This study presents the simulation results for 10.16 cm diameter and 7.62 cm thickness NaI(Tl) detector response, which is housed in a partially shielded scanning bed whole-body monitor (WBM), due to activity distributed in the axial cavities provided in the Indian reference BOMAB phantom. Experimental detection efficiency (DE) for axial cavity activity distribution (ACAD) in this phantom for photon emissions of ¹³³Ba, ¹³⁷Cs and ⁶⁰Co is used to validate DEs estimated using Monte Carlo code FLUKA. Simulations are also carried out to estimate DEs due to uniform activity distribution (UAD) as in the standard BOMAB phantom. The results show that the DE is ~3.8 % higher for UAD when compared with ACAD in the case of ⁴⁰K (1460 keV) and this relative difference increases to ~7.0 % for ¹³³Ba (~356 keV) photons. The corresponding correction factors for calibration with Indian phantom are provided. DEs are also simulated for activity distributed as a planar disc at the centre of the axial cavity in each part of the BOMAB phantom (PDAD) and the deviations of these DEs are within 1 % of the ACAD results. Thus, PDAD can also be used for ACAD in scanning geometry. An analytical solution for transmitted mono-energetic photons from a two-dimensional slab is provided for qualitative explanation of difference in DEs due to variation in activity distributions in the phantom. The effect on DEs due to different phantom part dimensions is also studied and lower DEs are observed for larger parts.