Real-world evidence · Neurology evidence generation
Turning messy real-world data into evidence people can trust.
I'm Sonick. For eight years I've designed and delivered studies from Claims and EHR data, from patient journeys and treatment patterns to biomarker research, for global pharmaceutical clients.
- 8 years
- in real-world evidence
- 5
- publications & conference abstracts
- 10+
- RWE studies delivered
- 6
- person team led
01 · SOURCE
Claims and EHR
02 · CURATE
Abstraction, QC, fit-for-purpose data
03 · ANALYSE
Cohorts, endpoints, statistics
04 · EVIDENCE
Publications & client decisions
What I do
Evidence generation from real-world data, from the first feasibility question to publication.
RWE study design & analysis
Retrospective studies with Kaplan-Meier and Cox models, interrupted time series, regression and concordance analysis.
Treatment patterns & persistence
Switching, persistence and discontinuation, healthcare resource use and real-world endpoints such as relapse and infections.
AI-enabled chart abstraction
LLM pipelines that turn clinical notes and lab reports into analysis-ready variables, validated for PPV, sensitivity and F1.
RWD for trial feasibility
Turning eligibility criteria into cohort definitions and attrition funnels to find and match patients for studies.
Selected work
ALZHEIMER'S · BIOMARKERS · CTAD
Diagnostic pathways to anti-amyloid therapy
Compared blood-based biomarkers, PET and CSF in routine care, found care gaps and the fastest route to anti-amyloid therapy, and mapped the patient journey from cognitive assessment to biomarker results.
ALZHEIMER'S · PERSISTENCE
Anti-amyloid therapy persistence
Kaplan-Meier analysis of real-world time to discontinuation, stratified by biomarker status, to find what drives patients to stop treatment.
MULTIPLE SCLEROSIS · TREATMENT PATTERNS
Disease-modifying therapy in real-world MS care
Led a retrospective study comparing DMT classes on switching, healthcare resource use and relapse, with real-world endpoints extracted from clinical notes by a validated LLM pipeline.
RESPIRATORY · ISPOR EUROPE 2023
COVID-19's impact on RSV and influenza
Interrupted time series models quantifying how RSV and influenza incidence changed before and after COVID-19 in the US and EU-5.
PYTHON · SYNTHETIC EHR · PERSONAL PROJECT
What real-world data can and can't tell you about trial eligibility
5,000 synthetic patients vs. a Phase 3 Alzheimer's protocol: 14.6% pass the checks an EHR can make; 5.5% are fully eligible.
COMMERCIAL ANALYTICS · CLAIMS
Launch planning & prescriber targeting
Treatment-pattern dashboards for an oncology portfolio, forecast inputs from claims, and HCP segmentation from prescription and distribution data.
- Medical & pharmacy claims
- EHR
- Clinical notes & lab reports
- HCP surveys
- MarketScan
- Flatiron
- IQVIA
- SAS
- R
- Python
- SQL
- Databricks
- Instant Health Data
- LLM extraction
Experience
Eight years across a health-tech start-up, a CRO, a global pharma company and analytics consulting.
Education
PG Diploma in Applied Statistics, IGNOU (2020–2022)
B.Tech, Electrical & Electronics Engineering, Amity University (2014–2018)
Recognition
Lilly Star Award, Eli Lilly
SPOT Award, Mu Sigma
- May 2026 – PresentAssociate Director, Clinical Trial Enrolment & Implementation Sciences
NeuroDiscovery AI · Led a six-person team matching patients to Alzheimer's and epilepsy studies from real-world data, and AI-enabled chart abstraction.
- Sep 2024 – May 2026Senior RWE Analyst
NeuroDiscovery AI · Foundation hire who helped build the RWE function; Alzheimer's and MS studies.
- Jul 2023 – Sep 2024Data Scientist II
Syneos Health · Time series, forecasting and prescriber analytics from claims and IQVIA data.
- Apr 2022 – Jul 2023Associate Manager
Eli Lilly · Oncology treatment-pattern analyses with MarketScan claims and Flatiron EHR.
- Dec 2018 – Mar 2022Trainee Decision Scientist → Apprentice Leader
Mu Sigma · 10+ RWE studies in cardiovascular and immunology from claims and EHR data.
Writing
Notes on real-world evidence. First posts coming soon.
Visit the blog