A community health team preparing for a field visit

Offline is not an edge case

Software written in a connected office treats a lost signal as an exception: a spinner, a retry, an error state. In a mobile screening camp several hours from the nearest town, connectivity is the exception. Everything the system does has to be true on a device that will not sync until evening.

That is an architectural commitment, not a feature. It decides how identity works, how duplicates are resolved and what happens when two workers record the same patient in the same afternoon.

A screening that finds something and loses the person is not a screening. It is a statistic.

The record is only half the job

  1. 01Capture the finding with enough identity to find the person again, which in practice means a phone number that belongs to someone who will still have it.
  2. 02Attach the next step at the moment of the finding: who calls, by when, and what happens if they cannot be reached.
  3. 03Close the loop visibly. A worker who never learns what happened to a referral stops believing the referral matters.
Health workers reviewing a day's records together
Detection is the part that gets funded. Follow-up is the part that decides whether detection was worth anything.

Respect the scarcest resource

In most outreach programmes the binding constraint is not equipment or even funding, it is the number of trained hours available in a district. Every form field is spent from that budget.

Design accordingly: fewer questions, defaults that reflect what is usually true, and no field collected because it might be interesting to somebody later.

Aashita EditorialOperations & Intelligence practice
Talk to us about this