Tag Archives: Voice Dictation

Signify Premium Insight: From ‘Base Camp’, Nuance Readies for Assault on Fully Ambient Solution Peak

Nuance Communications’ recent announcement of its new AI-automated clinical documentation app is the latest in a slew of innovations from the Microsoft-backed company. 

Nuance describes DAX Express as a ‘workflow integrated, fully automated clinical documentation application’, adding that it is the first solution combining ‘proven conversational and ambient AI with the newest and most capable generative AI technology, GPT4.’ 

In a voice dictation market that has often failed to live up to hype and (unrealistic) expectation, DAX Express potentially brings the holy grail of a ‘fully ambient solution’ one step closer.  

The Signify View 

DAX Express nudges the voice dictation market further along the track. Nuance now occupies a metaphorical ‘base camp’, the peak of fully ambient solutions now tantalisingly in view. Nuance occupies this rarefied space alone, its competitors still acclimatising to the relentless pace of the race.  

Nuance says DAX Express will generate draft clinical notes ‘automatically and securely in seconds from exam room or telehealth patient conversations for immediate clinical review after each patient visit’. It adds that clinicians ‘will benefit from the seamless capabilities of Dragon Medical One, DAX and DAX Express, which are tightly integrated with EHRs, beginning from pre-visit through post-encounter, reducing cognitive burdens and addressing staff shortages’. 

If true, it will be music to the ears of a healthcare industry for whom the reality of voice dictation in the EHR workflow has yet to scale serious heights. Despite rapid technological advances and the promise that innovations like DAX Express hold in transforming EHR workflows, Nuance (and its rivals) must still convince sceptics. And there are many. 

Trust Issues 

Buyers and users Signify Research speaks to confirm that many clinicians still manually type up their own notes in the absence of what they see as viable alternatives. Some complain of ‘unacceptable’ delays between voice entry and order activation. Others state that Nuance’s Dragon software ‘isn’t strong enough inside Epic’ (a reference to Dragon’s integration in Epic’s EHR). 

Most voice dictation technology still requires substantial manual input, either by the physician or transcription service providers. Interviewees say automated transcription quality can be poor, and even when it is good, the transcript must still be ‘translated’ by a human into terminology that the EHR can decipher, and then manually coded into the EHR’s structured data fields. 

Hospital buyers say many staff will start using voice dictation tools with good intentions, but quickly fall back into familiar workflow habits and established methods, due to disappointment with solutions. And some clinicians are also reluctant to do away with medical secretaries, whose job is much more than simply typing up clinician notes. 

A fully ambient solution must not only address these barriers but also accurately capture the patient-physician conversation, and then produce ICD-10 and SNOMED codes for diagnoses, prescriptions and reimbursement claims. The real value here lies in turning ‘unstructured’ data from an encounter into ‘structured’ data in the EHR. No solution does that yet, although Nuance DAX brings it closer to fruition. 

Mission to Convert 

If anyone can convert the doubters and bludgeon their way through the barriers to acceptance of voice dictation solutions, Nuance is the strongest contender.  

Backed by Microsoft, it has the financial and technological muscle to bring sophisticated solutions to market. The company also picks its partners well – Nuance DAX is integrated with world-leading EHR vendors like Epic and Oracle Cerner and their massive installed bases. These relationships run deep: Seth Hain, a Senior Vice President of Research and Development at Epic recently said GPT4 had ‘shown tremendous potential for its use in healthcare, and it will be used to help physicians and nurses spend less time at the keyboard and help them investigate data in more conventional, easy-to-use ways.’ Such ringing endorsements reinforce Nuance’s influence and, once it comes to market, DAX Express will enter the field of play on very solid ground. 

Much Ado About Nothing? 

We wrote very recently about the prospects for AI-powered start-ups (see this Insight) in this market. We concluded that they would struggle to compete with Nuance or other leading vendors like 3M and Dolbey. Given Nuance’s snowballing dominance, perhaps the more pertinent question now is how much more of a wedge will DAX Express drive between Nuance and its nearest rivals?  

Much will depend on whether these rivals can leverage technology from the likes of Microsoft and Google via integrations, but that route is complicated as it would require using underlying technology from some of their biggest competitors. 

It will take something momentous to slow Nuance’s momentum. Having said that, DAX Express will not make its commercial debut until 2025, and that leaves the door slightly ajar for others to make their play. We’re also basing this assessment on the promise offered by DAX Express, which is partly based on the company’s marketing and the disruption, excitement and, in some cases, alarm that ChatGPT has created over the last couple of months.   

Could Google be one of the ‘others’ we refer to above? Not to be outdone by Microsoft/Nuance, it launched Bard, its answer to ChatGPT, just days after the DAX news. Google is actively developing products, via CareStudio and other initiatives, to try to streamline clinical note interpretation in the EHR, which it hopes can challenge Nuance in this market. Elements of CareStudio have now been integrated into MEDITECH’s EHR solution, although to-date this has focused on search functionality. However, this provides a potential springboard for Google that could challenge Nuance. Although not confirmed, it is reasonable to assume that Bard will be integrated into Care Studio if Google continues its strategy of enhancing the smart notes functionality of CareStudio. However, Google has only a limited installed base of legacy solutions and any disruptive influence from it, or other big tech, on Nuance and the voice dictation market at large, will be minimal for now. 

Dollars and Sense 

Although buyers and users we speak to are almost unanimous in their belief in the potential of ambient voice recognition, the technology will remain a relatively expensive ‘luxury’ for the foreseeable future, out of reach of many smaller hospitals and primary care practices. While there are myriads more price sensitive marketplace App Store solutions for bolt-on to an EHR, nothing beats the power of EHR integration and the ability to capture unstructured data. Deep integrations with companies such as Epic, and its ‘Hey Epic’ functionality, offer this and strengthen Nuance’s competitive position on this front.  

Cautious Optimism 

Even as DAX Express gears up to enter the market in 2025, and despite the impressive strides Nuance has made in the market to date, sentiment in the wider voice dictation market in 2023 can be best described as cautiously optimistic. 

There is an inevitability about the journey towards fully ambient.  Even those buyers and users that Signify Research has interviewed on this topic via its Decision Maker Research, whose experiences with voice solutions have been less positive to date, almost unanimously acknowledge the technology’s emerging capabilities – whether that’s ambient listening or voice activated querying and searching – or, its benefits in reducing physician burnout. 

Amid the chatter (pun intended!) surrounding voice dictation, expectations were raised to unrealistic levels. The healthcare industry must shoulder some of the blame for falling for the hype and succumbing to peer pressure. 

But perhaps, at last, the hype and hyperbole are justified. In DAX Express and GPT4 technology, Nuance is considerably upping the ante. As it waits at ‘base camp’ ready for its assault to a fully ambient solution, it is proving that, in this case at least, words are speaking as loudly as actions.