Sohar Health

Today, we’re chatting about Sohar and how it’s used at your company. Before we begin, could you give a brief overview of the company and your role there?

I’m the CTO of a virtual behavioral healthcare company. My role encompasses overseeing product and engineering as well as managing traditional IT, clinical informatics, and data. Our company specializes in offering telehealth services to patients for psychiatry and therapy.

What was the need that drove you to look for a product like Sohar?

We needed to improve our manual process for verifying benefits, which had led to delays and inaccuracies. There was a lot of lead time before new patients could be seen, and there were frequent misunderstandings of patient financial responsibilities, as well as unexpected insurance issues that cropped up right before appointments.

When you were evaluating different vendors, what were some of the requirements you had in mind?

We prioritized breadth of payer support and accuracy. It was important to have a solution with a manual fallback option for instances when automatic resolution wasn’t possible. This would ensure the vendor, not our team, managed manual processes. We also sought a fast turnaround time and strong service-level agreements covering uptime and accuracy.

Were there other vendors that you looked at, and how did they compare?

We evaluated several vendors. One was Nirvana Health, which offered a concept similar to what we needed, but their solution didn’t fit our use case perfectly, and their API wasn’t as compatible with our system as Sohar’s. We considered using Change Healthcare or Availity, but the integration seemed too labor intensive, given that this isn’t a core product for us—it’s necessary but not a differentiating factor that justifies significant engineering resources. Additionally, we examined Candid’s built-in benefits verification, but it seemed more like a wrapper over Change. It would have either depended on Candid prioritizing it, or it would’ve required us to invest in development as if we were working directly with Change Healthcare. We also looked at another vendor, but they only serviced Medicaid, which doesn’t apply to our practice.

How was the sales process with Sohar?

It was straightforward. We engaged in extensive negotiations to tailor their product to our specific needs. Sohar was collaborative and accessible throughout the process. They accommodated our request to adjust the pricing model to per active patient per month, which fit better into our budgeting framework than the standard per-API-call model.

How was the onboarding and setup process?

It involved close collaboration between our engineering product teams and their CTO. They addressed gaps in our initial specifications, tailoring the product to our requirements. We had competing priorities internally, which impacted the smoothness of the onboarding process, but Sohar proved flexible and responsive when their implementation was our top priority. The integration into our systems was extensive, but we gained initial access to Sohar within a day. Our team could manually run checks by logging into Sohar’s interface and entering information for automatic or manual verifications through Sohar’s portal or support team as needed.

You mentioned you didn’t go with another solution because it was basically a wrapper on top of basic eligibility check requests. What’s the incremental product lift that Sohar is providing here?

Sohar provides significant value in handling the intricacies of eligibility and benefits verification in behavioral health. They have expert knowledge of the X12 service codes for each payer and understand the specific codes required by different plans. This is particularly crucial because behavioral health benefits mapping is complex, with various carve-outs and benefit managers.

When we refer a patient to a healthcare provider, we focus on ensuring the provider will be covered. We have a solid understanding of which CPT codes to bill based on the patient’s encounter. Importantly, Sohar takes a strategic approach in dealing with payers. Even though a provider may be scheduled to see the patient, Sohar knows that simply stating this may not yield precise benefit information. They adjust the data to obtain the most accurate benefit details, drawing on our established collaboration that defines the care we provide. Sohar identifies potentially relevant service codes and, through experience, discerns which code is most accurate for each payer and plan in order to secure the correct benefit information up front.

Do you know if it’s a very complex rules engine, if they’re making a lot of phone calls, or if it’s a machine-learning architecture?

The specifics of their architecture haven’t been detailed to me, but it’s likely that they will integrate a machine-learning component eventually due to the high volume of data involved. Initially, there was a need for numerous phone calls as they were establishing their information base with historical data provided by us, but it appears that the volume of calls has significantly decreased. This data management and acquisition is precisely why we engaged their services.

Do you know where they use the automated solution versus where they require humans to make a phone call or visit a portal?

It depends on whether the payer can provide accurate benefits information through an API. If they can, then automation is used. For payers and plans that don’t support this, a manual approach is necessary. Response times vary, and we have service-level agreements in place regarding latency.

How would you characterize strengths and weaknesses of a product like Sohar?

The strengths of a product like Sohar hinge on its accuracy in providing patients with reliable information, its speed in delivering results, and cost efficiency regarding engineering resources for updates. It’s critical for the system to maintain high-percentage accuracy to ensure that the vast majority of patients receive correct information. Moreover, keeping up with the annual changes instigated by payers, although the specific changes are not detailed, challenges the system to be agile and responsive, which speaks to the robustness of its updating mechanisms.

How does Sohar perform?

They typically perform well, although performance can vary by payer. Accuracy issues sometimes arise with certain payers providing incorrect information, a trend we’ve also observed with our manual phone call data. To address this, Sohar has integrated with Candid to ingest claims data from adjudication, identifying discrepancies and adjusting the interpretation of payer responses. We opted for direct integration between Sohar and Candid, eliminating the need for our engineers to create a separate integration.

Have you found the platform to be pretty reliable?

Yes, the platform has been very reliable. Even when we accidentally sent way more queries than usual in a short period of time, it didn’t go down at all.

How would you characterize the developer experience and documentation?

The developer experience with Sohar’s API, despite some gaps in the documentation, has been positive overall. Direct support was available when needed. Implementation was slightly challenging since Sohar supports both automated and manual checks, and we needed a single flow. However, the asynchronous webhook API, while less common in healthcare and slightly more complex, better met our requirements and ultimately provided a satisfactory developer experience.

Do you have any tactical advice for folks who want to build Sohar into their platform?

It’s essential to maintain a tight scope. Even for a seemingly straightforward task like verification of benefits, the application can quickly expand to include various features, such as patient responsibility information and proactive patient notifications. The key is to have a clear definition of the minimum viable product. Unlike typical third-party single-API implementations, which often involve a single call, this process can easily become complex with potential integrations across numerous systems. I’d suggest resisting that temptation and focusing on the core functionality initially.

How was the integration with Candid?

It was pretty easy. The main aspect revolved around a three-way negotiation concerning access levels, as we needed mutual data sharing between two third parties. We discussed and agreed upon what we would approve and what was technically possible. My engineering team wasn’t required to participate; my involvement was limited to writing policies and the approval process.

How is the account management and support from Sohar?

Our experience has been quite positive. We’ve mainly dealt with their leadership team, which has provided excellent attention to our needs. Recently, an account manager was assigned specifically to our account, but our engagement dates back to before this addition, so the CTO still regularly attends these calls. Our longstanding relationship might play a role in this level of service, and it could differ for newer clients.

Do you feel like you made the correct decision in choosing Sohar?

Yes, I believe it was. They almost entirely meet my needs. My primary goal was to make health insurance straightforward. While they don’t completely solve this issue, they greatly simplify the process of verifying eligibility, which enables us to do automatic checks.

Do you see any areas of growth for Sohar?

We’ve been talking with them about insurance discovery, which has been tested and shows promise, especially for practices dealing with frequent insurance inaccuracies. While not directly impacting our practice, the possibility of expanding Sohar’s services to areas beyond behavioral health could prove beneficial. Sohar being able to address the complexities of behavioral health billing suggests they could handle physical health billing as well.

Do you have any advice for buyers who are currently thinking through eligibility/verification of benefit solutions?

It’s essential to understand your minimum viable product requirements and the pricing model. Be aware that low per-call fees can add up, especially if each patient requires multiple calls. Additionally, consider the necessity of manual checks. For some practices, manual checks are important, but for others, especially if you have a high accuracy rate with a large insurer, they may not be necessary and could be an unnecessary expense. Conversely, for patients insured by smaller insurers where automated data may not be accurate, investing in manual verification could be highly beneficial. This varies depending on your practice’s specific needs.