Stuck in the Middle
Clinical analytics solutions are “sandwiched” between (1) a variety of shifting data sources; and (2) ever-evolving business rules that dictate how data is modeled to be consumed by apps. The result is expensive, non-scalable “ETL Churn,” or lengthy and ongoing data integration work to extract, transform, and model data (Figure 1).
With the number of data sources ready to explode, and widely-adopted standards for key business rules yet to emerge – this issue is here to stay for the near-term. Longer-term (5+ years), we expect some complexity to subside, as de-identified aggregate data sources begin to inform the clinical evidence, downstream markets begin to benefit from standards set by early adopters, and new innovative architectures emerge for managing heterogeneous structured & unstructured data sets needed for VBR.