Applying Next Best Action Technology in Pharmaceutical Services
When the term NBA is used, most of us think of the National Basketball Association. But in the technology realm, NBA is quickly becoming known for Next Best Action — a sub-class of artificial intelligence (AI) technology. NBA technology is increasingly being used by businesses to improve customer engagement, respond to target audience needs and interests and advance corporate objectives.
Companies like Amazon and Netflix have effectively applied NBA to support a range of customer marketing objectives such as acquisition, up-selling, cross-selling and retention.
It’s time for the pharmaceutical services industry to start widely applying NBA technology to realize some of the significant benefits it can bring.
Let’s start with the basics…
NBA technology solutions consist of five key components:
1. Data – NBA technology makes use of all data including structured data, like patient and HCP profiles and transactions history, and unstructured data, like voice data or conversations with an agent/nurse; patient communication through emails and apps; or social and external input and images. Incorporating newly recommended actions and responses back into the data pool is a critical step in honing responses for subsequent NBAs.
2. Analytics – To create an optimal NBA model, you need analytics based on specific business objectives. For example, medication therapy initiation is based on a simple-state objective (yes/no), involving a look-alike type of analytics model such as logistic regression, random forest, or a neural network. For multi-state objectives like “optimal” adherence, optimization requires multi-step NBAs. For multi-state and more complicated optimization problems, using reinforcement-learning algorithms helps achieve higher efficiency and flexibility.
3. Customer/Patient-Triggered Personalization – NBA is initiated based on a customer or patient’s personal history data, generating specific recommendations for that person. Actions that universally apply for all patients or different segments/clusters of patients, or actions purely based on business rules are not considered true NBA. Another distinctive attribute of NBA is that actions are triggered or initiated by a “status change” for individual patients (e.g., AE-reported) or “recommended status change” (e.g., to change 30-day refill to 90-day to improve adherence).
4. Real Time – In both outbound and inbound situations, the agent/nurse/bot may need the ability to take the best action in real time based on new information, situation and/or historical data. Using instant natural language processing (NLP) and sentiment analysis interactive conversation can be transformed into text and groups classified based on characters and sentiment. To generate real-time NBAs on the phone, website and apps, a back-end analytics engine is required.
5. Verification – Verifying positive impact or objectives being met is key to ensuring the NBA approach being applied is optimum. Even if all of the above four components are satisfied, it doesn’t mean an NBA approach is automatically achieving a positive effect. Perhaps some social unstructured data wasn’t being included, or a suboptimal analytics engine was being used. In a practical sense, it’s more important to achieve a significant improvement in the approach rather than an open-ended pursuit of the absolute “best.” For example, if one NBA approach extended length on therapy (LOT) by 30%, it can be viewed as the “best” even though another approach may achieve a slightly higher extension.
For LOT, the gold standard for evaluating whether there is significant improvement through an NBA approach is to compare the average LOT in a 3–12-month window between a randomly selected TEST group with an NBA implemented and a randomly selected CONTROL group with no NBA implemented. The next option would be to try comparing the average LOTs in a 3–12-month window between a patient group with NBA implemented and a control group without NBA.
Following propensity matching methodology will ensure similar test and control groups are created based on profile and interaction characteristics.
Effective use of NBA technology will enable pharmaceutical services solutions providers to more fully respond to customer needs in real time and ensure actions taken support patients, providers and pharmaceutical company objectives.
Leveraging NBA in hub and inside sales call centers
Without an NBA capability in place, call center agents are reliant on standardized scripts or business rules training to guide their communications. These scripts or rules may fail to take into consideration the individual’s unique situation, historical behaviors, preferred communication channels and other relevant information. As a result, communications are less personalized, less efficient and less effective.
With NBA technology, hub, bot and service center agents become more productive, solve queries more efficiently and accurately. Customer experience improves as handling time gradually drops; customer retention increases as discussions become more targeted.
A reimbursement hub or inside sales call center can leverage NBA in an inbound call situation by guiding actions based on the historical transactions, behaviors, profiles, and the most recent call information. Real-time NBA is equally applicable to outbound communication as new information is immediately ingested during the engagement.
With both inbound and outbound communication — NBA capabilities optimize the customer experience and significantly improve program efficiency.
Some services claim NBA is implemented in their process, but the NBA might be simply business rules-based (even complicated rules), without solid predictive analytics or a machine learning engine supporting it. This is not considered true NBA.
The time to deploy NBA in pharma services is now
Especially in this current environment, there’s a tremendous need to cultivate and strengthen patient and HCP connections to improve prescription adherence and drive advocacy and education. NBA technology is a critically important advancement that can significantly improve reimbursement hub program efficiency and adherence as well as ROI for inside sales programs.
As market dynamics continue to evolve, biopharma companies are further increasing their use of clinical nurse educators to provide behavioral health support, education, and training to patients to improve outcomes. Given the complexity and availability of data, especially unstructured data, NBA is more important than ever.
About Conduent Pharmaceutical and Life Sciences
Conduent delivers mission-critical healthcare provider and patient engagement solutions to help pharmaceutical and life sciences organizations launch and support commercialized products. We support over 450 products and have nearly 30 years of experience helping patients access and remain on their medications. As healthcare evolves, our programs and technology will keep pace with these changes to continue bringing patients together with life-altering treatments. By leveraging AI-powered discovery and analytics, Conduent empowers clients to have meaningful engagements with patients and HCPs that optimize their experiences and contribute to improved health.
Contact us today at email@example.com to discover how next best action technology can improve the experience and outcomes for your patients.