Recently, virtual reality (VR) technologies started gaining momentum in surgical simulation-based training by allowing clinicians to practice their skills before performing real procedures. The design of such simulators is usually focused on the primary operative tasks to be taught, but little attention is paid to secondary tasks that the user needs to perform, such as changing his/her point of view when manipulating the surgical instruments. More particularly, it is not clear how to design appropriate interaction techniques for those tasks, and how the fidelity of these interactions can impact the user's performance on such systems. In this paper, we compare two viewpoint changing techniques having two different levels of interaction fidelity during needle insertion in a semi-immersive VR (SIVR) biopsy trainer. These techniques were designed based on observing clinicians performing actual biopsy procedures. The first technique is based on tracking the user's head position (high interaction fidelity), while the second technique is touch-based with the user utilizing his/her non-dominant hand fingers to manipulate the point of view on a touch screen (moderate interaction fidelity). A user study was carried out to investigate the impact of the interaction fidelity of the viewpoint changing task (secondary task) on the user's performance during the needle insertion task (main task). Twenty-one