Using Speech Analytics to Increase Positive Healthcare Outcomes

AI and machine learning can help turn call-center conversations into actionable improvement strategies

Published: Tuesday, January 17, 2023 - 13:02

Listening to customers is critical for healthcare organizations to ensure they’re delivering high-quality care to their patients. Sure, the traditional methodology of doing so via surveys can increase customer retention and profitability. But much like evolving from analog to digital, there’s a better way to listen to patients.

Surveys don’t work. The responses nearly always lack context, and asking an already frustrated patient to answer a few questions skews the results.

Conversational intelligence—that is, using speech analytics to listen at scale—tells more than any written survey and is a must for all patient-focused healthcare organizations. While conversations happen in many places throughout a healthcare organization, contact centers already record customer conversations, making them the perfect place to apply conversational intelligence.

Let’s dig deeper into three ways conversational intelligence enables healthcare organizations—especially their contact centers—to use patients’ voices in the boardroom to increase positive outcomes.

Supplement survey information 

Unsolicited, raw feedback can supplement survey data, which is vital to highlighting social determinants of health (SDOH) that affect a patient. SDOH is defined by the U.S. Department of Health and Human Services as “conditions in the places where people live, learn, work, and play that affect a wide range of health and quality-of-life risks and outcomes.” While a survey can tell you if a patient had a satisfactory or unsatisfactory experience, you’re still missing the “why”—the SDOH—unless the patient chooses to add context.

Daily conversations in contact centers offer a powerful resource to unleash that valuable context. These conversations include Q&A and often an unsolicited story from the caller. Those shared stories provide the full story and background of patient experiences. If leaders can access that information, they can establish more effective remediation plans and build health equity.

But these valuable, qualitative data are nearly impossible to process and analyze using conventional data methods and tools. Artificial intelligence (AI), including natural language processing (NLP) and machine learning (ML) tools, converts unstructured information into meaningful, readable data.

In basic terms, teams input key target phrases and sentiments they’re interested in tracking. Audio conversations are converted to text, and the ML classifies the text by intent, sentiment, and topic to extract the specified information.

The AI can then correlate key phrases with specific contexts. An organization that wants to know whether its customers are frustrated with a new appointment-scheduling process would input keywords like “confusing,” “didn’t understand,” or “wasn’t sure what number to push,” to see how many people find the new system challenging. This approach allows organizations to get more granular data than what surveys typically capture.

Health systems continue prioritizing SDOH, but typical patient surveys can’t adequately capture these data. Conversational intelligence can. And healthcare organizations need these insights to deliver a better patient experience.

Take, for example, one organization hoping to refine its appointments and scheduling processes. SDOH factors, such as lack of transportation, affected patients’ ability to make appointments. One person—or even a team—couldn’t process thousands of interactions to find specific SDOH. But AI and ML could.

The organization used this technology to process 24,000 interactions and individually evaluate nearly 2,000, also using AI, to pinpoint specific SDOH affecting patients’ access to appointments or scheduling. Almost 10 percent of the processed calls identified patient concerns about job loss, cost barriers, and a lack of reliable transportation as top roadblocks and frustrations.

These data highlighted barriers preventing patients from arriving on time for their appointments. Thanks to these insights, leadership took steps to:
• Empower their call center teams to share patient information with Social Services, providing more holistic support for their patient population
• Offer opportunities for research and tracking to provide an authentic, measurable effect on vulnerable patient populations within the community
• Reassess training and resources to help referral coordinators better assist patients with identified SDOH

Unlock customers’ unsolicited feedback 

Surveys source feedback from patients that are either motivated or feel obligated to provide it. Typically, these patients have had either a terrific or terrible experience, or they are going through the motions to complete the survey. This type of solicited input just can’t paint a complete picture of what it’s been created to measure—including the experience throughout a patient’s entire healthcare journey.

Unsolicited feedback gathered by call centers adds an insightful look into what’s top of mind for a patient because they’re not just responding to questions you’ve already asked. Patients aren’t responding to questions in a survey. They’re sharing their challenges, identifying road bumps, and telling you what they want you to hear. That insight is huge because it identifies those pain points they’re experiencing. Once leaders understand those pain points, they can develop strategies to address and mitigate them.

Here’s a typical scenario: A pharmacy updates its phone menu, but the new directions confuse one of the elderly patients who’s more comfortable with how things used to be. She calls her healthcare provider because she isn’t sure how to navigate the new menu and wants to express her frustration. The patient explains that she’s struggling to understand what number to push to refill her prescription because there seem to be too many new steps.

As it turns out, she isn’t the only patient unhappy with the new system. In fact, AI and ML have captured data from hundreds of calls and created an aggregated report that suggests other patients representing multiple demographics are confused by the update, even if they didn’t use exactly the same words to describe the experience. Leadership receives this information and raises concerns with the proper teams that tweak the system update.

This is just one example showing how unsolicited feedback—assisted by AI and ML—holds tremendous value for healthcare organizations striving to improve their customer experience.

Develop more informed action plans

We can all agree that maintaining the status quo and opting for the path of least resistance is usually the easiest approach. But is it the best strategy? Probably not. Managing, analyzing, and gathering insights manually is impossible in an industry generating 30 percent of the world’s data volume. Yet without data, healthcare organizations can’t make decisions or develop accurate action plans.

Fortunately, AI and ML play a key role in unlocking insights in unstructured or conversational data, which comprise about 80 percent of data collected in the healthcare industry. This technology structures conversations via sorting and topic identification to tease out relevant data and make it actionable.  Once the information lives in one centralized database, it becomes easier to analyze and build more accurate models as more data are added over time. It’s impossible for a person—or team of people—to run analytics at the required scale and speed without AI and ML technology.

The technology also helps organizations handle unstructured healthcare data more effectively by:
• Optimizing data storage
• Identifying and eliminating unnecessary or redundant data
• Classifying data
• Maintaining data confidentiality
• Assigning access levels

By elevating patient voices and authenticity at scale, leaders are empowered to strive for excellence: developing more accurate, informed action plans for improving customer satisfaction; resolving pain points along the healthcare journey; and identifying areas of improvement and training for call center agents.

About The Author

Amy Brown

Amy Brown is the founder and CEO of Authenticx, a software platform that analyzes and activates patients’ voices at scale to reveal transformational opportunities in healthcare.

Amy built her career as a rising executive in the healthcare industry, during which time she advocated for underserved populations, led and mobilized teams to expand healthcare coverage to thousands of Indiana residents, and learned the nuances of corporate operations.

In 2018, Amy decided to leverage her decades of industry experience to tackle healthcare through technology. She founded Authenticx with the mission to bring the authentic voice of the patient into the boardroom and increase positive healthcare outcomes.