There are a number of different analytical tools that comprise an overarching Population Health Management toolkit. These tools help set priorities, highlight opportunities and identify the cohorts and individuals with the greatest capacity to benefit.
The core tools are those below. Click tile for more information
Whole Population Segmentation
Whole population segmentation refers to the process of dividing an entire population into distinct and homogeneous groups based on specific criteria and characteristics.
The purpose of population segmentation is to better understand the diverse needs, behaviours and preferences of different groups within a population in order to develop and implement targeted strategies and interventions.
More on Population Segmentation here
There are a number of ways of segmenting a population.
One such methodology is Johns Hopkins Healthcare (JHHC) Patient Needs Groups (PNGs). More on PNGs here
Risk Stratification
Risk stratification involves dividing populations into different groups or levels based on their level of risk.
The goal is to identify and prioritise high-risk individuals so that appropriate interventions can be taken in order to mitigate those risks and/or allocate resources (finance and human) effectively.
Sollis has been a long-term collaborator with Johns Hopkins Healthcare (JHH) on risk stratification.
We have extensive knowledge and expertise working with the JHH ACG® System and have fully integrated ACGs within the Sollis™ Population Health Analytics Platform. Learn more about the ACG® System here
https://www.hopkinsacg.org/about-the-acg-system/
- Risk modelling involves statistical analysis of historic events
- Every member of a population is assigned a probability of experiencing a specific outcome in the future
- Risk models should ideally always present not only the risk score, but also make available the risk attributes that form that score (e.g. age, medical history, hospital activity etc.) so we know why individuals are at a certain risk level (“drivers of risk”)
- Stratification partitions the distribution of risk scores in any population
- This is crucial in determining thresholds of risk at which specific actions should (or shouldn’t) be taken, and which resources should be targeted
- Stratification will only ever consider one outcome at a time
- Modelling the risk within a population and then stratifying the population supports the process of refining the cohort that the programme will focus on.
Predictive Modelling
Predictive modelling refers to the application of statistical models to make predictions about future health outcomes. It involves developing models that can learn patterns and relationships from historical data to predict the likelihood of specific events or conditions occurring for individual patients or populations.
Sollis has been a long-term collaborator with Johns Hopkins Healthcare (JHH) on predictive modelling.
We have extensive knowledge and expertise working with the JHH ACG® System and have fully integrated ACGs within the Sollis™ Population Health Analytics Platform.
Learn more about the ACG® System here
https://www.hopkinsacg.org/about-the-acg-system/
- Risk modelling involves statistical analysis of historic events
- Every member of a population is assigned a probability of experiencing a specific outcome in the future
- Risk models should ideally always present not only the risk score, but also make available the risk attributes that form that score (e.g. age, medical history, hospital activity etc.) so we know why individuals are at a certain risk level (“drivers of risk”)
- Stratification partitions the distribution of risk scores in any population
- This is crucial in determining thresholds of risk at which specific actions should (or shouldn’t) be taken, and which resources should be targeted
- Stratification will only ever consider one outcome at a time
- Modelling the risk within a population and then stratifying the population supports the process of refining the cohort that the programme will focus on.
Cohort Identification
Cohort identification is a key component of population segmentation. It refers to the process of dividing a population into distinct groups or characteristics based on certain shared characteristics or experiences.
Cohort identification aims to identify and understand the characteristics, behaviours and needs of different groups within a population.
Impactability
Risk stratification and population segmentation typically focus on identifying population groups that have a high risk of experiencing an adverse event.
The success of risk stratification depends on not just identifying those at most risk of an adverse event, but rather in identifying those who are at most risk and most likely to respond positively to a specific intervention.
Impactability therefore defines the degree to which different sub-populations will benefit from a range of interventions.
Case Finding
In Population Health Management, case finding refers to the process of identifying individuals within a population (or cohort) that satisfy certain criteria.
The goal of case finding is to proactively identify individuals who may benefit from early intervention, treatment or preventative measures.
Enrolling patients onto specific and targeted intervention programmes is a key component of preventative and anticipatory care.