Details
About the Reviewer
Reviewer Organization
Reviewer Tech Stack
Other Products Considered
Summary
Product Usage: Nabla is used as a transcription tool for virtual conversations with patients in order to generate organized notes for medical records.
Strengths: The tool is efficient in transcribing patient-doctor conversations and automatically transforming them into well-structured clinical notes, ultimately saving time for clinicians.
Weaknesses: AI limitations mean Nabla cannot perform a mental status exam without human input, nor generate the assessment and plan section unless the clinician visibly dictates it.
Overall Judgment: Despite some technological limitations, Nabla provides significant time advantages and boosts conversation focus with patients, thereby improving patient care.
Review
So today we’re chatting about Nabla and how it’s used at your company. Before we jump into that, could you give a brief overview of the company and your role there?
Sure. We are a virtual psychiatry and addiction care practice. Currently, we’re dedicated to serving our home state, with a specific mission of being the go-to telehealth option for mental health and addiction support in rural and frontier communities within this region. However, we’re eager to broaden our reach to other states within the next six months. My role is co-founder and CEO of the company, and I’m also a part-time clinician.
What drove your need for a product like Nabla?
We weren’t actively looking for a product like Nabla. We were already familiar with AI and had been using ChatGPT for other purposes, such as generating patient letters and documents. One of us stumbled upon a post on LinkedIn about Nabla, so we decided to look it up. We discovered that it was easy to try out with a free Chrome plugin and started using it to complete some notes. Since we were able to explore it on our own in a free version, test it, and play around with it, it was an easy decision. We eventually reached out to Nabla to inquire about the cost, and when we found out the pricing, it was quite affordable compared to similar products we were aware of.
What were the key requirements you used to evaluate Nabla and its competitors?
We’re a startup, three and a half years in, operating on a bootstrap budget. Cost effectiveness is a top priority for us. Some other products out there are just too expensive, costing up to a thousand bucks a month per provider, which is completely out of our range. So we focused on finding an affordable option that would meet our needs.
Initially, when we started using Nabla, it was a bit raw. However, the Nabla team was fantastic in their responsiveness. Whenever we reached out to them with our requirements and suggestions for change, they were quick to address them. This level of engagement kept us interested in their product. If we had to pay for the initial version, we probably would have passed on it. But Nabla’s ongoing commitment to improving their product showed us that they genuinely wanted to work on it. Their responsiveness, user-friendliness, and enthusiasm all played a part in keeping us engaged. If it had been a big, unresponsive company with a fixed price, it’s likely we would have moved on. But Nabla’s team seemed like genuinely great people, passionate about their work and dedicated to doing the right thing. The positive nature of our interactions with them carried us forward in our partnership.
Were there any additional vendors you evaluated once you did look deeper into Nabla?
Yeah, we briefly considered other options like the Microsoft/Nuance combo, but their pricing was ridiculously high. We also looked into DeepScribe, but their process of using your phone first and then uploading and transcribing had too many steps and was not cost-effective for us. We made a few phone calls to them, but either the cost was way off or the user experience was not up to par. That’s why we ultimately decided to stick with Nabla.
How would you characterize the sales process with Nabla?
The sales process was great, not pushy or complicated. We didn’t require much selling because we approached them as soon as we realized how user-friendly it was. We tested the free version for about a month before making the purchase.
How was the onboarding and setup process?
Nabla is a super easy Chrome plugin. It’s user-friendly and doesn’t have any unnecessary barriers. Unlike other healthcare software companies, it doesn’t require multiple meetings, and they don’t try to sell you on something you don’t want. They’re transparent and easy to work with.
When I started using the free version of Nabla, there were probably just two of us using it. I ended up being the most active user, using it for about five hours a week for four weeks before deciding to buy. During that time, I had interactions with their team and gave them feedback on how they could improve, especially in terms of language for mental health. They were very responsive and easy to engage with.
The software automatically updates, and with the plugins, I don’t have to worry about manually updating anything. I can see that there have been improvements over time, but I don’t have to handle any updates myself. I simply click on the icon in the upper right-hand corner of my Chrome, and it launches without any worries.
What is the use case you’re using Nabla for?
You can use Nabla for virtual visits or in-person visits. I haven’t tried it for in-person visits, as we’re a telehealth company. There are two types of encounters, initial encounters and follow-ups, and I use Nabla for both. They have different buckets or frameworks for specialties like psychiatry, psychology, and internal medicine. This ensures that Nabla understands the technical language for each specialty. They have also developed different note structures. In medicine, different people use different structures, like simple SOAP notes or longer notes that break out histories and more details. You can easily pick which one you want to use. There have been times when Nabla didn’t capture enough information, and other times when it captured too much, but now it captures just the right amount of relevant information for the note. They have definitely improved in that aspect.
What are some of Nabla’s strengths and weaknesses?
In terms of strengths, I find it impressive how Nabla can take a conversation transcript and turn it into a note. It captures the raw data of the conversation efficiently. Those who use Nabla are highly enthusiastic about it and amazed by the amount of time it saves them. Surprisingly, there are some benefits I hadn’t considered until I started using it myself. First, I can be more focused on interacting with the patient during our conversation, as I no longer have to worry about taking notes. I can have this conversation knowing that it’s all being recorded and will be included in the documentation. This allows me to be fully present with the patient. Second, in a 30-minute appointment, I used to rush toward the end to ensure I had enough time to take down all the notes myself and compose the entire note. Now, I no longer have to do that. Consequently, I’m able to engage with the patient for longer periods while remaining present with the patient. Over time, I believe this has a positive cumulative effect on patient care as well.
The main weaknesses I see are more related to technology than to Nabla specifically. In mental health, AI is not yet able to perform a mental status exam accurately without being able to consider facial expressions, tone of voice, and other nonverbal cues. This means that the exam still needs to be done by a human. Additionally, the assessment and plan section of a clinical note, which involves formulating the case and outlining the patient’s condition and plan, cannot be generated by Nabla unless the clinician spells it out verbally. I may not explicitly share my thoughts with the patient during the conversation, but I include them as an addendum when creating the note later.
There are some workarounds for that weakness, though. One option is to keep Nabla open after ending the call and vocalize my thoughts then. The other option is to simply type my thoughts in an addendum. We have discussed with the Nabla team how the product handles certain aspects of the mental status exam, such as noting the absence of any mention of suicide. However, it cannot determine the patient’s mood or affect without it being explicitly mentioned.
Overall, I’m very satisfied with what Nabla offers, as it saves me approximately five minutes per patient, which adds up to a lot of time. I only need to spend around two minutes adding additional notes.
Can you describe the workflow you use for Nabla?
Having a tool like this, separate from the EHR, provides more flexibility in my workflow. Here’s how it typically goes: I open my EHR and my note, then click on a button to launch the telehealth visit or video. When the patient appears on video, I open Nabla and start transcription. As I have the conversation, it transcribes in real-time. I can choose to see the transcription or minimize the window. The transcription labels the speaker as “Patient” or “You.” Whether the visit is short or long, the entire conversation is transcribed.
Once the patient hangs up, I click “stop transcribing” on the Nabla window. Then I click “generate note,” which takes about 10 seconds. It turns the transcription into a well-organized clinical note. You can choose the structure you prefer, and it’s easy to copy the note with a button click. It automatically puts it on the clipboard, allowing me to paste it directly into my EHR field.
I prefer this as a plugin rather than integrated because sometimes the video doesn’t work and I have to call the patient using an online voice-over-internet phone call, like Doximity’s phone function. I can continue the transcription over the phone and seamlessly switch between different communication tools. When I switch windows, Nabla asks if I want to continue the transcription there. It allows me to maintain the same transcription regardless of the communication tool I’m using. If I want to initiate a note with a phone call, I can also do that using an online phone tool. This has been helpful when the patient’s camera doesn’t work or their signal is weak.
From what I’ve observed, Nabla relies on two sources of input: the signal from the remote location and the microphone on my computer. It grabs information from wherever the computer detects input and output.
Are there some of your clinicians who are not using Nabla?
We try to give clinicians as much autonomy as possible, allowing them to use Nabla in their own preferred way. We have over 20 providers, and we were surprised that not all of them immediately adopted it. Some individuals feel uneasy about AI and have concerns about where the information goes and who has access to it. They worry about their own privacy and that of their patients. We respect that. On the other hand, there are those who feel comfortable with using Nabla and appreciate the time-saving benefits it offers. They prioritize efficiency over needing to have complete control. This variation in attitudes toward new technology is common, with some people being early adopters, some falling in the middle, and others resisting change altogether. With time, like with EHRs, we anticipate that more of our providers will embrace Nabla as they witness the time it saves for those already using it.
How has Nabla responded to your feedback? How would you characterize their account management and support team?
We gave feedback to the Nabla team, and they were very responsive. One aspect of the feedback was about the structure of the notes. We discussed what information should be included and where it should be placed. Essentially, we wanted a template from Nabla that would help organize the information effectively. They were very responsive to this and made the necessary adjustments.
Another aspect of the feedback was about the amount of information included in the notes. It was important to filter out unnecessary details and only include the most relevant information. For example, if a patient is depressed and their dog dies, we don’t need to know the dog’s name, breed, and color. The key information for us is that the dog’s death worsened the patient’s depression. Nabla was receptive to this feedback, and they’ve gotten very good at including only the clinically relevant information in the notes.
From a reliability and stability standpoint, have you faced any bugs or issues with Nabla?
In the beginning, it was taking Nabla a really long time to generate the note. It would sometimes take 30-60 seconds, and providers would get impatient. Sometimes it wouldn’t even generate the note and it would freeze. This was all within the first month of using it. But now, I haven’t experienced any of those problems. The note generation only takes about 10 seconds now.
How is the UI of the plugin?
It’s not fancy, but I like it. It’s simple and doesn’t have extra features, which is what makes it nice to use. It has everything it needs and nothing more.
Do you think you made the right decision in moving forward with Nabla?
Yes, I’m happy we made the decision. It has worked out well and saved our clinicians time. The clinicians who use it are satisfied with the system, and I believe it has actually improved patient care in those cases.
Are there any areas for growth you see for Nabla?
I think other companies are also exploring this idea, but one interesting possibility is, after generating the notes, suggesting CPT codes based on the transcript. Since the necessary information is already present in the transcript, it seems feasible. This upfront coding assistance could be quite helpful, and it shouldn’t require much extra work.
Do you have any advice for someone else who is looking into making a decision about a product like this?
I would suggest including additional clinicians on the team and in the decision-making process. This will make it easier to get buy-in up front. Instead of having only leadership making decisions and telling everyone, “We’re using this product,” it’s better to get input from frontline users. That’s a good approach for any product, try to involve frontline users in the decision-making process from the beginning.