For the eleventh year, Med Ad News has chosen new Pharmaceutical Marketing Ventures to Watch that could change the way pharmaceutical products are marketed and sold.
This past October, the Med Ad News staff began its annual search for the future of pharmaceutical marketing. We sought out young companies, spin-offs, offerings, and ventures to profile that are providing the most innovative and interesting products, services, or marketing opportunities to pharmaceutical companies and the healthcare community. This year’s three profilees are all exercises in technology with a human touch – a comprehensive AI engine built for life sciences that began its life as a way to win at Jeopardy, a digital platform to connect IBS patients with a difficult-to-find therapy, and a partnership bringing together creative experience with data tools to more quickly diagnose and treat patients with rare diseases. Here are Med Ad News’ newest Pharmaceutical Marketing Ventures to Watch.
Cognitive Core is Intouch Group’s artificial intelligence engine, custom-built for pharma. The offering powers a variety of AI activities for Intouch clients – chatbots, patient adherence programs, Veeva digital sales aid rep interactions – and that list is rapidly growing as brand managers seek out new ways to harness AI.
“Cognitive Core started out as an experiment back in 2012,” says Abidur Rahman, Intouch’s senior director of innovation and new technology development. “The experiment was to prove that robust AI systems could be built without massive super-computers like IBM Watson. The news of Watson competing on Jeopardy had just come out. In fact, the first version of Cognitive Core was named ‘Sherlock,’ and its primary capability was playing Jeopardy. I led a small team of developers to build a working prototype of Sherlock during a 24-hour hackathon, where Sherlock won the grand prize.”
With that win in the team’s pocket, they began looking for ways to make the technology relevant to pharma clients. “An opportunity presented itself when one of our clients became interested in creating an AI-based virtual assistant,” Rahman says. “We went back to the drawing board and discovered that when it comes to implementing natural language processing (NLP) and intelligent agents such as chatbots in pharma, existing technologies and platforms were not adequate. So, we took the lessons from Sherlock and built a pharma-centric AI platform that could be trained on pharma-specific data and used to provide NLP, predictive intelligence, content analysis for adverse events, pharma-centric workflow automation, and much more.”
In its first iteration, used to accompany a new drug launch in 2012, Intouch’s AI could already take typed text questions from an online form, deconstruct them into every possible variation, evaluate its known knowledge base for the set of content predicted to have the highest probability of answering the question, and deliver the result back to the user, all in a matter of seconds.
“Although complex in the behind-the-scenes functionality, the first iteration appeared to be very simple in its implementation,” says David Windhausen, executive VP, Intouch Solutions. “On the surface, it appeared to be just an FAQ section on a website. But behind the scenes, it was utilizing machine learning to understand a vast knowledge base of content; natural language processing to parse through all of the variants of a question; and predictive analytics to ascertain the best response to give back to the user. All the while using the ‘language of pharma’ to communicate about the disease state and to understand and track any instances of an adverse event.”
Since its initial launch, Intouch leaders say the platform has undergone major areas of advancement in the type of content it supports, the number of data sources that can be part of its knowledge base, the number of systems it can be integrated with to track user interactions, the methods used to train the machine learning aspect of the system, its own natural language processing engine, and the types of end-user experiences that can be created. Today, Cognitive Core has the ability to integrate with patient support solutions to track information regarding its interactions, Salesforce systems like Veeva to help direct the actions of sales reps in a manner that predicts the needs of HCPs, and a multitude of user experiences, from online banners to zero-UI interfaces such as Alexa, Siri, and Google Assistant.
Of course, building a machine learning-based system like Cognitive Core specifically for pharma requires a thorough understanding of how content is created and regulated in the space, the data security and privacy constraints, and empathy for how end-users will interact with this system to address their healthcare questions and concerns.
“To address the needs for regulatory compliance, we began by creating a proprietary AI engine so that we had complete insight into how CC would answer questions and could demonstrate the algorithm to pharma-compliance bodies,” Windhausen says. “We have also worked directly with pharma medical review boards to create specific workflows for managing and approving content at the speed required to support the system’s ongoing training needs.”
The data security aspects of the system take into consideration all levels of data security needs, from the management of personally identifiable information to HIPAA requirements, and can be customized to the specific business needs and implementation.
“And finally, CC has the ability to understand the user’s persona in order to respond to their personal needs and can even customize the response based on the channel of interaction,” Windhausen says. “For instance, a response to an inquiry via text can be kept short with perhaps a relevant link to other information, but a response via an Echo Show can be done via video.”
Each implementation of Cognitive Core is customized to the exact client need. CC can be used to power something as simple as an online FAQ page but can also be used as a platform to create a multichannel, integrated experience for patient services or field sales management. The engine is “trained” with the specific content and knowledge required for each application. For a patient services use case, it may be trained with all the specific information pertaining to the disease state or the brand. CC can also be integrated with patient support platforms to give end users the ability to track disease state adherence and journey information, alert their personal care network of their progress, or even just provide the patient with the date of their next appointment. As a platform for field sales team support, CC can be integrated with platforms like Veeva Systems to provide sales reps with a more in-depth view of their upcoming schedule, the ability to record a call, or just act upon suggested courses of action.
“Cognitive Core is meant to integrate with other platforms and client interfaces,” Rahman told Med Ad News. “In some cases, it can connect to Salesforce and gather relevant information, predict next events and provide user-friendly information over email, SMS, phone or any other connected devices. In other cases, it can remain in the background and, for example, provide nurses with data and robotic automation for an existing lengthy process.”
To support all this functionality, Intouch has a dedicated Cognitive Core team that includes AI developers, content analysts and creators, workflow automation experts as well as supporting teams that handle adverse events, product complaints and more.
“Under the hood, Cognitive Core is inspired by many academic papers in AI and algorithms that have been developed at universities over several decades,” Rahman says. “However, much of Cognitive Core is built from the ground up to ensure that it works efficiently and accurately in a strict regulatory environment.”
According to Windhausen, clients have four primary questions to consider when building a solution with Cognitive Core.
“First, how will the system be trained? In other words, what is the source of the knowledge by which CC will be trained to interact with users. This can be very structured in nature and could consist of specific answers to specific questions, or a structured decision tree-based flow to the intended use case, but could also be a completely unstructured content repository that requires CC to build its own complex network of how the data is tied together.
“Second, how will the system understand what the users are asking? This is where CC’s natural language processing engine is customized to meet the specific needs of the end users based on the intended use case, disease state, and audience.
“Third, what systems need to be integrated with CC in order to track user experiences over time and maintain a centralized view of the end-user/customer? These could include patient support systems and sales force management systems, for example.
“Fourth, what are the supported channels that will comprise the end-user experience? Based on the channels, we may want to consider that CC be trained with specific content so that it can respond appropriately depending on the channel being used.”
One of the key goals in the development of the Cognitive Care platform was to ease the barrier to entry for the consumer. “In order for us to do that, we ensure that Cognitive Core can be implemented on existing consumer devices rather than asking the consumer to buy a new device,” Rahman says. “We also want to make it a seamless experience for the end user.”
For example, a patient conversation can start with an SMS message from Cognitive Core in the morning, and then the conversation can continue over an Alexa-enabled device during breakfast, and then over the web in the afternoon. Each channel receives the appropriate content to deliver from Cognitive Core. For example, an SMS message ideally should be limited in length, so the SMS version of content is more concise than the content delivered via Alexa.
“CC has been built in a manner that completely separates the end-user channel from the machine learning and natural language processing aspects of the core platform,” Windhausen says. “This allows for complete flexibility to build out user experiences across multiple technologies and channels. For instance, CC can be used to power traditional chatbot interfaces on a website, but it can also be used to respond to similar conversations via SMS text messaging. CC can be used to create an interactive user experience that is integrated directly into a display banner advertisement or a social media-based ad. More sophisticated use cases may employ multiple user experiences across several connected channels which could include interactions via Siri, Echo, or other zero-UI systems.”
In fact, one of the early use cases for Cognitive Core was as a Siri-like assistant for a patient population. “We trained our platform to do exactly that,” Rahman told Med Ad News. “The most interesting challenge was to ensure the speech-to-text capabilities were able to recognize uncommon drug and disease names and appropriately understand the end user’s context.”
And as the capabilities of AI speed forward, Cognitive Core is keeping up. “We have a robust platform roadmap for Cognitive Core,” Rahman says. “Due to fast-moving advances in technology and environment, our roadmap covers the next three years rather than five or 10 years. Our future vision of Cognitive Core is based on our anticipation of evolving needs in pharma and healthcare. Our use cases of AI implementation cover the patient and the caregiver, the healthcare providers and the backend users in life sciences. We don’t envision AI taking over jobs; rather, we think AI can augment and do things that are time-consuming or repetitive for humans. So, we look at existing functions that caregivers or HCPs provide and identify where AI might fit in best.”
Rahman believes that people will eventually move away from web browser-based data consumption to IoT-based data consumption, so Cognitive Core will have to grow accordingly. “Cognitive Core can already integrate with different types of IoT devices; however, our roadmap takes into account new technologies that are coming up and how we can adequately measure with regard to privacy and security when incorporating Cognitive Core,” he says.
According to Windhausen, Cognitive Core’s evolution will be to become even more “aware” of how and when to interact with users in a nearly unnoticeable manner. “The future of AI systems like CC is not to replace human intelligence but to augment our own intelligence in a way that allows us to achieve and understand at greater levels, all while avoiding obtrusiveness in how we interact with our surroundings.”
metaMe Health is one of a number of startup companies who are developing what are coming to be called “digiceuticals” – digital apps or platforms for the treatment of disease that pass through clinical trials and the FDA approval process and can be prescribed by physicians. In metaMe’s case the target disease is irritable bowel syndrome, and the company’s platform is called metaMe Connect, a way to connect patients to therapists online for hypnosis treatment, which has been shown to be effective for IBS in a number of clinical trials.
“What we’re looking to do is to actually treat irritable bowel syndrome independent of any other medication using gut-directed hypnosis,” says Daniel Bernstein, CEO of metaMe. “Gut-directed hypnosis was first discovered back in 1984, and then in 1995 a version of the message that the therapist gives of the hypnosis discussion with the patient was standardized, scripted, and copyrighted and is referred to as the North Carolina protocol. We’ve licensed that copyright and are in the process of digitizing this, so instead of a patient going to a therapist, a specially trained therapist in gut-directed hypnosis, they can now take these seven sessions at home whenever they want, whether that’d be in the evening, the morning, whenever they feel that it fits into their schedule.”
The great advantage of offering this hypnosis via a digital platform is that it will help solve a problem of access – according to metaMe executives, there are only about a hundred psychologists in the entire United States who are properly trained to give gut-directed hypnosis for IBS. “We’ve communicated with GI physicians, and they are very familiar with cognitive behavioral therapy or even gut-directed hypnosis, but they just don’t have anywhere to send their patients, to be able to provide that therapy,” Bernstein says. “So by offering this treatment in a virtual environment, digitizing it, making it prescription-only, we’re providing an option, an additional option for physicians to put into their armamentarium, if you would, to help treat these patients.”
Just like a traditional medicine, metaMe Connect is working its way through the clinical trial process. The company is planning a Phase III trial on its computer-based platform while continuing to work on the mobile app, and the results so far are encouraging.
“We’ve got pilot data from about 40 patients who have used metaMe Connect on their computer, with their internet browser, and the results are exactly the same as you would have expected them to have been if they’d been going to a therapist in person and receiving live therapy,” Bernstein told Med Ad News.
And carrying the parallel even further, if metaMe can earn approval for its browser-based medium, the company won’t have to go through the full trial process for the mobile medium; it will just have to show “bioequivalence.” “We’ll do the study on the PC version, the computer version, and if that works then all we have to do is show the compatibility, if you would, the bioequivalence, between the PC and mobile versions,” Bernstein says.
As so often is the case with life sciences startups, Bernstein came to found metaMe because of a personal connection to the target disease – he has IBS himself. He was undergoing treatment at Northwestern University, and as a part of his treatment routine had come across a gut-directed hypnosis therapist there. Having spent time at Disney and Activision developing children’s video games, Bernstein already had a background in digital platforms. So when he discovered gut-directed hypnosis and the limited number of trained therapists who could provide it, it was a short jump to an entrepreneurial “a-ha” moment.
“I thought, well, wait a minute, we could increase access to this therapy to so many more patients if we could do it in a digital fashion versus sending all patients to these 100 therapists scattered around the country,” Bernstein says.
metaMe will not be limiting itself to a single product for a single indication, either. The company has an exclusive license for the use of gut-directed hypnosis in the treatment of IBD and GERD too. And the company is also working on a platform for cognitive behavioral therapy, another treatment for IBS.
“Cognitive behavioral therapy is a much more fluid process,” Bernstein says. “It’s a little bit more difficult, probably it’ll take a little bit more machine learning and artificial intelligence to crack that one, but it’s been done in other areas and we think we can do it here for IBS as well. So we have four products effectively in the pipeline.”
Digiceuticals – or “prescription-only digital therapeutics,” to the literal-minded – are so new to the marketplace that the regulatory and payer framework around them is still very much under development. It remains unclear whether digiceuticals will be covered under medical or pharmacy benefits. But in either case metaMe hopes to offer a significant cost savings to payers.
“The live therapy sessions right now are covered under the medical benefit under CPT codes, and they’re expensive,” Bernstein says. “And the pharmaceuticals that are used to treat the condition of course are under the pharmacy benefit, and they’re very expensive. Our cost is essentially nothing, once it’s developed, because it just sits on a server. We can bring this out, put it in the pharmacy benefit, and bring significant savings. And so I have every reason to believe that payers and PDMs will look at this as an opportunity, to use our language here, not only to put it on formulary but potentially use it as a step ahead, because of the potential cost savings and the effectiveness.”
So how will metaMe make its money? “It’s a three-month therapy, 3½ months, technically, to take all seven courses and the practice that come along with it,” Bernstein says. “We’re anticipating charging a flat fee for the prescription, the same way you would for a 30-day supply of any other drug.”
But what the company’s leaders do not know yet, and what they will have an opportunity to consider once the trial results come in is, will there be a need for re-treatment, and how might that impact the revenue model? “Right now we’ve really only looked at one course of treatment, the 3 ½ months,” Bernstein says. “The vast majority of those patients in the live therapy sessions, about 80 percent of them, are effectively cured for upwards of a year. But for those that do have relapses, we want to provide the therapy again to see if it works. So now we might have a type of subscription model capability, but we’ll also have the evidence to go to the marketplace and say, ‘Look, if you do relapse within a certain period of time, you have x percent probability of having a successful treatment if you take it a second time.’”
What will metaMe look like in five years? When asked, Bernstein turns the question around.
“We’ll no longer be asked, ‘What is a digital therapeutic? How does it work?’” he says. “Digital therapeutics will be perfectly ordinary. Not everyone will be using one, but everyone will know. And metaMe will be known as one of the companies that made that happen.”
Scout/Rare Expertise joint venture
Announced this past September, the Scout/Rare Expertise joint venture aims to combine the marketing chops of pharma agency Scout and the rare disorder-identifying data tools of Rare Expertise to help rare disorder patients obtain accurate diagnoses and begin appropriate medical treatment more quickly.
According to leaders of the two partner companies, it currently takes an average of between five and seven years for rare diseases to be correctly diagnosed. And according to the National Organization for Rare Diseases, such disorders are inherently difficult to identify and there are relatively few experts in any of the thousands of known rare disorders, which collectively affect nearly 30 million Americans, more than half of them children. Physicians receive little training in rare diseases and seldom encounter them in practice.
On average, a patient will go to seven doctors – often being misdiagnosed multiple times along the way – before finally getting a correct diagnosis. And because rare diseases are degenerative and frequently life-threatening, the patient may be hospitalized numerous times during the journey and become progressively ill, or more ill. So clearly an unmet need exists to support such patients more effectively.
“Because many doctors never come face-to-face with a rare disease and there are fewer experts in the field – remember, there are more than 7,000 rare disorders – patients often see multiple doctors and go through multiple hospitalizations before finally getting a correct diagnosis,” says Raffi Siyahian, principal, Scout. “This takes a crushing toll on patients and their families and seriously delays treatment. It also puts a heavy burden on our already stressed healthcare system.”
And Scout and Rare Expertise bring plenty of, well, expertise to the table. Together the two companies have worked on more than 30 rare disease brands and 20 rare tumor types, providing a variety of services and targeting a variety of audiences.
“So there wasn’t a single ‘a-ha’ moment in our decision,” Siyahian told Med Ad News. “There was easy compatibility and mutual respect right from the start, recognition of our common goals and complementary offerings, and realization that, in partnership, we could provide more services to rare disease marketers than any other company.”
So what will such a partnership be able to offer rare disease brands? The leaders of the two companies say they will be able to go “beyond the traditional means of marketing.” What does that mean?
“It means taking multiple approaches to multiple audiences, including patients, caregivers, HCPs, patient advocacy groups, geneticists, and centers of excellence,” says Jack Davis, founding partner, Rare Expertise. “Companies can benefit from a number of services developed by Rare Expertise, including clinical trial planning and recruiting, a network of online influencers, and a database of patient journeys, while also getting access to Scout’s big-picture planning and deep knowledge of commercialization and market development for rare disease therapies. This combination of services and products gives rare disease companies a truly complete range of essential services in a one-stop marketing partner.”
And what is Rare Expertise’s secret sauce on the technology side? The company has identified 25 distinct approaches to address the rare disease patient identification and activation challenge. These include search engine-enabled predictive algorithms that can accelerate diagnosis, cognitive analytics/NLP platforms capable of mining large amounts of unstructured data, online influencer networks (the company’s Rare Nexus) with existing reach into patient communities, and even pattern recognition software capable of detecting abnormal facial morphology. The company’s Rare Pathways system uses proprietary algorithms and extensive rare disease experience to assemble a custom-tailored plan for each client inclusive of stakeholder matrix, behavioral objectives, strategic framework, tactical approaches, funding levels, and ROI assessments. And all this is catnip for more than just pharma brand clients; fewer misdiagnoses and misguided treatments will be wins for patients and (especially) payers too.
“We believe payers are focused on getting value for their members and having confidence that the resources they apply to managing rare diseases result in a meaningful difference in the lives of these patients,” says Jeff Sweeney, founding partner, Rare Expertise. “Our ability to engage rare disease patient populations and assess areas such as truly meaningful endpoints and parameters for health-related outcomes helps all stakeholders, payers included, understand the true contribution of these highly specialized therapeutic agents.”
Scout and Rare Expertise executives believe that their partnership is more than just a good fit – it’s well-timed too, as more pharma companies are developing and launching products for rare diseases and the medical community and the public are becoming better educated. They also see an expanding number of stakeholders in rare diseases, all with their own needs and interests.
“Our immediate focus is on helping accelerate the time to diagnosis, but there are other needs and we’re equipped to meet them,” Siyahian says. “Maybe it’s help identifying suitable patients for a clinical trial. Maybe it’s the launch of a new therapy or market development for an existing therapy. Or maybe it’s relationship-building with patient groups and medical thought leaders. Our wide scope of expertise lets us grow with the market and meet the diverse needs of its diverse audiences.”
Unsurprisingly, as is often the case with folks who work in the rare disease space, the partners have personal connections to the “orphans” of medicine.
“I have a son who was born with a rare disease, so I know how frightening and frustrating it is to not be able to find accurate information and true experts for a given condition,” Sweeney says. “We’ve had access to some of the top institutions in the country, but we still made mistakes. And when you live this world, you realize very quickly the real challenges that face patients and their families.
“Our partnership is ultimately designed to support rare disease companies with a more complete range of services and support so they can be more successful, more quickly. And this means better care and better outcomes for rare disease patients and their families. This is what motivates me to work in rare diseases and why I’m excited by the promise of our new joint venture.”