Designing health information systems around decisions

Health information systems are essential for understanding how services are performing, identifying problems and guiding action. Yet many systems collect increasing amounts of data without necessarily making it easier for health workers and managers to decide what to do next.

The challenge is therefore not simply to collect more data or introduce better technology. It is to ensure that the information produced is relevant, timely and connected to decisions.

A strong health information system should answer not only what should we measure, but why do we need the information, who will use it and what action should it support?

Health information systems are often designed by first defining indicators, reporting forms or digital requirements.

A stronger starting point is to ask:

What decisions need to be made, by whom, and what information would help make those decisions?

A facility manager may need to know whether increasing patient attendance requires additional staff or medicines. A district team may need to identify facilities with declining vaccination coverage. A surveillance team may need to detect an unusual increase in cases early enough to investigate and respond. A national programme may need to know where performance is falling and where resources should be directed.

These are different decisions and require different information.

Once the decisions are clear, programmes can determine what information is actually needed.

Each indicator should have a purpose. It should be possible to explain who uses it, how often it is needed and what decision or action it supports.

This matters because health information systems tend to expand over time. New programmes, priorities and reporting requirements add indicators, while existing ones are rarely removed. Experience with large-scale health information systems shows how quickly datasets and reporting requirements can accumulate, increasing the burden on health workers without necessarily producing more useful information.

Prioritisation is therefore essential. A useful question is:

If we stopped collecting this information, what decision would we no longer be able to make?

If there is no clear answer, its continued collection should be reconsidered.

Not everyone needs the same information or the same level of detail.

Health workers need information to manage patients and services. Facility managers need it to organise staffing, medicines and service delivery. District teams need information for supervision and resource allocation. National programmes need information for planning, policy and performance monitoring.

A well-designed information system should therefore do more than move data from health facilities to a central database. It should ensure that useful information also flows back to the people who can act on it.

Digital platforms such as DHIS2 can make information easier to collect, analyse, visualise and share. But technology does not by itself create an effective health information system.

Before configuring a platform or building dashboards, programmes need clarity about what information is required, how it will be collected, who needs access to it and how it will support decisions.

Technology should therefore follow the purpose of the system rather than determine it. A sophisticated digital platform collecting information that is rarely reviewed or used remains primarily a reporting system.

Completeness, accuracy and timeliness are fundamental. But improving data quality should not depend only on periodic assessments or validation exercises.

Regular use of information is itself an important driver of quality.

When teams review results, compare performance, investigate unexpected changes and discuss what action to take, errors and inconsistencies become more visible. Health workers can also see more clearly why the information they collect matters.

Data quality and data use reinforce each other.

Health information needs change. Programmes evolve, services are reorganised, new priorities emerge and some indicators lose their relevance.

Periodic review is therefore an essential part of managing the system. It provides an opportunity to remove unnecessary indicators, simplify reporting and ensure that the information collected continues to serve a clear purpose.

For each major indicator, programmes can ask:

Who uses this information, for what purpose, and what would we lose if we stopped collecting it?

 
The objective is not simply to collect less data. It is to concentrate effort on information that has clear value for decision-making.