
Blog · HealthWatch
A health alert without an owner is just a red cell on a screen.
Command centres invest heavily in detecting problems. We think the harder and more important part is what happens in the minutes after something turns red.

Command centres invest heavily in detecting problems. We think the harder and more important part is what happens in the minutes after something turns red.
A cluster of fever cases appears in a block. The state screen turns it red. The district team can see it, the surveillance unit can see it, the mission office can see it. Each assumes one of the others is acting. A day later, someone asks what was done, and the honest answer is that everyone watched.
This is not a failure of detection. The signal was there. It is a failure of ownership, and it is common wherever many offices look at the same screen.
Large dashboards are often justified with the idea that everyone should see everything. Visibility is good, but it spreads responsibility thin. When ten people can see a problem, the likelihood that any single one feels it is theirs goes down. The more people a screen serves, the more it needs a rule about who acts.
Ownership also has to be specific. “The district” is not an owner; a named officer in the district is. “Surveillance” is not an owner; the person on the surveillance desk this morning is. Vague ownership feels safer in a meeting and fails in practice.
We think of an exception as a small record, not a colour. At minimum it should hold:
| Field | Why it matters |
|---|---|
| What was detected, with evidence | So the owner does not have to rebuild the case |
| A suggested action | So the first step is clear, even if the owner changes it |
| One named owner | A district or hospital officer, not a department |
| Time assigned and time closed | So delays are visible, not discovered later |
| How it was closed | So the review can learn from it |
Once exceptions look like this, the screen changes character. It stops being a wall of colours and becomes a short list of things specific people have to do, with a clock running on each.
District officers are rarely sitting in front of the state screen. They are in meetings, at facilities, on the road. Asking them to log in to a dashboard to find their own exceptions is asking for delay. The decision should reach them on the phone they already carry, with the numbers for their district and a clear choice to approve or act. And the approval should be recorded with the time and the person, so the state office does not have to call to find out.
HealthWatch ranks exceptions such as capacity shortfalls, clusters and delays by risk to life, and assigns each to a district or hospital owner, with a target of under five minutes from exception to owner. Clusters flagged against baseline arrive with an action and an owner. Each district gets a morning brief on WhatsApp with its own numbers and exceptions and decisions that close in one tap; approvals are logged with time and owner. Helpline and ambulance cases sit on the same exception queue, and every exception is kept with owner, time and closure for the weekly review and replay.
For the detection side, read disease cluster detection: seeing outbreaks early, and for the district routine, the monsoon disease surveillance playbook. The rest is in HealthWatch guides and tools.
The decisions themselves, clinical and administrative, stay with the officials responsible. The point is that someone specific knows it is theirs, and everyone else can see that they do.
When many offices can see the same alert, each may assume another is acting. Without one named owner, shared visibility spreads responsibility thin.
What was detected with evidence, a suggested action, one named owner, the time assigned and closed, and how it was closed.
On the phone they already carry, with their district’s numbers and a decision they can approve, logged with time and owner.
See it on your own data. HealthWatch — Outbreaks and ICU load, one screen. Book a 30-minute working session with an engineer.