Book chapter
Crossover and Self-Controlled Designs in Clinical Research
Medical Uses of Statistics, pp 83-103
1992
Abstract
In crossover studies, each patient receives two or more treatments in sequence, and in self-controlled studies each patient serves as his or her own control. These designs for clinical trials can produce results that are statistically and clinically valid with far fewer patients than would otherwise be required.
Before choosing a crossover design, an investigator should be prepared to demonstrate the absence of carry-over effects, guarantee low drop-out rates, and expect a stable underlying disease. Since a parallel study with covariate adjustment may still be a strong competitor, we recommend careful consideration of the trade-off between power and fragility.
We investigated the use of the crossover design in the 13 crossover studies that appeared in the New England Journal of Medicine during 1978 and 1979. We considered the following important features of design and analysis as they applied to these studies: the method for assigning initial treatment (only 7 of 13 studies used random assignment); the determination of when to switch treatments (10 of the 13 used a time-dependent rule, and 1 a less appropriate disease-state-dependent rule; in 2, this point could not be determined); blinding of the crossover point (in only 3 of the 13 studies was the crossover point concealed, but in 4 of the remaining 10 concealment was impossible); assessment of the effects of the order of treatments (included in only 1 of the 13 studies); and the use of at least minimally acceptable statistical analysis (11 of the 13 studies had such an analysis).
We also report briefly on 28 additional studies of a single treatment each. In these studies each patient served as his or her own control before or after treatment or both. The scientific issues were much the same as in crossover studies except that self-controlled comparisons of treatments tended to be less precisely designed and conducted and to focus on clinical problems and patient groups that are especially difficult to study.
The high degree of patient-to-patient variability and the small number of available patients necessitated the use of the repeated crossover design to provide credible information on treatment efficacy. Crossover designs are most appropriate in the study of treatment for a stable disease. Investigators commonly employ either of two types of crossover rules — one that calls for a switch in treatments after a specified length of time and one in which a crossover occurs when indicated by the clinical characteristics of the patient. Knowledge of a switch can influence treatment response or assessment or both, so that blinding the crossover point can reduce the influence of any subjective order effects. If the most favorable underlying conditions for a crossover study are met, pairing each patient with himself or herself and analyzing the data properly can markedly reduce the influence of patient characteristics on the treatment comparison.
Metrics
15 Record Views
Details
- Title
- Crossover and Self-Controlled Designs in Clinical Research
- Creators
- Thomas A. LouisW. Lavori PhilipBailar John C. (Author)Marcia Polansky
- Contributors
- John C. Bailar (Editor)Frederick Mosteller (Editor)
- Publication Details
- Medical Uses of Statistics, pp 83-103
- Publisher
- CRC Press
- Edition
- 2
- Number of pages
- 21
- Resource Type
- Book chapter
- Language
- English
- Academic Unit
- Epidemiology and Biostatistics
- Other Identifier
- 991022040669104721