Complexity in Epidemiology and Public Health. Addressing Complex Health Problems Through a Mix of Epidemiologic Methods and Data.

Public health and the underlying disease processes are complex, often involving the interaction of biologic, social, psychologic, economic, and other processes that may be nonlinear and adaptive and have other features of complex systems. There is therefore a need to push the boundaries of public health beyond single-factor data analysis and expand the capacity of research methodology to tackle real-world complexities. This article sets out a way to operationalize complex systems thinking in public health, with a particular focus on how epidemiologic methods and data can contribute towards this end. Our proposed framework comprises three core dimensions-patterns, mechanisms, and dynamics-along which complex systems may be conceptualized. These dimensions cover seven key features of complex systems-emergence, interactions, nonlinearity, interference, feedback loops, adaptation, and evolution. We relate this framework to examples of methods and data traditionally used in epidemiology. We conclude that systematic production of knowledge on complex health issues may benefit from: formulation of research questions and programs in terms of the core dimensions we identify, as a comprehensive way to capture crucial features of complex systems; integration of traditional epidemiologic methods with systems methodology such as computational simulation modeling; interdisciplinary work; and continued investment in a wide range of data types. We believe that the proposed framework can support the systematic production of knowledge on complex health problems, with the use of epidemiology and other disciplines. This will help us understand emergent health phenomena, identify vulnerable population groups, and detect leverage points for promoting public health.

Clinical characteristics and progression of COVID-19 confirmed cases admitted to a single British clinical centre-A brief case series report.

OBJECTIVES: In December 2019, a pneumonia-like illness was first reported in Wuhan-China caused by a new coronavirus named corona virus disease-2019 (COVID-19) which then spread to cause a global pandemic. Most of the available data in the literature is derived from Chinese cohorts and we aim to contribute the clinical experience of a single British clinical centre with the characteristics of a British cohort. DESIGN: A prospective case series. SETTING: A single clinical centre in the UK. METHODS: We have collected the demographics and medical characteristics of all COVID-19-positive cases admitted over 2-week period. All cases were diagnosed by PCR. RESULTS: Total of 71 COVID-19 patients were included in this case series. Majority of patients (75%) were ≥75 years old and 58% were men. Pre-existing comorbidities was common (85% of patients). Most patients presented with respiratory symptoms such as fever (59%), shortness of breath (56%) and cough (55%). Gastrointestinal symptoms were second-most common presenting compliant such as diarrhoea (10%) and abdominal pain (7%). Opacification in chest X-rays was demonstrated in 45% of patients. All patients received supportive treatment and no specific antiviral therapy was administered in this cohort. So far, 18 (25%) patients have fully recovered, 9 patients (13%) escalated to a higher level of care and 10 (14%) have died. Patients who died were non-significantly older than those who have recovered (78.0 vs 69.2 years, P = .15) but they had a significantly higher clinical frailty scores (5.75 vs 3.36, P = .005). CONCLUSION: This case series demonstrated that the characteristics of British COVID-19 patients were generally similar to what is published in literature, although we report more gastrointestinal symptoms at presentation. We have identified frailty as a risk factor for adverse outcome in COVID-19 patients and suggest that it should be included in the future vaccination recommendations.

Intellectualizing Medicine: A Reply to Commentaries on "Prediction, Understanding, and Medicine".

This article is a reply to two critics of my "Prediction, Understanding, and Medicine," published elsewhere in this journal issue. In that essay, I argued that medicine is best understood not as essentially a curative enterprise, but rather as one essentially oriented towards prediction and understanding. Here, I defend this position from several criticisms made of it.

Prediction, Understanding, and Medicine.

What is medicine? One obvious answer in the context of the contemporary clinical tradition is that medicine is the process of curing sick people. However, this "curative thesis" is not satisfactory, even when "cure" is defined generously and even when exceptions such as cosmetic surgery are set aside. Historian of medicine Roy Porter argues that the position of medicine in society has had, and still has, little to do with its ability to make people better. Moreover, the efficacy of medicine for improving population health has been famously doubted by historians and epidemiologists. The curative thesis demands that we have mostly been stupid, duped, or staggeringly hopeful, given that medicine has not until recently offered more than a handful of effective cures. I suggest, in this article, that the core medical competence is neither to cure, nor to prevent, disease, but to understand and to predict it. I argue that this approach does a better job than the curative thesis at explaining why not all medicine is concerned with curative efforts and that it enjoys historical support from the ancient entanglement of prophecy and medicine and from the fact that medicine thrived for centuries with almost no effective cures and continues to thrive today in various forms that are mostly without curative efficacy. I suggest that this approach grounds a fairer approach to alternative, traditional, and other medical practices, as well as some fresh lessons for the development of mainstream medicine.

Pandemic response strategies and threshold phenomena.

This paper critically evaluates the Suppression Threshold Strategy (STS) for controlling Covid-19 (C-19). STS asserts a "fundamental distinction" between suppression and mitigation strategies, reflected in very different outcomes in eventual mortality depending on whether reproductive number R is caused to fall below 1. We show that there is no real distinction based on any value of R which falls in any case from early on in an epidemic wave. We show that actual mortality outcomes lay on a continuum, correlating with suppression levels, but not exhibiting any step changes or threshold effects. We argue that an excessive focus on achieving suppression at all costs, driven by the erroneous notion that suppression is a threshold, led to a lack of information on how to trade off the effects of different specific interventions. This led many countries to continue with inappropriate intervention-packages even after it became clear that their initial goal was not going to be attained. Future pandemic planning must support the design of "Plan B", which may be quite different from "Plan A".

Can Robots Do Epidemiology? Machine Learning, Causal Inference, and Predicting the Outcomes of Public Health Interventions.

This paper argues that machine learning (ML) and epidemiology are on collision course over causation. The discipline of epidemiology lays great emphasis on causation, while ML research does not. Some epidemiologists have proposed imposing what amounts to a causal constraint on ML in epidemiology, requiring it either to engage in causal inference or restrict itself to mere projection. We whittle down the issues to the question of whether causal knowledge is necessary for underwriting predictions about the outcomes of public health interventions. While there is great plausibility to the idea that it is, conviction that something is impossible does not by itself motivate a constraint to forbid trying. We disambiguate the possible motivations for such a constraint into definitional, metaphysical, epistemological, and pragmatic considerations and argue that "Proceed with caution" (rather than "Stop!") is the outcome of each. We then argue that there are positive reasons to proceed, albeit cautiously. Causal inference enforces existing classification schema prior to the testing of associational claims (causal or otherwise), but associations and classification schema are more plausibly discovered (rather than tested or justified) in a back-and-forth process of gaining reflective equilibrium. ML instantiates this kind of process, we argue, and thus offers the welcome prospect of uncovering meaningful new concepts in epidemiology and public health-provided it is not causally constrained.

Can you lock down in a slum? And who would benefit if you tried? Difficult questions about epidemiology's commitment to global health inequalities during Covid-19.

The initial response to the Covid-19 pandemic was characterised by swift "lockdowns," a cluster of measures defined by a shared goal of suppressing Covid-19 and a shared character of restricting departure from the home except for specific purposes. By mid-April 2020, most countries were implementing stringent measures of this kind. This essay contends that (1) some epidemiologists played a central role in formulating and promulgating lockdown as a policy and (2) lockdowns were foreseeably harmful to the Global Poor, and foreseeably offered them little benefit, relative to less stringent measures. In view of the widespread commitment to reducing global health inequalities within the profession, this should prompt reflection within the epidemiological community and further work on pandemic response measures more appropriate for the Global Poor.

What could possibly go wrong? - A heuristic for predicting population health outcomes of interventions.

Austin Bradford Hill offers a general heuristic for causal inference in epidemiology, but no general heuristic for prediction is available. This paper seeks to identify a heuristic for predicting the outcome of interventions on population health, informed by the moral context of such interventions. It is suggested that, where available, robust predictions should be preferred, where a robust prediction is one which, according to the best knowledge we are currently able to obtain, could not easily be wrong. To assess whether a prediction is robust, it is suggested that we ask why the predicted outcome will occur, rather than any other outcome. Firstly, if, according to our current knowledge, we cannot identify the likeliest ways that the other outcomes could occur, then the prediction is not robust. And secondly, if, according to our current knowledge, we can identify the likeliest other outcomes but we are unable to say why our predicted outcome will occur rather than these, then, again, our prediction is not robust. Otherwise, it is robust. The inaccurate but memorable short version of the heuristic is, "What could possibly go wrong?"

Conceptual and methodological issues in epidemiology: An overview.

In 2010 a series of workshops on philosophical and methodological issues in epidemiology was held at the University of Cambridge. The papers in this volume arise from those workshops. This paper represents an effort to identify, in broad brush, some of the major conceptual and methodological issues in epidemiology, which form the basis of an emerging focus on the philosophy of epidemiology.

Corrigendum to "COVID-19 in South Africa" [Global Epidemiology 2 (2020) 100034].

[This corrects the article DOI: 10.1016/j.gloepi.2020.100034.].

Health as a Secondary Property.

In the literature on health, naturalism and normativism are typically characterized as espousing and rejecting, respectively, the view that health is objective and value-free. This article points out that there are two distinct dimensions of disagreement, regarding objectivity and value-ladenness, and thus arranges naturalism and normativism as diagonal opposites on a two-by-two matrix of possible positions. One of the remaining quadrants is occupied by value-dependent realism, holding that health facts are value-laden and objective. The remaining quadrant, which holds that they are non-objective but value-free, is unexplored. The article endorses a view in the latter quadrant, namely, the view that health is a secondary property. The article argues that a secondary property framework provides the resources to respond to the deepest objections to a broadly Boorsean account of natural function, and so preserves the spirit, though not the letter, of that account. Treating health as a secondary property permits a naturalistic explanation-specifically, an evolutionary explanation-of the health concept, in terms of the assistance such a concept might have provided to the survival and reproduction of those organisms that had it. (This approach is completely distinct from evolutionary and aetiological accounts of natural functions.) This provides the explanation, missing from Boorse's account, for the fact that function is determined with reference to the contribution to the goals of survival and reproduction, relative to the age of the sex of the species, rather than some other equally natural goals or reference classes. 1Introduction2Two Ways to Disagree about Health3Secondary Properties4Health as a Secondary Property5Conclusion.

Causality and causal inference in epidemiology: the need for a pluralistic approach.

Causal inference based on a restricted version of the potential outcomes approach reasoning is assuming an increasingly prominent place in the teaching and practice of epidemiology. The proposed concepts and methods are useful for particular problems, but it would be of concern if the theory and practice of the complete field of epidemiology were to become restricted to this single approach to causal inference. Our concerns are that this theory restricts the questions that epidemiologists may ask and the study designs that they may consider. It also restricts the evidence that may be considered acceptable to assess causality, and thereby the evidence that may be considered acceptable for scientific and public health decision making. These restrictions are based on a particular conceptual framework for thinking about causality. In Section 1, we describe the characteristics of the restricted potential outcomes approach (RPOA) and show that there is a methodological movement which advocates these principles, not just for solving particular problems, but as ideals for which epidemiology as a whole should strive. In Section 2, we seek to show that the limitation of epidemiology to one particular view of the nature of causality is problematic. In Section 3, we argue that the RPOA is also problematic with regard to the assessment of causality. We argue that it threatens to restrict study design choice, to wrongly discredit the results of types of observational studies that have been very useful in the past and to damage the teaching of epidemiological reasoning. Finally, in Section 4 we set out what we regard as a more reasonable 'working hypothesis' as to the nature of causality and its assessment: pragmatic pluralism.

Risk relativism and physical law.

In two 1959 papers, one coauthored, Jerome Cornfield asserts that 'relative' measures are more useful for causal inference while 'absolute' measures are more useful for public health purposes. In one of these papers (the single-authored one), he asks how epidemiology should respond to the fact that its domain is not a highly 'articulated' one-it is not susceptible to being subsumed under general laws. What is the connection between these issues? There has recently been some backlash against 'risk relativism', and Charles Poole has recently dismantled the mathematical argument for the first claim. However the problem with 'Cornfield's Principle' seems to go much deeper. The whole attempt to partition measures into absolute and relative is fundamentally mistaken. Why, then, has it seemed so appealing? Perhaps one reason is the influence that early education in the physical sciences continues to exert on the way epidemiologists think, and their response to the low articulation of their domain of study.

Causation and prediction in epidemiology: a guide to the "methodological revolution".

There is an ongoing "methodological revolution" in epidemiology, according to some commentators. The revolution is prompted by the development of a conceptual framework for thinking about causation here referred to as the Potential Outcomes Approach (POA), and the mathematical apparatus of directed acyclic graphs that accompanies it. But over and above the mathematics, a number of striking theses about causation are evident, for example: that a cause is something that makes a difference; that a cause is something that humans can intervene on; and that causal knowledge enables one to predict under hypothetical suppositions. This is especially remarkable in a discipline that has variously identified factors such as race and sex as determinants of health, since it has the consequence that factors of this kind cannot be treated as causes either as usefully or as meaningfully as was previously supposed. In this paper I seek to explain the significance of this movement in epidemiology, to understand its commitments, and to evaluate them.

Epidemiological evidence in law: a comment on Supreme Court Decision 2011Da22092, South Korea.

This paper offers a commentary on three aspects of the Supreme Court's recent decision (2011Da22092). First, contrary to the Court's finding, this paper argues that epidemiological evidence can be used to estimate the probability that a given risk factor caused a disease in an individual plaintiff. Second, the distinction between specific and non-specific diseases, upon which the Court relies, is shown to be without scientific basis. Third, this commentary points out that the Court's finding concerning defect of expression effectively enables tobacco companies to profit from the efforts of epidemiologists and others involved in public health to raise awareness of the dangers of smoking.