Social and Other Data

Claims and clinical data are the two pillars, but a complete value-based data landscape includes several other feeds. Some add crucial context; one is the most operationally valuable data source in all of population health. This lesson covers the rest of the landscape.

Social data

Much of what drives cost and outcomes, housing instability, food insecurity, transportation, isolation, appears in neither claims nor clinical records. Social data tries to fill that gap:

• Screening data, captured when providers ask about social needs, increasingly with standardized codes, but recorded inconsistently.
• Area-level indices, deprivation and vulnerability scores attached to a patient’s address, a workable proxy for stratification with the caveat that a neighborhood average is not a person.

Social data is sparse and imperfect, but as the Medicaid course showed, it is often the missing driver behind the utilization the other sources can only describe.

ADT feeds: the highest-value feed

Worth remembering: the admission, discharge, and transfer (ADT) feed is the most operationally valuable data source in population health. It is a near-real-time notification that a patient has entered or left a hospital, anywhere in the connected network. That timeliness is what makes it actionable: it triggers the transitional-care work that prevents readmissions while the window is still open. Claims tell you a hospitalization happened months ago; an ADT feed tells you it is happening now. For an analytics function, standing up reliable ADT alerts is often the single highest-return data project.

Supplemental sources

Several other feeds round out the picture:

• Pharmacy data, often more timely than medical claims, valuable for adherence and drug spend.
• Lab feeds, direct results from labs, filling the clinical gaps in claims.
• Health information exchange data, clinical records pulled from other organizations (the interoperability module).
• Patient-reported data, outcomes and experience the patient reports directly.

Assembling the landscape

The analytics job is to fuse all of this, complete-but-late claims, rich-but-walled clinical data, sparse social context, real-time ADT alerts, and supplemental feeds, into one coherent view of each patient and the whole population. No single source is sufficient; the value comes from the combination. The next module is about the hardest part of that combination: moving data across the organizational boundaries that keep it apart.

Key takeaways

• Social data (screening and area-level indices) supplies the missing drivers behind utilization, though it is sparse and imperfect.
• ADT feeds are the highest-operational-value source, triggering timely transitional care in near real time.
• Pharmacy, lab, exchange, and patient-reported feeds round out the landscape; value comes from fusing all sources, not any one.