This paper investigates the challenges of bodily communication during video-based clinical consultations. While previous works describe the lack of eye contact and gestures over video, it is unclear how these limitations impact the course of a clinical consultation, particularly in a domain like physiotherapy where the focus is on improving body movements and functioning. To contribute to this understanding, we conducted observations of 10 naturally occurring video and face-to-face consultations for physiotherapy. We found that clinicians rely on a variety of incidental bodily cues and fine-details of body movements to assess and examine the patient. These bodily cues were noticeable during face-to-face consultations; however, a variety of bodily cues got missed over video. Consequently, video consultations became conversational where the clinicians used verbal conduct to get a fair understanding of the patient's health. To guide design of future video consultation systems, we reflect on our understanding as 4 design sensitivities: Visual Acuity, Field-of-view, Clinical Asymmetries, and Time Sequence.