As is done by various international agencies to arrive at x%/Gy for various populations.
Radiation epidemiology uses several basic study designs, such as cohort studies, case control studies, nested case control studies, ecological studies, cross sectional studies. Risks are expressed in various forms, such as relative risk (RR), absolute risk (AR), Odds ratio (OR), and its excess (ERR, EAR and EOR). The dose response data points are fitted to various statistical models (e.g., pure liner, linear quadratic, pure quadratic, nonparametric, categorical, threshold), and its slope is used to derive risk per dose (e.g., ERR/Gy). Sometimes risks are also expressed as point estimates, e.g., ERR at 1 Gy.
If risk estimate is aimed for a certain type of radiation (e.g., gamma-rays), estimates may need to be adjusted for other types of radiation (neutrons) or medical exposures (e.g., X-rays). Other potential adjustments include sex, age at exposure, attained age, birth cohort, smoking, alcohol intake, socioeconomic status, obesity, diabetes, depending on the disease outcome of interest. These estimates are made not only from a single study, but also from a series of relevant studies (so-called meta-analysis).
There are various types of the methods for lifetime risk estimates (for calculating the risk, over a lifetime, that an individual will develop, or die from, a specific disease caused by an exposure), such as REIC or REID (risk of exposure induced cancer or death), LAR (lifetime attributable risk, approximation of REIC or REID), ELR (excess lifetime risk), LLE (loss of life expectancy), and DALY or DALE (disability adjusted life years or expectancy). For effective dose for a representative population, the International Commission on Radiological Protection uses nominal risk coefficient, i.e., lifetime risk estimate averaged over sexes and over ages at exposure.
As more individualized approach, the use of LAR has recently been proposed for some medical exposures of patients. The information on the relevant paper that came out two months ago and the presentation made two months ago by the same group is provided below.
https://www.ncbi.nlm.nih.gov/pubmed/29064376
http://www.icrp.org/docs/2017fuk/3%20Mattsson%20Presentation.pdf
You may be interested in the following new paper
Article History of radiation detriment and its calculation methodolo...