Summary
Understanding the links between exposures and health outcomes is crucial in epidemiology. Deterministic and stochastic models guide understanding of the link and allow learning more about the chance, non-causal, and causal outcomes. Sir Austin Bradford Hill’s nine criteria for a systematic approach help conduct accurate observational analysis. This text will discuss and evaluate the models and associated types of Hill’s criteria to analyze the complexities of causal inference in public health research.
Deterministic and Stochastic Models in Epidemiology
The deterministic model concerns the relationship between exposure and outcome in terms of predictability. According to Dadlani et al. (2020), when exposure occurs, the outcome will always follow. For example, if a person is infected with the bacterium Vibrio, they will always develop cholera if left untreated.
A stochastic model is more concerned with the randomness or probability of disease causation. The outcomes are uncertain with exposure, but the probability does not change. For instance, smoking can cause lung cancer, but not every smoker develops the problem even though the risks are high.
Chance association usually occurs when no real relationship exists between the exposure and the outcome. In this case, both can appear at random (Westreich, 2019). For example, some studies may say eating ice cream may be associated with the following death. However, in reality, it does not mean that one causes the other. A non-causal association shows a connection between exposure and outcomes, but no direct cause-and-effect mechanism (Westreich, 2019).
For instance, there is an association between carrying a lighter and having lung cancer, but the use of the lighter does not cause health issues. In this case, the following action of smoking related to the use of a lighter can cause cancer in the future. Finally, causal association concerns the relationship between exposure and outcome. For example, eating too much sugar can cause diabetes.
Criteria of Causality
The nine main concepts defined by Sir Austin Bradford Hill’s Criteria of causality are strength, consistency, specificity, temporality, biological gradient, plausibility, coherence, experiment, and analogy. According to Shimonovich et al. (2021), the stronger the association between exposure and outcome, the more likely it is to be causal. Moreover, the association should be consistent to ensure the coverage of a broader population with specific needs and desires.
The exposure must precede the outcome temporally, and there should be a dose of linked responses. Furthermore, the association should be biologically plausible and grounded in relevant scientific knowledge, thereby making the existing knowledge more coherent. Support from experimental evidence should not be neglected, and analogical observations should be considered to strengthen the argument for causality.
Examples of Epidemiological Causation Studies
Epidemiology studies consistently report a strong association between the Zika virus and pregnancy. The study found the risk of virus spread to be high, and causality is established in this case (Awadh et al., 2017). There is a consistency in the report of the increased incidence of microcephaly following Zika virus outbreaks, making the relationship between Zika and microcephaly more causal.
Additionally, microcephaly affects pregnant women following Zika virus exposure. Therefore, the subsequent occurrence of microcephaly strengthens the evidence of causality. The dose-response relationship is where higher levels of Zika virus exposure are associated with an increased risk of microcephaly.
The incidence of microcephaly is higher in areas with a greater prevalence of Zika virus infection, indicating a biological gradient. The association between Zika virus and microcephaly is consistent with existing knowledge in the field, and experimental studies support this association (Awadh et al., 2017). TORCH infections define the analogies with other exposures known to cause similar outcomes. Biological evidence supports the association between the Zika virus and microcephaly, with plausible mechanisms that affect neural cells and establish causality. Finally, experimental studies have corroborated the epidemiological findings by demonstrating the neurotropic effect of the Zika virus.
References
Awadh, A., Chughtai, A. A., & Dyda, A. (2017). Does Zika virus cause microcephaly – Applying the Bradford Hill viewpoints. Current Outbreaks, 1-13.
Dadlani, A., Afolabi, R. O., & Jung, H. (2020). Deterministic models in epidemiology: From modeling to implementation. Quantitative Biology.
Shimonovich, M., Pearce, A., & Katikireddi, S. V. (2021). Assessing causality in epidemiology: Revisiting Bradford Hill to incorporate developments in causal thinking. European Journal of Epidemiology, 36, 873-887.
Westreich, D. (2019). Epidemiology by design: A causal approach to the health sciences. Oxford University Press.