>> The path dependence theory has its origins in economics (David 1985; Arthur 1994) and was subsequently taken up by political science (Pierson 2000) and organization, management and information systems research (Singh et al. 2015). The “Berlin Path Model” (Phases I to III in the figure), which explains the emergence of organizational path dependence, is presented in more detail below (Sydow et al. 2009).
At the heart of the model is the assumption that unintended consequences of decisions combined with self-reinforcing mechanisms lead to a state that may be suboptimal but difficult to change. This can be done at various levels of analysis, for example with a view to the development of extremely stable sector and professional boundaries in the German healthcare system on a macro level through to the establishment of actually inefficient prescribing practices among service providers on a micro level (Reuter 2012). Path dependency arises in three phases: In the first phase there are a large number of options for action. The choice of one of these options for action – triggered by a critical event at a “critical juncture” – then leads to the action of one or more self-reinforcing mechanisms. This initiates the second phase in which the self-reinforcing mechanisms take effect and increasingly limit possible alternative courses of action for actors and organizations. This ultimately leads to the third phase of lock-in or hyperstability. A path has emerged here that makes choosing alternative courses of action virtually impossible. A fourth phase then characterizes attempts and their consequences to overcome the rigidity that has arisen through measures (“path break strategies”) (see Fig. 1).
Existing path research knows four mechanisms that can lead to self-reinforcement of a path without the need for further impulses or interventions from outside. Which includes:
1) Coordination effects (“the status quo is the least
coordination effort involved"),
2) complementarity effects (“the way everything is, it fits together best”),
3) Learning effects (“first-order learning stabilizes existing patterns”) and
4) adaptive expectations (“self-fulfilling prophecies, especially when several people look at future developments”).

Within the framework of path theory, the assumption is that conscious actions by actors such as B. Health managers do require path dependence; but this was not necessarily intended. Their behavior can certainly seem sensible from an individual perspective. Only the interaction of a large number of decisions and effective mechanisms then cause the outlined effects of self-dynamic processes. Path dependence can therefore arise “behind the backs” of actors and be difficult to notice. At the same time, an existing path dependency can last for a while, e.g. B. under constant conditions, represent a particularly efficient situation, at least for individual stakeholders. Both make it difficult to break path dependency.

Path break strategies (see Phase IV in the figure) are discussed as a possible way out of path dependencies. They should be applied at the latest when the path dependency no longer represents an efficient state. Path-breaking strategies ideally override self-reinforcing mechanisms and can require the creation of a new path (path creation). The literature discusses, for example, discursive practices, regulatory interventions and the change in existing power structures (Schreyögg et al. 2003).
However, the prospects of success of these and other conceivable path-breaking strategies have so far been relatively less researched than the development and effects of path dependence.
Potentially path-breaking events (e.g. large-scale political reforms, the corona pandemic) could be another trigger for a possible path break (Sydow et al. 2020). However, health systems are characterized by a high level of complexity and many interdependencies, which can make it extremely difficult to break through self-reinforcing mechanisms once they have been established. Although path breaks in the healthcare system have been documented (Singh et al. 2015; Stache/Sydow 2022), they also represent an enormous design challenge for the German healthcare system (Schreyögg 2004; Gersch/Sydow 2017).

The theory of path dependence does not just describe the hyperstability of a system as an actual state. Rather, it is a process theory that focuses on the path to this state - or its breaking through. This theoretical approach can therefore explain the stability or change of social and technical systems. At first glance, path theory therefore appears to be extremely attractive for illustrating entrenched conditions and resistance to change in healthcare at a technological, organizational and systemic level. Path dependence is assumed to be the cause of a sector separation that continues to exist in Germany (e.g. Auschra/Sydow 2021), slow progressing digitalization (SVR 2021) or a general rigidity towards innovation (Gersch/Sydow 2017).
Path theory is therefore able to explain developments in health systems ex post and at different levels of analysis. Political scientists in particular have used the theory to illustrate mechanisms of resistance to change in entire health systems (e.g. Wilsford 1994). This shows the framework conditions under which health care organizations operate and the associated effects, such as the inertia of health care compared to (health) politically or economically desirable change. Economists have been able to show, for example, that the care of diabetes patients is stabilized in an undesirable way by learning effects and adaptive expectations at the level of the individual actors (Reuter 2012).
At the same time, the theory can also develop a forward-looking character. This happens when it is used to identify self-reinforcing mechanisms, for example to develop forecasts for suspected standardization processes or perspectives for a potentially undesirable path emergence or a future path break. An example here would be the question of which mechanisms, such as coordination and complementarity effects in combination with adaptive expectations, lead to a lack of interoperability of technological standards in the German healthcare system (Stegemann/Gersch 2019). Overcoming the lock-in in provider-specific specifications and isolated data silos could be achieved if a suitable combination of regulatory framework and economic incentives for the use of international standards create new momentum. As a result, every decision would be based on a new, changed expectation, which could lead to a sufficient “critical mass” of new solution implementations, especially with simultaneous support through appropriate infrastructural announcements and measures. However, it remains unclear whether this will be achieved by linking the current KHZG (2021) to compliance with international standards, a gematik change of course to announce a TI 2.0 and the establishment of an interoperability council (“INTEROP Council”: gematik 2022) as well as changing the funding conditions of the innovation fund (Joint Federal Committee 2022) can be more successful than previous attempts (Gersch/Sydow 2017).
Surprisingly, there are so far few studies from the field of health services research that use path theory (Auschra/Sydow 2021). This may be because the theory is primarily suitable for explaining processes retrospectively and requires complex empirical investigations (see below). However, path dependence theory is not only used at the health system level. It is shown that self-reinforcing mechanisms can inhibit the reorganization of hospitals (Krantz/Knoch 2013) or, at the organizational level, hinder the implementation of care innovations (e.g. Ilott et al. 2016). Few studies apply path dependence theory in the context of individual indications, for example Reuter (2012) for diabetes or Matsuoka and Yamaguchi (2011) for dementia.

Measuring or proving path dependence is associated with methodological challenges (see also Sydow et al. 2012; Sydow et al. 2020). It is therefore obvious to assume a path dependence if strong resistance to change is observed. However, this assumption has so far rarely been supported by the representation of the causes that lead to path dependence - critical events and self-reinforcing mechanisms (Auschra/Sydow 2021). In most cases, this can only be achieved through a detailed, often retrospective analysis - which is very time-consuming, especially when examining path dependencies at a systemic level. For example, identifying a critical event that led to path dependence is systematically difficult in retrospect. This is particularly the case if the event happened a long time ago and research results depend on the memory of certain key people. In any case, the analysis of path dependencies requires longitudinal data collection and, if possible, addressing multiple levels.
The path dependence theory is hardly suitable for evaluating specific innovations to improve care. Nevertheless, the strength of this theory lies in the fact that it takes a reflective look at the historical framework that makes the implementation of such an innovation difficult, thereby enabling a well-founded analysis of the well-known puzzles: why are changes so difficult, why are they difficult in terms of course and The result is so unpredictable and why do change projects so often fail? <<

Arthur, W. (1994): Increasing returns and path dependence in the economy. Ann Arbor, MI: University of Michigan Press.
Auschra, C./Sydow, J. (2021): Path dependence and integrated care. In: Amelung, V./Stein, V./Suter, E./Goodwin, N./Nolte, E./Balicer, R. (eds.): Handbook Integrated Care. Cham: 119–129.
David, PA (1985): Clio and the economics of QWERTY. In: American Economic Review 1985, 75, 2: 332–337.
gematik (2022): Interoperability in healthcare. In: https://www.gematik.de/telematikarchitektur/iop (accessed on April 24, 2022).
Federal Joint Committee (2022): G-BA Innovation Fund. In: https://innovationsfonds.g-ba.de (accessed on April 24, 2022).
Gersch, M./Sydow, J. (2017): The innovation fund from the perspective of innovation research. In: Amelung, VE/Eble, S./Hildebrandt, H./Knieps, F./Lägel, R./Ozegowski, S./Schlenker, R.-U./Sjuts, R. (eds.): Innovation Fund. Impulses for the German healthcare system. Berlin: 58–64.
Ilott, I./Gerrish, K./Eltringham, SA/Taylor, C./Pownall, S. (2016): Exploring factors that influence the spread and sustainability of a dysphagia innovation: An instrumental case study. In: BMC Health Services Research 2016, 16, 1: 1–14.
KHZG (2021): KHZG – The Hospital Future Act. In: https://khzg.de (accessed on April 24, 2022).
Krantz, U./Knoch, M. von (2013): Path dependencies as obstacles to reorganization processes in hospitals: Using the example of coordination effects. In: Healthcare 2013, 75, 10: 658–659.
Matsuoka, H./Yamaguchi, H. (2011): Path dependence in social and psychological risk factors for dementia. In: Dementia & Neuropsychologia 2011, 5, 1: 2–7.
Pierson, P. (2000): Increasing returns, path dependence, and the study of politics. In: American Political Science Review 2000, 94, 2: 251–267.
Reuter, A. (2012): The resistance to change of medical practices in a dynamic field of care. A process-oriented analysis of medical prescribing practices in the treatment of diabetics (type 2) in the period from 1993 to 2009. Berlin: Doctoral dissertation. Free University of Berlin.
Schreyögg, G./Sydow, J./Koch, J. (2003): Organizational Paths - From Path Dependence to Path Creation? In: Management Research 2003, 13: 257–294.
Schreyögg, J. (2004): Path dependence of decision-making processes using the example of the German healthcare system. In: WiSt - Economic Studies 2004, 33, 6: 359–363.
Singh, R./Mathiassen, L./Mishra, A. (2015): Organizational path constitution in technological innovation: Evidence from rural telehealth. In: MIS Quarterly 2015, 39, 3: 643–665.
Stache, F./Sydow, J. (2022): Breaking a path by creating a new one – How organizational change boosts children's cancer care. In: Organization Studies 2022, 43: in print.
Stegemann, L./Gersch, M. (2019): Interoperability – Technical or economic challenge? In: it - Information Technology 2019, 61, 5-6: 243–252.
SVR (2021): Digitalization for health. Goals and framework conditions of a dynamically learning health system: Report 2021. Bern: Hogrefe Verlag.
Sydow, J./Schreyögg, G./Koch, J. (2009): Organizational path dependence: Opening the black box. In: Academy of Management Review 2009, 34, 4: 689–709.
Sydow, J./Schreyögg, G./Koch, J. (2020): On the theory of organizational path dependence: Clarifications, replies to objections, and extensions. In: Academy of Management Review 2020, 45, 4: 717–734.
Sydow, J./Windeler, A./Müller-Seitz, G./Lange, K. (2012): Path constitution analysis. A methodology for understanding path dependence and path creation. In: Business Research 2012, 5, 2: 155–176.
Wilsford, D. (1994): Path dependency, or why history makes it difficult but not impossible to reform health care systems in a big way. In: Journal of Public Policy 1994, 14, 3: 251–283.