Video-Observed TB Treatment App with Offline Upload for Patients and Treatment Supporters

The problem: six months of daily pills

Tuberculosis is curable, but treatment means taking medicine every day for months. Missing doses lets the disease come back and can lead to drug resistance, which is far harder to treat. Pakistan has one of the larger TB burdens in the world, and its programme, like many others, has relied on directly observed treatment: a health worker or treatment supporter watches the patient take each dose.

In practice, daily observation is hard. Patients work long hours, live far from clinics or travel. Supporters cannot visit everyone every day. Doses go unobserved, and nobody knows which ones were actually taken. Supporters spend much of their time travelling rather than supporting, and patients who need help with side effects or motivation may not be noticed until they stop treatment altogether.

What already exists and the gap

• Directly observed treatment is the standard approach in TB programmes, with in-person observation.
• Commercial video-observed therapy platforms are used by health departments in several countries, with apps for patients and review dashboards.
• Medication reminder apps remind patients but cannot show that a dose was taken.

The gap is an open video-observed treatment app designed for the conditions many patients in Pakistan face: low-cost Android phones with little storage, patchy mobile data, shared phones and privacy concerns, with a review workflow simple enough for busy treatment supporters.

Inspiration

"Smartphone-enabled video-observed versus directly observed treatment for tuberculosis: a multicentre, analyst-blinded, randomised, controlled superiority trial" by Alistair Story and colleagues, published in The Lancet in 2019, compared smartphone video observation with in-person observation and found video observation improved observed adherence. Your project takes the same model of care and engineers it for low-cost phones and poor connections.

What you build

• Patients get a reminder, record a short video of taking their medicine and see that it has been sent, or that it will send when signal returns.
• Treatment supporters review the day's videos in minutes, mark each dose and contact patients who missed one.
• Clinicians see adherence over time for each patient and are alerted to repeated missed doses.

The patient side must be very simple: one big button, clear progress, and nothing lost if the phone dies mid-upload. Many patients will use a shared or borrowed phone, so the app needs a PIN and must not leave videos in the phone's gallery.

The technically interesting part

• Reliable offline upload. A dose video recorded in a village with no signal must reach the server days later, exactly once, even if the phone restarts. WorkManager, a durable queue and idempotent uploads are the core engineering.
• Small files. Compress video on the phone to a size that cheap data packages can carry, while keeping it clear enough to see the pill being taken. Measure the trade-off.
• Privacy. Health videos are sensitive. Encrypt at rest on the phone and on the server, keep them out of the gallery, and delete them after review.
• Review speed. Supporters may review dozens of videos a day. Design the review screen for speed: autoplay, keyboard or one-tap decisions, and a queue sorted by risk.

How to evaluate it

• Upload reliability: in tests with simulated and real poor connections and phone restarts, share of doses that arrived exactly once.
• File size and clarity: average size per video and reviewers' ability to confirm the dose.
• Review time: seconds per video for a supporter.
• Usability: a short trial with volunteers on low-cost phones, recording problems and missed reminders.

Suggested plan

• Weeks 1 to 3: read the Lancet trial, talk to someone in a TB programme about the workflow, design screens.
• Weeks 4 to 7: recording, compression, encrypted storage and the upload queue.
• Weeks 8 to 10: supporter review screen, messaging and summaries.
• Weeks 11 to 13: reliability tests and volunteer trial.
• Weeks 14 to 15: report and demo.

Path to a product

TB programmes, NGOs and private providers running treatment support need tools that fit their patients' phones and connections, and the same app works for other long treatments such as hepatitis C. An open app with paid hosting and support for programmes is realistic. The project shows serious Android engineering for real users, a strong portfolio piece.

Scope warning

Test only with volunteers acting as patients unless you have a clinical partner and ethics approval; never collect real patient videos in a student project without both. Do not build automatic dose verification in the main scope; a human reviews every video. Keep the app to Android.

Learn what you need on Sefism

• Mobile development roadmap: background work, storage and offline design.
• Mobile development cheat sheet: quick reference while you build.
• Android Developers: Develop for Android: CameraX, WorkManager and encryption.
• Guide to App Architecture (Android): structuring the offline-first data layer.
• OWASP Mobile Application Security (MASVS and MASTG): protecting sensitive health data on the phone.
• Firebase Documentation: if you use Firebase for storage and authentication.

Get help with this idea

The risk is an upload queue that loses doses on real networks. Tauseef can help you design the offline flow, the reliability tests and the conversation with a TB programme. If you are new here, create a free account first.

• Discuss this idea with a mentor
• Send a query to Tauseef
• Book a 1:1 session