Healthie

Today, we’re talking about Healthie and how it’s being used at your company. Before that, could you give a brief overview of your company and your role there?

I’m one of the co-founders at a virtual clinic; we treat folks who have obesity, high blood pressure, high cholesterol, and diabetes. We do that through full-stack clinical care. We’ve got a provider group, do telemedicine consultations, order labs, and write prescriptions. To complement the clinical side, we provide lifestyle, i.e. fitness and nutrition support, including a dietician team on staff. I look after product, patient experience, as well as data, among other things.

Everything we do is in the fee-for-service paradigm.

When did you purchase Healthie and how long have you been using it in production?

We bought Healthie in Q4 of 2022. We had a bit of a jumpy start where we almost went with Elation but then changed our mind and chose Healthie. We started using it in production in Q1 of 2023, so have been using it for about six months.

I’m curious what use cases you’re using with Healthie. What are the various workflows they’re enabling?

Initially, we used Healthie forms for patient intake. Patients signed up through paid ads or provider referrals. They filled out a form, and scheduled a consultation with us, and then had a clinical consultation through Healthie’s Zoom features. This is all enabled through their patient app, or their client app, as they call it. We have since moved to FormSort for forms and use Morf to push the data into Healthie. At one point, we also used chat through the client app, but we’ve built our own client app that no longer uses Healthie chat. Besides forms and chat, there are some other features we don’t use. We don’t use programs or care plans. We don’t really use the charts or metrics, we do our own analysis in BigQuery; we have a data person who does that. We don’t use goals either.

We also use the Healthie API to push patient-generated health data into the journal feature. Our model requires us to send people connected devices (blood pressure monitors, scales, etc.) and our service ingests that data that patients record with the devices, and pushes it into Healthie using their API. We use a bunch of features related to the ‘actions’ tab within a patient profile – personal information, insurance, diagnoses, allergies, etc. On the provider side, they manage appointments and chart notes within Healthie. They use the charting templates to log clinical encounters. The task management feature helps them manage their day-to-day. And we also use documents for storing lab results.

In terms of integrations, we use the Dosespot integration for ordering medications. We struggled for a long time to get labs working with the Change Healthcare integration. We’ve since moved to a different provider for that. Although very recently, the Change Healthcare integration launched, and we’ve begun testing that. Finally, we are in the process but have yet to use the Candid integration. We’ve been seeing patients without billing, which you can do in a Medicare context and submit claims later. All signs suggest that that integration will work well.

Which vendors did you consider when choosing Healthie and how did you evaluate them?

We were ultimately deciding between Healthie, Athena, Elation, and Canvas. Those are the ones that made sense for our care model.

Our rubric included usability (clinician experience), compliance (specifically ONC certification for Medicare). When we made this decision, Medplum was not certified and we didn’t want to build everything at the beginning. Besides these factors, we wanted to know how good the support would be, because we were going to need to chat with customer service and support quite a lot.

These were the non-functionality elements we considered, in order of importance. We also accounted for some baseline considerations – cost, scalability, and the API.

In terms of functionality, visits and scheduling were really important for us because those are the bread and butter of a fee-for-service healthcare company. We wanted to understand how those would work, alongside lab ordering, prescribing, vitals (which Healthie calls journals), task management, patient communications, and billing.

Could you break down which of Healthie’s features are standouts, and which don’t work well for you?

If I step through each of the features I just mentioned, I think their concept for visits and appointments is pretty good. The main issue with it is that it only works with the Healthie app or through their API, which is not trivial to use. It’s far more complicated than we thought initially, based on our cursory assessment. It’s also very fragile, several production issues with pages and timeouts that cause us to worry about vital integrations. When we were contracting, the vision we had for Healthie was, “Wouldn’t it be great to have an EHR that we didn’t need to build, allowing us to focus on the patient experience and build the patient app?” What we ultimately realized is that there is no off-the-shelf EHR that does what we need it to do, so we were going to need to build our own one over time.

This is why we’re very interested in Medplum – you can do that, you do not have to do everything. We wanted to see how far we could get and how long we could go without having to do that because you buy what you can’t build initially. Eventually, you build everything when you become a successful company. With Healthie, first and foremost, we needed to get rid of their app because we were building a consumer brand and it just wasn’t going to cut it. We thought it would be easier to not use their app and interface with their scheduling, form, and chat features, and just rebuild the consumer app on our own. What we didn’t fully interrogate – and I think other vendors don’t have great solutions for this either – is that the clinician experience with some of those features (particularly chat) didn’t cut it. So, we ended up building our own separate app. Now that’s built on top of Medplum, where we manage chat. We’re now going to iteratively go through everything we do in Healthie and put it in Medplum instead, eventually turning Healthie off.

Having done this with chat and forms, we’re going to do tasks next. I’m not a provider, so take what I say with a grain of salt, but it’s pretty difficult to actually manage a complicated set of tasks for a large panel in Healthie. I don’t think our providers would be comfortable with us continuing to ramp up our scale with the quality of task management that exists with Healthie. We might then do the same with appointments, labs, or prescriptions.

It sounds like tasks are not scalable on the Healthie platform, for your providers.

I’ll add that this is all specific to our context. Different companies might have different experiences. For us, the provider experience for chat, for example, was a little janky. If someone in the provider team went away, it wasn’t easy to move and assign chats to another provider to pick up. Tasks had similar issues. Role-based access is the root of it – it affects tasks and messages. How would our operations work if someone on our team went on vacation? How would we manage the work that they had to do without giving others access to their accounts?

Other features had issues too. With prescribing in Dosespot, there were huge gaps in aspects of how Healthie and Dosespot spoke to one another. That made the integration there less useful than you would want it to be for things, specifically in terms of safety around medications.

Does this have to do with Prescription Drug Monitoring Programs (PDMP) or controlled substances? Or might the issue be something else?

No, it’s not about controlled substances, it’s more about the provider experience, and our belief that the product is going to deliver a safe and 10 star patient experience. There isn’t a proper two way sync between Dosespot and Healthie, which causes data inconsistencies (which can cause safety issues), there isn’t a working integration with Change Healthcare for labs, so we’re sending folks to Quest and Labcorp, and manually entering their labs into Healthie. Ultimately, we made the decision that the clinician experience is the number one lever in moving the patient experience. We don’t want our clinicians to quit, and the system isn’t working out for us because it’s not tightly integrated with the important parts of care, so there is a lot of manual work, the UI is quite buggy, and people get frustrated using it. All these little things could have been worked with, but in the long run, what is the most sustainable thing for us? We would have loved to go for longer – maybe a year more – using Healthie, while we focused on the consumer app. But we had to make the hard decision to not do that. So, we’ll probably continue to use Healthie until we’ve built the core elements of our provider experience into Medplum. It might take longer than we think, particularly with billing and charting, but that’s the plan.

What is the process of moving over from Healthie to Medplum? How are you porting things over, from a migration perspective?

We are going to stop using Healthie tasks and start using Medplum’s implementation of tasks that we’ve built. Messages have, from day one, been in Medplum. Data porting will be interesting, but Medplum has great documentation on how to do that. You have to figure out what in the previous system maps to which FHIR resources in Medplum. Then, you write these to-FHIR and from-FHIR API endpoints. You run batches to move all the data over. We haven’t thought deeply about the data migration, and we will also need to figure out the workflow migration.

For scheduling, we’re using Acuity, rather than Healthie now. Scheduling might become tough because of licensing, credentialing, etc. That’s going to be a lot more work than I think we give it credit for. Today, we have providers that are all licensed and credentialed in the same place, so it’s not a problem. But we will have to figure that out going forward.

It sounds like Healthie is not doing what it needs to for your providers. What made you go with Healthie as a provider-first tool?

I will say that Healthie is working today for what we needed it to do originally. We could actually have continued to use the client app to run our clinic. Some of the patient app features they have are great – people love them. From a consumer perspective, chat functions fine. It’s not like it’s entirely not working. We are running our business on it, and there are a good number of patients that are being seen in Healthie right now. But as we look to expand, we’re not confident it will be able to do that.

The way we made the decision was as follows. We looked at Elation, Athena, Canvas, Healthie, and Akute. We first dropped Akute because of missing certifications. I don’t think Akute had SOC2 certification at that point in time, and it didn’t have insurance-billing workflows. It felt like an interesting option, closer to Medplum, but we had to drop it. The next thing we did was cost modeling. We tried to assess what our usage would be, and how much it would cost us at 2,000, 10,000, and 100,000 patients on a monthly basis. Healthie was considerably cheaper. Once we had done that cost assessment, we reviewed the options again and dropped Canvas, which had a number of missing features. While Canvas had the most differentiated charting feature set from a provider’s perspective, it was pretty expensive and missed some key pieces of functionality that we needed. We also dropped Athena at this point. Athena and Elation felt like they were in the same vein and fairly different from Healthie, given Healthie’s origins and where they came from. But Athena is just super expensive. We also conducted a lot of reference checks and heard some negative things about Athena’s API. At that point, our clinical team also tested the different EHRs, and our legal counsel also reviewed the companies. Ultimately, it came down to Healthie and Elation.

Healthie felt far more usable but we didn’t factor into consideration that it was pretty buggy. We thought labs and meds in Healthie were going to be far better than they ended up being. Scalability-wise, Elation had the slight edge. We very nearly went with Elation. Finally, we chose Healthie for simplicity, ease of use, flexibility in partnership (openness to talking about changes and whether it felt like a partnership), and cost. In terms of functionality, we didn’t really think that there was any major difference that would have us pick one or the other.

Healthie has a big focus on integrating with different third parties and bringing them into the app. I’d love to get your perspective on the ecosystem as a whole.

I think I’m in the camp of a solution being more opinionated on workflows, which is why we’re going to build everything in Medplum. Philosophically, the platform ecosystem approach makes sense. It’s all about execution, though. With the issues with Dosespot, they’re working to fix them, but it’s time-consuming because their business is booming, and they’re getting pulled in so many directions. What that means is that we haven’t really used many of the elements in the marketplace. I think if it was like Shopify, where the platform and plugins were amazing, it would be great. But it’s never like that – and not just with Healthie. The process is like stitching together a Frankenstein, but I’m not sure what that ultimately does for the provider experience.

I think the bottom line of our Healthie assessment is that we thought the benefits outweighed the risks of a less mature EHR. We had planned to use it for some time and then revisit what we would do. We thought there was a possibility it could be great and it ended up being okay. We set out with very high expectations, but we also set ourselves up with the ability to pivot away, if we needed to, and that’s where we’re going.

From your perspective, what sort of providers or what sort of customers would be ideal for Healthie?

Healthie started off not being an EHR – it was primarily for non-clinical workflow management. So those customers, I’m sure, are finding it great because the journals, the metrics, and the charting notes would be pretty solid. If you didn’t want to build and own that end-to-end patient experience yourself and were looking for a patient app that worked seamlessly with the EHR, it would work for you. If you didn’t write medications or labs, or you did but in a pretty tight way, it would be great for you.