Executive Summary

By now, executives across industries recognize the value of leveraging data to make business decisions. Yet, so much of that data remains locked away in various systems and spreadsheets, making it difficult to use, unless you’re a data analytics expert. This is why machine learning lives up to its hype.

When used correctly, machine learning technology creates useful, actionable insights that enable you to make smarter, more objective decisions for your organization. For Health Systems, data generated from machine learning has the potential to transform the revenue cycle into an economic growth engine. To maximize the benefits of machine learning, you must be ready to experiment.

The Value of Experimentation

Here’s how. By running multiple targeted experiments, you can identify which factors impact certain revenue cycle activities, such as how you engage patients during the billing process and how that affects payment rates. These experiments test how changes to those activities impact your organization’s bottom line. Based on the outcomes of those experiments, i.e. the data generated, machine learning can adjust and optimize your revenue cycle activities accordingly. Without experimentation and machine learning, your efforts at fine-tuning would be based on little more than gut-feeling and guesswork.

Experimentation Can Impact Your Bottom Line

Data from experimentation and machine learning will tell you exactly where adjustments are needed within your revenue cycle. Some of these adjustments might seem small, but when compounded, they can have a significant impact on your Health System’s bottom line.

To illustrate how small of a change we’re talking, consider this example: in one experiment with a Patientco client, we tested eBill content and found that a change as small as adding an exclamation point to the call-to-action impacted payment rates by 3 percent. This is just one instance of how a small change can noticeably impact the bottom line, so imagine how other adjustments could help your organization.

Experimenting with your healthcare organization’s patient engagement methods is an ideal place to start. This is because there are several changes you can implement to your patient outreach efforts. With machine learning, those changes can then be tested on a large scale against a control group to understand how they affected patients’ engagement levels and payment rates.




Experiment with Outreach Tactics to Optimize Patient Engagement

There are two components of patient engagement to look at:

  • Getting the patient’s attention with the right outreach
  • Getting the patient to act on that outreach

You must get the first step right in order to succeed with the second.

Oftentimes, revenue cycle staff interact with patients through a channel like mailed paper statements or eBills. The channel of engagement is usually determined based on a patient’s expressed preference, i.e. the patient opts in for eBills instead of paper statements. But, what if you could take into account whether that patient is actually opening their eBills? With data analytics and machine learning, you can.

If a patient hasn’t opened their last two eBills, your organization could try an alternative method of engagement, like sending a text notification to the patient. Not only does the method of engagement matter, the content within the financial communication matters too. In other words, when a patient opens their billing statement, which factors influence whether or not they complete a payment?

The ideal engagement pattern and messaging will likely vary for different patients with different bill balances, but machine learning allows you to accommodate for those variables and fine-tune outreach efforts accordingly.

Test the Payment Experience

Machine learning also empowers a better, more tailored payment experience. Let’s look at two crucial factors of the payment experience. 1) How easy it is to pay and 2) how relevant or flexible the payment options are for the patient. These are important because patients are likely to move on without making a payment if it’s too difficult to pay or if there’s no way they can afford their bill.

Consider your healthcare organization’s payment flows. For instance, what are the steps a patient has to take to make a payment upon receiving an eBill?

Experimenting and making small adjustments to the payment flow can be the difference between a completed or abandoned payment. It can also impact the ratio of self-service payments compared to staff-assisted payments. For example, reducing the number of clicks a patient has to make to complete a payment can increase payment rates. It can also boost the number of self-service payments.

Tailor the Payment Experience

Additionally, consider the payment options offered to patients. Fine-tuning these options for patients based on their consumer data influences how many of them pay. Should you offer financing options to patients? How many of those patients should have access to a tailored financing plan?

Experiments, along with the data generated from machine learning, enable your Health System to answer those questions and ultimately, optimize the payment experience and target the right payment options to the right patients. In fact, one Patientco client saw a +10x increase in patient adoption of payment plans simply by adjusting payment plan thresholds.

Experimenting with your revenue cycle processes can seem overwhelming, especially without a data expert on hand. However, the smallest of changes can have a massive impact on your Health System’s bottom line.

With machine learning and a trusted partner, your organization can automatically test and re-test changes to the payment process. This, in turn, will help continually optimize your revenue cycle operations. If you can make the billing process as painless as possible, patients will remember that. This positive experience will set your healthcare organization up for more return visits. By doing so, you’ll transform your revenue cycle into an economic growth engine with more horsepower than ever before.

Download as a PDF