Lessons learned from multi-site federated analysis in VALO NSCLC pilot study

Summary

The VALO Nordic OMOP implementation lessons learned exercise represents the first systematic assessment of OMOP Common Data Model (CDM) implementation experiences across multiple Nordic healthcare systems.

Conducted in September 2025 following the VALO metastatic non-small cell lung cancer (NSCLC) pilot study, this assessment captured experiences from ten respondents across five countries: Norway, Finland and Denmark as active data partners, and Sweden and Iceland as observers.

The assessment employed a structured survey instrument with six thematic sections, complemented by operational insights from the IQVIA coordinating team. Response rates ranged from 69% (Finland) to 95% (Sweden). This mixed-methods approach combined quantitative challenge assessments with qualitative experiential narratives to document technical challenges, data availability gaps, governance requirements and resource timelines for future network implementations.

Key takeaways

• OMOP implementation is a multi-year organisational transformation, not a short-term project (Norway reports ~3 years).
• Data readiness follows three tiers:
  • tier 1 (diagnoses, medications) is OMOP-ready
  • tier 2 (labs) requires mapping
  • tier 3 (biomarkers, PROs) is absent.
• Technical infrastructure compatibility, not OMOP conceptual understanding, is the primary implementation challenge.
• Final analytical cohorts represent only 10-20% of initial population estimates after applying eligibility criteria.
• Governance knowledge is fragmented across roles; cross-functional teams are essential from project inception.
• Vocabulary versioning requires network-level coordination to ensure semantic consistency across federated sites.
• Timeline estimates should be multiplied by 1.5-2.0× to account for multi-site coordination complexity.