The broadening of lines by Stark effect is widely used for inferring electron density and temperature in plasmas. Stark-effect calculations often rely on atomic data (transition rates, energy levels,...) not always exhaustive and/or valid only for isolated atoms. In this work, we first present a recent development in the detailed opacity code SCO-RCG for K-shell spectroscopy. The approach is adapted from the work of Gilles and Peyrusse. Neglecting non-diagonal terms in dipolar and collision operators, the line profile is expressed as a sum of Voigt functions associated to the Stark components. The formalism relies on the use of parabolic coordinates and the relativistic fine structure of Lyman lines is included by diagonalizing the hamiltonian matrix associated to quantum states having the same principal quantum number n. The SCO-RCG code enables one to investigate plasma environment effects, the impact of the microfield distribution, the decoupling between electron and ion temperatures and the role of satellite lines (such as Li-like 1snℓn ′ ℓ ′ −1s 2 nℓ, Be-like, etc.). Atomic structure calculations have reached levels of accuracy which require evaluation of Breit interaction and many-electron quantum electro-dynamics (QED) contributions. Although much work was done for QED effects (self-energy and vacuum polarization) in hydrogenic atoms, the case of an arbitrary number of electrons is more complicated. Since exact analytic solutions do not exist, a number of heuristic methods have been used to approximate the screening of additional electrons in the self-energy part. We compare different ways of including such effects in atomic-structure codes (Slater-Condon, Multi-Configuration Dirac-Fock, etc.).