Skip to content

Entrepreneurship

Entrepreneurship and Commercial Viability

Overview

Adverse drug-drug interactions (DDIs) are a major cause of severe medical complications and a significant reason behind inefficiencies in treatments worldwide. We are deeply convinced that SEPiA has the potential to minimize the risk of adverse DDIs for safe medication use by not only empowering patients and healthcare providers with AI-driven insights but also by bringing the power of AI prediction to the world pharmacovigilance community and providing accurate and impartial DDI knowledge to pharmaceutical companies during the drug design process - for the benefit of humanity. 

Background

Studies estimate that around 17% of adult American citizens take five or more prescription drugs, meaning the United States alone accounts for more than 56 million polypharmacy patients1. The elderly in particular are impacted by polypharmacy, as chronic ailments develop with age. Studies put the percentage of polypharmacy patients among this age demographic as ranging from 40 to 60%2. These numbers are the result of an increase in drug prescriptions, especially among senior citizens,  over the last 20 years. In 2000, the prevalence of polypharmacy in the general population was 8%, constituting an average annual percent increase of 2.9%1. Furthermore, this rate stands to grow significantly, as nations around the globe are undergoing a demographic shift towards generally older societies. This trend projects a strain on the healthcare-system that it is ill-equipped to handle in its current form.

Market Analysis

SEPiA targets three different markets.

Pharmacovigilance Market

The projected compounded annual growth rate (CAGR) figures of the pharmacovigilance market are ranging from 7.0% to 13.1% depending on the source, highlighting the growing importance of pharmacovigilance3,4,5. Its market size is valued at $6.7 billion in 2022 and is expected to grow to $14,85 billion in 2028 at the most conservative estimate3. The market is divided into three major segments:

  1. Pharmaceutical Companies

  2. Biotechnology Companies

  3. Medical Device Manufacturers

Pharmaceutical companies currently possess the greatest market share, however, biotechnology companies are projected to grow faster than the other two segments, presenting a future in which the market is primarily served by biotechnology companies3. This trend indicates that biotechnological services such as SEPiA are flourishing in the present and the future. We believe that SEPiA has the opportunity to become a major player in the long-term.  

Drug Design Market

The drug design market, often referred to as the drug discovery market, is expected to grow rapidly in the coming years. The most conservative estimates foresee a CAGR of 8,20 % with a market size value of $85.81 billion by 2030 ranging to a CAGR of 9,2 % from 2023 to 2032 and an expected size of $133.11 billion by 20326,7 . Comparing AI in the drug design market to the overall market paints an incredible picture: In 2022, AI is estimated to account for around $1.5 billion of the market and is expected grow with a CAGR of 30.1% until 2032, putting the projected market size in 2032 at $20.9 billion8. These market conditions present a great opportunity, and SEPiA possesses tremendous potential to become a valuable asset in drug design.  

Drug Interaction Checker Market

There is no reliable market data available yet; however, the demand for medical consulting tools is expected to grow with telemedicine systems, online pharmacists, health technology companies, and research centers making drug interaction reviews available to patients through their websites and mobile apps9. 

Competitive Analysis

Pharmacovigilance Market

DDI information is crucial in the Pharmacovigilance market as it allows identifying potential safety issues with drugs. It’s an integral part of ensuring drug safety and monitoring DDIs, driving the demand 3. According to our market research, the pharmacovigilance community is mainly served by a limited number of companies such as Certara Drug Interaction Solutions that rely on a manually curated collection of qualitative and quantitative human information. SEPiA distinguishes itself by being AI-based, bringing all the benefits explained in previous paragraphs. Additionally, we intend to collect first-hand patient data on medication plans. Combining patient data with AI allows for the analysis of macro-trends in drug consumption, providing immeasureable value to the pharmacovigilance market.

Drug Design Market

DDI information helps pharmaceutical scientists in the design process of new drugs. Companies such as BioPharma Services and Certara offer a number of services from trial research to pre-clinical and clinical development. A number of companies have focus on applying AI in this context, predicting potential side effects in the process. These companies include Insilico Medicine and Atomwise, among others. We believe that SEPiA can be relevant in this manner, however, its unique potential lies in drug repurposing. SEPiA's ability to predict emergent side effects cannot only be used to anticipate adverse drug reactions but also potential beneficial effects. This puts SEPiA in the unique position to potentially find usage in drug repurposing.

Drug Interaction Checker Market

There are multiple drug interaction checkers on the market. These include services by Drugs.com, Medscape, WebMD, and Drugbank. The following table presents a comparison between these services and SEPiA.

Service Target Audience Database Information Unique Selling Point
Drugs.com Patient + Professionals Over 24,000 prescription drugs10 Severity, sources, therapeutic duplication warnings, and drug/food-interactions drug/food - interactions
Medscape Professionals No insights Severity, recommendations recommendations
WebMD Patients No insight severity
Drugbank Professionals No insight Severity, sources, extended description Limited to 5 drugs total, commerical API for full access
SEPiA Professionals Manually created, based on multiple publically available databases Confidence score Use AI model to make predictions for more than two interacting drugs

SWOT-Analysis

In order to make an assessment of our situation, we conducted a Strengths-Weaknesses-Opportunities-Threats-analysis (SWOT-Analysis). We looked at internal (Strengths, Weaknesses) and external (Opportunities, Threats) factors promoting or impeding our success.

SEPiA SWOT-Analysis

Image title
Fig. 1 | SWOT-Analysis

Business Model Canvas

The Business Model Canvas allows for a comprehensive overview of a business model. We believe that the depicted model allows us to provide the greatest value to all stakeholders involved in combating adverse DDIs.

SEPiA Business Model Canvas

Image title
Fig. 2 | Business Model Canvas

Strategy

In order to make our vision a reality, we have established a close partnership with the Department of Clinical Pharmacology at the University Hospital Aachen in Germany. This collaboration allows us to gather input from key medical stakeholders to design and deliver a practical, easy-to-use SEPiA engagement app that is able to empower the university medical community and the patient. This is to capture the relevant patient data for enriching our model to be in a position to analyze and optimize at the level of the individual patient for their individualized medication plan.

Additionally, we aim for a partnership with 2-3 major nursing home providers for their essential inputs during the app design to fully capture the core processes and requirements. 

How we envision the pilot phase

The SEPiA engagement app will be white labeled and piloted in the selected care giving facilities as well as at the Pharmacology unit at the university of Aachen. Once the pilot is successfully validated and has proven its value, we aim to introduce our solution to

  1. University hospitals: we will present the outcomes of the model to the respective clinical scientists through the network of RWTH Aachen.

  2. Nursing homes: we conduct a controlled roll out to additional nursing home providers for a systematic medication management and monitoring.

Through this approach we build up our relevance amongst the expert community and drive tangible impact for patients. This will give us the credibility and relevance for to support the pharmaceutical companies in their drug design phase with highly relevant DDI insights as well as to bring the power of our model to the Pharmacovigilance with impartial and accurate insight DDI services.

The following illustration depicts this strategy separated into phases. Currently, we find ourselves in phase one.

SEPiA Strategy

Image title
Fig. 3 | Strategy

Team

The SEPiA team consists of seven highly motivated students at the Technical University of Munich (TUM) and Ludwig-Maximilians University (LMU). The team's expertise lies primarly in engineering; We are six bioinformatics students and one information systems student. As such, we are confident in our ability to produce high-quality models and deploy them, however, we recognize that our team lacks the diverse range of skillsets required to enjoy commercial success. We are in need of business and legal expetise for our journey in the future.
In order to remedy these shortcomings, we scheduled a meeting with Silicon Valley-based venture capitalist Mani Honigstein of HoneystoneVC. He provided us with valuable lessons and expressed interest in joining the project in an advisory role. HoneystoneVC have a proven trackrecord of supporting startups in their journeys and are a crucial partner to any startup in their portfolio. Our meeting paves the way for us to enter their ecosystem, allowing us to rely on their intricate knowledge on operating a startup.

Long-term Vision, Implications, and Concerns

SEPiA's impact on healthcare could be enormous. In an ideal world, SEPiA provides value to patients by analyzing, optimazing, and personalizing medication plans. This not only reduces the risk of adverse DDIs but also supports patients in achieving a healthier lifestyle. Additionally, SEPiA plays a vital role in drug design by relaying accurate information about whether or not the novel drug interacts with any other drug on the market. Furthermore, SEPiA facilitates drug repurposing by selecting for beneficial DDIs. Lastly, SEPiA serves the pharmacovigilance market by providing impartial information on trends in drug consumption based on AI analysis. The combination of these use cases allows SEPiA to contribute to the health of millions of people.

There are, however, concerns with the project. AI's application in the context of patient data is a difficult subject, as privacy concerns arise. We need to establish a protocol on how to work with proprietary patient data, taking a patient-centric view into account, that is in full compliance with GDPR.
Another point of contention is the model's capacity for failure. As SEPiA uses inference to arrive at its predictions, a chance exists that it might be wrong. Additionally, SEPiA's nature as a blackbox model makes it impossible for humans to comprehend its reasoning patterns. These concerns could become a hindrance to the adoption of the model in clinical settings, as, even though the model might be more accurate than a professional, patients are prone to trust the professional over the model. We have to emphasize that SEPiA is to be used in a supporting role to a healthcare professional, not replace one.
We also need to account for biases in the data. If a specific group is over- or underrepresented in the training data, the model might become biased in favor or against that group. Seeing as we aim to apply SEPiA in a medical environment, such failures could result in catastrophic consequences. It is therefore of utmost importance to curate a balanced dataset and to always be vigilant concerning biases. Again, it needs to reiterated that SEPiA should always be placed under supervision by a healthcare professional.
Overreliance on SEPiA could become a problem as well. Due to the reasons mentioned above, we believe that every prediction made by the model should be validated by a healthcare professional. However, this requires vigilance on the professional's part. If they accepted the model's output without checking its truthfulness, a chance for failure arises.

We believe these concerns to be of vital importance in our success. If we do not account for them, we cannot deliver on our vision. In our work, we are mindful of these concerns and address them by collaborating closely with potential SEPiA users during and after development of the product.

  1. Wang, Xiaowen, et al. “Prevalence and Trends of Polypharmacy in U.S. Adults, 1999-2018.” Global Health Research and Policy, U.S. National Library of Medicine, 12 July 2023, www.ncbi.nlm.nih.gov/pmc/articles/PMC10337167/. ↩↩

  2. Oktora, Monika P, et al. “Trends in Polypharmacy and Dispensed Drugs among Adults in the Netherlands as Compared to the United States.” PloS One, U.S. National Library of Medicine, 22 Mar. 2019, www.ncbi.nlm.nih.gov/pmc/articles/PMC6430511/#pone.0214240.ref004. ↩

  3. “Pharmacovigilance Market Size and Share [2023 Report].” Pharmacovigilance Market Size And Share [2023 Report], www.grandviewresearch.com/industry-analysis/pharmacovigilance-industry#:~:text=The%20global%20pharmacovigilance%20market%20size,7.0%25%20from%202023%20to%202030. Accessed 10 Oct. 2023. ↩↩↩↩

  4. “At 13.1% Cagr, Pharmacovigilance Market Size [2021-2028] to Reach USD 14.85 Billion: Industry Size, Share, Growth, Revenue & Forecast Report.” GlobeNewswire News Room, Fortune Business Insights, 20 Dec. 2021, www.globenewswire.com/en/news-release/2021/12/20/2354972/0/en/At-13-1-CAGR-Pharmacovigilance-Market-Size-2021-2028-to-Reach-USD-14-85-Billion-Industry-Size-Share-Growth-Revenue-Forecast-Report.html ↩

  5. “Pharmacovigilance Market - Size, Share & Trends Report, 2032.” Global Market Insights Inc., www.gminsights.com/industry-analysis/pharmacovigilance-market. Accessed 10 Oct. 2023. ↩

  6. “Drug Discovery Market Size, Share, Trends, Opportunities & Forecast.” Verified Market Research, 5 Apr. 2023, www.verifiedmarketresearch.com/product/drug-discovery-market/. ↩

  7. “Drug Discovery Market (by Drug Type: Small Molecule and Large Molecule; by End User: Pharmaceutical Companies, CRO, Others; by Technology: High Throughput Screening, Pharmacogenomics, Combinatorial Chemistry, Nanotechnology, Other Technologies) - Global Industry Analysis, Size, Share, Growth, Trends, Regional Outlook, and Forecast 2023 – 2032.” Precedence Research, www.precedenceresearch.com/drug-discovery-market. Accessed 10 Oct. 2023. ↩

  8. “Ai in Drug Discovery Market Size & Share Forecast Report - 2032.” Global Market Insights Inc., www.gminsights.com/industry-analysis/ai-in-drug-discovery-market#:~:text=AI%20in%20Drug%20Discovery%20Industry,will%20accelerate%20the%20market%20growth. Accessed 10 Oct. 2023. ↩

  9. Lingaiah, Thamineni Bheema, et al. “Developing a Desktop Application for Drug-Drug Interaction Checker Ordered for Chronic Diseases in Ethiopian Hospitals Pharmacy - BMC Pharmacology and Toxicology.” BioMed Central, BioMed Central, 6 June 2022, bmcpharmacoltoxicol.biomedcentral.com/articles/10.1186/s40360-022-00576-4. ↩

  10. “Prescription Drug Information.” Drugs.Com, Drugs.com, www.drugs.com/. Accessed 11 Oct. 2023. ↩