Health systems are complex due to the numerous elements involved and their interconnected relationships. They evolve over time and in response to external environments. This complexity often leads to unintended consequences when health policies overlook the feedback and relationships between system elements.

To effectively manage and analyse this complexity, researchers need to adopt a ‘systems thinking’ approach. This approach emphasises the connections and relationships between system elements as part of a larger, evolving system. One method to achieve this is through Causal Loop Diagrams (CLDs).

The Power of Causal Loop Diagrams

CLDs provide a visual representation of the relationships between system elements and their interactions. This allows researchers to understand what drives problematic system behaviour. They are particularly useful in resource-constrained health systems, where investments need to be well-targeted. Surprisingly, the use of CLDs has been limited in health systems research in low- and middle-income countries.

CLDs help us understand what actions or mechanisms drive behaviour in a system. They illuminate desirable or undesirable behaviour and can identify spill-over effects of actions or interventions to wider parts of the system.

When designing a CLD study, researchers need to consider the time frame of interest, the boundary of the issue, and the level of system aggregation. The goal should always be to use CLDs to map key structural drivers for a given behaviour or problem of interest, not to map the feedback that drives behaviour in the entire, wider health system.

CLDs can be generated using a variety of data sources, including primary and secondary data. The method chosen for CLD development depends on the purpose of the research and data requirements. The developed CLD then needs to be validated to minimise any unconscious bias that may have been introduced by the researcher during development or misinterpretation of data.

The Potential of CLDs in Health Systems Research

CLDs have the potential to significantly improve health systems research in low- and middle-income settings. By providing a visual representation of the relationships between system elements, they allow researchers to better understand and manage the complexity of health systems. The challenge now is to increase the use of CLDs in these settings.