Public-health institutions are surrounded by data: routine service statistics, surveys, surveillance systems, laboratory records, programme reports and, increasingly, digital tools that capture information in real time. Volume is rarely the problem. The challenge is knowing when that data is ready to guide a decision, and how much confidence it deserves.
Leaders do not need to be statisticians to make good use of evidence. They do need to ask the right questions. These five are a practical place to start.
1. What question is this data actually answering?
Data is collected for specific purposes, and it answers some questions far better than others. Routine service data may show how many people used a service, but not how many needed it. A survey may describe a population well at one point in time, but say little about change since then.
Before relying on a figure, it helps to ask what it was designed to measure and whether that matches the decision at hand. Many misunderstandings begin with data being asked to answer a question it was never built for.
2. How are the key terms defined?
Indicators that sound simple often depend on careful definitions. What counts as a case, a visit, a completed treatment or a covered population? Definitions can differ between programmes, sites and years. When they shift quietly, trends can appear or disappear for reasons that have nothing to do with health.
Leaders can ask for definitions to be stated alongside the numbers, and for any changes to be flagged clearly.
3. How complete and timely is it?
Missing reports, delayed submissions and gaps in coverage can change a picture considerably. A district that appears to be performing well may simply be reporting less. Understanding completeness and timeliness helps leaders judge whether differences are real.
Simple measures, such as the share of expected reports received and when they arrived, should accompany any dashboard used for decisions.
4. Can the result be reproduced?
If an analysis were repeated next month by a different analyst, would it produce the same result? Reproducible workflows, with documented steps from raw data to final figures, protect institutions from errors and make it possible to update evidence quickly when new data arrives.
This is not only a technical matter. It determines whether evidence remains useful after the person who produced it moves on.
5. What would change our minds?
Good decisions acknowledge uncertainty. It is worth asking what additional evidence would lead to a different choice, and whether that evidence can be gathered in time. This keeps decisions open to learning and helps institutions decide where further investment in data is worthwhile.
Building institutions that decide well
These questions are most powerful when they become habits across an institution, not just at the leadership table. That requires analysts who can explain their methods, systems that keep data organised and accessible, and teams trained to interpret evidence with confidence.
At Praxis Global Institute, our research and data work focuses on exactly this: clear definitions, reproducible methods and evidence that institutions can trust and reuse. If your institution wants to strengthen how it moves from data to decisions, let us talk.
Simon Aseno
Simon Aseno is President and CEO of Praxis Global Institute, where he leads work connecting research, technology, education and advisory to strengthen people, systems and institutions.
Facing a similar challenge in your institution?
We work with teams on research, data, digital systems, and learning that hold up in practice.
