Integrations

Most patients are already wearing the sensor

A watch on the wrist, a phone in the pocket, or a monitor on the shelf. CareSaathi reads from the phone's own health data, from connected devices and wearables, and — when a device connects to nothing at all — from a photo of its display. Nothing here asks your patients to buy new hardware.

The phone in their pocket

Apple Health and Android Health Connect

Both platforms are where a patient's existing apps and wearables already deposit their vitals. Reading from them means a programme can start with zero procurement.

Android Health Connect

Supported

Android's system health store. Any app the patient already uses that writes vitals into it becomes a source for their care team — no new device, no pairing, no account to link.

Vitals we read
Blood pressureBlood glucoseWeightOxygen saturationTemperatureHeart rate

Typical sources: Samsung Health, smartwatch and band companion apps, many glucometer and BP apps

Readings sync in the background and carry the name of the app that wrote them, so a clinician always knows the origin.

Apple Health & Apple Watch

Supported

Apple Watch measures heart rate through the day and oxygen saturation in the background. That data lands in Apple Health, and CareSaathi reads it — verified on real hardware, not assumed.

Vitals we read
Heart rateOxygen saturationBlood pressureBlood glucoseWeightTemperature

Typical sources: Apple Watch, Omron Connect on iOS, and other apps that publish to Apple Health

Readings carry the device or app that recorded them, so a clinician can tell an Apple Watch measurement from a cuff. Patients who prefer no app at all can still send readings over WhatsApp.

Browse by measurement

Pick a category to see the devices

Clinical measurements first — the ones a care programme is actually built on.

How to read this list:Live nowworking in production todayOn requestenabled for your organisation on requestIn developmentcommitted, not yet shippedEvery model listed here works today by photo capture or manual entry, whatever its connection status.

Clinical measurements

Wellness & activity — context between clinical readings

Models shown are representative of each category rather than the complete integration list — support spans 700+ devices, apps and platforms, manufacturers are added over time, and a few require their own approval before use. We confirm the exact list for your programme during onboarding.

Four ways a reading arrives

Source-neutral by design

The same pipeline accepts all four, normalizes them, and checks them against the same rules. A programme never fails because a patient owns the wrong brand of cuff.

📶

Connected & automatic

Best for adherence

Bluetooth, cellular and cloud-synced devices push readings on their own. The patient measures; nothing else is asked of them. No vendor credentials ever live on the patient’s phone.

📷

Photographed

Works with any device

The patient photographs their monitor’s display — in the app or on WhatsApp — and AI reads the numbers into the chart as a preliminary reading, with the image kept for a clinician to confirm.

⌨️

Entered by hand

Always available

A metric-specific form with the unit fixed by policy, so a value can’t arrive in the wrong scale. Always available, and it works offline.

🏥

Measured in clinic

One timeline

Readings taken at a visit are recorded by staff in the portal and sit on the same timeline, so a patient’s trend doesn’t break every time they come in.

What that buys you

No hardware vendor holds the programme hostage

🧭

Vendor-neutral

The programme is defined by metrics and rules, not by a hardware contract. Devices can be swapped without rewriting a single care plan.

New devices without an app release

Integrations are added centrally. A device that becomes available appears for your patients with no update to install.

📐

Normalized to canonical units

mg/dL or mmol/L, kg or lb — everything is converted to one canonical unit before a rule ever sees it, with the original value kept.

🔍

Full provenance on every reading

What measured it, how it arrived, when it was observed and when it reached us — stored separately, so a late upload never rewrites history.

Common questions

Before you check your inventory

What if the patient’s device isn’t on the list?

It almost certainly still works. Any device with a display can be captured by photo or manual entry, and both are first-class paths — a reading that arrives by photo is stored, charted and rule-checked exactly like one from a Bluetooth cuff. Tell us the model and we’ll confirm whether an automatic integration exists.

Does the patient need a smartphone?

No. Cellular-enabled devices send readings over their own connection with no phone involved, which is often the right answer for elderly patients. Otherwise a phone, a shared bedside device, or WhatsApp all work.

Do you sell the devices?

We don’t. Hospitals buy from whichever distributor they already use, or patients use what they own — which is usually cheaper and always faster than a procurement cycle.

How accurate is reading a value from a photo?

Good enough to be useful, not good enough to be trusted blindly — which is why it isn’t. Photo-derived readings are stored as preliminary and marked suspect until a clinician confirms them against the image, and implausible values are flagged rather than charted.

Can a patient use more than one device?

Yes, and a reading that arrives twice — once from the device and once from a wearable — is de-duplicated so it counts once.

Tell us what your patients use

Send us the models already in your wards and homes, and we'll confirm exactly how each one would connect.