Population Health Analytics

We’ve recently been highlighting some of the common myths and assumptions about population health analytics and the Johns Hopkins ACG^® System. In our first post in this series, we looked at misconceptions around proactive care, frailty and the idea that the ACG System is “just a risk model.”

Today — in part two of our three-part series —we’re exploring more myths and setting the record straight.

Myth: The ACG System is a U.S. model.

Because the ACG System originated at Johns Hopkins University in Baltimore, some assume it is designed primarily around U.S. data and health care practices. In fact, the system is far from U.S.-centric.

The ACG System is used in more than 20 countries worldwide. In England, it has been embedded across health care systems since 2009 and is currently used by organizations covering more than a quarter of the population.

Importantly, the ACG System isn’t a one-size-fits-all solution. Its predictive models, classifiers and mapping tables are regularly reviewed and recalibrated to reflect the realities of local health systems around the world.

The ACG team’s guiding philosophy is simple: models must be localized to be meaningful. This involves adapting to differences in coding practices, system structures, policies, and resources, ensuring it delivers actionable insights in every context.

Myth: The ACG System only processes one coding system.

Another common misconception is that the ACG System is rigid, built to process a single coding standard, and therefore limited in scope. In reality, the system is a true ‘polyglot’ in the world of health analytics — designed to work with and translate a wide range of coding systems and ontologies. It has built-in mapping tables and logic to harmonize data from SNOMED CT, ICD (various versions), ATC, DM+D and more.

This flexibility allows the ACG System to adapt to different countries, data sources and clinical environments. Whether using primary care, secondary care or pharmacy datasets, the system can provide consistent and meaningful insights.

In short: the ACG System is fluent in the languages of global health care, making it a true global platform for population health analytics.

Myth: The ACG System requires clean and perfect data.

It’s a common assumption that population health analytics solutions require pristine data to produce reliable results. In reality, health datasets are often messy — with missing values, inconsistent coding and incomplete patient histories.

The ACG System is specifically designed to handle this real-world complexity. It includes robust logic to accommodate missing or incomplete data while still generating meaningful insights. By producing aggregated clinical markers rather than relying only on individual diagnostic codes, it is less sensitive to gaps or variability in coding.

This resilience allows the ACG System to reliably identify key measures such as multimorbidity, even when working with imperfect datasets. Rather than demanding a level of data cleanliness that rarely exists, the ACG System works with the data that is available — delivering robust, actionable outputs that organizations can trust.

Myth: The ACG System is a ‘black box’.

Some analytics tools are described as ‘black boxes’, processing data behind the scenes and producing results with little visibility into the underlying methodology. The ACG System is different: its processes, logic and models are designed to be fully transparent and understandable to users.

Its ongoing connection to Johns Hopkins University ensures that research continuously drives ongoing software development and is published in peer-reviewed journals. The ACG System’s methodologies, classification logic and risk model designs are well-documented and accessible to users. Users have access to detailed documentation and guidance, and the Johns Hopkins team provides training to help organizations understand how outputs are generated.

This transparency builds trust. Health care organizations can understand, explain and justify results produced by the ACG System, rather than relying on an opaque score.

Myth: Data must be sent to Johns Hopkins for processing.

Finally, a common concern relates to data governance. Some assume that using the ACG System means sending sensitive patient information to Johns Hopkins in the U.S. for processing.

This isn’t the case. The ACG System is deployed on premises and hosted locally within each organization. All patient data remains on site and under the full control of the organization using the system. This fully complies with local data protection requirements and governance frameworks.

Rather than requiring any overseas data transfer, the ACG System is designed to integrate seamlessly into existing IT and governance infrastructures — keeping patient information secure.

To learn more about how the ACG System can support your organization, visit hopkinsacg.org or contact us at acginfo@jh.edu. If you are a current ACG System customer, please reach out to your Account Manager.