Design of Retrospective and Case-Control Studies in Oncology
Retrospective studies allow researchers to evaluate outcomes in a real-world setting at reduced costs compared with prospective trials, and have long-established use in surgical oncology. In retrospective studies, the study sample is generated from secondary or pre-existing data, which precludes randomization. As a result, the potential for unique and significant biases exists and these can limit the applicability and generalizability of the findings. This chapter is intended to serve as a guide for conducting retrospective research studies. Topics covered include internal and external validity; types of biases; sampling and matching techniques, including propensity score matching; missing data; and special considerations for common study designs. Examples from the surgical oncology literature are provided.
KeywordsRetrospective cancer research Case-control studies
- 1.Gay J. Clinical study design and methods terminology. 2010. http://people.vetmed.wsu.edu/jmgay/courses/glossclinstudy.htm. Accessed 1 May 2017.
- 13.The Cochrane Collaboration. Glossary. 2017. http://community-archive.cochrane.org/glossary/5#letterv. Accessed 1 May 2017.
- 15.Cook TD, Campbell DT. The design and conduct of quasi-experiments and true experiments in field settings. In: Dunnette MD, editor. Handbook of industrial and organizational psychology, vol. 223. Amsterdam: Elsevier; 1976. p. 336.Google Scholar
- 20.York RO. Conducting social work research: an experiential approach. London: Pearson College Division; 1998.Google Scholar
- 21.Aschengrau A, Seage GR. Essentials of epidemiology in public health. Burlington, MA: Jones & Bartlett Learning, LLC; 2013.Google Scholar
- 22.Weiss NS. Clinical epidemiology: the study of the outcome of illness. Oxford: Oxford University Press; 1996.Google Scholar
- 26.Schatman ME, Campbell A, Loeser JD. Chronic pain management: guidelines for multidisciplinary program development. Boca Raton, FL: CRC Press; 2007.Google Scholar
- 27.Cochran WG, Rubin DB. Controlling bias in observational studies: a review. Sankhyā. 1973;35:417–46.Google Scholar
- 34.Gu XS, Rosenbaum PR. Comparison of multivariate matching methods: structures, distances, and algorithms. J Comput Graph Stat. 1993;2(4):405–20.Google Scholar