Drain Registry — Every Drain, Every Visit

Watch

The 48-second tour

Demo with fictional patients on a local copy of the app. The scores, trend summaries and validation figures are the app’s real output.

The Problem

Drains outlive the paperwork

Patients with drains were routinely lost to follow-up. When a patient walked in, or rang, there was no reliable way to say what the drain output had been doing for the previous three days — and every case that ended quietly was a data point that never made it into a study.

What I Built

One record per drain, from placement to removal

Each patient can have several drain episodes, and each episode has its own visits. At placement, seven factors — drain size, the patient’s condition, first-day output, indication, collection type, a complex procedure and any fistula — give a PRIDIT-S score (Predictive Risk Index for Drain Indwelling Time with drain Size) and a predicted indwelling window. Every follow-up records output, condition, labs, aspiration, flushing, imaging, cultures and the plan, and the episode timeline shows the trend at a glance. The dashboard lists every drain past its predicted window, and an episode can be closed out with its outcome in two clicks.

For Patients

Patients log their own daily output

Between visits, patients can now record their drain output themselves. For an ongoing drain, staff create a private link and send it on WhatsApp in one tap. The patient opens it on their phone with no account or login. The page is in English and Urdu side by side, so most patients can use it in whichever they read. Each day they enter the date, the volume in ml, whether the drain was flushed and an optional note. Saving the same date again corrects that day’s entry. The entries appear straight away in the episode’s Patient Daily Log. So when a patient rings or walks in, the last few days of output are already on screen.

Technical Highlights

Built so the record stays trustworthy

One score, computed on the server

The server calculates and stores the PRIDIT-S score when an episode is created; the form shows a matching preview while the factors are entered.

Baselines can’t drift

Placement factors are locked once saved. A clinician can record an override with a reason, but the calculated score is never overwritten.

Validation that computes itself

Every completed episode is compared live with its predicted window — within, earlier or longer — so the validation covers all historical data with no manual entry.

Removal closes the loop

Recording a visit where the drain came out resolves the episode and sets its end date automatically.

Repeat drains, kept apart

A patient → episode → visit model means a second drain in the same patient gets its own score, timeline and outcome.

A link, not a login

The patient link opens one drain’s daily log and nothing else: no name, record number or clinical data. It stops taking entries when the episode closes, and staff can replace it or switch it off at any time.

Ready for analysis

Cohort charts, a per-episode trend summary and a full-dataset CSV export turn the registry straight into a research dataset.

Stack

Under the hood

Frontend	React + Vite, installable PWA with automatic updates
Backend	Plain PHP API, one file per resource, JSON Web Token authentication
Database	MySQL — patients, drain episodes, visits, comorbidities and patient daily logs
Hosting	Shared hosting on Hostinger

Status

Where it stands

Live in its second version and in active use at Hearts International Hospital. The clinical work and the research have become the same activity: a year on, the dataset needed to validate PRIDIT-S is already in place.

Drain Registry — every drain, every visit