 3, 130–140. [Google Scholar] [CrossRef]
  2. Katsaliaki, K.; Mustafee, N. Applications of simulation within the healthcare context. J. Oper. Res. Soc. 2011, 62, 1431–1451. [Google Scholar] [CrossRef] [PubMed]
  3. Fone, D.; Hollinghurst, S.; Temple, M.; Round, A.; Lester, N.; Weightman, A.; Roberts, K.; Coyle, E.; Bevan, G.; Palmer, S. Systematic review of the use and value of computer simulation modelling in population health and health care delivery. J. Public Health 2003, 25, 325–335. [Google Scholar] [CrossRef] [PubMed]
  4. Forrester, J.W. System dynamics—The next fifty years. Syst. Dyn. Rev. J. Syst. Dyn. Soc. 2007, 23, 359–370. [Google Scholar] [CrossRef]
  5. Homer, J.B.; Hirsch, G.B. System dynamics modeling for public health: Background and opportunities. Am. J. Public Health 2006, 96, 452–458. [Google Scholar] [CrossRef]
  6. Sterman, J. Business Dynamics; Irwin/McGraw-Hil: New York, NY, USA, 2010. [Google Scholar]
  7. Sterman, J.D. System dynamics modeling: Tools for learning in a complex world. Calif. Manage. Rev. 2001, 43, 8–25. [Google Scholar] [CrossRef]
  8. Henderson, S.G.; Biller, B.; Hsieh, M.H.; Shortle, J.; Tew, J.D.; Barton, R.R.; Brailsford, S. Tutorial: Advances and challenges in healthcare simulation modeling. In Proceedings of the 2007 Winter Simulation Conference, Washington, DC, USA, 9–12 December 2007; pp. 1436–1448. [Google Scholar]
  9. Roberts, S.D. Tutorial on the Simulation of Healthcare Systems. In Proceedings of the Winter Simulation Conference, Phoenix, AZ, USA, 11–14 December 2011; pp. 1408–1419. [Google Scholar]
  10. Mingers, J.; White, L. A review of the recent contribution of systems thinking to operational research and management science. Eur. J. Oper. Res. 2010, 207, 1147–1161. [Google Scholar] [CrossRef] [Green Version]
  11. Mustafee, N.; Katsaliaki, K.; Taylor, S.J. Profiling literature in healthcare simulation. Simulation 2010, 86, 543–558. [Google Scholar] [CrossRef]
  12. Stahl, J.E. Modelling methods for pharmacoeconomics and health technology assessment. Pharmacoeconomics 2008, 26, 131–148. [Google Scholar] [CrossRef]
  13. Mielczarek, B. Review of modelling approaches for healthcare simulation. Oper. Res. Decis. 2016, 26, 55–72. [Google Scholar] [CrossRef]
  14. Cilenti, D.; Issel, M.; Wells, R.; Link, S.; Lich, K.H. System Dynamics Approaches and Collective Action for Community Health: An Integrative Review. Am. J. Community Psychol. 2019, 63, 527–545. [Google Scholar] [CrossRef] [PubMed]
  15. Allahi, F.; De Leeuw, S.; Sabet, E.; Kian, R.; Damiani, L.; Giribone, P.; Revetria, R.; Cianci, R. A review of system dynamics models applied in social and humanitarian researches. In Proceedings of the World Congress on Engineering 2018 (WCE 2018), London, UK, 4–6 July 2018. [Google Scholar]
  16. Kunc, M.; Mortenson, M.J.; Vidgen, R. A computational literature review of the field of System Dynamics from 1974 to 2017. J. Simul. 2018, 12, 115–127. [Google Scholar] [CrossRef]
  17. Torres, J.P. System Dynamics Review and publications 1985–2017: Analysis, synthesis and contributions. Syst. Dyn. Rev. 2019, 35, 160–176. [Google Scholar] [CrossRef]
  18. Gul, M.; Guneri, A.F. A comprehensive review of emergency department simulation applications for normal and disaster conditions. Comput. Ind. Eng. 2015, 83, 327–344. [Google Scholar] [CrossRef]
  19. Aboueljinane, L.; Sahin, E.; Jemai, Z. A review on simulation models applied to emergency medical service operations. Comput. Ind. Eng. 2013, 66, 734–750. [Google Scholar] [CrossRef]
  20. Salmon, A.; Rachuba, S.; Briscoe, S.; Pitt, M. A structured literature review of simulation modelling applied to Emergency Departments: Current patterns and emerging trends. Oper. Res. Health Care 2018, 19, 1–13. [Google Scholar] [CrossRef]
  21. He, L.; Madathil, S.C.; Oberoi, A.; Servis, G.; Khasawneh, M.T. A systematic review of research design and modeling techniques in inpatient bed management. Comput. Ind. Eng. 2019, 127, 451–466. [Google Scholar] [CrossRef]
  22. Zhang, X. Application of discrete event simulation in health care: A systematic review. BMC Health Serv. Res. 2018, 18, 687. [Google Scholar] [CrossRef]
  23. Atkinson, J.-A.; Wells, R.; Page, A.; Dominello, A.; Haines, M.; Wilson, A. Applications of system dynamics modelling to support health policy. Public Health Res Pr. 2015, 25, e2531531. [Google Scholar] [CrossRef]
  24. Günal, M.M.; Pidd, M. Discrete event simulation for performance modelling in health care: A review of the literature. J. Simul. 2010, 4, 42–51. [Google Scholar] [CrossRef]
  25. Brailsford, S.C.; Carter, M.W.; Jacobson, S.H. Five decades of healthcare simulation. In Proceedings of the 2017 Winter Simulation Conference (WSC), Las Vegas, NV, USA, 3–6 December 2017; pp. 365–384. [Google Scholar]
  26. Scholz, S.; Mittendorf, T. Modeling rheumatoid arthritis using different techniques—A review of model construction and results. Health Econ. Rev. 2014, 4, 18. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  27. Saville, C.E.; Smith, H.K.; Bijak, K. Operational research techniques applied throughout cancer care services: A review. Health Syst. 2019, 8, 52–73. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  28. Mustafee, N.; Katsaliaki, K.; Gunasekaran, A.; Williams, M.D.; Fakhimi, M.; Probert, J. Operations research within UK healthcare: A review. J. Enterp. Inf. Manag. 2013, 26, 21–49. [Google Scholar] [CrossRef]
  29. Long, K.M.; Meadows, G.N. Simulation modelling in mental health: A systematic review. J. Simul. 2018, 12, 76–85. [Google Scholar] [CrossRef]
  30. Liberati, A.; Altman, D.G.; Tetzlaff, J.; Mulrow, C.; Gøtzsche, P.C.; Ioannidis, J.P.A.; Clarke, M.; Devereaux, P.J.; Kleijnen, J.; Moher, D. The PRISMA Statement for Reporting Systematic Reviews and Meta-Analyses of Studies That Evaluate Health Care Interventions: Explanation and Elaboration. PLoS Med. 2009, 6, e1000100. [Google Scholar] [CrossRef]
  31. Brüssow, H. What is health? Microb. Biotechnol. 2013, 6, 341–348. [Google Scholar] [CrossRef]
  32. Health Care. Available online: https://en.wikipedia.org/wiki/Health_care (accessed on 20 November 2019).
  33. Steffen, G.E. Quality medical care: A definition. JAMA 1988, 260, 56–61. [Google Scholar] [CrossRef]
  34. Health System. Available online: https://en.wikipedia.org/wiki/Health_system (accessed on 20 November 2019).
  35. Vanderby, S.; Carter, M.W. An evaluation of the applicability of system dynamics to patient flow modelling. J. Oper. Res. Soc. 2010, 61, 1572–1581. [Google Scholar] [CrossRef]
  36. Rashwan, W.; Abo-Hamad, W.; Arisha, A. A system dynamics view of the acute bed blockage problem in the Irish healthcare system. Eur. J. Oper. Res. 2015, 247, 276–293. [Google Scholar] [CrossRef] [Green Version]
  37. Rashwan, W.; Ragab, M.; Abo-Hamad, W.; Arisha, A. Evaluating policy interventions for delayed discharge: A system dynamics approach. In Proceedings of the 2013 Winter Simulation Conference: Making Decisions in a Complex World, Washington, DC, USA, 8–11 December 2013; pp. 2463–2474. [Google Scholar]
  38. Grida, M.; Zeid, M. A system dynamics-based model to implement the Theory of Constraints in a healthcare system. Simulation 2019, 95, 593–605. [Google Scholar] [CrossRef] [Green Version]
  39. Lane, D.C.; Monefeldt, C.; Rosenhead, J.V. Looking in the wrong place for healthcare improvements: A system dynamics study of an accident and emergency department. J. Oper. Res. Soc. 2000, 51, 518–531. [Google Scholar] [CrossRef]
  40. Brailsford, S.C.; Lattimer, V.A.; Tarnaras, P.; Turnbull, J.C. Emergency and on-demand health care: Modelling a large complex system. J. Oper. Res. Soc. 2004, 55, 34–42. [Google Scholar] [CrossRef]
  41. Cooke, D.; Rohleder, T.; Rogers, P. A dynamic model of the systemic causes for patient treatment delays in emergency departments. J. Model. Manag. 2010, 5, 287–301. [Google Scholar] [CrossRef]
  42. Kumar, S. Modeling patient flow operation of a US urban county hospital. Technol. Health Care 2011, 19, 247–260. [Google Scholar] [CrossRef] [PubMed]
  43. Taylor, K.; Dangerfield, B. Modelling the feedback effects of reconfiguring health services. J. Oper. Res. Soc. 2005, 56, 659–675. [Google Scholar] [CrossRef] [Green Version]
  44. German, J.D.; Miña, J.K.P.; Alfonso, C.M.N.; Yang, K.-H. A study on shortage of hospital beds in the Philippines using system dynamics. In Proceedings of the 2018 5th International Conference on Industrial Engineering and Applications (ICIEA), Singapore, 26–28 April 2018; pp. 72–78. [Google Scholar]
  45. Wolstenholme, E.; Monk, D.; McKelvie, D.; Arnold, S. Coping but not coping in health and social care: Masking the reality of running organisations beyond safe design capacity. Syst. Dyn. Rev. 2007, 23, 371–389. [Google Scholar] [CrossRef]
  46. Abdel-Hamid, T.K. Modeling the dynamics of human energy regulation and its implications for obesity treatment. Syst. Dyn. Rev. J. Syst. Dyn. Soc. 2002, 18, 431–471. [Google Scholar] [CrossRef] [Green Version]
  47. Frerichs, L.M.; Araz, O.M.; Huang, T.T.-K. Modeling social transmission dynamics of unhealthy behaviors for evaluating prevention and treatment interventions on childhood obesity. PLoS ONE 2013, 8, e82887. [Google Scholar] [CrossRef]
  48. Liu, S.; Osgood, N.; Gao, Q.; Xue, H.; Wang, Y. Systems simulation model for assessing the sustainability and synergistic impacts of sugar-sweetened beverages tax and revenue recycling on childhood obesity prevention. J. Oper. Res. Soc. 2016, 67, 708–721. [Google Scholar] [CrossRef]
  49. Powell, K.E.; Kibbe, D.L.; Ferencik, R.; Soderquist, C.; Phillips, M.A.; Vall, E.A.; Minyard, K.J. Systems thinking and simulation modeling to inform childhood obesity policy and practice. Public Health Rep. 2017, 132, 33S–38S. [Google Scholar] [CrossRef]
  50. Roberts, N.; Li, V.; Atkinson, J.-A.; Heffernan, M.; McDonnell, G.; Prodan, A.; Freebairn, L.; Lloyd, B.; Nieuwenhuizen, S.; Mitchell, J. Can the Target Set for Reducing Childhood Overweight and Obesity Be Met? A System Dynamics Modelling Study in New South Wales, Australia. Syst. Res. Behav. Sci. 2019, 36, 36–52. [Google Scholar] [CrossRef] [Green Version]
  51. Sabounchi, N.S.; Hovmand, P.S.; Osgood, N.D.; Dyck, R.F.; Jungheim, E.S. A novel system dynamics model of female obesity and fertility. Am. J. Public Health 2014, 104, 1240–1246. [Google Scholar] [CrossRef] [PubMed]
  52. Jalali, M.; Rahmandad, H.; Bullock, S.; Ammerman, A. Dynamics of obesity interventions inside organizations. In Proceedings of the 32nd International Conference of the System Dynamics Society, Delft, The Netherlands, 20–24 July 2014. [Google Scholar]
  53. Meisel, J.D.; Sarmiento, O.L.; Olaya, C.; Lemoine, P.D.; Valdivia, J.A.; Zarama, R. Towards a novel model for studying the nutritional stage dynamics of the Colombian population by age and socioeconomic status. PLoS ONE 2018, 13, e0191929. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  54. Chen, H.-J.; Xue, H.; Liu, S.; Huang, T.T.K.; Wang, Y.C.; Wang, Y. Obesity trend in the United States and economic intervention options to change it: A simulation study linking ecological epidemiology and system dynamics modeling. Public Health 2018, 161, 20–28. [Google Scholar] [CrossRef]
  55. Abidin, N.Z.; Zulkepli, J.H.; Zaibidi, N.Z. The effect of food portion sizes on the obesity prevention using system dynamics modelling. In AIP Conference Proceedings, Proceedings of Statistics and Operational Research International Conference; American Institute of Physics: College Park, MD, USA, 2014; Volume 1613, pp. 436–443. [Google Scholar]
  56. Allender, S.; Owen, B.; Kuhlberg, J.; Lowe, J.; Nagorcka-Smith, P.; Whelan, J.; Bell, C. A community based systems diagram of obesity causes. PLoS ONE 2015, 10, e0129683. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  57. Miller, R.L.; Levine, R.L.; McNall, M.A.; Khamarko, K.; Valenti, M.T. A dynamic model of client recruitment and retention in community-based HIV prevention programs. Health Promot. Pract. 2011, 12, 135–146. [Google Scholar] [CrossRef]
  58. Dangerfield, B.C.; Fang, Y.; Roberts, C.A. Model-based scenarios for the epidemiology of HIV/AIDS: The consequences of highly active antiretroviral therapy. Syst. Dyn. Rev. J. Syst. Dyn. Soc. 2001, 17, 119–150. [Google Scholar] [CrossRef]
  59. Kok, S.; Rutherford, A.R.; Gustafson, R.; Barrios, R.; Montaner, J.S.; Vasarhelyi, K. Optimizing an HIV testing program using a system dynamics model of the continuum of care. Health Care Manag. Sci. 2015, 18, 334–362. [Google Scholar] [CrossRef] [Green Version]
  60. Lebcir, R.M.; Atun, R.A.; Coker, R.J. System Dynamic simulation of treatment policies to address colliding epidemics of tuberculosis, drug resistant tuberculosis and injecting drug users driven HIV in Russia. J. Oper. Res. Soc. 2010, 61, 1238–1248. [Google Scholar] [CrossRef] [Green Version]
  61. Atun, R.A.; Lebcir, R.M.; McKee, M.; Habicht, J.; Coker, R.J. Impact of joined-up HIV harm reduction and multidrug resistant tuberculosis control programmes in Estonia: System dynamics simulation model. Health Policy 2007, 81, 207–217. [Google Scholar] [CrossRef]
  62. Lounsbury, D.W.; Schwartz, B.; Palma, A.; Blank, A. Simulating patterns of patient engagement, treatment adherence, and viral suppression: A system dynamics approach to evaluating HIV care management. AIDS Patient Care STDs 2015, 29, S55–S63. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  63. Zou, X.; Xu, Y.; Chen, W.; Xia, Y.; Liu, Y.; Gong, C.; Ling, L. Strategies to control HIV and HCV in methadone maintenance treatment in Guangdong Province, China: A system dynamic modeling study. Subst. Abuse Treat. Prev. Policy 2018, 13, 1. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  64. Shariatpanahi, S.P.; Jafari, A.; Sadeghipour, M.; Azadeh-Fard, N.; Majidzadeh-A, K.; Farahmand, L.; Ansari, A.M. Assessing the effectiveness of disease awareness programs: Evidence from Google Trends data for the world awareness dates. Telemat. Inform. 2017, 34, 904–913. [Google Scholar] [CrossRef]
  65. Barber, P.; López-Valcárcel, B.G. Forecasting the need for medical specialists in Spain: Application of a system dynamics model. Hum. Resour. Health 2010, 8, 24. [Google Scholar] [CrossRef] [Green Version]
  66. Lodi, A.; Tubertini, P.; Grilli, R.; Mazzocchetti, A.; Ruozi, C.; Senese, F. Needs forecast and fund allocation of medical specialty positions in Emilia-Romagna (Italy) by system dynamics and integer programming. Health Syst. 2016, 5, 213–236. [Google Scholar] [CrossRef]
  67. Wu, M.-H.; Yu, J.-Y.; Huang, C.-H. Theoretical system dynamics modeling for Taiwan pediatric workforce in an era of national health insurance and low birth rates. Pediatr. Neonatol. 2013, 54, 389–396. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  68. Vanderby, S.A.; Carter, M.W.; Latham, T.; Feindel, C. Modelling the future of the Canadian cardiac surgery workforce using system dynamics. J. Oper. Res. Soc. 2014, 65, 1325–1335. [Google Scholar] [CrossRef]
  69. Samah, A.A.; Wah, L.K.; Desa, M.I.; Majid, H.A.; Azmi, N.F.M.; Salleh, N.; Bakar, A.A.; Manual, A. Decision Support System Using System Dynamics Simulation Modelling for Projection of Dentist Supply. In Proceedings of the 2014 International Conference on Computer Assisted System in Health, Kuala Lumpur, Malaysia, 19–21 December 2014; pp. 22–25. [Google Scholar]
  70. Ansah, J.P.; De Korne, D.; Bayer, S.; Pan, C.; Jayabaskar, T.; Matchar, D.B.; Lew, N.; Phua, A.; Koh, V.; Lamoureux, E. Future requirements for and supply of ophthalmologists for an aging population in Singapore. Hum. Resour. Health 2015, 13, 86. [Google Scholar] [CrossRef] [Green Version]
  71. Taba, S.T.; Atkinson, S.R.; Lewis, S.; Chung, K.S.K.; Hossain, L. A systems life cycle approach to managing the radiology profession: An Australian perspective. Aust. Health Rev. 2015, 39, 228–239. [Google Scholar] [CrossRef]
  72. De Silva, D. How many doctors should we train for Sri Lanka? System dynamics modelling for training needs. Ceylon Med. J. 2017, 62, 233–237. [Google Scholar] [CrossRef] [Green Version]
  73. Ishikawa, T.; Fujiwara, K.; Ohba, H.; Suzuki, T.; Ogasawara, K. Forecasting the regional distribution and sufficiency of physicians in Japan with a coupled system dynamics—Geographic information system model. Hum. Resour. Health 2017, 15, 64. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  74. Abas, Z.A.; Ramli, M.R.; Desa, M.I.; Saleh, N.; Hanafiah, A.N.; Aziz, N.; Abidin, Z.Z.; Shibghatullah, A.S.; Rahman, A.F.N.A.; Musa, H. A supply model for nurse workforce projection in Malaysia. Health Care Manag. Sci. 2018, 21, 573–586. [Google Scholar] [CrossRef] [PubMed]
  75. Morii, Y.; Ishikawa, T.; Suzuki, T.; Tsuji, S.; Yamanaka, M.; Ogasawara, K.; Yamashina, H. Projecting future supply and demand for physical therapists in Japan using system dynamics. Health Policy Technol. 2019, 8, 118–127. [Google Scholar] [CrossRef]
  76. Anderson, G.; Horvath, J. The growing burden of chronic disease in America. Public Health Rep. 2004, 119, 263–270. [Google Scholar] [CrossRef] [Green Version]
  77. Bernell, S.; Howard, S.W. Use your words carefully: What is a chronic disease? Front. Public Health 2016, 4, 159. [Google Scholar] [CrossRef] [Green Version]
  78. De Andrade, L.; Lynch, C.; Carvalho, E.; Rodrigues, C.G.; Vissoci, J.R.N.; Passos, G.F.; Pietrobon, R.; Nihei, O.K.; de Barros Carvalho, M.D. System dynamics modeling in the evaluation of delays of care in ST-segment elevation myocardial infarction patients within a tiered health system. PLoS ONE 2014, 9, e103577. [Google Scholar] [CrossRef] [Green Version]
  79. Tejada, J.J.; Ivy, J.S.; King, R.E.; Wilson, J.R.; Ballan, M.J.; Kay, M.G.; Diehl, K.M.; Yankaskas, B.C. Combined DES/SD model of breast cancer screening for older women, II: Screening-and-treatment simulation. IIE Trans. 2014, 46, 707–727. [Google Scholar] [CrossRef]
  80. Liew, H.P. Potential Roles of Screening Mammography and Clinical Breast Examination in Affecting Breast Cancer Trends in Nebraska. Health Behav. Policy Rev. 2018, 5, 56–64. [Google Scholar] [CrossRef]
  81. Palma, A.; Lounsbury, D.W.; Schlecht, N.F.; Agalliu, I. A system dynamics model of serum prostate-specific antigen screening for prostate cancer. Am. J. Epidemiol. 2015, 183, 227–236. [Google Scholar] [CrossRef] [Green Version]
  82. Hallberg, S.; Claeson, M.; Holmström, P.; Paoli, J.; Larkö, A.-M.W.; Gonzalez, H. Developing a simulation model for the patient pathway of cutaneous malignant melanoma. Oper. Res. Health Care 2015, 6, 23–30. [Google Scholar] [CrossRef]
  83. Kenealy, T.; Rees, D.; Sheridan, N.; Moffitt, A.; Tibby, S.; Homer, J. A ‘whole of system’approach to compare options for CVD interventions in counties Manukau. Aust. N. Z. J. Public Health 2012, 36, 263–268. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  84. Lich, K.H.; Tian, Y.; Beadles, C.A.; Williams, L.S.; Bravata, D.M.; Cheng, E.M.; Bosworth, H.B.; Homer, J.B.; Matchar, D.B. Strategic planning to reduce the burden of stroke among veterans: Using simulation modeling to inform decision making. Stroke 2014, 45, 2078–2084. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  85. Recio, A.; Linares, C.; Díaz, J. System dynamics for predicting the impact of traffic noise on cardiovascular mortality in Madrid. Environ. Res. 2018, 167, 499–505. [Google Scholar] [CrossRef] [PubMed]
  86. Loyo, H.K.; Batcher, C.; Wile, K.; Huang, P.; Orenstein, D.; Milstein, B. From model to action: Using a system dynamics model of chronic disease risks to align community action. Health Promot. Pract. 2013, 14, 53–61. [Google Scholar] [CrossRef]
  87. Kunc, M.; Kazakov, R. Competitive dynamics in pharmaceutical markets: A case study in the chronic cardiac disease market. J. Oper. Res. Soc. 2013, 64, 1790–1799. [Google Scholar] [CrossRef]
  88. Brittin, J.; Araz, O.M.; Nam, Y.; Huang, T.-K. A system dynamics model to simulate sustainable interventions on chronic disease outcomes in an urban community. J. Simul. 2015, 9, 140–155. [Google Scholar] [CrossRef]
  89. Kang, H.; Nembhard, H.B.; Ghahramani, N.; Curry, W. A system dynamics approach to planning and evaluating interventions for chronic disease management. J. Oper. Res. Soc. 2018, 69, 987–1005. [Google Scholar] [CrossRef]
  90. Zare Mehrjerdi, Y. A dynamic systems approach to weight related health problems. Int. J. Qual. Reliab. Manag. 2013, 30, 571–589. [Google Scholar] [CrossRef]
  91. Sugiyama, T.; Goryoda, S.; Inoue, K.; Sugiyama-Ihana, N.; Nishi, N. Construction of a simulation model and evaluation of the effect of potential interventions on the incidence of diabetes and initiation of dialysis due to diabetic nephropathy in Japan. BMC Health Serv. Res. 2017, 17, 833. [Google Scholar] [CrossRef] [Green Version]
  92. Milstein, B.; Jones, A.; Homer, J.B.; Murphy, D.; Essien, J.; Seville, D. Peer reviewed: Charting plausible futures for diabetes prevalence in the United States: A role for system dynamics simulation modeling. Prev. Chronic. Dis. 2007, 4, 1–8. [Google Scholar]
  93. Jones, A.P.; Homer, J.B.; Murphy, D.L.; Essien, J.D.; Milstein, B.; Seville, D.A. Understanding diabetes population dynamics through simulation modeling and experimentation. Am. J. Public Health 2006, 96, 488–494. [Google Scholar] [CrossRef] [PubMed]
  94. Montazemi, A.R.; Pittaway, J.; Saremi, H.Q. Assessment of Ubiquitous Healthcare Information Systems Benefits. In Proceedings of the AMCIS, Lima, Peru, 12–15 August 2010; p. 121. [Google Scholar]
  95. Guariguata, L.; Guell, C.; Samuels, T.A.; Rouwette, E.; Woodcock, J.; Hambleton, I.R.; Unwin, N. Systems Science for Caribbean Health: The development and piloting of a model for guiding policy on diabetes in the Caribbean. Health Res. Policy Syst. 2016, 14, 79. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  96. Ansah, J.P.; Koh, V.; Chiu, C.-T.; Chei, C.-L.; Zeng, Y.; Yin, Z.-X.; Shi, X.-M.; Matchar, D.B. Projecting the Number of Elderly with Cognitive Impairment in China Using a Multi-State Dynamic Population Model. Syst. Dyn. Rev. 2017, 33, 89–111. [Google Scholar] [CrossRef]
  97. Thompson, J.P.; Riley, C.M.; Eberlein, R.L. Modelling for Insight: The Case of Dementia in Singapore. Syst. Res. Behav. Sci. 2014, 31, 227–235. [Google Scholar] [CrossRef]
  98. Hwang, W.-S.; Choi, Y.-H. Socio-economic impact of the mHealth adoption in managing diabetes. Technol. Anal. Strateg. Manag. 2019, 31, 791–802. [Google Scholar] [CrossRef]
  99. Homer, J.; Hirsch, G.; Minniti, M.; Pierson, M. Models for collaboration: How system dynamics helped a community organize cost-effective care for chronic illness. Syst. Dyn. Rev. J. Syst. Dyn. Soc. 2004, 20, 199–222. [Google Scholar] [CrossRef]
  100. Brailsford, S.C.; Desai, S.M.; Viana, J. Towards the holy grail: Combining system dynamics and discrete-event simulation in healthcare. In Proceedings of the Winter Simulation Conference, Baltimore, MD, USA, 5–8 December 2010; pp. 2293–2303. [Google Scholar]
  101. Chahal, K.; Eldabi, T. Applicability of hybrid simulation to different modes of governance in UK healthcare. In Proceedings of the 2008 Winter Simulation Conference, Miami, FL, USA, 7–10 December 2008; pp. 1469–1477. [Google Scholar]
  102. Viana, J. Reflections on two approaches to hybrid simulation in healthcare. In Proceedings of the Winter Simulation Conference 2014, Savannah, GA, USA, 8–11 December 2014; pp. 1585–1596. [Google Scholar]
  103. Brailsford, S.C.; Hilton, N.A. A Comparison of Discrete Event Simulation and System Dynamics for Modelling Health Care Systems; University of Southampton: Southampton, UK, 2001. [Google Scholar]
  104. Laker, L.F.; Torabi, E.; France, D.J.; Froehle, C.M.; Goldlust, E.J.; Hoot, N.R.; Kasaie, P.; Lyons, M.S.; Barg-Walkow, L.H.; Ward, M.J. Understanding emergency care delivery through computer simulation modeling. Acad. Emerg. Med. 2018, 25, 116–127. [Google Scholar] [CrossRef] [Green Version]
  105. Marshall, D.A.; Burgos-Liz, L.; IJzerman, M.J.; Crown, W.; Padula, W.V.; Wong, P.K.; Pasupathy, K.S.; Higashi, M.K.; Osgood, N.D. Selecting a dynamic simulation modeling method for health care delivery research—Part 2: Report of the ISPOR Dynamic Simulation Modeling Emerging Good Practices Task Force. Value Health 2015, 18, 147–160. [Google Scholar] [CrossRef] [Green Version]
  106. Gunal, M.M. A guide for building hospital simulation models. Health Syst. 2012, 1, 17–25. [Google Scholar] [CrossRef]
  107. Morgan, J.S.; Belton, V.; Howick, S. Lessons from mixing OR methods in practice: Using DES and SD to explore a radiotherapy treatment planning process. Health Syst. 2016, 5, 166–177. [Google Scholar] [CrossRef] [Green Version]
  108. Abdelghany, M.; Eltawil, A.B. Individual versus integrated simulation techniques in healthcare applications. In Proceedings of the 2014 IEEE International Conference on Industrial Engineering and Engineering Management, Selangor Darul Ehsan, Malaysia, 9–12 December 2014; pp. 1214–1218. [Google Scholar]
  109. Abdelghany, M.; Eltawil, A.B. Linking approaches for multi-methods simulation in healthcare systems planning and management. Int. J. Ind. Syst. Eng. 2017, 26, 275–290. [Google Scholar] [CrossRef]
  110. Chahal, K.; Eldabi, T.; Young, T. A conceptual framework for hybrid system dynamics and discrete event simulation for healthcare. J. Enterp. Inf. Manag. 2013, 26, 50–74. [Google Scholar] [CrossRef]
  111. Djanatliev, A.; German, R. Prospective healthcare decision-making by combined system dynamics, discrete-event and agent-based simulation. In Proceedings of the 2013 Winter Simulation Conference: Simulation: Making Decisions in a Complex World, Washington, DC, USA, 8–11 December 2013; pp. 270–281. [Google Scholar]
  112. Zulkepli, J.; Eldabi, T. Towards a framework for conceptual model hybridization in healthcare. In Proceedings of the 2015 Winter Simulation Conference (WSC), Huntington Beach, CA, USA, 6–9 December 2015; pp. 1597–1608. [Google Scholar]
  113. Townshend, J.R.P.; Turner, H.S. Analysing the effectiveness of Chlamydia screening. J. Oper. Res. Soc. 2000, 51, 812–824. [Google Scholar] [CrossRef]
  114. Evenden, D.; Harper, P.R.; Brailsford, S.C.; Harindra, V. System Dynamics modeling of Chlamydia infection for screening intervention planning and cost-benefit estimation. IMA J. Manag. Math. 2005, 16, 265–279. [Google Scholar] [CrossRef]
  115. Evenden, D.; Harper, P.R.; Brailsford, S.C.; Harindra, V. Improving the cost-effectiveness of Chlamydia screening with targeted screening strategies. J. Oper. Res. Soc. 2006, 57, 1400–1412. [Google Scholar] [CrossRef]
  116. Viana, J.; Brailsford, S.C.; Harindra, V.; Harper, P.R. Combining discrete-event simulation and system dynamics in a healthcare setting: A composite model for Chlamydia infection. Eur. J. Oper. Res. 2014, 237, 196–206. [Google Scholar] [CrossRef]
  117. Teng, Y.; Kong, N.; Tu, W. Optimizing strategies for population-based chlamydia infection screening among young women: An age-structured system dynamics approach. BMC Public Health 2015, 15, 639. [Google Scholar] [CrossRef] [Green Version]
  118. Chalmers, J.; Ritter, A. Subsidising patient dispensing fees: The cost of injecting equity into the opioid pharmacotherapy maintenance system. Drug Alcohol Rev. 2012, 31, 911–917. [Google Scholar] [CrossRef] [Green Version]
  119. Karanfil, Ö.; Barlas, Y. A dynamic simulator for the management of disorders of the body water homeostasis. Oper. Res. 2008, 56, 1474–1492. [Google Scholar] [CrossRef]
  120. Hosseinichimeh, N.; Wittenborn, A.K.; Rick, J.; Jalali, M.S.; Rahmandad, H. Modeling and estimating the feedback mechanisms among depression, rumination, and stressors in adolescents. PLoS ONE 2018, 13, e0204389. [Google Scholar] [CrossRef]
  121. Wittenborn, A.K.; Rahmandad, H.; Rick, J.; Hosseinichimeh, N. Depression as a systemic syndrome: Mapping the feedback loops of major depressive disorder. Psychol. Med. 2016, 46, 551–562. [Google Scholar] [CrossRef]
  122. Lee, K.A.; Dziadkowiec, O.; Meek, P. A systems science approach to fatigue management in research and health care. Nurs. Outlook 2014, 62, 313–321. [Google Scholar] [CrossRef] [PubMed]
  123. Apostolopoulos, Y.; Lemke, M.K.; Barry, A.E.; Lich, K.H. Moving alcohol prevention research forward—Part I: Introducing a complex systems paradigm. Addiction 2018, 113, 353–362. [Google Scholar] [CrossRef] [PubMed]
  124. Apostolopoulos, Y.; Lemke, M.K.; Barry, A.E.; Lich, K.H. Moving alcohol prevention research forward—Part II: New directions grounded in community-based system dynamics modeling. Addiction 2018, 113, 363–371. [Google Scholar] [CrossRef] [PubMed]
  125. Li, L.; Yu, Y. Feedback complexity analysis on community health service based on system dynamic. In Proceedings of the 2012 International Symposium on Management of Technology (ISMOT), Hangzhou, China, 8–9 November 2012; pp. 71–75. [Google Scholar]
  126. Lingard, H.; Turner, M. Promoting construction workers’ health: A multi-level system perspective. Constr. Manag. Econ. 2017, 35, 239–253. [Google Scholar] [CrossRef]
  127. Sönmez, S.; Apostolopoulos, Y.; Lemke, M.K.; Hsieh, Y.-C.J.; Karwowski, W. Complexity of occupational health in the hospitality industry: Dynamic simulation modeling to advance immigrant worker health. Int. J. Hosp. Manag. 2017, 67, 95–105. [Google Scholar] [CrossRef]
  128. Faezipour, M.; Ferreira, S. A system dynamics perspective of patient satisfaction in healthcare. Procedia Comput. Sci. 2013, 16, 148–156. [Google Scholar] [CrossRef] [Green Version]
  129. Lyons, G.J.; Duggan, J. System dynamics modelling to support policy analysis for sustainable health care. J. Simul. 2015, 9, 129–139. [Google Scholar] [CrossRef] [Green Version]
  130. Newell, B.; Siri, J. A role for low-order system dynamics models in urban health policy making. Environ. Int. 2016, 95, 93–97. [Google Scholar] [CrossRef]
  131. Friel, S.; Pescud, M.; Malbon, E.; Lee, A.; Carter, R.; Greenfield, J.; Cobcroft, M.; Potter, J.; Rychetnik, L.; Meertens, B. Using systems science to understand the determinants of inequities in healthy eating. PLoS ONE 2017, 12, e0188872. [Google Scholar] [CrossRef] [Green Version]
  132. Metcalf, S.S.; Northridge, M.E.; Widener, M.J.; Chakraborty, B.; Marshall, S.E.; Lamster, I.B. Modeling social dimensions of oral health among older adults in urban environments. Health Educ. Behav. 2013, 40, 63S–73S. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  133. Schröttner, J.; Herzog, A. Development of a system dynamics model for cost estimation for the implantation and revision of hip joint endoprosthesis. In Proceedings of the XII Mediterranean Conference on Medical and Biological Engineering and Computing, Chalkidiki, Greece, 27–30 May 2010; pp. 566–569. [Google Scholar]
  134. Djanatliev, A.; German, R.; Kolominsky-Rabas, P.; Hofmann, B.M. Hybrid simulation with loosely coupled system dynamics and agent-based models for prospective health technology assessments. In Proceedings of the Winter Simulation Conference, Berlin, Germany, 9–12 December 2012; pp. 1–12. [Google Scholar]
  135. Goh, Y.M.; Love, P.E.; Stagbouer, G.; Annesley, C. Dynamics of safety performance and culture: A group model building approach. Accid. Anal. Prev. 2012, 48, 118–125. [Google Scholar] [CrossRef] [PubMed]
  136. Vallerotonda, R.; Leva, A.; Silvia, M.A. Modeling and Training: How System Dynamics is Usable in OSH and MAH Frameworks. Chem. Eng. Trans. 2018, 67, 325–330. [Google Scholar] [CrossRef]
  137. Edaibat, E.A.; Dever, J.; Stuban, S.M. System dynamics simulation modeling of health information exchange (HIE) adoption and policy intervention: A case study in the State of Maryland. Oper. Res. Health Care 2017, 12, 60–70. [Google Scholar] [CrossRef]
  138. Merrill, J.A.; Deegan, M.; Wilson, R.V.; Kaushal, R.; Fredericks, K. A system dynamics evaluation model: Implementation of health information exchange for public health reporting. J. Am. Med. Inform. Assoc. 2013, 20, e131–e138. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  139. Alkayid, K.; Hasan, H.; Meloche, J.A. Simulating information exchanges to investigate the utility of public health web sites. Transform. Gov. People Process Policy 2009, 3, 271–288. [Google Scholar] [CrossRef]
  140. Barlas, Y. Formal aspects of model validity and validation in system dynamics. Syst. Dyn. Rev. J. Syst. Dyn. Soc. 1996, 12, 183–210. [Google Scholar] [CrossRef]
  141. Koelling, P.; Schwandt, M.J. Health systems: A dynamic system-benefits from system dynamics. In Proceedings of the Winter Simulation Conference, Orlando, FL, USA, 4 December 2005; p. 7. [Google Scholar] [CrossRef] [Green Version]
Ijerph 17 05741 g001 550
Figure 1. Chart of the selection strategy.
Figure 1. Chart of the selection strategy.
Ijerph 17 05741 g001
Ijerph 17 05741 g002 550
Figure 2. Included papers per year.
Figure 2. Included papers per year.
Ijerph 17 05741 g002
Ijerph 17 05741 g003 550
Figure 3. Published papers before and after the screening process.
Figure 3. Published papers before and after the screening process.
Ijerph 17 05741 g003
Ijerph 17 05741 g004 550
Figure 4. The core part of the stock-flow diagram for the patient flow model.
Figure 4. The core part of the stock-flow diagram for the patient flow model.
Ijerph 17 05741 g004
Ijerph 17 05741 g005 550
Figure 5. The core part of the causal loop diagram for patient flow.
Figure 5. The core part of the causal loop diagram for patient flow.
Ijerph 17 05741 g005
Ijerph 17 05741 g006 550
Figure 6. The core part of the stock-flow diagram for obesity.
Figure 6. The core part of the stock-flow diagram for obesity.
Ijerph 17 05741 g006
Ijerph 17 05741 g007 550
Figure 7. The core part of the stock-flow diagram for workforce demand.
Figure 7. The core part of the stock-flow diagram for workforce demand.
Ijerph 17 05741 g007
Ijerph 17 05741 g008 550
Figure 8. The core part of the stock-flow diagram for chronic diseases.
Figure 8. The core part of the stock-flow diagram for chronic diseases.
Ijerph 17 05741 g008
Ijerph 17 05741 g009 550
Figure 9. The core part of the stock-flow diagram for chlamydia.
Figure 9. The core part of the stock-flow diagram for chlamydia.
Ijerph 17 05741 g009
Ijerph 17 05741 g010 550
Figure 10. Quality assessment of the reviewed papers (legend: good quality > 80%, 80% > medium quality > 70%, 70% > bad quality > 65%, 65% > very bad quality).
Figure 10. Quality assessment of the reviewed papers (legend: good quality > 80%, 80% > medium quality > 70%, 70% > bad quality > 65%, 65% > very bad quality).
Ijerph 17 05741 g010
Ijerph 17 05741 g011 550
Figure 11. Time horizons among the selected papers.
Figure 11. Time horizons among the selected papers.
Ijerph 17 05741 g011
Ijerph 17 05741 g012 550
Figure 12. The map of the co-occurrence of terms in titles and abstracts of included papers.
Figure 12. The map of the co-occurrence of terms in titles and abstracts of included papers.
Ijerph 17 05741 g012
Ijerph 17 05741 g013 550
Figure 13. The heatmap in terms of titles and abstracts of included papers.
Figure 13. The heatmap in terms of titles and abstracts of included papers.
Ijerph 17 05741 g013
Table
Table 1. Review articles related to simulation modeling in healthcare and published from 2000 to 2019.
Table 1. Review articles related to simulation modeling in healthcare and published from 2000 to 2019.
ReferenceArea of ApplicationSimulation MethodNumber of Papers Included
[1]HealthcareOperations research approaches342
[2]HealthcareDifferent simulation methods251
[3]Healthcare delivery and population healthSimulation techniques182
[11]HealthcareDifferent simulation methods201
[13]HealthcareDifferent simulation methods232
[16]All fieldsSystem dynamics800
[17]All fieldsSystem dynamics1441
[18]Emergency departmentsDifferent simulation methods106
[19]Emergency medical services (EMS)Simulation models24
[20]Emergency departmentsDifferent simulation methods254
[21]Inpatient bed managementModeling techniques92
[22]HealthcareDiscrete-event simulation211
[23]Health policySystem dynamics6
[24]HealthcareDiscrete-event simulationNot mentioned
[25]HealthcareSimulation methodologiesNot applicable
[26]Rheumatoid arthritisModeling techniques58
[27]Cancer careOperations research approaches90
[28]UK healthcareOperations research approaches142
[29]Mental healthDifferent simulation methods160
Table
Table 2. The first set of keywords used in the present review.
Table 2. The first set of keywords used in the present review.
RowSet
Test set 1(Health OR health care OR medical care)
Test set 2(System dynamics OR SD)
Search 1#1 AND #2
Test set 3(Disease OR illness OR treatment OR injury OR morbidity)
Search 2#3 AND #2
Table
Table 3. The second set of keywords used in the present review.
Table 3. The second set of keywords used in the present review.
RowSet
Test set 1(System dynamics OR SD)
Test set 2(Disorder OR stress)
Test set 3(Drug OR medication)
Test set 4(Human body OR human behavior)
Search 3#1 AND (#2 OR #3 OR #4)
Table
Table 4. Classification of the papers included in the present review.
Table 4. Classification of the papers included in the present review.
CategorySubcategoryNumber of Included Papers
Aging and populationHealthcare demand3
Evolution of population1
Communicable diseasesHIV/AIDS, tuberculosis11
Chlamydia5
Hepatitis C2
Influenza2
Ebola virus1
Foot-and-mouth disease1
Middle East Respiratory Syndrome (MERS) Coronavirus (CoV)1
Severe acute respiratory syndrome coronavirus (SARS-CoV)1
Other infectious diseases11
Cost and price in healthcarePharmaceutical fees1
Cost–benefit analysis1
Insurance economics1
Patient dispensing fees1
Performance-based payment system1
Disorder and stressObesity14
Depression3
Body Water Homeostasis1
Fatigue1
Post-traumatic stress disorder (PTSD)1
Workplace stress1
Drugs and medicationsAlcohol misuse3
Opioid use and misuse2
Nicotine product1
Pain medicine1
Use/abuse of drugs1
Healthcare operationsPatient flow17
Emergency department7
Hospital3
Patient service centers2
Behavior of nurses1
Clinical knowledge1
Clinical workforce1
Neonatal care services1
Outpatient clinics1
Patient access to general practice1
Radiotherapy department1
Referrals system1
Telehealth1
Healthcare systemsSystem improvement4
Mental health care2
Urban health2
Community health services2
Construction workers’ health1
Curative and preventive services1
Healthcare delivery systems1
Healthcare affordability and accessibility1
Immigrant worker health1
Military psychological health system1
National healthcare system1
One Health concept1
Public health1
Rural minority health1
Sustainability in healthcare1
Sustainable health care1
Youth health1
Other healthcare systems4
Human body and behaviorBody weight1
Functional loss1
Healthy eating1
Lower back pain (LBP)1
Oral health care2
Medical treatment and devicesChemotherapy1
Childhood immunization1
Disease awareness1
Eyecare services1
Hip joint endoprosthesis1
Home hemodialysis (BASIC-HHD)1
Knee implants1
Maintenance procedure1
Mobile stroke units1
Prospective Health Technology Assessment (ProHTA)1
Non-communicable diseasesDiabetes7
Cancer4
Cardiovascular disease4
Dementia3
Intellectual disabilities (ID)2
Asthma1
Cognitive impairment1
End-stage renal disease1
Hernia1
Other chronic diseases7
Healthcare process and policyHealth interventions5
Long-term care policy2
Care planning1
Clinical decision thresholds1
Lean deployment strategies1
Policy implementation1
Strategic planning1
Successful health programs1
Workforce planning1
Safety in healthcareOccupational safety and health2
Risk of adverse events1
System safety1
Simulation in healthcareSimulations comparison7
Hybrid modeling6
Introduction to system dynamics1
Specialty in neonatology, obstetrics, gynecologyInfant mortality4
Cesarean delivery1
Supply and demand in healthcareWorkforce demand12
Supply chain3
Care supply and demand1
Blood supply chain1
Internal service supply chains1
Medical demand1
Need for facilities1
Technology and information in healthcareElectronic health records4
Health information exchange (HIE)3
Cybersecurity1
Health information technology1
Health information systems1
Sharing of information1
Table
Table 5. List of quality criteria.
Table 5. List of quality criteria.
RowQuality Criteria
1Presenting clear objectives
2Presenting clear scenarios and interventions
3Presenting clear outcomes variables by graphs, charts, or tables
4Describing the development of an SD model/framework or presenting a detailed model/framework
5Presenting and explaining model parameters
6Improving the quality of data by using stakeholders’ engagement, surveys, interviews, and databases
7Validating models
8Presenting clear results
Legend: Score is from 0 (not mentioned) to 2 (fully described). Rows 1, 4, 5, 6, and 8 applied qualitative studies. Rows 1 to 8 applied quantitative studies.

Share and Cite

MDPI and ACS Style

Davahli, M.R.; Karwowski, W.; Taiar, R. A System Dynamics Simulation Applied to Healthcare: A Systematic Review. Int. J. Environ. Res. Public Health 2020, 17, 5741. https://doi.org/10.3390/ijerph17165741

AMA Style

Davahli MR, Karwowski W, Taiar R. A System Dynamics Simulation Applied to Healthcare: A Systematic Review. International Journal of Environmental Research and Public Health. 2020; 17(16):5741. https://doi.org/10.3390/ijerph17165741

Chicago/Turabian Style

Davahli, Mohammad Reza, Waldemar Karwowski, and Redha Taiar. 2020. "A System Dynamics Simulation Applied to Healthcare: A Systematic Review" International Journal of Environmental Research and Public Health 17, no. 16: 5741. https://doi.org/10.3390/ijerph17165741

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop