Opus (defined in RFC 6716) is the audio codec at the center of WebRTC — it is mandatory to implement, so any compliant endpoint can speak it, which makes it the default for real-time voice. Its standout property is range and resilience: it operates across a very wide bitrate span, roughly 6 to 510 kbps, smoothly adapting from a thin voice trickle to high-fidelity audio, and it ships with built-in tools for surviving bad networks, including forward error correction (FEC) that embeds recovery data in-band and packet-loss concealment that masks gaps the ear would otherwise hear.
For clinical calls this matters more than any video feature, because audio is the channel that actually carries the consultation. Opus stays intelligible on connections so degraded that video has already frozen or been switched off, which is exactly why an audio-only fallback is a realistic and clinically acceptable mode rather than a failure state — a clinician can still take a history and counsel a patient on voice alone.
The practical lever is configuration. Opus has features that are not always optimally enabled by default: in-band FEC trades a little bitrate for resilience to packet loss, and DTX (discontinuous transmission) saves bandwidth by sending little during silence. Tuning these for the lossy, variable networks patients actually use — home Wi-Fi, cellular, rural links — improves real-world consults more than most video adjustments. The common mistake is pouring effort into video quality while leaving audio on naive defaults, then being surprised that the part patients complain about most is hearing each other clearly.
