In today’s value-based care environment, long-term post-acute analytics (LTPAC) have become essential for skilled nursing facilities (SNFs) to succeed, stay competitive and achieve the objectives of the Triple Aim for participation in value-based care initiatives. Predictive analytics are changing healthcare with insights that help clinicians identify patients at risk for adverse events and poor health outcomes. Analytics from clinical data provide decision-making support to facilitate the implementation of care pathways and evidence-based best practices to manage clinically complex patients with multiple comorbidities.
High-performing predictive algorithms go beyond the identification of single risk factors. They also consider interdependent factors that contribute to the outcome. For example, predictors such as weight and conditions like diabetes and hypertension, especially when multiple comorbidities are present, tend to negatively impact a wound’s likelihood of healing.
Predictive algorithms can be created leveraging data from both EHR and Minimum Data Set (MDS) data. Predictive analytics derived from EHR and MDS data are best used in a complementary way, providing comprehensive insights to guide patient care. In this blog, we examine how EHR- and MDS-based analytics each bring value and the potential synergy that exists when both are used together.
The Merits of EHR Data-Based Analytics
EHR data-based analytics are primarily used to identify and manage acute changes in a patient’s condition so a clinician can further assess the patient and intervene to manage and prevent complications. For example, a patient develops a fever, declines slightly in activities of daily living self-performance, and has decreased fluid intake. An EHR-based predictive algorithm may identify these descriptive indicators and determine that the patient is at risk for readmission to the hospital within the next few days. Once alerted to this change in condition and imminent risk, the clinician can perform an assessment so the care team can intervene quickly to treat the patient and mitigate the risk.
There are some drawbacks to EHR-based analytics. Because post-acute EHR data is not yet standardized and is proprietary to each EHR vendor, the size of the data sets and benchmarking capabilities may be limited. Much EHR data is not structured (e.g., progress notes are largely free text), which is challenging to use in predictive models. Developing algorithms may require the use of additional technologies, such as optical character recognition (OCR), to identify predictors – and they may be less reliable as a result. In addition, analytics is not a core competency for most post-acute EHR vendors.
The Merits of MDS Analytics Data
MDS-based patient-level analytics are used to manage the risk of adverse outcomes through the interdisciplinary assessment and care planning process. MDS analytics can help clinicians avoid adverse events like falls, pressure ulcers, and rehospitalization by developing individualized care planning interventions based on a patient’s risk factors. While it may be possible to work on eliminating some risk factors, others may need to be managed to prevent complications.
Unlike EHR data, MDS data is highly structured and standardized. Driven by comprehensive clinical and functional assessments of the patient, the data allows the development of valid and reliable predictive models that are clinically relevant with robust benchmarking capabilities. Analytics-focused companies have a core competency in data science and analytics.
The Complementary Partnership of EHR- and MDS-Based Analytics
While EHR data and MDS-based analytics are effective individually, they’re not mutually exclusive when it comes to providing the best patient care possible. When used together, clinicians get the right analytics, at the right time, for the right purpose, which ultimately results in the right decisions for patients.
A great example of this is the way these analytics can be used in a complementary way to prevent rehospitalization, a key performance outcome for SNFs. A patient with a chronic lung disease in the SNF for a short-term rehabilitation stay following a hospitalization is identified to be at moderate risk for readmission by the MDS-based predictive algorithm with predictors including COPD, CHF, oxygen therapy, and incontinence.
The interdisciplinary care team considers these factors when developing the patient’s care plan, with a focus on managing them and preventing an acute exacerbation of her chronic illness. During the patient’s stay, she develops a mild cough and worsening shortness of breath. An EHR-based predictive algorithm identifies that she is now at high risk for readmission to the hospital and alerts the care team. A nurse assesses the patient, notifies her physician, and interventions are quickly ordered to manage the acute exacerbation of symptoms in the SNF to prevent a rehospitalization.
When used together, these two types of analytics provide a much greater opportunity to reduce rehospitalization. A recent case study on Archcare (see below), the continuing care community of the Archdiocese of New York, does a masterful job of showcasing this in a real-world example.
The Next Step
If you’re already using EHR data or MDS analytics individually, great! However, if you want to get the most analytics power from all of your data, we’d encourage you to explore technology-driven solutions to activate this complementary relationship and achieve better patient outcomes and even greater value.
Net Health does a masterful job of providing easy-to-integrate, clearly actionable, and highly effective solutions. At Net Health, analytics is a core competency and our data science and analytics experts are industry thought leaders. Choosing the right analytics partner is key to realizing the value analytics can bring. If you’d like to learn more about our PointRight Post-Acute analytics solutions and how they complement your EHR-based analytics, we’d encourage reaching out to schedule a demo and consultation today.