Drilling for analytics

Being able to predict an outcome before beginning a task or activity can help guarantee a positive experience.

Getting ready to embark on your morning commute? Check Google Maps or Waze to determine the fastest route. Trying to decide on the best Thai restaurant within five miles of your sister’s house in Atlanta? Use Yelp or Trip Advisor to improve the chances you’ll have a scrumptious meal and superb service.

Predictive analytics offer real-time data that can provide key information you need to make an informed decision. So why don’t we use predictive analytics to determine patterns and forecast outcomes?

According to Cameron Cox III, MHA, FACMPE, president, MSOC Health, Chapel Hill, N.C., it’s imperative for practice managers and administrators to rely on hard data to validate their decisions: “If you operate from a level of subjectivity all the time, it’s going to be a constant world of ‘I feel like.’” It’s much better to operate from a position of “I know.”

This is what happens when we only work from the first, or descriptive, level of the analytics continuum, which is where most management lives. According to Cox, descriptive analysis is used to address our daily toils, which can be captured in end-of-month reports or spreadsheets and measured with tools such as dashboards, scorecards and business reports.

“What we look at every month are charges, payments and adjustments … and we might look at production data by doctor,” Cox says. “But again, all we’re looking at is descriptive, what happened … It’s really not taking us to another level.”

If something goes wrong in our practice, we may take the next step in our analysis by diagnosing the problem. Tools such as trending reports, pivot tables and key metric indicator (KMI) comparisons help determine why collections have been so low, why the no-show rate spiked this month or why there have been an aberrant number of frustrated patient phone calls during the past week.

“This is where we get shifted and pushed up,” Cox says. “So, diagnostic is ‘why did it happen?’ Typically, diagnostic is a catalyst-driven approach.”

But only focusing on the first two steps of the analytics continuum will not get us to where we need to be to help our practice.

By taking the next step, predictive analysis, we can forecast what may happen in a specific situation. To understand what’s coming down the pike, we need to take a deep dive into the data.

There is an abundance of data in healthcare, but as Cox notes, it needs to be used effectively. Instead, we get mired in descriptive and diagnostic analysis, only reacting to what comes our way each day, week and month.

“We’re not taking our data and trying to look forward on it,” Cox says. “What do we really know about our patients? Do we know how many women visit our office, Millennials, Boomers? What do we do with our data to learn more about our practice to predict the patterns going forward?”

By mining our practice’s data sets, we can predict outcomes such as patient utilization, no-shows and risk scoring for chronic diseases. Similarly, forecasting can be used in Microsoft Excel to not only reveal predictive activity but substantiate it as well. With strategic planning, it’s essential to utilize metrics and formulate objectives. But this isn’t possible if we don’t have the information to back it up.

“If you cannot find your points in time going forward to validate the decisions you have been making, then you really don’t have anything,” Cox says.

Not only do we need to be able to predict our experience, we also need to be able to optimize it. That’s where we can take predictive analysis a step further by using data to calculate the best course of action for our practice.

“A lot of times doctors think in terms of revenue; they don’t think in terms of profitability,” Cox says. This can be problematic when they don’t consider the expenses associated with the services they provide.

Tools such as Excel’s Solver can demonstrate how expenses affect practices and how practices can make use of resources to maximize profitability. By keying on expenses, bandwidth, constraints and other variables, practices can conduct a maximizing analysis.

Conversely, the same Excel tool can be used to conduct a minimizing analysis to determine staffing levels, for example. “Our job as managers and administrators is to allocate resources appropriately,” Cox says. With Solver, “you can find the minimal number of the optimal point.”

Whether you use Solver or other forms of linear programming, or employ decision modeling, prescriptive analysis can make a huge difference in helping to optimize your practice.

“Our days of just looking at the descriptive and the diagnostic have passed,” Cox says. “It’s now up to us to start taking our talents and actually going to a higher level.”

For practice managers, analytics tools can help them maximize and use their resources effectively, which has become a big challenge in healthcare, given stagnant reimbursement rates for some practices.

“A lot of us are a one-trick pony; it’s price [times] volume. It’s what we do,” Cox says. “It’s hard for us to compete with a health system that can make money in different areas.”

In the end, it’s most important to get creative when determining the type of data you want to measure. “I encourage you to go outside the box,” Cox says. “Don’t stay in the box we’ve always been in, which is, ‘I’m looking at my charges, I’m looking at my payments and I’m looking at my adjustments’ … Begin to look for diagnostic data and actually take it one level more, to predictive data.”

Did you know?

Excel Solver is an add-in program that can help optimize your business model by carrying out what-if analyses. It can be used to ascertain the maximum or minimum value of one cell by adjusting the value of other cells. For example, it can examine the type of patients a physician should see to maximize ROI or determine appropriate staffing levels for a medical practice. For a quick video tutorial on Solver, click here.