To make informed choices, the care-at-home industry needs to base its decisions on evidence. Business intelligence has traditionally taken the form of quarterly or yearly reports that monitor a defined set of key performance indicators (KPIs), but today’s software-backed tools work continuously and at lightning speed. The mountains of data that organizations produce using their electronic medical record (EMR), combined with available market data, will help reveal valuable patterns and trends.
Building an analytics dashboard
How do organizations optimize and scale business intelligence? Here are five basic steps to consider when developing dashboards to assess your operational, financial and quality components within the care at home industry.
1. Define: The first step is to identify the end goal of a report. Keep the goal broad enough that it will allow you to measure a variety of metrics, but narrow enough to prevent overwhelming data.
2. Assign team: Assign a team to own measuring metrics. Depending on your organization size, you may have additional departments, such as education and training, intake and budgeting. Define champions within each department (these can be subject matter experts) to drive the goal, as well as the metrics within.
3. Define metrics: Remember, garbage in is garbage out. Metrics that 1) drive success and value, 2) are easy to understand and 3) lead operators to implement corrective action concurrently will stand the test of reliable data.
4. Map connections: You know the goal, who needs the information and what they should measure. Now it’s time to find where that information lives. Determine the data source for each metric, being as specific as possible. Examples:
- Internal: EMR, budget
- External: Published data and vendor sources
5. Compare data: You’ve done all the legwork, now it’s time to build your dashboard. Determine time periods for data comparisons: Week over week, month over month, etc. Some data may not have enough information to compare day over day. Data can be filtered multiple ways for comparisons. For example, break it down by payer source or branch to gain additional insights. It is important that a dashboard allows for multiple comparisons points.
Compiling the results
As you blend data from various connectors, be sure to follow data cleanliness standards so your team can trust what they’re seeing. The end report should be visual and easy to read and have been vetted through quality improvement (QI) before presenting to a larger group.
Look for inconsistencies. If there are discrepancies, your team needs to research, validate and correct the metrics (the single source of truth), which will evolve over time.
Management styles: Reports vs. exception
Organizations need to be cautious of information overload (the excess of information available to a person aiming to complete a task or make a choice), typically resulting in a poor decision being made – or none at all. This “analysis paralysis” is common in the fast-paced world of healthcare operations. Leaders are inundated with metrics from multiple sources that are not always actionable or concurrent. Too often this leads to no action and poor outcomes.
To reduce the risk of analysis paralysis, organizations should focus on managing by reports and by exception in three key areas: operations, finance and quality.
Management by reports communicates business results, issues and risk and is critical for directing a business. Including key performance metrics will provide density and textual information.
Management by exception is a style that focuses on identifying and handling cases that deviate from the norm. This enables management to practice by exception, narrowing focus to problem areas and creating concurrent actionable items to improve outcomes.
Putting these management styles to use
Both management by reports and by exception factors in two types of retrospective analytics found in business intelligence:
- Descriptive analytics: Answers the question, “What happened?” and should be done first. Often referred to as traditional BI, descriptive analytics acts as the baseline and is where all organizations should start. Textual information is typically transformed into pie charts, bar charts or line graphs.
- Diagnostic analytics: Answers the question, “Why did it happen?” This style is comprised of drill-down data discovery to identify correlations. These diagnostics provide information without cluttering the dashboard and offer valuable data for multiple stakeholders.
It is through proactive analytics that the care at home industry can improve employee and patient engagement, relationship management and clinical care. The two forms of proactive analytics are:
- Predictive analytics: Answers the question, “What is likely to happen?” This style utilizes a variety of statistical techniques from data mining, predictive modeling and machine learning. Both historical and current data is used to make predictions.
- Prescriptive analytics: Answers the question, “What should be done to prevent it from happening again?” This advanced level uses techniques, such as complex event processing, recommendation engines, simulation and machine learning, and provides insights on how to optimize business practices.
These insights can help a company choose a course of action in a matter of minutes.
Only once you understand the behaviors behind essential metrics can your organization move performance toward excellence.
Photo: Nuthawut Somsuk, Getty Images