The ACG System is continuously evolved and maintained through ongoing research and development by the Johns Hopkins Center for Population Health IT (CPHIT). CPHIT is located within the Johns Hopkins Bloomberg School of Public Health, and works to distinguish the ACG System as a unique analytical toolkit through their work – including regularly publishing in top peer-reviewed journals. Today, we’re sharing an article recently published in the health care administration journal Medical Care by Dr. Klaus Lemke, a senior member of the ACG System R&D team at CPHIT. This article describes the development of Patient Need Groups (PNGs), the ACG System’s framework for segmenting patients based on their health care needs—a unique, proprietary method by which the ACG System supports users in targeting actionable patient groups. Read the full article here.

The development of the PNG framework began with Dr. Lemke and other Johns Hopkins researchers reviewing millions of health insurance claims from Medicare, Medicaid and employer-sponsored insurance. Most population segmentation frameworks have focused on older populations and complex diseases that are high cost, with less attention given to other population segments and levels of health care need. The goal was to develop a population segmentation approach relevant to the full human lifespan, inclusive of children, pregnant women, adults and the elderly.

Five principles guided the development of the PNG framework:

1. It must be relevant to all age groups
2. It should create separate categories for pregnant women because of their unique health needs
3. The number of categories should be limited to as few as possible
4. Each category should be meaningful to clinicians, health system administrators and payers
5. Together, the categories should array all types and levels of health needs and serve as a framework for population and person-based analyses that integrate additional available risk information, such as predictive health care utilization risk scores and socio-behavioral related factors

Using these principles, the team created 11 mutually exclusive PNG categories. Each category represents patients with similar levels and patterns of comorbidity, and similar needs for health care services. The PNG categories range from those with little or no health care utilization (PNG 01) to those with multiple complex and/or chronic conditions and who are frail (PNG 11). The team then undertook an initial evaluation of the validity of this framework, assessing clinical, utilization, expenditure and other need characteristics of persons within each PNG. They also explored how a need-based segmentation framework like PNGs interrelates with the ACG System’s predictive risk model to identify high-need and high-risk patients.

How can PNGs help your organization improve your population health strategy? PNGs allow you to design and direct care improvement programs based on individual needs. For less-complex PNG groups, the focus is on appropriate lifestyle management and prevention of worsening disease. For highly complex patients, the focus is on management of serious disease, care coordination, medication adherence and/or prevention of future hospitalizations. In summary, PNGs allow ACG System users to tailor interventions and resource allocation by directing resources optimally toward patients who have a capacity to benefit from health care services.

Click here to read the full article. And learn more about patient segmentation and PNGs here.

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 an ACG customer, reach out to your Account Manager.