Jain, A. K., Topchy, A., Law, M. H., & Buhmann, J. M. (2004, August). Landscape of clustering algorithms. In Proceedings of the 17th International Conference on Pattern Recognition, 2004. ICPR 2004. (Vol. 1, pp. 260-263). IEEE.

The advent of machine learning and deep learning approaches in the past decade has not only advanced existing approaches and introduced a host of new methods such as deep clustering methods. This article provides the historical timeline for some of these developments. Some of these methods are also increasingly being used for health and epidemiological data, for example deep embedded clustering (DEC) used with critical care data in this paper by de Kok et al. (2024) , I’ve provided some examples below. A good review of recent clustering methods in this space by Ezugwu et al. (2022).

Ezugwu, A. E., Ikotun, A. M., Oyelade, O. O., Abualigah, L., Agushaka, J. O., Eke, C. I., & Akinyelu, A. A. (2022). A comprehensive survey of clustering algorithms: State-of-the-art machine learning applications, taxonomy, challenges, and future research prospects. Engineering Applications of Artificial Intelligence, 110, 104743.

Collins, L. M., & Lanza, S. T. (2009). Latent Class and Latent Transition Analysis (1st ed.). John Wiley & Sons, Ltd. https://doi.org/10.1002/9780470567333

“Latent Class and Latent Transition Analysis is a comprehensive guide for identifying unobserved subgroups within a population using categorical data. Designed for researchers and students in social, behavioral, and health sciences, the book covers latent class and latent transition analysis techniques, which are used to infer hidden patterns within groups based on survey responses or other observed variables. It begins with foundational concepts and progresses to advanced topics like longitudinal latent class models, parameter restrictions, and multi-group analysis. Each method is presented with both theoretical background and practical applications, including examples of real-world data analysis. Chapters conclude with key takeaways, and the book offers online resources with data sets and specialized software tools (Proc LCA and Proc LTA in SAS) to help readers apply and experiment with the methods discussed. This resource is ideal for advanced coursework in categorical data analysis or for researchers using latent variable modeling.“

Hagenaars, J. A., & McCutcheon, A. L. (2002). Applied Latent Class Analysis. Cambridge University Press. https://www.cambridge.org/core/books/applied-latent-class-analysis/30C364913C52083262DD7CE5A2E05685

“Applied Latent Class Analysis is a hands-on guide for researchers seeking practical applications of latent class modeling to uncover hidden subgroups within data. This book is structured to support applied work, moving beyond theoretical explanations to focus on implementing latent class analysis (LCA) for real-world research challenges. The chapters cover essential methods, such as clustering and measurement models, and extend to sophisticated applications, including causal analysis, dynamic models, and handling missing data. Each section combines theoretical insights with concrete examples and step-by-step instructions on conducting analyses in SAS, providing readers with a toolkit for navigating complex datasets. The book is particularly valuable for social and behavioral researchers who need guidance on translating latent class techniques into empirical insights, with each chapter including practical case studies and applications that demonstrate the versatility of LCA in various research contexts. With empirical examples and tips on best practices, Applied Latent Class Analysis is an essential resource for those aiming to enhance their methodological skill set and apply LCA effectively in their own work. The book is also a tribute to Clifford C. Clogg, whose work laid foundational principles in the field.”

“Inferential measures such as standard errors and confidence intervals should not be used to describe the variability of characteristics, and significance tests should be avoided in descriptive tables. Also, P values are not an appropriate criterion for selecting which confounders to adjust for in analysis; even small differences in a confounder that has a strong effect on the outcome can be important.”

Vandenbroucke, J. P., Elm, E. V., Altman, D. G., Gøtzsche, P. C., Mulrow, C. D., Pocock, S. J., … & Strobe Initiative. (2007). Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. Annals of internal medicine, 147(8), W-163.

Should there be p-values in Table 1 of controlled trials (RCT)? P-values were traditionally used to assess for imbalanced baseline covariates in RCTs. However, this practice is increasingly ceasing in favor of qualitative description or no description of baseline imbalances. Randomization distributes known and unknown confounding factors, and any differences seen will be due to chance. Showing p-values can actually give a misleading picture of these differences.

“The authors of this debate are obviously more (Baethge) or less enthusiastic (Stang), with the former advocating the presentation of CIPs [covariate imbalance p values], its careful use as a screening tool, and its interpretation within the context of each study, while the latter emphasizes the dangers of misuse. The aim of this debate was to further trigger the discussion of the role of NHST in biomedical research that uses randomization… For the detection and judgment of imbalances between the study groups, it remains important that descriptive statistics of the groups (categorical characteristics: percentage values; continuous characteristics: eg, mean values and SDs) are presented. Whether a baseline imbalance is meaningful or not depends on subject matter knowledge. For example, it is clinically relevant in a stroke prevention study if 30% diabetics are in one arm of the study and only 15% are diabetics in the other arm, regardless of the p value, as diabetes mellitus is a very relevant risk factor for stroke.”

Stang, A., & Baethge, C. (2018). Imbalance p values for baseline covariates in randomized controlled trials: a last resort for the use of p values? A pro and contra debate. Clinical epidemiology, 531-535.

Are there any alternatives to the use of hypothesis testing for assessing balance? Multiple alternatives have been proposed which avoid hypothesis testing and p-values. These include tools that can assess balance graphically and numerically.

“In any study where all observed covariates were not fully blocked ahead of time, balance should be checked routinely by comparing observed covariate differences between the treated and control groups. Any statistic that is used to evaluate balance should have two key features: (a) it should be a characteristic of the sample and not of some hypothetical population and (b) the sample size should not affect the value of the statistic.”

Imai, K., King, G., & Stuart, E. A. (2008). Misunderstandings between experimentalists and observationalists about causal inference. Journal of the Royal Statistical Society Series A: Statistics in Society, 171(2), 481-502.

“MatchIt contains several tools to assess balance numerically and graphically. The primary balance assessment function is summary.matchit(), which is called when using summary() on a MatchIt object and produces several tables of balance statistics before and after matching. plot.summary.matchit() generates a Love plot using R’s base graphics system containing the standardized mean differences resulting from a call to summary.matchit() and provides a nice way to display balance visually for inclusion in an article or report. plot.matchit() generates several plots that display different elements of covariate balance, including propensity score overlap and distribution plots of the covariates. These functions together form a suite that can be used to assess and report balance in a variety of ways.”

https://cran.r-project.org/web/packages/MatchIt/vignettes/assessing-balance.html

A great literature base or reference library is vital to help with writing and developing ideas; however, getting started in a new area or developing interdisciplinary / transdisciplinary work can be challenging. A literature base is the first step in achieving a comprehensive review, systematic review, or meta-analysis. Literature reviews are essential for constructing an introduction section of a manuscript, discussion section, or covering information within a book chapter. Sometimes finding the various sources and filtering out articles to synthesize can be a daunting task. Now there are several tools to build a solid literature base. The first step is to get started with a comprehensive literature review or even a systematic literature review and meta-analysis.

There are many bibliographic research tools available out there. Many use simple co-citation analysis to identify a network of relevant articles, while others use more powerful algorithms to generate and visualize their networks. Below, is a selection of some of the more unique tools, representing a slice of tools available. These can be used to build a base reference library. They are user-friendly and many have re-occurring updates. You can play around with these features to figure out which tool works best for you.

*This list is in alphabetical order, not by ranking. All videos provided by websites cited.

CoCites: www.cocites.com
CoCites uses a browser-based extension for Chrome and Firefox which adds an information box for papers in PubMed. It provides information about the numbers of times a paper is cited and its co-citated articles. The tool uses a co-citation similarity network to rank relevant papers, currently limited to 100 most recently published. This has been found to be effective in evaluation studies of the tool.

Connected Papers: www.connectedpapers.com
Connected Papers is a literature visualization tool, with a search engine connected to multiple databases including PubMed, Semantic Scholar and arXiv. Other features include:
i. A graph that details the lead author, year, and strength of the connection based on co-citation and bibliographic coupling to the original paper
ii. Recommends relevant articles, identifying articles that are brand new and articles that do not cite each other but may be related as it does not rely on a co-citation mechanism
iii. Once the graph developed, the articles in the graph can be downloaded, and the prior papers (likely to be most critical in the field) and derivative works (likely to be reviews) tab can also be downloaded
The demonstration below (image provided by Connected Papers) shows the website’s recent collaboration with arXiv, where every paper in arXiv.org links to a Connected Papers Graph.

Inciteful: inciteful.xyz
Inciteful.xyz is an academic article network finder with a fast, user friendly interface. It features various ways to engage with the “seed papers” users’ input. The tool’s powerful seeding mechanism allows for further refining the network of papers that are most relevant in multiple rounds. It also allows for filtering of the network by keywords, by distance and by year. As well, for adding additional papers to the network manually. The tool provides a range of metrics useful in analyzing the publication landscape including:
i. Similar papers
ii. Most important papers in the network as identified by pagerank
iii. Recent papers by the top 100 authors
iv. Most important recent paper with a ranking
v. Top authors in the area for the papers identified
vi. Institutions most published in the network
vii. Top journals for the research area

JSTOR Labs Text Analyzer: jstor.org/analyze
JSTOR Text Analyzer is a program that extracts and analyzes text in an article. It can help build better key terms for a systematic review and papers related to the example article analyzed. Following submission, the tool analyzes terms that are explicit or implied in the text and highlights relevant topics. It then generates a list of recommended topics that can be filtered by year, type, and accessibility.

Litmaps: www.litmaps.co
Litmaps website develops maps based on key terms and articles, which can be selected on the site or uploaded from a citation manager. There are multiple ways to create a map based on the articles selected or key terms, which shows the selected articles’ relatedness. The maps tab features allow users to develop their maps. The explore tab allows users to find articles based on the original map. The systematic tab is a new feature coming soon to develop a systematic search process.

Open Knowledge Maps: www.openknowledgemaps.org
Open Knowledge Maps promotes research discoverability by showing an overview of topics and relevant concepts using the 100 relevant papers. Retrieving literature with high meta-data quality, the maps cluster papers based on key terms used in the field.

Vosviewer: www.vosviewer.com
VOSViewer is a text mining and article network software that must be downloaded. The software allows users to build various maps based on authors, journals, universities, conferences, and key terms. These maps also have functions for customized clustering, how the maps can be visualized, and an analysis function. The tool has been extensively reviewed, validated, and is highly cited in literature. It features:
i. Rich range of data sourcesfrom Web of Science and Pubmed to OpenCitations and WikiData
ii. Mapping tool and visualization network with labeling
iii. Clustering of networks based on co-authorship, co-citations, and bibliographic coupling
iv.

Researchrabbit: www.researchrabbit.ai
ResearchRabbit is a newly-developed literature discovery tool which provides a personalized experience for users. Currently early access is available by request. The features include:
i. Finding earlier, latest, and similar works based on key article(s) provided to build a collection that can be saved if you need to come back to it
ii. A literature map based on the data provided
iii. A list of references from selected papers can also be added to the collection
iv. Ability to export the collection to a reference manager
v. Email updates based on your collections about articles that may be of interest

Amsterdamska, O. (2005). Demarcating epidemiology. Science, technology, & human values, 30(1), 17-51.

Who are we?

‘Who are we?’ is a question that every introduction to epidemiology course grapples with. In “To Advance Epidemiology” Stallones (1980) formulates the ‘theory of epidemiology’ as an axiom and its two corollaries: that “Disease does not distribute randomly in human populations”, but variations occur across time and space, and in response to variations in causal factors. A more vague version of this definition appears in Wikipedia today.

But who are we really? Epidemiologists are well aware that we are not a basic or clinical science, but we rely on the basic sciences and clinical sciences to understand disease processes. We are not a social science, demography, or economics and yet so many our theories and methods dealing with populations and data collection are borrowed from these fields. We’re in a dependent relationship with the field of biostatistics— they could do without us, I’m not sure we could survive without them. We’re neither public health nor in public policy and yet the knowledge we generate finds its way there. Some have pinned this identity crisis on the radical transformation of the field from its inception to its present state.

Traditional versus modern?

Among these, Pearce’s article “Traditional Epidemiology, Modern Epidemiology, and Public Health” (1996) has split epidemiology in two eras: ‘traditional’ versus ‘modern’. The ‘traditional approach’, the epidemiology as practiced by Jon Snow and his contemporaries in mid-19^th century, was motivated by public health and studies of populations in situ— focusing on outcomes as a result of “processes and structures”. These ‘traditional’ epidemiological studies tackled structural problems. The tools used were demographic tools, and the outcomes of interest and interventions were almost exclusively at the level of the population. On the other hand, ‘modern epidemiology’, as demarcated by Pearce considers itself a science and embraces reductionist, positivist approaches. Modern epidemiologic studies are often disjointed from a context; subjects are not populations in situ but increasingly individuals—and even organs, DNA, molecules. The ‘gold standard’ from which data is derived is the randomized clinical trial—an experiment—insofar as epidemiologists can conduct a scientific experiment— and the ideal intervention is directed to the individual.

Much like those who lament the invasion of science in humanities (“Crimes Against Humanities: Now science wants to invade the liberal arts. Don’t let it happen.”) Pearce warns this has already happened to epidemiology, and perhaps explains the identity crisis we face today. In Pearce’s view, modern epidemiology has morphed into a field which privileges what it views as evidence generated from ‘modern science’, that is modern biomedicine and modern biostatistics, to evidence generated from all other modes of inquiry—including but not limited to science that is now considered passé or too ‘undeveloped’ or irrelevant (“too political”). Pearce makes a compelling case that epidemiology is now dismissing and disregarding the very questions and methods that on which the field was founded—and which only this field can answer. This we are to understand, has emerged only in recent years—a kind of usurpation of scientism, heralded the latest and shiniest methods and approaches, that increasingly marginalized the core of what made epidemiology a distinct field.

Amsterdamska, however, paints a different picture, not one of a field usurped and neutered but one of a field that has re-invented itself again and again to survive.

Engagement or isolation?

Amsterdamska describes epidemiology as a field that has struggled with internal anxieties regarding its involvement with politics and administrative apparatuses and its non-laboratory based, inductive methods since its inception. At the turn of the century, with the advent of the germ theory of disease, society increasingly turned to biomedicine and experimental lab-based sciences to solve problems of health and disease. The biosciences saw great rapid expansion with the founding of National Institutes of Health (NIH) in the US and the Medical Research Council (MRC) in the UK (pg. 39). As ‘the sciences’ increasingly asserted their position in society as ‘epistemic authorities’, some epidemiologists (one notable one being Major Greenwood) may have seen a proverbial writing on the wall for their field: if epidemiology were not to be recognized as a ‘science’ as a field it could be left behind in investigations of disease, and eventually find its demise as a proto-science or mere historic curiosity—a new alchemy.

At the turn of the 20th century, these ‘threats’ translated into effort to demarcate the field and establish its terrain within the broad scope of sciences and create “boundary-work”. The internal anxieties outlined above began to manifest as tension between epidemiology and other emerging fields including biometrics/biostatistics and the biomedical sciences, in particular bacteriology. In biometricians, epidemiologists found adversaries who criticized both the methods of epidemiology and epidemiology-derived public policy and health efforts as lacking “scientific objectivity” (pg. 27). Early epidemiologists such as Arthur Newsholme, viewed this as an attempt of either (1) “usurpation” where biometricians lacking knowledge in disease processes tried to solve problems they didn’t understand using unnecessarily complex mathematical methods (2) “overstepping” by a field that was an “excellent servant but a very bad master” for health investigations . Both these responses reflected an endeavor to demote a competing field as subservient to epidemiology and thereby establish the field’s epistemic dominance.

The internal anxieties also manifested as what Amsterdamska finds persistent attacks on “bacteriologists” and “renegade epidemiologists” (pg. 32). Interestingly Amsterdamska finds no evidence that any active critiques of epidemiology from the quarters of the laboratory sciences ever took place. Nevertheless, epidemiologists such as William Hamer and F.G. Crookshank launched critiques of the field of bacteriology. No surprisingly, early on they also revered adamant opponents of germ theory such as Charles Creighton (Trostle, 1986). Amsterdamska attributes the scathing attacks on bacteriology by British epidemiologists as a ‘scapegoating’ to distract from epidemiology’s own failures, in for example minimizing the public health devastation caused by the 1918 influenza. As well, a pre-emptive act to, once again “limit any potential competition” (pg. 33).

Among these ‘renegade’ epidemiologists who were also subject of critiques, was the epidemiologist and biometrician Major Greenwood, who early in his career was heavily criticized for his statistical work by the community as producing epidemiologic work based on a “mathematical house of cards”  (pg. 28). Later on, Greenwood worked with bacteriologists including W.C.C. Topley and developed experimental epidemiology . This renegade action also earned him criticism from Crookshank who viewed these new methods as comparable to “attendance at cinemas to the study of human life” (pg. 38). Here we find evident a tension between those in the field itself regarding how the boundaries and identity of epidemiology could be asserted. Individuals such as Hamer and Crookshank took an isolationist approach in order to maintain field ‘purity’ and epistemic authority, whereas individuals such as Greenwood and later Wade Hampton Frost saw epistemic authority as being maintain only so long as epidemiology was closely engaging and growing with other sciences on an equal plane (pg. 35).

Cause of death: war or the bullet?

In late 19th century the field of epidemiology was distinguished by the ‘objects’ of its study: disease. Amsterdamska finds that in this early stage, and even with the early advent of germ theory, epidemiologists were not yet concerned with demarcating their field, as the information flowing from bacteriology was found useful and complementary with epidemiologic and public health efforts (pg. 21). During the interwar period epidemiologists began to refine the ‘object’ of their domain, as a specific ‘perspective towards disease’ (specifically infectious diseases) though at least two views on the subject existed (pg. 34).  The subtle difference between these two views was that Greenwood believed a convergent view where all fields were examining the same plane through different perspectives (“bird’s eye view”  of a city, versus a ground tour), whereas Crookshank believed a divergent view where different planes were being examined through different perspectives (“causes of war” versus “individual cause of death in a war”) (pg. 31).

However, the epidemiologic transition to chronic diseases presented a new challenge in that the etiologies and causal links were now complex. The ‘mono-causal’ flat planes of infectious diseases had turned into massive ‘multi-causal’ webs where a ‘single perspective towards disease’ was no longer sufficient. In the challenge posed by chronic diseases the ‘object’ of epidemiology shifted from ‘disease’ to ‘methods’ investigating causal webs associated with health outcomes (pg. 42). One could argue that this transition made epidemiology as a field less defined than ever, as social epidemiologists, environmental epidemiologists etc. emerged whose proximal outcomes of interest often did not directly pertain to a disease process.

However, these methods, often adopted from other fields were used to create a coherent epidemiologic methodological framework, in which rigorous high-quality knowledge could be obtained. Epidemiology now had at its core a methodological framework, from its randomized control trials to the full integration of inferential statistics and hypothesis testing. The complexities of chronic diseases had led to the refinement of methods in the field to better understand multi-causality, including refined study designs and techniques to identify bias and confounding (Hanne, 2007).

The master or the emissary?

The ambition of early epidemiologists had been to establish epistemic authority and be recognized as a branch of science. The transformations that occurred during the transition to chronic diseases in many ways fulfilled this ambition, putting it irrefutably shoulder-to-shoulder with other progenies of the Baconian scientific method. However, it must be asked whether this transformation brings forth Theseus’ paradox: did epidemiology survive its transformation? Furthermore, is the “scientification” of epidemiology—held in such high esteem by the forefather of the field— a good thing?

In contrast to Pearce who lamented the increasing scientism in epidemiology, others fear the exact opposite phenomenon. In “The Rise and Fall of Epidemiology, 1950–2000 A.D.”, one of the foremost thinkers in the field, Kenneth Rothman, forecasted a decline in epidemiology, because it was becoming un-scientific (Rothman, 1981).The article lamented that while the field had made tremendous headways in mid-twentieth century, it would perhaps die in the new century leaving behind only “methods… to serve some future generation with sufficient curiosity to apply them”. The culprits of this inevitable downfall? non-experimental methods (observational studies, ecologic studies), and greater entanglement with governmental bureaucracy. In a way, Rothman’s concerns paralleled that of Major Greenwood and others— in that epidemiology cannot decline into a less ‘scientific’ state lest it be left behind in as a historic relic. This view has been countered by others such as Coleman in “Is epidemiology really dead, anyway?” (2007), have called epidemiology as vital and alive as ever—with observational studies proving a powerful scientific tool, and the field answering new and important questions about the human condition.

Whether it is more scientific or less scientific than it used to be— there is no doubt that epidemiology has transformed— head to head with medicine where from examining disease through descriptive pathology we’ve turned to the lens of prediction. Indeed, increasingly human health is no longer described in terms of symptoms and diagnoses, rather predictive statistics where every individual is at risk—as I once heard it described ‘we’re all sick, we just don’t know it yet.’ In examining these trends, early epidemiologists such as Crookshank, and perhaps even Greenwood who admittedly viewed statistics as “necessary but not sufficient” (pg. 36) , would perhaps find one of their great fears realized: that epidemiology has in fact been usurped by biostatistics and is no longer the master of its own domain, but an emissary linking fields together.

Where do epidemiologists belong?

In its nascent era epidemiologists were clinicians or specialists in other areas often working on behest of governmental institutions. Modern epidemiologists are PhDs trained in the field, and rather than being embedded in health departments they work in institutes of higher education, schools of medicine and/or schools of public health.  Amsterdamska details both the development of identity and work-boundary of epidemiology and the shift of epidemiology in England from the confines of governmental institutions to academic institutions. In summary, associated with changes in the disciplinary identity and boundaries of epidemiology as a field epidemiologist in the UK moved from being embedded within governmental institutions to academic institutions—a trend replicated elsewhere in the world.

As described by Amsterdamska “these changes in the definitions of disciplinary identity of epidemiology went hand in hand with changes in the institutional location of epidemiology, its professional organization, and its practical engagement in public health policy and administration” (pg.19). The construction of boundaries and demarcation of epidemiology, Amsterdamska argues, addressed two challenges faced by the field. First, the ‘real’ and ‘imagined’ challenges posed by fields such as statistics and bacteriology, which created anxiety regarding the encroachment into the domain of epidemiology by these other fields. Second, the methods used by the field, non-experimental and non-laboratory created anxiety regarding its place and relevance in the future of science.

Through demarcation, first of objects and later methods, epidemiologists sought to address the first concern: establish an identity and prevent the encroachment of other fields into their domain. Through, first criticism of shortcomings of other scientific fields’ methods and later incorporation of other field’s methods, epidemiologists sought to address the second concern: asserting the nature epidemiology as a science. It can be argued that the restructuring of identity and the new needs associated with it precipitated into a reactive move to a place that could serve these new needs. For example, academic settings would provide epidemiologists with the opportunity for greater “cooperation and harmonious development” with other academic fields—which Major Greenwood and other interwar epidemiologists conceived of as an immediate need in the discipline (pg. 34). However, it can also be argued that any true shift in disciplinary identity could not occur until there was a proactive move to a place outside the framework, not “stifled” by constraints of the administration and state (p. 41).

It seems that prominent epidemiologists such as Major Greenwood who played an instrumental role in “locating epidemiological research in… academic setting[s]” (p. 40) took a proactive rather than a reactive approach. Their driver for doing so is argued by Amsterdamska to have been two-fold: First, to provide epidemiology with “greater intellectual and institutional autonomy”; second, to permit for opportunities to move beyond the addressing only the practical daily concerns of public health in governance and allow for “the development of epidemiological theory” (p. 40). The association between epidemiology and politics and non-laboratory-based practices is described a “perennial worry” of epidemiologists (pg. 18), and the move outside the governmental apparatus seems to have assuaged these anxieties and provided the opportunity to construct and demarcate their identity.

Foucauldian discourse on power-knowledge and the creation of professional orders provide us with another perspective to view this shift in identity and place. Long before the transitions of the interwar period, epidemiology was already professionalizing as a field, with its own professional society in Britain established in 1850 (pg. 23). Through the Foucauldian lens, professionalization represents a shift in power from the state to the professional order. This is described as a governmentality/professionalism duality, whereby the profession becomes in itself an institution of power, resisting the encroachment of the state and government into its territory, and establishing its own power-knowledge dynamic (Adams, 2012).

Throughout its history epidemiology has been closely associated with public health practice and by extension the state. Foucault’s conceptualization of power-knowledge can be understood as a dynamic in which power and knowledge emerge from one another and reinforce one another. Early epidemiologic knowledge in a large part emerged because it was needed by the state (a monopoly of power); its questions and direction were shaped by the immediate needs of the state. In its early phase, the field is described by Amsterdamska as having been a “tool for monitoring” and “used to underpin the claims that state… should have the jurisdiction to control water quality” among other things (pg. 23). The majority of epidemiologists were not merely associated with the state, rather they worked within the framework of the state – for example in the context of the London General Board of Health responsible for investigating the Cholera epidemics in mid-19^th century.

From this perspective, then, the changes in the disciplinary identity of epidemiology and its academic professionalization occurred not to assert epistemic authority not merely in the domain of science, but also to re-center power to the field. As its own institution, epidemiology is subject to its own professional rules (‘methods’ discussed earlier) its own order, norms, and questions of interest. The physical movement into institutions of learning may have been the most practical way to allow for epidemiology to develop as its own independent institution, shifting the power-knowledge dynamic away from the state and into the discipline itself.

Epidemiologists and other population health researchers have made considerable developments in bridging intersectionality theory with epidemiological methods over the past decade. The following list attempts to reflect the dynamic and non-linear nature of this endeavor and is organized as follows. The first set of articles are conceptual, reviewing the why and how of incorporating an intersectional lens into epidemiological research. The second set of articles are methodological, presenting new or novel applications of existing analytic approaches to studying the distribution and determinants of health and disease. When possible, I mention published commentaries and responses to these articles to highlight the ongoing challenges involved in “quantifying” intersectionality and to emphasize the fact that no single method is inherently intersectional. Finally, because intersectionality theory is first and foremost a critical theory with the goal of enacting transformative social change, the third set of articles are focused on examining the process of epidemiological knowledge production itself as a means of reinforcing or challenging systems of power, as well as how epidemiological knowledge can be used to promote intersectional health equity. Of course, this list is nowhere near exhaustive, and is influenced by both extant interpretations of intersectionality theory in the field of epidemiology and my own social location, academic training, and worldview… read more »

Praxis Articles

Shim JK. Understanding the routinised inclusion of race, socioeconomic status and sex in epidemiology: The utility of concepts from technoscience studies. Sociol Heal Illn. 2002;24(2):129 – 150. 

In this sociological article, Shim reviews the use and conceptualization of social identities within epidemiological research and argues that the conflation of social identity with individual-level biology and/or behavior in mainstream epidemiology (re)produces hegemonic (and problematic) notions regarding differences between social groups. She offers a critique of the multifactorial model of disease causation, positing that the model ignores both the intersectional nature of social identities and the social and historical contexts in which disease occurs, and raises important questions about the process of epidemiological knowledge production (e.g., “To what extent are epidemiologists deliberately conscious of and concerned about the meanings that specific measures of race, class and sex/gender embody?” (p. 143)). Overall, Shim’s article is an important read as it encourages epidemiologists to view our work as both a scientific and social product. 

Access article on Google Scholar »

Wemrell M, Merlo J, Mulinari S, Hornborg AC. Contemporary epidemiology: A review of critical discussions within the discipline and a call for further dialogue with social theory. Sociol Compass. 2016;10(2):153–171. 

Wemrell and colleagues offer a critical review of historical debates within epidemiology, with a focus on critiquing the hegemony of “risk factor epidemiology”. In particular, risk factor epidemiology is argued to erase within-group heterogeneity, ignore social and historical contexts, and prioritize scientific objectivity and neutrality over public health-promoting social change. They conclude by arguing for a greater incorporation of social theory (including intersectionality theory) into epidemiological research, particularly with respect to studying the social determinants of health and health disparities, in order to produce epidemiological knowledge that better promotes health equity. 

Access article on Google Scholar »

Breilh J. Latin American critical (‘Social’) epidemiology: New settings for an old dream. Int J Epidemiol. 2008;37:745–750. 

Breilh’s article provides an important critique of the “scientific discrimination” in mainstream epidemiological and public health research, whereby the scientific contributions from Latin America are under-appreciated and under-acknowledged relative to contributions from Western countries. As a means of challenging this form of structural academic bias, he outlines the objectives and major contributions of “Latin American critical epidemiology”, illustrating how the field has “constructed a sound institutional and academic platform from which to exercise a democratic projection of science and mold an alternative public health movement” (p. 749). Breilh’s article is an essential read for epidemiologists looking to engage in intersectional scholarship as it encourages us to ask who is “at the table” of epidemiological knowledge production (and why), and what the implications are with respect to producing knowledge to promote intersectional health equity.

Access article on Google Scholar »

Ng E, Muntaner C. A critical approach to macrosocial determinants of population health: Engaging scientific realism and incorporating social conflict. Curr Epidemiol Reports. 2014;1:27–37. 

Ng and Muntaner argue for the advancement of macrosocial epidemiology, defined as the study of macro-level factors, processes, and institutions (e.g., globalization, macroeconomics) on the population patterning and determinants of health and disease. Although they do not explicitly discuss intersectionality theory, macrosocial epidemiology’s emphasis on structural power and social justice is directly in-line with the theory’s core tenets and critical bent, making this read helpful for thinking about how intersectionality and social epidemiological theories can inform and build upon one another. 

Access article on Google Scholar »

Inhorn MC, Whittle KL. Feminism meets the “new” epidemiologies: toward an appraisal of antifeminist biases in epidemiological research on women’s health. Soc Sci Med. 2001;53(5):553–67. 

Similar to Ng and Muntaner’s argument for macrosocial epidemiology, Inhorn and Whittle put forth their vision of a “feminist epidemiology”, characterized by a strong engagement with critical social theory (including intersectionality theory), an examination of the researcher’s positionality, a focus on structural power, and a grass-roots, participatory approach. 

Access article on Google Scholar »

Cho S, Crenshaw KW, McCall L. Toward a field of intersectionality studies: Theory, applications, and praxis. Signs (Chic). 2013;38(4):785–810. 

Last but certainly not least, this article from Cho and colleagues provides a comprehensive overview of extant intersectional scholarship in an effort to define and distinguish a field of “intersectionality studies”. Although not specific to epidemiology, the authors “Template for a Collaborative Intersectionality” can provide epidemiologists with important guidelines for incorporating intersectionality into their work and foster a more critical, interdisciplinary, and social justice-oriented epidemiology.

Access article on Google Scholar »

Epidemiologists and other population health researchers have made considerable developments in bridging intersectionality theory with epidemiological methods over the past decade. The following list attempts to reflect the dynamic and non-linear nature of this endeavor and is organized as follows. The first set of articles are conceptual, reviewing the why and how of incorporating an intersectional lens into epidemiological research. The second set of articles are methodological, presenting new or novel applications of existing analytic approaches to studying the distribution and determinants of health and disease. When possible, I mention published commentaries and responses to these articles to highlight the ongoing challenges involved in “quantifying” intersectionality and to emphasize the fact that no single method is inherently intersectional… read more »

Methodological Articles

Veenstra G. Race, gender, class, and sexual orientation: Intersecting axes of inequality and self-rated health in Canada. Int J Equity Health. 2011;10(3):1–11. 

Veenstra’s article is one of the first to explicitly present a quantitative analytic approach for incorporating intersectionality theory into epidemiological research. He uses statistical interaction (i.e., two- and three-way cross-product terms between social identity variables in multivariable models) to assess whether multiply marginalized groups are more likely to report poor self-rated health relative to singly- or non-marginalized groups, drawing on intersectionality core tenets of directionality, simultaneity, multiplicativity, and multiple jeopardy. Although the use of statistical interaction to assess intersectionality is increasingly critiqued by social epidemiologists, this article and those informed by it helped motivate the development of the methods introduced in the following articles. 

Access article on Pubmed »

Jackson JW, Williams DR, VanderWeele TJ. Disparities at the intersection of marginalized groups. Soc Psychiatry Psychiatr Epidemiol. 2016;51(10):1349–1359. 

Building off the aforementioned critiques of using statistical interaction to assess intersectionality, this article from Jackson, Williams, and VanderWeele argues in favor of evaluating additive-scale interaction (i.e., a situation in which the combined effect of two factors differ from the sum of each factor’s individual effect) as an alternative quantitative analytic approach. They present a novel additive-scale interaction method that is specific to intersectionality, the joint disparity and its decomposition, and illustrate how it and related measures can be used to quantify excess (i.e., intersectional) risk of disease among groups at the nexus of two marginalized social identities. Technical appendices provide equations for applying the methods to continuous and binary outcomes. 

Access article on Pubmed »

Commentaries:

Schwartz S. Commentary: on the application of potential outcomes-based methods to questions in social psychiatry and psychiatric epidemiology. Soc Psychiatry Psychiatr Epidemiol. 2017;52(2):139–42. 

Jackson JW. Explaining intersectionality through description, counterfactual thinking, and mediation analysis. Soc Psychiatry Psychiatr Epidemiol. 2017;52(7):785–93.

Wemrell M, Mulinari S, Merlo J. Intersectionality and risk for ischemic heart disease in Sweden: Categorical and anti-categorical approaches. Soc Sci Med. 2017;177:213–22. 

The majority of intersectionality-informed epidemiological research adopts an intracategorical or intercategorical orientation; much less attention has been directed to developing quantitative methods consistent with anticategorical intersectionality (see McCall, 2005 listed above). Wemrell and colleagues help to fill this gap by introducing an innovative method for conducting quantitative anticategorical intersectionality-informed research based on the epidemiological concept of discriminatory accuracy (DA). Specifically, the authors contend that quantifying the ability of social identity-based categories to distinguish between those with and without a health outcome of interest (as is typically done in epidemiological research) can demonstrate the large degree of heterogeneity within these categories, and thus what anticategorical intersectionality describes as the “simplifying social fictions” (McCall, 2005, p. 1773) of social identities. They outline their analytic approach (using area under the receiver-operating characteristic curves) and provide guidance for intersectionality theory-consistent interpretations.

Access article on Pubmed »

Evans CR, Williams DR, Onnela JP, Subramanian SV. A multilevel approach to modeling health inequalities at the intersection of multiple social identities. Soc Sci Med. 2018;203:64–73. 

In this pioneering article, Evans and colleagues introduce a novel modeling approach for intercategorical intersectionality-informed epidemiological research that is referred to in subsequent articles as Intersectional Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (MAIHDA). Briefly, Intersectional MAIHDA models are multilevel models in which individuals are nested within groups defined by intersecting social identities, which enables investigations of within- and between-group heterogeneity and the ability to identify and quantify interaction (i.e., intersectional) effects for all groups. The authors outline methodological and theoretical advantages of the method over “conventional” intersectional models (i.e., fixed effects models with cross-product terms between social identity variables) and provide an illustrative example. 

Access article on Pubmed »

Commentaries: 

Merlo J. Multilevel analysis of individual heterogeneity and discriminatory accuracy (MAIHDA) within an intersectional framework. Soc Sci Med. 2018; 

Evans CR. Adding interactions to models of intersectional health inequalities: Comparing multilevel and conventional methods. Soc Sci Med. 2019;221:95–105. 

Evans further develops the Intersectional MAIHDA method by investigating whether adding additional dimensions of identity into the models reveals or explains away intersectional effects. She also compares findings from Intersectional MAIHDA models to those obtained from conventional intersectional models and discusses plausible statistically-based reasons for the observed differences. Stata code for all analyses are provided in the supplementary materials. 

Access article on Pubmed »

Commentaries:

Bauer GR. Math versus meaning in MAIHDA: A commentary on multilevel statistical models for quantitative intersectionality. Soc Sci Med. 2019; 112500. 

Evans CR, Leckie G, Merlo J. Multilevel versus single-level regression for the analysis of multilevel information: The case of quantitative intersectional analysis. Soc Sci Med. 2019;245:112499. 

Evans CR. Reintegrating contexts into quantitative intersectional analyses of health inequalities. Heal Place. 2019;60:102214. 

Another Intersectional MAIHDA methods article, Evans outlines how to incorporate social contexts (e.g., schools and neighborhoods) into the models. She also suggests alternative conceptualizations of contexts for quantitative intersectionality-informed research, with the goal of motivating epidemiologists to better attend to socio-structural processes in their analyses.

Access article on Google Scholar »

Scheim AI, Bauer GR. The Intersectional Discrimination Index: Development and validation of measures of self-reported enacted and anticipated discrimination for intercategorical analysis. Soc Sci Med. 2019;226:236–45. 

In the only measurement-specific article on this list, Scheim and Bauer develop and validate the Intersectional Discrimination Index (InDI), a measure of attribution-free anticipated and enacted discrimination to be used in intercategorical intersectionality-informed epidemiological research. The authors include the final version of the index within the paper, provide guidance regarding its scoring, and outline an analytic approach for using the index to assess discrimination as social determinant of intersectional health inequities in a companion article (listed below). 

Access article on Google Scholar »

Commentaries:

Harnois CE, Bastos JL. The promise and pitfalls of intersectional scale development. Soc Sci Med. 2019;223:73–6. 

Bauer GR, Scheim AI. Methods for analytic intercategorical intersectionality in quantitative research: Discrimination as a mediator of health inequalities. Soc Sci Med. 2019;226:236–45. 

In the companion article to that listed above, Bauer and Scheim illustrate a novel method for analytic (i.e., process-oriented) intersectionality-informed epidemiological research. They start with a helpful overview of extant analytic approaches for incorporating intersectionality theory into epidemiological research and highlight persisting methodological challenges, with an emphasis on research aiming to conduct causal analyses. They then introduce their method, a causal mediation analysis (adapted from VanderWeele’s three-way decomposition method and informed by the potential outcomes framework) that allows for heterogeneity of the mediated effect across groups defined by intersecting social identities, and provide an illustrative demonstration complete with equations and SAS code. 

Access article on Google Scholar »

Commentaries:

Evans CR. Modeling the intersectionality of processes in the social production of health inequalities. Soc Sci Med. 2019;226:249–53.

Jackson JW, VanderWeele TJ. Intersectional decomposition analysis with differential exposure, effects, and construct. Soc Sci Med. 2019;226:254–9. 

Richman LS, Zucker AN. Quantifying intersectionality: An important advancement for health inequality research. Soc Sci Med. 2019;226:246–8. 

Bauer GR, Scheim AI. Advancing quantitative intersectionality research methods: Intracategorical and intercategorical approaches to shared and differential constructs. Soc Sci Med. 2019;226:260–2.

Social identities (i.e., gender, race, ethnicity, class, sexual orientation, age, etc.) are a key construct within epidemiological research, used to define populations and study the patterning of health and disease. And yet, epidemiologists have conceptualized identity in varied, shifting, and oftentimes problematic ways. There have been numerous calls to shift from a discrete and individualized framing of identity to a more nuanced and contextualized one, reflecting broader debates in epidemiology between theories of disease distribution that locate risk within the biology and behaviors of individuals and those that locate risk further upstream (Krieger, 2001). 

Perhaps because of its potential to improve upon extant conceptualizations of identity as well as other core population health constructs, there has been a growing interest in intersectionality theory among epidemiologists. Originating in Black feminist activism and scholarship as a way to explain the unique forms of discrimination experienced by Black women (Collins, 1990; Crenshaw, 1989), intersectionality is a theoretical framework primarily concerned with the multidimensional, interlocking nature of social inequities. Within this framework, social identities are conceptualized as mutually constitutive, jointly shaping lived experience through their historical and ongoing relationships with systems of power. This has several implications for epidemiology: adopting an intersectional perspective could lead to a more nuanced (and valid) understanding of the population patterning of disease, a heightened focus on social determinants, and more structurally oriented interventions. However, there are well-articulated challenges involved in translating a theory aimed at describing complex realities (versus generating hypotheses or predictions) into a quantitative science such as epidemiology, as well as important questions over epistemological consistency.  

Thankfully, epidemiologists and other population health researchers have made considerable developments in bridging intersectionality theory with epidemiological methods over the past decade. The following list attempts to reflect the dynamic and non-linear nature of this endeavor and is organized as follows. The first set of articles are conceptual, reviewing the why and how of incorporating an intersectional lens into epidemiological research. The second set of articles are methodological, presenting new or novel applications of existing analytic approaches to studying the distribution and determinants of health and disease. When possible, I mention published commentaries and responses to these articles to highlight the ongoing challenges involved in “quantifying” intersectionality and to emphasize the fact that no single method is inherently intersectional. Finally, because intersectionality theory is first and foremost a critical theory with the goal of enacting transformative social change, the third set of articles are focused on examining the process of epidemiological knowledge production itself as a means of reinforcing or challenging systems of power, as well as how epidemiological knowledge can be used to promote intersectional health equity. Of course, this list is nowhere near exhaustive, and is influenced by both extant interpretations of intersectionality theory in the field of epidemiology and my own social location, academic training, and worldview. 

Conceptual articles

Bowleg L. The problem with the phrase women and minorities: Intersectionality-an important theoretical framework for public health. Am J Public Health. 2012;102(7):1267–73. 

In one of the first articles to explicitly advocate for an intersectional approach to epidemiological-related research, Bowleg provides a compelling case for adopting intersectionality theory as a critical public health framework, focusing on its potential to “[reframe] how public health scholars conceptualize, investigate, analyze, and address disparities and social inequality in health” (p. 1267). After providing a brief history of intersectionality and outlining its core tenets, she reviews the major theoretical and methodological challenges associated with incorporating the theory into public health research; however, she stresses that having an “intersectionality-informed stance” is more important than methodological refinement. Bowleg concludes by outlining five ways in which this intersectionality-informed stance will benefit public health, ranging from increased understanding of how intersecting social identities influence the patterning and determinants of health disparities to improved interventions and surveillance efforts.     

Access article on Pubmed »

Bauer GR. Incorporating intersectionality theory into population health research methodology: Challenges and the potential to advance health equity. Soc Sci Med. 2014;110(10–17). 

Another formative article, Bauer starts similarly by articulating the potential benefits of incorporating intersectionality theory into the broader field of population health sciences. She provides a detailed and nuanced description of the associated methodological challenges, with a focus on measurement- and analytic-related concerns, and offers guidelines for addressing them. The most useful aspect is Bauer’s discussion of how tensions within intersectional scholarship may be understood in the context of population health research. For example, questions such as “who” is intersectional (e.g., all social locations or only those with multiple marginalized identities) and whether all identities are relevant/intersectional in all contexts are addressed in terms of their significance for understanding the distribution and determinants of health and disease. She also considers how intersectionality meshes with epidemiologic theories, particularly Nancy Krieger’s ecosocial theory, with implications for the validity and social value of epidemiological research. 

Access article on Pubmed »

Gkiouleka A, Huijts T, Beckfield J, Bambra C. Understanding the micro and macro politics of health: Inequalities, intersectionality & institutions – A research agenda. Soc Sci Med. 2018;200:92–8. 

Gkiouleka and colleagues review the theoretical and methodological underpinnings of using intersectionality as an analytic tool for studying health inequities and offer two research approaches relevant to epidemiology: “situational intersectionality” (i.e., focusing on the specific social identities relevant to a given research question, while exploring potential heterogeneity in such relevance across groups) and “institutional imbrication” (i.e., operationalizing institutions and other macro-level factors typically studied discreetly in epidemiology (e.g., neighborhoods and laws/policies) as intersectional). The article concludes with recommendations for measurement and analysis when using these research approaches and outlines seven actions that researchers can take to make their work more intersectional. 

Access article on Pubmed »

Bowleg L. When Black + lesbian + woman ≠ Black lesbian woman: The methodological challenges of qualitative and quantitative intersectionality research. Sex Roles. 2008;59:312–25. 

While not specific to epidemiology or even to population or public health, this earlier article by Bowleg nonetheless provides an excellent overview of measurement-, analytic-, and interpretation-related challenges to conducting intersectionality-informed research. For each of these domains, she provides specific guidelines for addressing the noted challenges, although she emphasizes that contextualization is paramount to defining intersectional scholarship. 

Access article on Google Scholar »

McCall L. The complexity of intersectionality. Signs (Chic). 2005;30:1771–800. 

This seminal article in intersectional scholarship laid the groundwork for how epidemiologists began to operationalize the theory for quantitative research. McCall presents three orientations for intersectionality-informed research, which represent a continuum of approaches to managing the complexity of multiple intersecting identities: “anticategorical intersectionality” (critique and deconstruct social identities), “intracategorical intersectionality” (reveal nuance within a particular social location), and “intercategorical intersectionality” (document inequalities across many social locations). After reviewing the underlying assumptions and analytic guidelines for each orientation, she calls for an expansion of methodologies used to answer intersectionality-informed research questions so as to “fully engage with the topics and issues of intersectionality” (p. 1774); on a personal note, I find this a strong argument for the incorporation of intersectionality theory into epidemiology! 

Access article on Google Scholar »

Evans CR. Modeling the intersectionality of processes in the social production of health inequalities. Soc Sci Med. 2019;226:249–53. 

This article is a comprehensive overview of the types of intersectionality-informed epidemiological and population health studies. Evans provides a useful classification system for this growing body of research, distinguishing specific/intracategorical from comprehensive/intercategorical and descriptive from analytic studies, providing researchers with a common language to describe their objectives and analyses. She also suggests specific epidemiologic theories that are particularly well-aligned with intersectionality theory (and could thus be used in conjunction, advancing both fields) and outlines future research directions.  

Access article on Google Scholar »

Directed Acyclic Graphs, or DAGs, are commonly used in Epidemiology to communicate the assumptions that are being made about the structure of the causal network being studied, and can be used to guide our analytic strategy by facilitating the identification of non-causal pathways which could bias the estimates of targeted effect estimands. The following is a list of resources that are useful in learning about DAGs and their applications:

Pearl J. Causal diagrams for empirical research. Biometrika 1995;82:669–88.

Access article on Google Scholar »

In this article, Pearl introduces a framework utilizing DAGs to identify and communicate what assumptions are necessary to produce causal effect estimates from nonexperimental data.

Greenland S, Pearl J, Robins JM. Causal Diagrams for Epidemiologic Research. Epidemiology 1999; 10:37–48.

Access article on Google Scholar »

Greenland, Pearl and Robins provide an introduction to the use of DAGs in Epidemiology. This includes an overview of terminology, application, and a discussion of how this framework expands the traditional epidemiologic criteria for confounding.

Glymour MM. Using causal diagrams to understand common problems in social epidemiology. Methods Soc. Epidemiol., 2006, p. 387–422.

Access article on Google Scholar »

In Chapter 16, Glymour provides an overview for the application of DAGs in Epidemiology with an emphasis on their use in causal inference and in social epidemiology. Glymour also provides discussion of related contemporary issues such as: natural experiments, identification of direct and indirect effects, and the use of indicator variables for handling missing data.

Hernán MA, Robins JM. Graphical Representation of Causal Effects. Causal Inference What If, 2020, p. 69–82.

Access article on website »

Chapter 6 provides an introduction to the use of DAGs in Epidemiology for causal inference, with an emphasis on linking DAGs to the counterfactual framework.

Westreich D, Greenland S. The table 2 fallacy: Presenting and interpreting confounder and modifier coefficients. Am J Epidemiol 2013; 177: 292–8.

Access article on Google Scholar »

Westreich and Greenland use DAGs to introduce and illustrate the Table 2 Fallacy (i.e., the fallacy that all estimated coefficients for a given model can be interpreted as estimating the same type of effect). They also discuss strategies that can be employed to avoid misinterpretations that can result from the Table 2 Fallacy.

Textor J, Hardt J, Knüppel S. DAGitty: A graphical tool for analyzing causal diagrams. Epidemiology 2011;22:745.

Access article on Google Scholar »

DAGitty is a free, user-friendly, online tool that allows you to easily build your DAG. The application is then able to identify the minimally sufficient adjustment set necessary for estimating the total effect or direct effect. It can also identify instrumental variables that are available in the causal network, and it provides a list of testable assumptions based on the causal network (http://www.dagitty.net/dags.html).

Segregation of epidemiology into chronic and infectious diseases has led to a neglected area in public health – the interface between chronic disease and infectious disease.

Choi, B. C., Morrison, H., Wong, T., Wu, J., & Yan, Y. P. (2007). Bringing chronic disease epidemiology and infectious disease epidemiology back together. Journal of Epidemiology & Community Health, 61(9), 802-802.

In the past month or so I’ve been seeing a lot of dialogue on #epitwitter about the lack of infectious disease training in many epidemiology programs. In an informal poll of 1000+ it seemed like many in epidemiology programs don’t receive little to no infectious disease training as part of their core training. The name ‘epidemiology’ calls back the era at the founding of the field when infectious diseases were the main health ‘epidemics’ impacting the population at large. But increasingly, infectious diseases have become a relic of a bygone era, as by far most epidemiologist are focused on chronic diseases.

However, we’ve seen in recent decades many infectious diseases not only interact with chronic diseases, but are the underlying causes of chronic diseases. Infectious disease epidemiologists in many ways continue to be at the forefront of epidemiology, integrating some of the most cutting edge tools into their field: predictive modelling and forecasting, complex spatial analysis, genomic mapping etc. As Choi et al. (2007) describe elegantly, many tools developed by infectious disease researchers may be of utility to understand ‘infectiousness’ in other fields in absence bacterial or viral agents.

Fortunately in absence of structured programs, there are a lot of good resources out there. For example, to be able to conduct my research on tuberculosis (TB) transmission, I had to self-teach a lot of infectious disease epidemiology through text-books, online resources, etc. Below I’ve listed ten courses for self-teaching at the introductory level, all freely available online.

Open courses in infectious disease epidemiology

Epidemics in Western Society Since 1600 (Yale)
https://oyc.yale.edu/history/hist-234

A well-rounded understanding of epidemiology requires a dive into the past. After all, many of the infectious diseases prevalent in the world today, such as tuberculosis, have been with us historically– and as look to eradicate these diseases– those who don’t remember the past, as they say, are doomed to repeat it. This course, offered from a historian’s perspective, covers history stretching from the plague and the social reactions to it, the development of the sanitary movement and germ theory, historical impact of infectious diseases (such as TB, malaria, polio) and re-emergence of these diseases in the modern context. Course recordings from 2010 and sample mid-term and final exams are provided.

John Snow and the Cholera Epidemic of 1854 (Harvard University via edX)
https://www.edx.org/course/predictionx-john-snow-and-the-cholera-epidemic-of

This course focuses on the a pivotal point in the history of infectious disease epidemiology: the Cholera epidemic of 1854 that swept through London. The course explores the role of one of the legendary figures of epidemiology, John Snow, in understanding how cholera was being spread and how it could be stopped. John Snow’s endeavor is perhaps the most well-known example of what is called ‘shoe-leather epidemiology’, going into the field with your feet to investigate what is happening. John Snow knocked on doors and created maps, and this course includes interactive mapping tools and timelines that allow students to engage with his investigation as it unfolded.

Epidemiology of Infectious Diseases
http://ocw.jhsph.edu/index.cfm/go/viewCourse/course/EpiInfectiousDisease/…

This is a comprehensive course in infectious disease epidemiology, covering the basics of the field including definitions, the principles of disease surveillance and outbreak investigation, and types of epidemiologic studies used to investigate outbreaks. The course also includes information about transmission dynamics, technical presentations on modelling and molecular epidemiology which many other courses in this list don’t cover quite as well. Although this is a comprehensive offering, one drawback is that the course is from 2006 and may not include the most up to date information. Nevertheless, highly recommend this for any students in epidemiology looking to extend their training.

Outbreaks and Epidemics (Johns Hopkins University via Coursera)
https://www.coursera.org/learn/outbreaks-epidemics

This short course takes a more hands-on approach to epidemic dynamics. Over the course of four modules, students are challenged to apply epidemiological and statistical concepts to infectious disease problems. The course examines concepts such as risk difference and relative difference, attack rates, reproductive numbers and other important metric useful in understanding scale of epidemics. Finally, the course applies these concepts to not just infectious disease outbreaks worldwide as many other courses, but to non-infectious disease epidemics such as the opioid epidemic in the US.

Epidemics – the Dynamics of Infectious Diseases (Penn State University via Coursera)
https://www.coursera.org/learn/epidemics

Over eight modules, this course provides a detailed review of the dynamics of infectious diseases from how they emerge to how they spread through environmental and social networks including concepts such as breaking transmission chains and ‘super-spreaders’. The basic ecology of infectious diseases, including vector-pathogen interaction and how the characteristics of these agents impacts the patterns of disease incidence is examined. The course uses case studies of pandemic influenza, childhood diseases such as measles and malaria as case studies. Finally, the course touches on drug resistance and how it impacts efforts to contain many re-emerging infectious diseases.

Epidemics, Pandemics and Outbreaks (University of Pittsburgh via Coursera)
https://www.coursera.org/learn/epidemic-pandemic-outbreak

While we often think of infectious diseases at the smallest scale—the level of the patient. However, infectious diseases do not recognize borders, and much of how infectious diseases can spread into pandemics is driven by policies at the global, national and local levels. The strength of this course is its focus on policy—how legal and public health systems respond to outbreaks and pandemics. One of the modules on local countermeasures includes extensive discussion ethical and pragmatic issues around disease reporting, travel restrictions, quarantine. Another module focuses on global health security. examining how globalization has altered the patterns of disease transmission, and how nations work together (or not) to detect and respond to infectious disease threats.

Principles of Infectious Disease Epidemiology Online Training Course (Missouri Department of Health)
https://health.mo.gov/training/epi/

For those interested in outbreak investigation, this course developed by the Missouri Department of Health has a fantastic series of modules on the topic. The course covers public health surveillance, including how to evaluate and improve surveillance systems. Then it dives deep into understanding and organizing and interpreting epidemiologic data generated from surveillance and investigation systems. Finally, the course dives into the principles of outbreak investigation—including interviewing skills, preparing investigation reports, with each step taught working through outbreak case studies. To access the course you need to register on the site, but the course is open to public and a great learning experience.

Epidemics I and II (The University of Hong Kong via edX)
https://www.edx.org/course/epidemics-i
https://www.edx.org/course/epidemics-ii

These two courses examine infectious disease epidemiology in a comprehensive manner. Offered by the University of Hong Kong, the courses have a special focus on infectious diseases in the East Asian context. The first course starts by examining the ecology and evolution of infectious disease, moving on to diseases such as Ebola and Zika and how epidemics emerge. Finally using SARS as a model, the first course evaluates identification of novel diseases, and the use of informatics including genetic data to understand the trajectories of epidemics. The second course elaborates and expands on some of these discussions, with a strong module on infectious disease modeling, examining dynamic and structural uncertainty in models of disease spread. The course also includes a supplemental module on COVID-19, as the course designers are involved in active research in this disease. This module will include “study results, final size estimation for super-spreading clusters, the Chinese experience for global preparedness, and more.”

Plagues, Pestilence and Pandemics: Are You Ready? (Griffith University via Futurelearn)
https://www.futurelearn.com/courses/plague-pestilence-pandemic

This very short course is a useful non-technical review of some of the current and emerging global health pandemics (including SARS and Ebola). The course examines outbreak responses and how transmission chains can be broken. The course also explores bacterial resistance, including healthcare associated infections and the threat they present.

Lessons from Ebola: Preventing the Next Pandemic (Harvard University via edX)
https://www.edx.org/course/lessons-from-ebola-preventing-the-next-pandemic-2

Ebola virus disease (EBV), a hemorrhagic fever with mortality rates ranging between 20-90%, was first identified in 1976. The disease re-appeared throughout the next four decades in small clusters. However, in 2014 the WHO declared a major outbreak in Guinea which over the course of the following year expanded to many neighboring countries across West Africa. While that outbreak subsided, newer Ebola outbreaks have appearing in the intervening years and as of last year the WHO has declared Ebola a global health emergency. This course takes a retrospective examination of the 2014 Ebola outbreak, focusing on the many aspects of the disease including the presentation and transmission patterns. As well the course examines the numerous challenges that surfaced in containment efforts at the local level, and failure of international efforts to prevent the spread of the disease. The deep dive into this pandemic provides insight into what went wrong, and how we could avoid the same pitfalls in the future.

Additional: live courses focused on ongoing corona virus pandemic (COVID 19)

COVID-19: Tackling the Novel Coronavirus (The London School of Hygiene & Tropical Medicine via Futurelearn)
https://www.futurelearn.com/courses/covid19-novel-coronavirus

In December of 2019, the WHO was informed about a small cluster of cases of pneumonia of unknown origin in Wuhan, China. In the two months following, the disease agent was identified as a new coronavirus (COVID-19) and rapidly spread across the world, infecting hundreds of thousands in a matter of weeks and declared as a worldwide pandemic. At the moment of writing this summary (3/20/2020) this pandemic has lead to states of emergency across many countries, travel bans, and shut down of non-essential services. This course, starting on Mar 23, 2020 will offer the latest information about COVID-19 presented by experts in the field, including the spread mechanisms for the disease the public health measures necessary to contain it.

Science Matters: Let’s Talk About COVID-19 (Imperial College London via Coursera)
https://www.coursera.org/learn/covid-19

This course (starting on March 19, 2020) is a more in-depth review of COVID-19 covering topics including the basic reproduction number of the disease, its case fatality, phylogenetic analysis, and clinical presentation. The course also offers a review of the economics and social impact of the current outbreak, with the latest materials as events unfold. Finally, the development of a vaccine is the target goal for any new disease, and COVID-19 is no different. this course contains module on the process of vaccine development during real-time epidemics.

If you have suggestions for additional resources, please comment below or email me!

Last updated: 3/21/2020