Methodological quality of pharmacogenetic studies: issues of concern

Stat Med. 2008 Dec 30;27(30):6547-69. doi: 10.1002/sim.3420.

Abstract

Pharmacogenetic studies investigate association between genetic variants(double dagger) and drug efficacy and toxicity. Regardless of the extent of current knowledge regarding the human genome(double dagger) and relatively low genotyping costs, the value of pharmacogenetic studies in improving understanding of variable drug response remains contentious. Failure to replicate initial significant findings in subsequent studies is one reason for this and problems with methodological quality has been cited as contributing to this lack of replication. Here, concerns raised in the literature regarding the methodological quality of genetic association studies in general are discussed with particular consideration given to how applicable they are to the conduct and reporting of pharmacogenetic studies. Some of these issues such as sample size and quality of genotypes(double dagger) have been given significant attention in the literature already whereas others, such as dealing with missing genotypes and assumptions regarding mode of inheritance, have been discussed very little. Issues of quality believed to be uniquely relevant to pharmacogenetic studies are also considered. These discussions close with an assessment of how well the issues are dealt with in pharmacogenetic studies published over recent years and, overall, methodological quality is found to be poor. A checklist is provided of issues to be addressed whenever a future pharmacogenetic study is embarked upon. Referring to this checklist will improve the chances of replication and instil confidence in the integrity of future pharmacogenetic studies. It will also serve as a tool in assessing methodological quality of studies included in a systematic review of pharmacogenetic studies.

MeSH terms

  • Empirical Research
  • Genetic Variation
  • Genome, Human
  • Genotype
  • Humans
  • Models, Statistical
  • Pharmacogenetics*
  • Polymorphism, Single Nucleotide
  • Reproducibility of Results
  • Research Design / standards*
  • Sample Size