Skip to main content

Table 2 Criteria for quality assessment of genetic association studies

From: Influence of genetic variants on toxicity to anti-tubercular agents: a systematic review and meta-analysis (protocol)

Issue Assessment criteria
1. Choosing the genes/single-nucleotide polymorphisms (SNPs) to genotype Was a literature review undertaken and the findings summarised?
Are reasons given for choosing the genes and SNPs genotyped?
If reasons include previous association studies, are key details from these provided?
If reasons include functional studies are supporting data provided?
Is method to adjust for multiple testing described?
Are precise p values provided for all associations?
2. Sample size What is the sample size?
Are details given of how the sample size was calculated?
Are details given of the a priori power to detect effect sizes of varying degrees?
3. Study design What is the study design?
If study is case–control, are the two groups clearly defined?
If study is case-control, were they genotyped in mixed batches?
4. Reliability of genotypes Is the genotyping procedure described?
Are the primers described?
Were quality control methods used and described?
Were findings from quality control methods, if used, described?
Are any genotype frequencies previously reported quoted?
Were genotyping personnel blinded to outcome status?
If human inference required, was this independently undertaken by at least two people?
5. Missing genotype data Is extent of missing data summarised?
Where extent is summarised are reasons for missing data given?
Are checks for missingness at random reported?
Are missing genotype data imputed?
Does paper quote number of patients contributing to each analysis?
If paper does quote number of patients contributing to analyses, does this agree to sample size?
6. Population stratification Are tests undertaken for cryptic population stratification?
If so, are results quoted?
Is cryptic population stratification adjusted for in the analyses?
7. Hardy–Weinberg equilibrium (HWE) What test is undertaken to check for HWE?
Where test undertaken, is p value threshold applied to determine deviation from HWE quoted?
Where test undertaken, are SNPs deviating from HWE highlighted?
Where test undertaken, and some SNPs found to deviate, are steps taken to explore deviation from HWE reported?
Where test undertaken, and some SNPs found
to deviate, are deviating SNPs excluded from further analysis?
8. Mode of inheritance Is a specific mode of inheritance assumed? If so which?
Is justification provided for assumptions made regarding mode of inheritance (if no mode or a specific mode is assumed)?
If no mode of inheritance is assumed does the paper explain limitations of this?
If several analyses undertaken under different assumptions, are they adjusted for multiple testing?
9. Choice and definition of outcomes Does the paper clearly define all outcomes investigated?
Is justification provided for the choice of outcomes?
Are results shown for all outcomes mentioned?
10. Adherence to treatment Is adherence to treatment measured?
If adherence is measured, are adjustments for non-adherence made in the analyses?
  1. HWE Hardy–Weinberg equilibrium, SNP single-nucleotide polymorphism