Using data dashboards to track NCD outcomes at the community level

A useful dashboard does more than display charts. It helps a health team see whether people are receiving appropriate care, where gaps are emerging and which local service can take action. For non-communicable diseases, that means connecting measures of diabetes, obesity and cardiovascular health with the daily work of primary care, community health workers, hospitals and aged-care services.

The 2023 APEC Conference on Promoting Community-based Non-Communicable Diseases Integrated Care Model brought these issues together, with a particular focus on ageing populations and integrated prevention. Its official resources cover the agenda, speakers, venue, registration and hybrid participation, as well as downloadable presentation slides, virtual backgrounds and the program book. Together, they offer a useful foundation for organisations considering how to measure community-level improvement.

Australian teams have a familiar phrase for focusing attention on the right issue: “track the horse”. Applied to health data, it means looking beyond national averages and examining the communities, pathways and conditions where outcomes vary. The challenge is to turn that principle into a clear, trusted and workable dashboard without creating another reporting burden or exposing personal information.

From national averages to community signals

A national rate can hide considerable differences between suburbs, age groups and service settings. A community dashboard should identify relevant populations, establish a baseline and show change over time. It can then explore which parts of the care pathway contribute to that change, rather than leaving the figures without an operational purpose.

The official conference agenda provides a useful starting point for this work because it places measurement within a wider discussion of prevention, referral and integrated care. Suitable indicators might include blood-pressure control, completion of diabetes reviews, access to weight-management support or follow-up after a cardiovascular event. Each indicator should have a clearly defined population, data source, reporting period and owner.

Presentation matters as much as selection. A simple trend line may reveal sustained improvement, while a map or service-level view may show where an apparent gap is concentrated. Small numbers must be handled carefully, and apparent changes should be checked against clinical knowledge, service redesign or changes in data collection before teams draw conclusions.

Choosing measures that support action

Dashboards should contain measures that can trigger a decision. A waiting time becomes useful when a team knows whether it will review staffing, alter triage or contact people at risk. A low screening rate may prompt outreach, but only if the dashboard identifies the eligible population, the gap and the service responsible for responding.

A balanced set of indicators normally includes prevention, service access and health outcomes. Process measures show whether care happened, while outcome measures show whether care is associated with better health. For chronic conditions, some results take time to appear, so teams should use near-term service measures alongside longer-term clinical measures rather than expecting every improvement to be visible immediately.

Community participation helps prevent a dashboard from becoming an administrative artefact. Consumers, carers, Aboriginal health workers and clinicians can identify which distinctions are meaningful and which information might cause misunderstanding. A useful working session ends with named actions, not merely agreement that a chart looks attractive.

Connecting the Australian health data landscape

Australia has many capable data sources, but they are distributed across general practices, pathology services, pharmacies, public hospitals, Medicare, aged-care providers and local health districts. Primary Health Networks help coordinate primary care within their regions, while Local Health Networks oversee much of the public hospital system. My Health Record can support information exchange where patients and providers have chosen to use it.

No single dashboard should connect every source on day one. A pragmatic approach starts with a small, well-defined pathway, documents where each measure comes from and establishes a refresh schedule. ABS population data can provide context, while local service data can reveal access gaps. Speaker perspectives can help teams compare approaches and avoid assuming that a model developed in one setting will transfer unchanged to another.

Data definitions require particular attention in a federated system. Teams should distinguish between a confirmed diagnosis, a self-reported condition and a recorded risk factor. They should also record whether missing data means that a service was not provided, the result was not available or the person was outside the cohort.

Making inequity visible without stereotyping

A dashboard can show whether outcomes differ by age, location, socioeconomic circumstances, disability, language background or Aboriginal and Torres Strait Islander identity. Disaggregation can reveal barriers that an overall average conceals. It can also produce misleading conclusions, so interpretation must consider local knowledge and the limits of the underlying data.

Remote and regional communities require a different view of “access”. A service may be geographically close while remaining difficult to reach without transport, digital connectivity or culturally safe support. Programs in the Tiwi Islands, remote parts of western New South Wales and Central Australia demonstrate why travel distance alone does not capture the burden on patients and families. Community-controlled services must also retain authority over how their information is collected and used.

Teams can reduce these risks by involving local representatives in indicator design, reviewing small-number results in context and avoiding public league tables. Multilingual materials and accessible formats help ensure that data supports communication with people from culturally and linguistically diverse backgrounds, including those who use interpreters, Auslan or communication support workers.

Following referrals from first contact to follow-up

Referral dashboards are most valuable when they follow the complete handover between services. For a person with diabetes or cardiovascular disease, this may include a general-practice assessment, a community health post, a specialist appointment, medication changes and a subsequent clinical review. The dashboard should show whether each step was completed, where communication failed and whether the person received the intended follow-up.

A related discussion on community referral pathways illustrates why pathway design and measurement need to advance together. Useful measures can include referral acceptance time, the proportion of referrals closed with documented feedback, overdue reviews and avoidable hospital use. These measures should be interpreted alongside patient experience, since a short pathway that leaves people feeling unsupported is not a successful pathway.

Closed-loop communication also matters in remote settings. A referral recorded as “sent” should be distinguishable from one received, triaged and acted upon. Secure messages, clear escalation criteria and a named coordinator can prevent the kind of red tape that makes people repeatedly present to emergency departments because earlier follow-up never occurred.

Using the conference as a learning platform

The conference brought together perspectives from different health systems, creating opportunities to compare how community organisations, hospitals and policy bodies approach integrated care. A broader integrated care forum offers another setting for examining shared implementation issues, including data, workforce development and governance. Such exchanges are most useful when participants bring a real pathway and specific questions rather than general examples.

Hybrid participation widened access to delegates who could not travel to the venue. The event’s registration, venue details and online access information supported both in-person and remote participation, while downloadable materials allowed participants to revisit technical content. Program books and presentation slides can also become the basis for internal discussions after the event.

For Australian teams, the strongest lesson is to treat the dashboard as a shared improvement tool. Clinicians need to see clinical meaning, community representatives need assurance about governance, and managers need evidence of value. A dashboard that serves all three groups is more likely to remain active than one created solely to satisfy a reporting cycle.

Governing data for trust and longevity

Privacy cannot be added after the dashboard is built. The Australian Privacy Principles provide a national reference, while state and territory laws, health-records requirements and the My Health Record framework may also apply. Organisations should obtain appropriate legal and privacy advice, establish access controls and clarify the circumstances in which identifiable information can be used.

Small datasets need special care because combinations of location, condition and age can identify individuals. Teams can often use aggregated or de-identified information, but governance should cover re-identification risk, export permissions, retention and secure disposal. Community representatives should have a meaningful role in these decisions, particularly where data concerns services they manage.

A durable dashboard also needs operational ownership. A data dictionary, version-controlled definitions, an audit trail and a scheduled review prevent figures from becoming deceptively precise. Conference reporting on closing reflections reinforces how important shared commitments are to future progress: dashboards matter when leaders act on what they reveal and communities can see how the information is used.

Choose one community care pathway, define three meaningful measures and appoint a clinical lead, data steward and community representative to oversee the first review. Use the conference materials and speaker resources to refine the design, then schedule the next review before the dashboard is launched. Make the first action visible, assign responsibility and return to the figures at regular intervals; a dashboard should drive a decision, not simply describe one.