AboutPharma and Medical Devices – Editorial staff
Big data is not an option. In other words, in order to do business and grow, no company can restrain from an intelligent use of data. This assumption obviously applies to pharmaceutical companies as well. Experts in Sudler – the WPP Health & Wellness’ specialised network in brand healthcare strategy and communication – are well aware of this scenario and developed Orioon, a tool specifically tailored to the healthcare industry needs. Based on an innovative analytical approach, the system was designed to support strategic decisions in the life science segment by combining human understanding and advanced artificial intelligence technologies. Developed in partnership with SAS – an international business analytics solution provider – Orioon has the unique capability to work on both “big” and “small” heterogeneous, structured and unstructured, qualitative and quantitative data.
By evaluating the correlations among the information gathered from social networks, websites, scientific publications, clinical trials, etc. – in compliance with privacy regulations – the Orioon system provides an analytical support to the expertise of Sudler’s consultants in developing insights based on an approach that includes several key features. It’s a combination of advanced data analytics methods and targeted consultancy services specially aimed to improve communication and marketing strategies. “The idea to create such an innovative tool specifically tailored to the pharmaceutical industry originated a couple of years back, when we understood that more and more data were becoming available in this segment, both related to patents, medical specialists and other stakeholders as well as to therapies and trials,” explained Maurizio Mioli, CEO of Sudler International Milano & Zürich. “However, in order to apply advanced analytics methods to healthcare, big data are not enough. It is necessary to include an indepth analysis of qualitative data from smaller population samples. To make informed decisions in this area, it is critical to acquire a combination of online research, market surveys, health databases, opinion leaders’ insights, etc. Artificial intelligence has the unique ability to help us achieve this goal; we therefore found an ideal partner in SAS, both in terms of expertise in the field of AI and machine learning, as well as for their specific expertise in healthcare.”
Initial testing was performed with Italian customers. A real-case application of the tool with an Italian pharmaceutical company was presented during the meeting “Artificial intelligence applied to data analysis: guiding strategic decisions in the pharmaceutical business”, promoted by Sudler, WPP Health & Wellness, SAS, and Kantar Health in cooperation with AboutPharma. The first stage consisted in a workshop with the client (undisclosed here) to identify their business needs.
In this particular case, the topic at hand was understanding why a product (here “brand X”) was not performing as expected in terms of prescriptions, and identifying the obstacles that prevented the full adoption of the product.
Special focus was put on the interaction between patients and medical specialists, also in view of the fact that patients play an important role in influencing therapeutic decisions within the product’s indicated disease. “A]er feeding the initial data (outcomes of quantitative and qualitative market surveys targeting doctors specialised in the disease under examination) we then included patient information, as also prompted by the tool,” explains Paolo Mistrorigo, Head of Data Analytics & Strategy at Sudler. “We then gathered online conversations among patients on public pla^orms and analysed them with natural language processing techniques, which allow processing and understanding of natural language.” The analysis of these conversations unveiled the most relevant topics related to the disease and available therapies, including treatment with brand X. “We therefore identified common topics among medical specialists and patients, thus understanding their relevance both in general terms and for the medical specialists who most prescribed the product under evaluation,” added Mistrorigo. “Through the use of a proprietary algorithm, we integrated data from various sources to make them comparable and allow them to interact.”
By applying various analysis levels that contributed to understanding relevant signs in which doctors and patients’ product evaluations would align or disagree, the system simulated the potential impact of possible strategic communication, marketing and educational actions on brand X for the two stakeholder groups. Through human understanding and customised consultancy, it was also possible to identify which actions had a greater chance of providing a positive impact on the sales performance. “We will soon be able to see the outcomes of the project carried out with this specific company,” said Sudler’s CEO.
According to Sudler’s Head of Data Analytics, technologies are extremely versatile. “Of all the technologies involved in this specific case and others that could be applied in different contexts, Orioon can make use of more advanced technologies than AI, ranging for example from machine learning to basket analysis, from natural language processing to voice recognition, among others.” In addition to this specific example, Sudler’s analytical approach can also be implemented in other scenarios. According to Mioli, these include “for instance the area of volumetric estimation, through the analysis of wide-ranging data collections, such as distribution and sales data; or the analysis of therapeutic pathways with the objective to improve compliance and minimise mistakes and waste. This innovative approach could even help identifying the most suitable channel, with a multichannel approach, based on the messages that need to be implemented and the target stakeholders. And last but not least, the tool could prove to be a valuable resource in the clinical se`ng to choose the most appropriate therapeutic approach.”
When designing this first application of the system, Sudler and SAS strongly focused on the role played by patients in health decisions, especially in view of the new increasingly informed and “engaged” role that these stakeholders are taking on. This is one of the reasons why the new approach specially focuses on qualitative data. As Guendalina Graffigna – associate professor within the Psychology Department of the Cacolica University in Milan – explained, “it is imperative to understand that there is a subjective and emotional dimension in the patient’s consumption of health services, drugs and therapies that can hardly be reduced to behavioural or cognitive data. The analysis needs to integrate these qualitative features, and scientific models must be identified for their interpretation. This is the challenge that big data analysis will most likely have to face in healthcare in the future.”
Orioon’s strengths include “a greater understanding of pharmaceutical market instances and paIerns. By combining data from various sources, effective insights can be obtained, for instance, to outline a company’s positioning compared to competitors, or to understand how doctors and patients perceive diseases, companies, and products,” explained SAS Innovation Consultant Francesco Rainini, who contributed to designing this innovative approach. In addition, by focusing on doctors it is possible to gather input to becer understand “which medical specialists are keener on welcoming conventional communication actions from pharmaceutical sales representatives and which ones would rather prefer digital channels, as well as how they would prefer to be contacted.”
Another advantage of this system is the possibility to constantly monitor data. The tool is designed to get self-training from the updates received and to regularly reformulate them. For example, once a strategy is selected, the system can monitor its effects on the target stakeholders. These longitudinal investigations allow for constant fine-tuning of the adopted strategic measures in order to maximise the expected outcomes.”
Orioon developed against a background in which AI solutions play an increasingly important role in guiding business chioces. “In the past decade, this evolution was driven by the increased volume of available data, new computing capabilities, and the availability of increasingly more sophisticated algorithms,” highlighted SDA Bocconi Professor of Practice Decision Sciences & Business Analytics Renata Trinca Colonel during the meeting. “Not all companies will make use of big data, in part because of a cultural delay in understanding their potential. Those who already chose this solution, however, are enjoying extensive competitive advantages and will probably profit from these solutions even more in the future.”
Artificial intelligence, big data
Sudler, WPP Health & Wellness, SAS, Kantar Health
© 2018 Health Publishing & Services S.r.l. – All rights reserved | JUNE 2018 | No. 159