Home Integrating Hill’s classical considerations with modern causal inference methods in observational studies: a ‘How-Questions’ framework
Article
Licensed
Unlicensed Requires Authentication

Integrating Hill’s classical considerations with modern causal inference methods in observational studies: a ‘How-Questions’ framework

  • José R. Banegas EMAIL logo , Javier Muñoz-Laguna , Francisco F. Caballero , Pilar Guallar-Castillón , Esther López-García , Auxiliadora Graciani , Verónica Cabanas , Javier Damián , Rosario Ortolá , Miguel A. Royo Bordonada and Fernando Rodríguez-Artalejo
Published/Copyright: August 2, 2024
Become an author with De Gruyter Brill
Submit Manuscript Author Information
Epidemiologic Methods
From the journal Epidemiologic Methods Volume 13 Issue s2

Abstract

Context

Modern causal inference methods – although core to epidemiological reasoning – may be difficult to master and less intuitive than Hill’s classical considerations. We developed a ‘How-Questions’ (HQ) framework to integrate Hill's classical considerations with modern causal inference methods in observational studies.

Methods

First, we extracted the main causal considerations from contemporary philosophy of science: characteristics of empirical associations, universality, depth, and degree of corroboration of a theory. From these, we developed a HQ framework based on six domains formulated as questions: (1) how valid?, (2) how time-ordered?, (3) how big?, (4) how shaped?, (5) how replicable?, and (6) how explainable? Then, we qualitatively checked whether Hill's classical considerations and key selected modern causal inference methods were compatible with the HQ framework. Lastly, as a proof-of-concept, we applied the HQ framework to two observational studies of current topics in epidemiology.

Findings

Both Hill’s considerations and key selected modern causal inference methods were compatible with the six domains of the HQ framework. (1) The how-valid domain is addressed by considering the same internal validity issues in Hill’s and modern methods, namely confounding, selection and measurement biases; modern methods use more formalized techniques, including quantitative bias analyses/sensitivity analyses (QBA/SA). (2) The how-time-ordered domain is addressed by considering reverse causation in Hill’s; modern methods may use G methods within the context of longitudinal data analyses and time-varying exposures. (3) The how-big domain is addressed by strength of association in Hill’s; modern methods first consider estimands and may use QBA/SA to assess robustness of effect estimates. (4) The how-shaped domain is represented by biological gradient in Hill’s; modern methods may use generalized propensity scores to estimate dose-response functions. (5) The how-replicable domain is addressed in Hill’s by consistency of study findings with existing evidence; modern methods may use triangulation of different study designs and consider generalizability and transportability concepts. (6) The how-explainable domain is addressed by biological plausibility in Hill’s and by mediation/interaction analyses in modern methods. The application of the HQ framework to two observational studies provides a proof-of-concept and suggests its potential usefulness to integrate Hill’s considerations with modern causal inference methods.

Perspective

We found that the six dimensions of the HQ framework integrated Hill’s classical considerations with modern causal inference methods for observational studies. Apart from its potential pedagogical value, the HQ framework may provide a holistic view for the causal assessment of observational studies in epidemiology.

Keywords: causal inference; epidemiological methods; Hill’s criteria for causality; modern methods; observational studies
Corresponding author: José R. Banegas, MD, MPH, PhD, Department of Preventive Medicine and Public Health, Universidad Autónoma de Madrid and CIBERESP, Arzobispo Morcillo 4, 28029, Madrid, Spain, E-mail: 

Acknowledgement

We thank professor Kenneth J. Rothman for wise advice on some causal inference readings and encouraging us to complete and publish this manuscript.

  1. Research ethics: Not applicable.

  2. Informed consent: Not applicable.

  3. Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission.

  4. Competing interests: The authors state no conflict of interest.

  5. Research funding: This work was supported by Fondo de Investigación Sanitaria grant 22/1164 (Instituto de Salud Carlos III and FEDER/FSE). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

  6. Data availability: Not applicable.

References

1. Elwood, M. Critical appraisal of epidemiological studies and clinical trials, 4th ed. Oxford: Oxford University Press; 2017.10.1093/med/9780199682898.001.0001Search in Google Scholar

2. US Public Health Service. Smoking and health, report of the advisory committee to the surgeon general of the US public health service, PHS publ. No. 1103. Washington, D.C.: U.S. Government Printing Office; 1964.Search in Google Scholar

3. Hill, AB. The environment and disease: association or causation? Proc Roy Soc Med 1965;58:295–300. https://doi.org/10.1177/003591576505800503.Search in Google Scholar

4. Susser, M. Causal thinking in the health sciences. Concepts and strategies of epidemiology. Oxford: Oxford University Press; 1973.Search in Google Scholar

5. Susser, M. What is a cause and how do we know one? A grammar for pragmatic epidemiology. Am J Epidemiol 1991;133:635–48. https://doi.org/10.1093/oxfordjournals.aje.a115939.Search in Google Scholar PubMed

6. Hill, AB, Hill, ID. Bradford Hill's principles of medical statistics, 12th ed. London: Edward Arnold; 1991.Search in Google Scholar

7. Lilienfeld, AM. Epidemiological methods and inferences in studies of non-infectious diseases. Publ Health Rep 1957;72:51–60. https://doi.org/10.2307/4589685.Search in Google Scholar

8. Bradford Hill, A. Statistical methods in clinical and preventive medicine. Edinburgh: E.&S. Livingstone; 1962a.Search in Google Scholar

9. Bradford Hill, A. The statistician in medicine. Alfred watson memorial lecture. J Inst Actuar 1962b;88:178–91. https://doi.org/10.1017/s0020268100014980.Search in Google Scholar

10. Morabia, A. Hume, Mill, Hill, and the sui generis epidemiologic approach to causal inference. Am J Epidemiol 2013;178:1526–32. https://doi.org/10.1093/aje/kwt223.Search in Google Scholar PubMed PubMed Central

11. Rothman, KJ, editor. Causal inference. Chesnut Hill, MA: Epidemiology Resources I; 1988.Search in Google Scholar

12. Greenland, S, editor. Evolution of epidemiologic ideas. Chesnut Hill, MA: Epidemiology Resources I (E.R.I.); 1987.Search in Google Scholar

13. Rothman, KJ, Greenland, S. Causation and causal inference in epidemiology. Am J Public Health 2005;95:S144–50. https://doi.org/10.2105/ajph.2004.059204.Search in Google Scholar

14. Pearl, J, Mackenzie, D. The book of why. The new science of cause and effect (2018). London: Penguin Random House; 2019.Search in Google Scholar

15. Lash, TL, VanderWeele, TJ, Haneuse, S, Rothman, KJ. Modern epidemiology, 4th ed. Philadelphia, PA: Wolters Kluwer; 2021.Search in Google Scholar

16. Weed, DL, Gorelic, LS. The practice of causal inference in cancer epidemiology. Cancer Epidemiol, Biomarkers Prev 1996;5:303–11.Search in Google Scholar

17. Weed, DL. On the use of causal criteria. Int J Epidemiol 1997;26:1137–41. https://doi.org/10.1093/ije/26.6.1137.Search in Google Scholar PubMed

18. Popper, KR. The open society and its enemies (1965). Addenda (facts, standards, and truth: a further criticism of relativism. 1961). Princeton, NJ: Princeton University Press; 2020.Search in Google Scholar

19. Popper, KR. Logik der Forschung. 10 Auflage (1994). Tübingen: J.C.B. Mohr, 1994; 1995 (last version in English: Popper KR. The Logic of Scientific Discovery, 1980. London: Routledge).Search in Google Scholar

20. Bunge, M. Philosophy of science, volume 2: from explanation to justification. Revised edition (1998). New York: Routledge; 2017b.10.4324/9781315126388Search in Google Scholar

21. Bunge, M. Philosophy of science, volume 1: from problem to theory. Revised edition (1998). New York: Routledge; 2017a.10.4324/9781315126371Search in Google Scholar

22. Goldstein, M, Goldstein, I. The experience in science. An interdisciplinary approach. New York: Plenum Press; 1988.Search in Google Scholar

23. Braithwaite, RB. Scientific explanation. Cambridge: Cambridge University Press; 1953.Search in Google Scholar

24. Hempel, CG. Philosophy of natural science. New Jersey: Prentice-Hall; 1966.Search in Google Scholar

25. Bunge, M. Causality and modern science. 4th revised ed (2009), London: Routledge; 2017c.10.4324/9781315081656Search in Google Scholar

26. Popper, KR. The myth of the framework. In defence of science and rationality (1994). London: Routledge; 1996.Search in Google Scholar

27. Illari, P, Russo, F. Causality. Philosophical theory meets scientific practice. Oxford: Oxford University Press; 2014.Search in Google Scholar

28. Kleinberg, SA. Guide to finding and using causes. Sebastopol, CA: O’Reilly Media Inc; 2016.Search in Google Scholar

29. Popper, KR. Realism and the aim of science. Postscript to the logic of scientific discovery. vol. I. Ed by WW Bartley III (1982). London: Routledge; 2012.10.4324/9780203713969Search in Google Scholar

30. Abelson, RP. Statistics as principled argument. Hillsdale, NJ, USA: Lawrence Erlbaum Associates, Inc; 1995.Search in Google Scholar

31. Shimonovich, M, Pearce, A, Thomson, H, Keyes, KK, Katikireddi, SV. Assessing causality in epidemiology: revisiting Bradford Hill to incorporate developments in causal thinking. Eur J Epidemiol 2021;36:873–87. https://doi.org/10.1007/s10654-020-00703-7.Search in Google Scholar PubMed PubMed Central

32. VanderWeele, TJ. Hill’s causal considerations and the potential outcomes framework. Obs Stud 2020;6:47–54. https://doi.org/10.1353/obs.2020.0008.Search in Google Scholar

33. Hernán, MA, Robins, JM. Causal inference: what if. Boca Raton: Chapman & Hall/CRC; 2020. (Revised, april 2024).Search in Google Scholar

34. Mørch, LS, Skovlund, CW, Hannaford, PC, Iversen, L, Fielding, S, Lidegaard, O. Contemporary hormonal contraception and the risk of breast cancer. N Engl J Med 2017;377:2228–39. https://doi.org/10.1056/nejmoa1700732.Search in Google Scholar

35. Delgado-Velandia, M, Maroto-Rodríguez, J, Ortolá, R, Rodríguez-Artalejo, F, Sotos-Prieto, M. The role of lifestyle in the association between frailty and all-cause mortality amongst older adults: a mediation analysis in the UK Biobank. Age Ageing 2023;52:afad092. https://doi.org/10.1093/ageing/afad092.Search in Google Scholar PubMed

36. Infante-Rivard, C, Cusson, A. Reflection on modern methods: selection bias—a Review of recent developments. Int J Epidemiol 2018;47:1714–22. https://doi.org/10.1093/ije/dyy138.Search in Google Scholar PubMed

37. Lu, H, Cole, SR, Howe, CJ, Westreich, D. Toward a clearer definition of selection bias when estimating causal effects. Epidemiology 2022;33:699–706. https://doi.org/10.1097/ede.0000000000001516.Search in Google Scholar

38. Westreich, D, Edwards, JK, Lesko, CR, Cole, SR, Stuart, EA. Target validity and the hierarchy of study designs. Am J Epidemiol 2019;188:438–43. https://doi.org/10.1093/aje/kwy228.Search in Google Scholar PubMed PubMed Central

39. Van Smeden, M, Lash, TL, Groenwold, RH. Reflection on modern methods: five Myths about measurement error in epidemiological research. Int J Epidemiol 2020;49:338–47. https://doi.org/10.1093/ije/dyz251.Search in Google Scholar PubMed PubMed Central

40. Mansournia, MA, Etminan, M, Danaei, G, Kaufman, JS, Collins, G. Handling time varying confounding in observational research. BMJ 2017;359:j4587. https://doi.org/10.1136/bmj.j4587.Search in Google Scholar PubMed

41. Naimi, AI, Cole, SR, Kennedy, EH. An introduction to g methods. Int J Epidemiol 2017;46:756–62. https://doi.org/10.1093/ije/dyw323.Search in Google Scholar PubMed PubMed Central

42. Kahan, BC, Cro, S, Li, F, Harhay, MO. Eliminating ambiguous treatment effects using estimands. Am J Epidemiol 2023;192:987–94. https://doi.org/10.1093/aje/kwad036.Search in Google Scholar PubMed PubMed Central

43. Li, H, Wang, C, Chen, WC, Lu, N, Song, C, Tiwari, R, et al.. Estimands in observational studies: some considerations beyond ICH E9 (R1). Pharmaceut Stat 2022;21:835–44. https://doi.org/10.1002/pst.2196.Search in Google Scholar PubMed

44. Austin, PC. Assessing the performance of the generalized propensity score for estimating the effect of quantitative or continuous exposures on binary outcomes. Stat Med 2018;37:1874–94. https://doi.org/10.1002/sim.7615.Search in Google Scholar PubMed PubMed Central

45. Stuart, EA, Cole, SR, Bradshaw, CP, Leaf, PJ. The use of propensity scores to assess the generalizability of results from randomized trials. J Roy Stat Soc Stat Soc 2011;174:369–86. https://doi.org/10.1111/j.1467-985x.2010.00673.x.Search in Google Scholar

46. Dahabreh, IJ, Petito, LC, Robertson, SE, Hernán, MA, Steingrimsson, JA. Towards causally interpretable meta-analysis: transporting inferences from Multiple randomized trials to a new target population. Epidemiology 2020;31:334. https://doi.org/10.1097/ede.0000000000001177.Search in Google Scholar PubMed PubMed Central

47. VanderWeele, TJ. Explanation in causal inference. Methods for mediation and interaction. New York: Oxford University Press; 2015.10.1093/ije/dyw277Search in Google Scholar PubMed PubMed Central

48. Bunge, M. Mechanism and explanation. Philos Soc Sci 1997;27:410–65. https://doi.org/10.1177/004839319702700402.Search in Google Scholar

49. Cook, TD, Campbell, DT. Quasi-experimentation. Design & analysis issues for field settings. Boston: Houghton Miflin Company; 1979.Search in Google Scholar

50. Banegas, JR, Rodríguez Artalejo, F, Rey Calero, J. Popper and the problem of induction in epidemiology (abstract in English). Rev Esp Salud Publica 2000;74:327–39. https://doi.org/10.1590/s1135-57272000000400003.Search in Google Scholar

51. Lash, TL, Fox, MP, Fink, AK. Applying quantitative bias analysis to epidemiologic data, 1st ed. New York: Springer; 2009.10.1007/978-0-387-87959-8Search in Google Scholar

52. Rosenbaum, PR. Causal inference. Cambridge, Massachusetts: The MIT Press; 2023.10.7551/mitpress/14244.001.0001Search in Google Scholar

53. VanderWeele, TJ, Jackson, JW, Li, S. Causal inference and longitudinal data: a case study of religion and mental health. Soc Psychiatr Psychiatr Epidemiol 2016;51:1457–66. https://doi.org/10.1007/s00127-016-1281-9.Search in Google Scholar PubMed

54. Clare, PJ, Dobbins, TA, Mattick, RP. Causal models adjusting for time-varying confounding: a systematic Review of the literature. Int J Epidemiol 2019;48:254–65. https://doi.org/10.1093/ije/dyy218.Search in Google Scholar PubMed

55. Mathur, MB, VanderWeele, TJ. Methods to address confounding and other biases in meta-analyses: Review and Recommendations. Annu Rev Publ Health 2022;43:19–35. https://doi.org/10.1146/annurev-publhealth-051920-114020.Search in Google Scholar PubMed PubMed Central

56. Rosenbaum, PR. Does a dose-response relationship reduce sensitivity to hidden bias? Biostatistics 2003;4:1–10. https://doi.org/10.1093/biostatistics/4.1.1.Search in Google Scholar PubMed

57. Szklo, M, Nieto, FJ. Epidemiology. Beyond the basics, 4th ed. Burlington, MA: Jones & Barlett Learning; 2019.Search in Google Scholar

58. Browner, WS, Newman, TB, Cummings, SR, Brady, DG, Huang, AJ, Kanaya, AM, et al.. Designing clinical research, 5th ed. Philadelphia: Wolters Lluwer; 2023.Search in Google Scholar

59. Lawlor, DA, Tilling, K, Davey Smith, G. Triangulation in aetiological epidemiology. Int J Epidemiol 2016;45:1866–86. https://doi.org/10.1093/ije/dyw314.Search in Google Scholar PubMed PubMed Central

60. Hansford, HJ, Cashin, AG, Jones, MD, Swanson, SA, Islam, N, Dahabreh, IJ, et al.. Development of the TrAnsparent ReportinG of observational studies emulating a target trial (TARGET) guideline. BMJ Open 2023;13:e074626. https://doi.org/10.1136/bmjopen-2023-074626.Search in Google Scholar PubMed PubMed Central

61. Hernán, MA. Methods of public health research -strengthening causal inference from observational data. N Engl J Med 2021;385:1345–8. https://doi.org/10.1056/NEJMp2113319.Search in Google Scholar PubMed

62. Rubin, DM. Estimating causal effects of treatments in randomized and nonrandomized studies. J Educ Psychol 1974;66:688–701. https://doi.org/10.1037/h0037350.Search in Google Scholar

63. Greenland, S, Pearl, J, Robins, JM. Causal diagrams for epidemiologic research. Epidemiology 1999;10:37–48. https://doi.org/10.1097/00001648-199901000-00008.Search in Google Scholar

64. Glymour, MM, Greenland, S. Causal diagrams. In: Rothman, KJ, Greenland, S, Lash, TL, editors. Modern epidemiology, 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2008. Chapter 12.Search in Google Scholar

65. Joffe, M, Gambhir, M, Chadeau-Hyam, M, Vineis, P. Causal diagrams in systems epidemiology. Emerg Themes Epidemiol 2022;9:1. https://doi.org/10.1186/1742-7622-9-1.Search in Google Scholar PubMed PubMed Central

66. Pearl, J. Causality. Models, Reasoning, and Inference. 2nd ed. 2009 (reprinted with corrections, 2021). Cambridge: Cambridge University Press; 2022.Search in Google Scholar

67. Rothman, KJ. The wrong message from the wrong talk. Obs Stud 2020;6:30–2. https://doi.org/10.1353/obs.2020.0006.Search in Google Scholar

68. Brumback, BA. Fundamentals of causal inference with R. Boca Raton, FL: CRC Press, Taylor & Francis Group, LLc; 2022.Search in Google Scholar

69. Lipsky, AM, Greenland, S. Causal directed acyclic graphs. JAMA 2022;327:1083–4. https://doi.org/10.1001/jama.2022.1816.Search in Google Scholar PubMed

70. Livingstone, EH, Lewis, RJ. JAMA guide to statistics and methods. New York: McGraw Hill; 2020.Search in Google Scholar

71. Barnard-Mayers, R, Childs, E, Corlin, L, Caniglia, EC, Fox, MP, Donnelly, JP, et al.. Assessing Knowledge, Attitudes, and practices towards causal directed acyclic graphs: a qualitative research Project. Eur J Epidemiol 2021;36:659–67. https://doi.org/10.1007/s10654-021-00771-3.Search in Google Scholar PubMed PubMed Central

72. Vandenbroucke, JP, Broadbent, A, Pearce, N. Causality and causal inference in epidemiology: the need for a pluralistic approach. Int J Epidemiol 2016;45:1776–86. https://doi.org/10.1093/ije/dyv341.Search in Google Scholar PubMed PubMed Central

Supplementary Material

This article contains supplementary material (https://doi.org/10.1515/em-2023-0015).

Received: 2023-05-15
Accepted: 2024-07-15
Published Online: 2024-08-02

© 2024 Walter de Gruyter GmbH, Berlin/Boston

Articles in the same Issue

  1. Perspective
  2. Integrating Hill’s classical considerations with modern causal inference methods in observational studies: a ‘How-Questions’ framework
  3. Research Articles
  4. A compound representation of the multiple treatment propensity score with applications to marginal structural modeling
  5. Non-plug-in estimators could outperform plug-in estimators: a cautionary note and a diagnosis
  6. Item-level heterogeneous treatment effects of selective serotonin reuptake inhibitors (SSRIs) on depression: implications for inference, generalizability, and identification
  7. Estimation of the number needed to treat, the number needed to be exposed, and the exposure impact number with instrumental variables
  8. Bounds for selection bias using outcome probabilities
https://doi.org/10.1515/em-2023-0015
Keywords for this article
causal inference; epidemiological methods; Hill’s criteria for causality; modern methods; observational studies

Articles in the same Issue

  1. Perspective
  2. Integrating Hill’s classical considerations with modern causal inference methods in observational studies: a ‘How-Questions’ framework
  3. Research Articles
  4. A compound representation of the multiple treatment propensity score with applications to marginal structural modeling
  5. Non-plug-in estimators could outperform plug-in estimators: a cautionary note and a diagnosis
  6. Item-level heterogeneous treatment effects of selective serotonin reuptake inhibitors (SSRIs) on depression: implications for inference, generalizability, and identification
  7. Estimation of the number needed to treat, the number needed to be exposed, and the exposure impact number with instrumental variables
  8. Bounds for selection bias using outcome probabilities
Sign up now to receive a 20% welcome discount
Institutional Access
How does access work?
Have an idea on how to improve our website?
Please write us.
© 2025 De Gruyter Brill
Downloaded on 21.9.2025 from https://www.degruyterbrill.com/document/doi/10.1515/em-2023-0015/html
Scroll to top button