Bilimsel Araştırmalar Ne Kadar Bilimsel: Karıştırıcılar, Çarpıştırıcılar Ve Etki-Ölçüm Değiştiriciler

Abstract

Bilimsel araştırmaların başlıca amacı gerçeği bulmak olduğu halde pek çok araştırma sonuçlarının bunu başaramadığı, bazen dikkatsizlik ve özensizlik bazen acelecilik çoğunlukla da yöntem bilgisi yetersizliği yüzünden gerçeğin çok uzağında kaldığı görülmektedir. Bilim ve teknolojideki tüm gelişmelere rağmen halen tıpta nedeni ve tedavisi bilinmeyen onlarca hastalık bulunmakta, özellikle nedensellik konusundaki bilgilerimiz yerinde saymaktadır. Yayınlanan araştırmaların planlanmasından tasarımına, kullanılan yöntemlerden veri toplama biçimine, toplanan verilerin analizinden yorumuna kadar her aşamada yapılabilen sayısız hata ve yanlılıklar söz konusudur. Özellikle sistematik hata olarak da bilinen yanlılıkların araştırma sonuçlarını farkında olmadan bambaşka bir yöne çekebildiği bilinmektedir. Bu yanlılıkların en önemlileri arasında yer alana “karıştırıcılık”, “etkileşim” ve “çarpıştırıcılık” neden-sonuç ilişkilerinin incelenmesi sırasında çok karşılaşılan, fark edildiğinde kontrolü mümkün olan, fark edilmediğinde ise bulguları çarpıtarak değersizleştiren yanlılık kaynaklarıdır. Bunların ne oldukları, nasıl oluştukları ve nasıl kontrol edilebilecekleri konuları ilgili literatür ışığında özetlenmiştir.

Keywords

• Karıştırıcılık
• Çarpıştırıcılık
• Etkileşim
• Etki-ölçüm değişimi

How Much Scientific Are The Scientific Studies: Confounders, Colliders, And Effect-Measure Modifiers

Abstract

Although the main purpose of scientific studies is to find the truth, it is seen that the results of many studies cannot achieve this, and sometimes their findings are far from the truth due to carelessness, lack of rigorousness, hastiness and mostly to lack of methodology knowledge. Despite all the developments in science and technology, there are still dozens of diseases whose causes and treatments are unknown in medicine. There are numerous mistakes and biases that can be made at every stage, starting from the study plan and design, to the methods used and to the way of data collection from the analysis to the interpretation of the findings. It is well known that biases, also known as systematic errors, can lead research results in a completely different direction without being aware of it. Among the most important of these biases, "confounding", "interaction" and "colliding" are sources of bias that are frequently encountered during the investigation of cause-effect relationships. They can be controlled when noticed, and distort and devalue the findings when not noticed. This article summarizes the issues regarding what are they, how do they occur and how they can be controlled in the light of the relevant literature.

Keywords

• Confounders
• Colliders
• Interaction
• Effect-measure modifiers

References

1. Parfrey PS, Barrett BJ (eds.). Clinical Epidemiology Practice and Methods. Third Edition, Humana Press, 2021. https://doi.org/10.1007/978-1-0716-1138-8.
2. Porta M. Dictionary of epidemiology. 6th edition, IEA, Oxford University Press, 2014.
3. Michael M III, Boyce WT, Wilcox AJ. Biomedical Bestiary: An epidemiological guide to flaws and fallacies in the medical literature. Boston, Little, Brown, 1984:16.
4. Frerichs, R. R., Beeman, B. L., and Coulson, A. H. Los Angeles airport noise and mortality-Faulty analysis and public policy. Am. J. Public Health 1980;70:357.
5. Wayne W. LaMorte, Lisa Sullivan, Boston University School of Publich Health)
6. Pearce N, Greenland S. Confounding and Interaction. In: Ahrens W., Pigeot I. (eds) Handbook of Epidemiology. Springer, New York, NY, 2014. https://doi.org/10.1007/978-0-387-09834-0_10)
7. Kjeldsen SE, Os I. Assessing hypertension therapies: randomization or confounding by indication? Nat Rev Cardiol. 2020 Feb;17(2):73-74. doi: 10.1038/s41569-019-0313-z.
8. Kyriacou DN, Lewis RJ. Confounding by Indication in Clinical Research. JAMA. 2016;316(17):1818-1819. doi: 10.1001/jama.2016.16435.

9. Feenstra H, Grobbee RE, in't Veld BA, Stricker BH. Confounding by contraindication in a nationwide cohort study of risk for death in patients taking ibopamine. Ann Intern Med. 2001;134(7):569-72. doi: 10.7326/0003-4819-134-7-200104030-00010.
10. Griffith GJ, Morris TT, Tudball MJ vd. Collider bias undermines our understanding of COVID-19 disease risk and severity. Nat Commun 2020;11, 5749. https://doi.org/10.1038/s41467-020-19478-2.
11. Berkson J. Limitations of the application of four-fold table analyses to hospital data. Biometrics Bull 1946;2:47-53.
12. Simons D, Shahab L, Brown J, Perski O. The association of smoking status with SARS‐CoV‐2 infection, hospitalisation and mortality from COVID‐19: A living rapid evidence review with Bayesian meta‐analyses (version 7). Addiction [Internet]. 2020 Oct 2 [cited 2020 Oct 29];add.15276. https://onlinelibrary.wiley.com/doi/10.1111/add.15276.
13. Tattan-Birch H, Marsden J, West R, Gage SH. Assessing and addressing collider bias in addiction research: the curious case of smoking and COVID-19. Addiction. 2021;116(5):982-984. doi: 10.1111/add.15348.
14. Cole SR, Platt RW, Schisterman EF, Chu H, Westreich D, Richardson D, Poole C. Illustrating bias due to conditioning on a collider, International Journal of Epidemiology, 2010;39(2):417–420, https://doi.org/10.1093/ije/dyp334.
15. Catalogue of bias collaboration, Lee H, Aronson JK, Nunan D. Collider bias. In Catalogue Of Bias. 2019. https://catalogofbias.org/biases/collider-bias/)
16. Banack HR, Kaufman JS. The obesity paradox: understanding the effect of obesity on mortality among individuals with cardiovascular disease. Prev Med. 2014;62:96-102. doi: 10.1016/j.ypmed.2014.02.003.
17. Selikoff IJ, Seidman H, Hammond EC. Mortality effects of cigarette smoking among amosite asbestos factory workers. J Natl Cancer Inst 1980;65:507–513.
18. Steenland K, Thun M. Interaction between tobacco smoking and occupational exposures in the causation of lung cancer. J Occup Med 1986;28:110–118.
19. Correia KF, Dodge LE, Farland LV, Hacker MR, Ginsburg E, Whitcomb BW, Wise LA, Missmer SA. Confounding and effect measure modification in reproductive medicine research. Hum Reprod. 2020;35(5):1013-1018. doi: 10.1093/humrep/deaa051.
20. Marquis GS, Habicht JP, Lanata CF, Black RE, Rasmussen KM. Association of breastfeeding and stunting in Peruvian toddlers: an example of reverse causality. Int J Epidemiol. 1997;26(2):349-56. doi: 10.1093/ije/26.2.349.
21. Gaines LS, Slaughter JC, Schwartz DA, Beaulieu DB, Horst SN, Dalal RL, Scoville EA, Sandler RS, Kappelman MD. Does Reverse Causality Underlie the Temporal Relationship Between Depression and Crohn's Disease? Inflamm Bowel Dis. 2020;26(3):423-428. doi: 10.1093/ibd/izz123.
22. Kim TJ, von dem Knesebeck O. Income and obesity: what is the direction of the relationship? A systematic review and meta-analysis. BMJ Open. 2018;8(1):e019862. doi: 10.1136/bmjopen-2017-019862.