With seemingly endless amounts of data and the constant pressure to achieve “data-driven” healthcare improvement, health systems increasingly turn to data to solve problems and accelerate change. However, this onslaught of data within healthcare has left many organizations questioning whether they use their data effectively to achieve better outcomes.
To successfully apply data and track its success in data-driven healthcare improvement, Bryan Hinton, Health Catalyst Chief Technology officer, and TJ Elbert, Health Catalyst SVP and General Manager of Data, discuss the five data-centric mindset shifts health systems must adopt:
As cloud-based data platforms become more widely available to healthcare organizations, data interpretation is no longer the primary challenge. Instead, healthcare leaders need to focus on data orchestration: quickly combining data from different sources, then distributing the relevant data sets to appropriate end users throughout the organization.
An effective data platform should be able to orchestrate data, alleviating the burden of data growth because all the system’s data—no matter the volume—live in one place (the data platform). The next part of data orchestration is distributing the relevant data insights to team members so they can use the data in their everyday decision making. Effective data distribution is especially important for health systems balancing fee-for-service and value-based payment models with limited resources.
COVID-19 surfaced a critical need for real-time data from all data sources, both structured (e.g., claims and EHR data) and unstructured (e.g., clinician notes and imaging). As structured and unstructured data sources make the data technology landscape more complex, most organizations must manage both source types to gather timely data. Instead of relying on humans to manually compile data into spreadsheets, which has historically delayed access to real-time data, automation can increase access to real-time data, which is vital for organizations responding to health crises, including a pandemic, that require quick decisions in a short timeframe.
Health systems have historically been protective of their data largely due to privacy concerns. According to Hinton and Elbert, this protective mindset has become a roadblock to data sharing. The broader data landscape has used data to fuel growth and higher levels of efficiency without compromising quality, teaching the healthcare industry that data governance is less about control and more about data democratization.
In short, instead of withholding data from team members, leaders should enable teams with data along with the training they need to use it to make decisions. Giving team members visibility into data, data sources, and algorithms that derive analytic insight also helps team members trust the data, increasing the likelihood they will continue to use it. When team members feel confident using data, organizations achieve data-driven healthcare improvement at every level of the system, not just the top.
Internet of Things (IoT) data, namely consumer data from wearables, has vastly expanded the patient information that helps providers capture a more complete picture of a patient’s health. However, health systems can’t fully leverage this IoT data unless they’re using augmented intelligence (AI). Organizations have so many available data sets that it’s not possible for humans to compute them. With the rigor of AI, allows organizations can quickly make sense of IoT data and separate the value from the noise. AI is the key to making patient data actionable because it allows systems to take advantage of the breadth of IoT and patient-reported data, combine those insights with traditional data sources (e.g., EHR and claims data) for a comprehensive picture of patient health.
Healthcare leaders typically plug data into existing care models. Although this sounds like a sensible solution, organizations should take the opposite approach and alter their care models to fit their data sources. To take full advantage of a health system’s data and empower team members to work at the top of their license, leaders should identify the most relevant data for the appropriate end users, then alter their care models to support that data delivery. This data-first, instead of process-first, approach also prioritizes how end users, such as clinicians, need to receive data so they feel the most comfortable using it and can help reduce the all-too-common data burden that many clinicians experience.
Data is not going away. It’s only becoming more prevalent in healthcare. To make sense of the growing data and maximize this valuable resource, organizations need to think about data differently, starting with these five data-centric mindset shifts.
Would you like to learn more about this topic? Here are some articles we suggest: