Behavioral Health Provider Finder

Anthem, Inc. operates a publicly accessible provider finder. This tool provides members with a platform to search for various types of medical providers that accept their health insurance plan. To provide additional guidance, Anthem supplements the provider finder with a behavioral health specialty call center. Over the course of the pandemic, the call center fielded about 29,000 calls per month—of which I managed about 350. After completing an initial assessment, clinicians, including myself, utilized Anthem’s in-house provider finder to refer members to the best mental health and/or substance use providers in our network.

Problem

Anthem’s provider finder inaccurately returned zero available providers and did not properly filter for specific search criteria. Members often became disoriented while searching for the specific providers they believed they needed to find. Because of this, Anthem’s call centers have struggled to meet increasing demand, driven by these confused self-service users. To combat this, Anthem implemented several strategies. One strategy was for call center consultants to use competitors’ provider finders to augment their results. Another strategy was to hire additional associates dedicated to contacting providers and manually identifying availability. It was clear that these strategies were not sustainable long term.

Proposed Solution

I proposed that we improve our provider finder’s results so that fewer self-service members would request assistance from the call center. This would limit the strain on the call center as well as provide members with an independent tool to guide their search. Feedback during use regarding search criteria and education about different provider services would further improve members’ health literacy. This literacy would translate to better utilization of the correct types of services limiting wasted time and effort.

My Role

Due to taking a considerable amount of member-facing calls and requests, I was able to hear firsthand how the issues impacted members finding providers on their own. As an expert in searching with workarounds and knowing the correct terminology to aid in the search, it was still considerably difficult to fulfill every request. I reached out to team members who also voiced similar concerns and issues as to the negative compounding impacts the system caused for multiple levels of users. I reached out to several product teams to discuss member feedback and suggestions for improvements.

Methods

My initial task was heuristic and expert evaluation of the existing systems utilized by our members. In creating user personas based on real-life use cases, I selected common search criteria to highlight that in high density areas, the system continued to identify 0 providers matching the criteria. I showcased this on several platforms including desktop and app use. As a subject matter expert, I explained to the product teams the intricacies of behavioral healthcare and terminology to demystify the jargon used. I highlighted how competitors implement member-facing tools like safety planning and resources as well as broader suggestions with clarification on differences.

Results and Deliverables

I presented suggestions to the product teams that strongly focused on psychoeducation–providing information and resources to those seeking mental/behavioral health services. First, I identified dead–end searches for common criteria that provided little to no feedback or guidance. While using an AI interface within the in-house app, I demonstrated how the interface failed to identify safety plan measures for at risk members–those suffering possible acute symptoms such as suicidal ideation. The culmination of these issues resulted in the continual abandonment and negative feedback similar to what members were expressing.

Future Outlook

If searches provided additional feedback, members would have greater ability to select providers matching their needs, such as medication vs psychotherapy, and receive that care more readily. Members would then improve their own health literacy and have a sense of independence while using the tool. This would reduce the burden on the call center. Without the additional strain, call metrics would improve in areas such as faster answering times, more focus on acute cases, and greater member attention. While no immediate steps were taken, stakeholders did agree that there are areas of needed improvement across web and app usage. If searches provided additional feedback, members would have greater ability to select providers matching their needs, such as medication vs psychotherapy, and receive that care more readily. Members would then improve their own health literacy and have a sense of independence while using the tool. This would reduce the burden on the call center. Without the additional strain, call metrics would improve in areas such as faster answering times, more focus on acute cases, and greater member attention. While no immediate steps were taken, stakeholders did agree that there are areas of needed improvement across web and app usage.