Real-time information can go a long way in care management

The COVID-19 pandemic has revealed shortcomings in communicable disease management. At the heart of this issue is the lack of data management and analytics infrastructure to help with recovery and mitigation efforts.

“We need systems for tracking and managing workflows to support safely, effectively and efficiently bringing patients back into the hospital and clinic for these interventions,” says Jenifer Leaf Jaeger, MD, MPH, senior medical director, HealthEC, of the need to prevent delayed and deferred care.

Moving forward, medical practices and providers may want to consider establishing preparedness and response plans to help stave off reinfections, new strains of coronavirus and future pandemics, not to mention providing a uniform vaccine response.

A four-bucket solution

For HealthEC, the solution for communicable disease management can be broken down into four buckets:

Monitoring — Vigorous contact tracing; evolving internet-enabled devices, including thermometers and remote monitoring systems; and better access to data from HIE and other systems, all of which can provide more effective warning capabilities.
Preparation — Risk stratification and predictive modeling to assist in bringing assets and resources online, providing services to targeted populations and dynamic establishment of programs that mitigate the impact of outbreaks on the whole person, including preventing unwanted outcomes, such as domestic violence and food insecurity.
Response management — Real-time adaptive modeling to support resource allocation and deployment, as well as using data to determine the best ways to prevent spread and establish control measures, clinical interventions, care management protocols and human services programs.
Transition management — Modeling that helps states or localities confidentially loosen prevention and control measures.

Effectively managing response starts with monitoring and analyzing data, which is done on three levels: testing, cases and deaths. “Denominator data is critically important for determining prevalence rates, and in calculating both the attack rates among high-risk groups and identifying health disparities,” says Jaeger of the significance of this information. “This is contributed through the collection of data, clinical laboratory claims, all of which have to be linked with demographic and social determinants of health (SDoH) data elements.”

Denominator data includes demographics such as age, gender, ethnicity and race, education, income and ZIP code; and health data such as comorbidities, medications and vaccination history. Geospatial data, including overlaying providers, practices, pharmacies, high-risk and general population, and positive results, can then be added and linked to utilization and service data to help determine treatment, outcomes and gaps in care.

In aggregating this information alongside admissions/discharges/transfers (ADT), claims, labs and pharmacy data, HealthEC is able to track and risk stratify for patients at the state and local level. This enables drilling down to the patient level to determine hot spots, high-risk individuals and potential spread, which helps in allocating resources and providing real-time case updates.

“By collecting and analyzing this data in real time, we actually can contribute to identifying the leading edge within the spread by layering daily the new positive cases on top of our high-risk cases,” notes Jaeger. “City, county and state jurisdictions can then closely monitor high-risk and COVID-19 positive patients and track this spread.”

Personal health information data is then converted into de-identified data, which can be used to allay HIPAA security and portability concerns.

Risk stratification model

To develop its risk stratification model and identify high-risk patients, HealthEC uses the most recent guidelines from the Centers for Disease Control and Prevention (CDC) and the World Health Organization (WHO), along with other published research, collaboration with other organizations throughout the world and the organization’s robust data analytics capacity. Its model considers:

Age — 65 and older, with the highest mortality risk earmarked for those 80 and older.
Comorbid conditions — cardiac disease, including hypertension, coronary artery diseases and congestive heart failure; cancer; chronic kidney disease; COPD; obesity; and type 2 diabetes
Risk score (based on available data) — Johns Hopkins Prospective and Likelihood of Hospitalization scores; Hierarchical Condition Category (HCC); C3 Score (oncology patients)
When available and to support risk stratification and decision-making at the local level, HealthEC utilizes data on hospital admissions, ED visits, total cost of care, as well as the Johns Hopkins Frailty Assessment Calculator.

Objectives and deliverables

A COVID-19 module aids to determine objectives and deliverables. As Jaeger notes, it starts with a survey that helps classify high-risk members. Next, it incorporates COVID-19 test results to pinpoint individuals who tested positive. It then integrates ADT and other data to help determine patterns of spread, severity and variations in demographics. Finally, it helps support the management of individuals who are at risk of complications or are recovering from COVID-19. This aggregate data make it easier to provide dynamic population health management, establish advanced care management protocols, expedite the development of care plans, boost compliance monitoring, improve modeling of financial impacts and deliver more exact documentation and reporting of COVID-19’s impact on quality and other performance measures.

According to Sita Kapoor, chief informatics officer, HealthEC, once you have the clinical, claims and SDoH data, the goal “is to aggregate as much of the key components needed to predict as accurately as possible to create interventions and to create a workflow that will truly help the community clients to manage the patient population they are addressing.”

Along with predictive modeling, stratification outreach and care management, the data HealthEC aggregates provides the information needed to develop an SDoH referral program and an interventions program used to determine short- and long-term goals.

Intervention

Streamlining patient data is paramount for care managers when assessing vulnerable patients, enabling better engagement and interventions for the entire care team. The interventions workflow starts with a list of high-risk patients, which is uploaded into a population health management platform. HealthEC is then able to follow a five-step process to determine the level of intervention:

Stratification of high-risk patients to discern level of outreach and assessment.
An assessment survey produces a risk score and action plan.
Action plans are then managed within a patient’s care plan.
COVID-19 program recommendations are linked to interventions and goals, which trigger alerts for patient follow up.
Utilization and cost data for the COVID-19 cohort of patients is analyzed and tied to interventions.

When the intervention workflow is initiated, geo maps can be used to chart high-risk and COVID-19 positive patients by ZIP code, provider or facility. In addition, COVID-19 test results can be used to distinguish susceptible and infected members to provide them with ongoing care management, help with resource allocation and conduct cluster analysis.

Within the intervention workflow, patients are assigned a score between 0 to 9+ to determine actions to take. “Care management workflows need to ensure that patients receive necessary support and education to manage active or potential COVID-19 symptoms,” says Lyndsey Lord, senior vice president, technical operations, HealthEC. “And understanding their risk level and the right care for them is really key to success in this area, especially with so many patients having questions or needing care.”

Each risk score requires a different actionable intervention:
9+          Action 1: Emergency warning signs related/not related to COVID-19; seek immediate medical attention
5-8        Action 2: Stay at home, monitor symptoms and contact provider within 24 hours; attempt to prevent spread at home
2-4        Action 3: Stay at home, monitor symptoms and contact provider within 24-48 hours; attempt to prevent spread at home
0-1        Action 4: Stay at home, monitor symptoms and contact provider if symptoms worsen; attempt to prevent spread at home

Beyond patients, the benefits for providers are also far reaching. “Collectively, taking all of this data and synthesizing it allows providers to manage their capacity, and ensure that the patients are safely cared for when and how it's most appropriate for their current state,” says Lord.

The importance of addressing SDoH

With SDoH influencing up to 80% of health outcomes, a sound strategy is needed to help mitigate these conditions, which have become even more glaring during COVID-19. “When you think about the social isolation, the job and financial security … all of these pieces that have skyrocketed during COVID-19, and we see the dramatic impact that this can have on patient outcomes,” asserts Lord about the need to track this data.

In essence, the pandemic has brought health inequities to the forefront; therefore, creating a strategy to address SDoH can help better meet the needs of all patients. HealthEC does this by obtaining data — including demographic and socioeconomic variables, ethnicity and language variables, household unit profiling variables, etc. — from a series of assessment questions, combined with other data sets, which ensures the best actions and interventions are taken.

“We can boil this down to a really granular level, which allows us to do things such as geo mapping and provide greater insight to practices and providers on vulnerable pockets and vulnerable areas,” says Lord. “So as they create population health plans and community health plans, they have this added level of detail to consider in their process.”