Background
There is a growing recognition that most major threats to the public’s health - including cardiovascular disease, pulmonary disease, cancer, diabetes, mental health problems, HIV, substance abuse, violence, emerging infectious diseases, obesity, sedentary lifestyle, poor diet, sleep disorders, and more—are complex in the sense that each one arises from an intricate mix of behavioral, economic and social factors interacting with biological factors, as well as each other, over the lifespan and across an array of settings (e.g., home, school, workplace, neighborhood, etc.). For example, tobacco use and successful cessation are influenced by a host of interrelated factors, including: the tobacco product itself (e.g., percent free-base nicotine content, presence or absence of menthol/other flavoring, and other product constituents), the person (e.g., genetic predisposition), influences on the person (peer influence, media exposure - both tobacco promotion and health messages, cultural norms, prior tobacco exposure, availability and usage of pharmacotherapy, history of quit attempts, presence of workplace smoking bans), and the tobacco industry (product design, marketing, pricing; for a discussion of the myriad of factors in tobacco control, see NCI, 2007).
Such problems have typically been approached using correlation-based analytic methods (e.g., regression), which are useful for identifying linear relationships but are limited because of their inability to set up and test a web of causal relationships. While such methods can be valuable in providing detailed information about various aspects of the problem, used alone they are insufficient for addressing complex problems that are dynamic (i.e., change over time) and complex in terms of the large number of relationships in the system. Moreover, these methods are not designed to put all the pieces together for a big picture view.
Systems science methodologies provide a way to address complex problems, while taking into account the big picture and context of such problems. These methods enable investigators to examine the dynamic interrelationships of variables at multiple levels of analysis (e.g., from cells to society) simultaneously (often through causal feedback processes), while also studying the impact on the behavior of the system as a whole over time (Midgely, 2003). They are also useful for making implicit assumptions about complex phenomena explicit, which exposes gaps in knowledge about the problem. Moreover, simulation modeling can be used to generate “alternative futures” allowing decision makers (e.g., policy makers) to simulate the impact of various policy decisions and how they play out over time before actually putting them into practice (Sterman, 2006). For example, insights gained by the use of simulation models can help policy makers choose the most effective option among competing strategies when resources for combating the problem are limited. Systems science methodologies are also extremely useful for understanding why programs and interventions fail to have their intended effects (and in the worst cases magnify the problem; Sterman, 2000).
Systems science methodologies can also be used to refine and reform systems of care to enable planners to identify impediments to implementing proven innovations in everyday treatment and prevention practice. Dynamic models can facilitate the adoption of proven new therapeutic and business practices to ensure effective interface within existing complex systems of care. Decision tools and models can be developed to discover unanticipated effects of change on barriers to treatment and prevention services access, gaps in resource allocation, new training requirements, insufficient inter-organizational linkages, and numerous other factors affecting healthcare systems improvements.
Specific examples of systems science methodologies include, but are not limited to: systems dynamics modeling (Sterman, 2000), agent based modeling (Epstein, 2006), discrete event simulation (Banks et al., 2005), network analysis (Wasserman & Faust, 1994; Scott, 2000), dynamic microsimulation modeling (e.g., Mitton, Sutherland & Weeks, 2000), and Markov modeling (Sonnenberg & Beck, 1993). These techniques (among others) are particularly well-suited for understanding connections between a system’s structure and its behavior over time; anticipating a range of plausible futures based on explicit scenarios for action or inaction in certain areas; identifying unintended or counter-intuitive consequences of interventions; evaluating both the short- and long-term effects of policy options; and guiding investments in new research or data collection to address critical information needs. Such tools have proven heuristic power, typically integrating data from multiple prior studies and surveillance systems, and can offer innovative solutions to seemingly intractable problems. For example, systems modeling can enhance decision making and policy decisions by showing how to strike a more effective balance between treatment and prevention approaches.
Many system modeling methodologies are not new and indeed are now used routinely in fields such as corporate management, economics, engineering, physics, energy, ecology, biology, and others precisely because these methods add value compared with alternative techniques or unaided decision-making. System-oriented methods have been slower to diffuse in health-related behavioral and social science. Not surprisingly, as the appreciation for the complexity of many problems in the public health sphere has grown, there have been calls recently to address public health problems with systems science (Gerberding, 2007; Homer & Hirsch, 2006; Mabry, et al. 2008; Madon, et al., 2007; Milstein, 2008).
Systems Science and Health Resources
A compilation of systems science resources, including relevant literature, presentations, and video casts can be found on the systems science page
The Institute on Systems Science and Health (ISSH) provides investigators with a thorough introduction to selected systems science methodologies that may be used to study behavioral and social dimensions of public health. Participants in this annual, week-long Institute focus on one of three methodologies: agent-based modeling, system dynamics modeling, or network analysis. For further information (registration, track details, slide presentation, video casts, etc.) on the summer course on system science methodologies visit the ISSH home page
Behavioral and Social Sciences (BSSR) Systems Science Listserv
BSSR_Systems_Sci-L@list.nih.gov: The BSSR Systems Science listserv is a communication tool for keeping members apprised of relevant news and events; it is not a discussion board. The messages disseminated on this list are focused on those related to the intersection of behavioral and social sciences with systems science, particularly as this intersection relates to matters of health. For example, the listserv is used to announce lectures of interest, training opportunities, and funding announcements. If you or a colleague would like to subscribe to this list, simply email Patty Mabry at mabryp@od.nih.gov with full contact info, including name, title, degree, institutional affiliation, department, discipline, email address, and phone number, and she will be happy to add you to the listserv.
BSSR-Systems Science Listserv Policy
|