How Virtual Visits Capture Patient Vital Signs Without Wearables

Virtual visits have reshaped how health systems deliver care. Since 2020, televisit volumes have stabilized at levels that would have been unthinkable a decade ago, with most large health systems now running thousands of video appointments per week across primary care, behavioral health, chronic disease management, and specialty follow-ups.

But there is a persistent gap in these encounters that every clinician knows well: the absence of objective vital sign data. A provider conducting a virtual visit is essentially making clinical decisions without the physiological baseline that has been standard in every in-person encounter for over a century. Heart rate, oxygen saturation, heart rate variability, and stress indicators are simply missing from the clinical record.

That gap is now closing. Remote photoplethysmography, or rPPG, makes it possible to capture clinical-grade vital signs during a video visit using nothing more than the camera already built into a patient's phone or laptop. No wearables. No shipped devices. No additional hardware of any kind.

The Problem: Virtual Visits Without Vital Signs

When a patient walks into a clinic, the first thing that happens is a vitals check. Blood pressure, heart rate, oxygen saturation, temperature, and respiratory rate are captured before the provider enters the room. These measurements form the foundation of clinical decision-making, informing differential diagnoses, medication adjustments, and disposition decisions.

In a virtual visit, that foundation is absent. Providers rely on patient self-report ("I feel fine," "My heart has been racing"), visual assessment through the camera, and history review. While experienced clinicians can gather significant information through a skilled video interview, the lack of objective physiological data introduces uncertainty into every encounter.

This gap has real consequences. Providers report lower confidence in virtual assessments compared to in-person visits. Some clinicians reflexively schedule in-person follow-ups after televisits simply because they lack the vital sign data needed to make a definitive assessment. Patients with undetected tachycardia, hypoxemia, or elevated stress responses may be missed entirely.

Traditional remote patient monitoring (RPM) programs attempt to address this by shipping devices to patients -- pulse oximeters, blood pressure cuffs, connected scales. These programs serve an important role for high-acuity chronic disease management, but they are not a practical solution for the general virtual visit population. Device logistics are expensive, patient compliance is inconsistent, and the operational overhead of managing thousands of shipped devices is substantial.

How rPPG Works: The Science Behind Camera-Based Vitals

Remote photoplethysmography is built on a straightforward physiological principle. With each heartbeat, blood pulses through the capillary beds beneath the skin, producing subtle changes in skin color that are invisible to the human eye but detectable by a standard digital camera.

These micro-changes in reflected light correspond directly to cardiovascular activity. By analyzing the pixel-level color variations in a video stream of a patient's face, rPPG algorithms can extract the same blood volume pulse signal that a traditional contact-based pulse oximeter captures through the fingertip.

The technology pipeline works in several stages. First, the camera captures a continuous video stream of the patient's face during the virtual visit. Computer vision algorithms identify and track the regions of interest, primarily the forehead, cheeks, and nose where capillary density is high and skin is relatively uniform. Signal processing techniques then isolate the blood volume pulse from environmental noise, compensating for ambient lighting changes, patient movement, and camera artifacts.

From this extracted pulse signal, multiple vital signs can be derived simultaneously:

Heart Rate (HR): The fundamental pulse frequency, measured in beats per minute. rPPG-derived heart rate has been validated against ECG reference standards in peer-reviewed clinical studies, with accuracy comparable to contact-based pulse oximetry in controlled conditions.

Heart Rate Variability (HRV): The beat-to-beat variation in heart rate intervals, which reflects autonomic nervous system function. HRV is an increasingly important biomarker in cardiology, stress assessment, and chronic disease management. rPPG can extract time-domain HRV metrics from the inter-beat interval series.

Blood Oxygen Saturation (SpO2): By analyzing the differential absorption of light at multiple wavelengths, rPPG algorithms can estimate peripheral oxygen saturation. This is particularly relevant for respiratory conditions, post-COVID monitoring, and pulmonary disease management.

Stress Index: Derived from HRV analysis and sympathetic/parasympathetic balance indicators, the stress index provides a quantitative assessment of physiological stress that complements subjective patient reports.

The Patient Experience: Seamless and Invisible

From the patient's perspective, the vital sign capture during a virtual visit with rPPG should be essentially invisible. This is a critical design principle, and it is what differentiates the approach from alternatives that require patients to perform specific actions.

The patient joins their scheduled video appointment through whatever telehealth platform their health system uses. During the natural course of the visit, while the patient is facing the camera and speaking with their provider, the rPPG analysis runs in the background. The patient does not need to hold still, look at a specific point, or perform any measurement ritual. They simply have their appointment.

Within the first 30 to 60 seconds of the visit, while the medical assistant or nurse is conducting intake, vital signs are captured and made available. The provider sees the results in their clinical workflow before or during the encounter, just as they would see vitals captured by a medical assistant in an exam room.

Circadify's approach processes all data on the patient's device itself. The facial video analysis happens locally -- raw video frames are never transmitted to external servers. Only the derived vital sign values (numerical readings) are communicated to the clinical system. This on-device processing model addresses privacy concerns at the architectural level, rather than relying solely on policy controls.

For patients, there is no app to download (assuming the capability is integrated into the existing telehealth platform), no device to charge or calibrate, no instructions to follow, and no compliance burden. The technology meets patients exactly where they already are: sitting in front of a camera for a video appointment.

Clinical Validation and Accuracy

Health system leaders evaluating camera-based vitals rightly focus on clinical accuracy. The question is not whether the technology is interesting, but whether the measurements are reliable enough to inform clinical decisions.

rPPG heart rate measurement has the deepest evidence base. Multiple peer-reviewed studies have demonstrated strong correlation with ECG-derived heart rate across diverse populations, skin tones, and lighting conditions. The technology has matured significantly from early academic prototypes, with modern algorithms incorporating deep learning approaches that improve robustness in real-world conditions.

SpO2 estimation via rPPG is an area of active advancement. While the accuracy profile differs from contact-based pulse oximetry, rPPG-derived SpO2 provides clinically useful screening-level data that is substantially better than the alternative in a virtual visit, which is no oxygen saturation data at all.

HRV measurements derived from rPPG have been validated against chest-strap reference devices, with time-domain metrics showing strong agreement in controlled settings. The clinical utility of HRV in virtual care contexts is an evolving area, with growing evidence supporting its role in stress assessment, cardiac risk stratification, and autonomic dysfunction screening.

It is important to frame accuracy appropriately. The comparison point is not a clinical-grade bedside monitor in an ICU. The comparison point is the current state of virtual visits, where zero vital signs are captured. Even screening-level accuracy represents a transformative improvement in the clinical data available during a televisit.

Comparison to Traditional Remote Patient Monitoring

Health systems often ask how camera-based vitals relate to their existing RPM programs. The two approaches serve different but complementary roles.

Traditional RPM programs typically involve shipping connected devices (pulse oximeters, blood pressure cuffs, glucometers, scales) to patients enrolled in specific chronic disease management protocols. These programs generate longitudinal data between visits, support CMS RPM billing codes, and are essential for high-acuity population management.

However, traditional RPM has well-documented limitations. Device costs range from $50 to $300 per patient for initial provisioning. Shipping and logistics add operational overhead. Patient compliance rates for daily device use typically decline from 70-80% in the first month to 30-40% by month six. Device management (replacements, troubleshooting, returns) requires dedicated staff. And RPM programs are inherently limited to enrolled populations, typically a small percentage of a health system's total patient panel.

Camera-based vitals during virtual visits operate differently. There are no devices to ship, no compliance curves to manage, and no per-patient hardware costs. The technology applies to every patient who joins a video visit, not just those enrolled in a specific monitoring program. It captures vital signs at the point of the clinical encounter, providing data exactly when the provider needs it for decision-making.

The tradeoff is that camera-based vitals provide episodic measurements during visits, not continuous or daily monitoring between visits. For patients who need daily weight tracking for heart failure management or continuous glucose monitoring for diabetes, traditional RPM devices remain essential.

The practical result is that camera-based vitals and traditional RPM serve different segments of the patient population with minimal overlap. RPM serves the high-acuity enrolled population with longitudinal between-visit monitoring. Camera-based vitals serve the entire virtual visit population with point-of-care data during encounters.

Operational and Financial Considerations

For health system decision-makers, the operational model of camera-based vitals is notably different from device-based approaches.

There is no device inventory to manage. No shipping logistics. No patient-facing technical support for hardware issues. No device return processes. No asset tracking. The infrastructure requirement is software-based, integrating with existing telehealth platforms and clinical systems.

From a financial perspective, the cost model shifts from per-patient hardware costs to a software licensing model. This eliminates the variable cost per enrolled patient that characterizes RPM programs and makes it economically feasible to capture vitals for every virtual visit, not just for patients whose acuity justifies the cost of shipped devices.

Provider workflow impact is minimal when properly implemented. Vital signs appear in the same clinical workflow locations where providers already review vitals for in-person visits. No additional clicks, no separate application to open, no new workflow to learn. The data is simply present where it was previously absent.

Looking Forward

The convergence of computer vision, signal processing, and clinical workflow integration is making it practical to close the vital signs gap in virtual care. Health systems that have built substantial virtual care programs now have the opportunity to add objective physiological data to every video encounter without the operational burden of device-based approaches.

For health system CIOs, virtual care directors, and clinical informatics teams evaluating this capability, the key questions are straightforward: How does it integrate with your existing telehealth platform? What is the accuracy profile for your clinical use cases? How do the vital sign values flow into your EHR documentation? And what is the privacy architecture for facial video analysis?

The technology is not a replacement for comprehensive remote patient monitoring programs, in-person clinical assessment, or medical-grade diagnostic devices. It is a practical way to ensure that every virtual visit includes the baseline physiological data that providers need to deliver confident, well-documented clinical care.