ABSTRACT
The Scottish physician James Lind is widely credited with conducting the first controlled clinical trial. In 1747, in his capacity as ship's surgeon on HMS Salisbury, Lind selected 12 crew members afflicted by scurvy, and allocated them in pairs to six treatments. Self-control offers a different approach to resolving this conundrum. Self-control is the use of a study subject as his or her own control, sampled at different times: individuals are matched with themselves, and constitute individual-level strata. Self-controlled methods in epidemiology include the case-crossover method and the self-controlled case series method. Both involve only cases, that is, individuals who have experienced the event of interest. Self-controlled methods also suffer from limitations. Self-controlled methods can only aspire to estimate the second, within-individual effect, which in self-controlled case series (SCCS) analyses refers to as the relative incidence. The chapter also presents an overview of the key concepts discussed in the subsequent chapters of this book.
