Aetion, Inc.

"Rassen" can refer to unique, connectable chopsticks from Japan (Rassen chopsticks by Nendo), a family name for notable individuals like epidemiologist Dr. Jeremy Rassen or music director Greg Anthony Rassen , or the German/Dutch word for "race" or "breed," as seen in historical Nazi charts (Rassen der Erde ) or general translations. The meaning depends heavily on context, from innovative design to personal identity to historical or linguistic roots.123456

Specific Meanings & Contexts:

• Rassen Chopsticks (Design): A clever, single-piece chopstick design from Japan that separates for use and rejoins for storage, crafted with traditional lacquer techniques.
• People (Rassen Surname):
  • Dr. Jeremy Rassen: Co-founder of Aetion, focusing on AI and real-world evidence in healthcare.
  • Greg Anthony Rassen: A music director, orchestrator, and arranger for Broadway/cast albums.
  • Amy Rassen: A social worker and consultant in human services.
• Linguistic (German/Dutch): Rassen (plural of Rasse ) means "races" or "breeds," used in contexts like:
  • Rassen der Erde : A Nazi propaganda chart of "Races of the Earth".
  • Rassen- : A prefix meaning "racial" in Dutch/German.
• Zoology (Archaic): Merriam-Webster lists "rasse" (singular) as an older term for a type of small civet, also called the lesser civet.123456789

To understand what "rassen" means, you need to know if it refers to the chopsticks, a person, or is being used in a linguistic or historical sense, notes this Merriam-Webster page and this PONS dictionary page.12345678

Datavant, a leader in health data connectivity, has aggressively expanded its platform through major acquisitions and a 2021 merger with Ciox Health ($7B value). Recent 2025 acquisitions, including Ontellus (health records retrieval) and DigitalOwl (AI medical data analysis), established a new Legal & Insurance vertical to automate manual data exchange and analysis.12345

Key Recent Acquisitions & Strategic Moves (2025):

• DigitalOwl (Sept 2025): Acquired to integrate AI-driven medical data analysis, automated summaries, and predictive analytics, significantly strengthening their capability to process and interpret health records for legal and insurance sectors.
• Ontellus (Aug 2025): Acquired to create a new Legal & Insurance business vertical, automating the retrieval of clinical and billing data to speed up claim approvals and legal case resolutions.
• Aetion (May 2025): Acquired to enhance real-world evidence (RWE) capabilities, enabling life sciences to generate high-quality insights throughout the drug development lifecycle.12345

Previous Key Acquisitions:

• Ciox Health (2021): Merged to create a massive, neutral health data ecosystem, connecting data sets across EHRs, claims, and labs.
• Trace Data (2024): Acquired to expand GDPR-compliant data connectivity, particularly in Europe.
• Universal Patient Key (UPK) (2018): Acquired to secure, encrypted, patient de-identification technology.12345

Strategic Goal:
Datavant's acquisitions focus on building an end-to-end, secure, and privacy-compliant data infrastructure. These moves aim to reduce administrative burdens for healthcare providers (over 80,000 in their network) and create actionable AI-driven insights for insurance, legal, and life science customers.12

Major real-world data (RWD) companies include Flatiron Health (oncology data), Inovalon (claims and electronic health data), Datavant (data platform), and TriNetX (healthcare provider network data).These companies use RWD from sources like electronic health records (EHRs), claims databases, and wearables to generate insights for life sciences and healthcare.Other key players include Clarivate, Certara, and Veradigm.1234567891011

Oncology and clinical data1

• Flatiron Health: Specializes in oncology, using real-world data from EHRs to provide evidence on treatment patterns and outcomes.12
• Clarivate: Provides RWD, including medical claims and EHR data, to the life sciences and healthcare industries.1
• Veradigm: Offers a real-world evidence analytics platform that uses data from its network.1
• Certara: Uses modeling and advanced analytics on RWD to provide insights, particularly in drug development.12

Data platforms and networks1

• Datavant: A data platform company focused on healthcare, it integrates and connects data for analysis.1234
• TriNetX: Operates a federated network of RWD from healthcare providers, which is used to accelerate innovation.1
• Aetion: A Datavant company that builds software to analyze how medical innovations work in real-world settings.1

Analytics and insights

• Inovalon: Provides RWD from claims data, which offers visibility into the U.S. healthcare system.1
• Atropos Health: Uses RWD from anonymized patient records to help answer clinical questions and enable evidence-based medicine.1
• Cytel: Works with RWD for clinical trial design and analysis.1

Propensity weighting is a statistical method, mainly used in observational studies, to create balanced comparison groups by assigning weights to individuals , making them more similar to a randomized trial, thus reducing selection bias and helping estimate treatment effects accurately, often using Inverse Probability of Treatment Weighting (IPTW) where weights are 1/PS for treated and 1/(1-PS) for control, incorporating all subjects for a better picture.123

How it works

1. Calculate Propensity Scores : A model (like logistic regression) estimates the probability (propensity score) of each person receiving the treatment, based on their characteristics (covariates).
2. Assign Weights:
   • People in the treatment group get a weight of 1/P S 1 / cap P cap S 1/𝑃𝑆 (probability of treatment).
   • People in the control group get a weight of 1/(1−P S)1 / open paren 1 minus cap P cap S close paren 1/(1−𝑃𝑆) (probability of not receiving treatment).
3. Balance Groups : These weights adjust the influence of individuals, making the treated and control groups comparable in terms of observed characteristics, mimicking randomization.1234567

Key Benefits

• Reduces Bias : Controls for confounding variables that differ between groups in observational studies.
• Uses All Data : Unlike matching, it includes all subjects, preventing data loss, which is great for small samples.
• Estimates Treatment Effects : Helps estimate the true impact of an intervention when randomization isn't possible, aiming for average treatment effects (ATE) or average treatment effect on the treated (ATT).12345

Example

If a treatment group has fewer older people, propensity weighting gives more weight to the older people in that group and less to the younger ones, effectively increasing their "representation" to match the control group's age distribution, creating a fairer comparison.12