Key Takeaways

• The AHA GWTG Stroke registry measures care in minutes (last known well, door-to-needle, door-to-thrombectomy), but the timeline is scattered across EMS, ED, nursing and physician notes that don't always agree with each other.

• A difference of minutes in one timestamp can change whether a case meets a time-based benchmark, so accuracy depends on an abstractor reconciling conflicting evidence by hand.

• At White Plains Hospital, AI-guided abstraction cut stroke abstraction time by double digits and held validated accuracy above 98%, including on the hardest fields to locate manually.

• Faster abstraction clears backlogs and returns nurses and quality specialists to the improvement work the registry data is meant to drive.

Why GWTG Stroke Abstraction Is Both High-Value and High-Difficulty

Stroke care is measured in minutes. Last known well. Door to imaging. Door to needle. Door to thrombectomy.

Learning from those measures requires an abstractor to reconstruct what happened from a record created by many people during one of the most time-sensitive episodes in medicine. That makes Get With The Guidelines Stroke abstraction both highly valuable and unusually demanding.

This is the second post in a series looking at what registry-specific abstraction actually demands. For the bigger picture on why registries are more valuable than ever and where the strain is coming from, see Clinical Registries Are More Valuable Than Ever. The Work Behind Them Is Under Strain.

GWTG Stroke Turns Individual Charts Into a System-Wide View of Care

The American Heart Association's Get With The Guidelines Stroke program is used by more than 80% of US hospitals. Participating organizations use the registry to:

• Benchmark performance

• Support certification and state designation requirements

• Identify where stroke care can improve

The value comes from standardization. When hospitals use the same definitions and measures, a stroke program can compare its performance across sites and against national peers. Quality leaders can see whether a delay is isolated or recurring, whether a process change produced the intended result, and where variation warrants a closer look. Consistent comparison begins before the data reaches the registry.

Why GWTG Stroke Abstraction Starts With a Scattered Timeline

The stroke timeline is scattered across the medical record:

• EMS documentation may contain the first account of symptom onset

• The emergency department records arrival and triage

• Nursing notes capture assessments and medication administration

• Physician notes explain clinical decisions

• Radiology, pharmacy and procedural documentation add their own timestamps

The same event can appear more than once, with times that differ by a few minutes or much more. One note may reflect when the event occurred. Another may reflect when it was documented. Some entries are rounded. Others are copied forward. The abstractor has to determine which source the registry specification calls for and how to handle conflicts.

Last known well, the last time the patient was confirmed symptom-free, can be especially difficult. It may be documented only in narrative text as a family member's recollection. Treatment decisions may depend on a chain of events described across several notes rather than one discrete field. Even when a timestamp is available as structured data, the clinical meaning needed for abstraction may sit somewhere else.

The abstractor reconstructs the clinical story one piece of evidence at a time.

How Small Chart Abstraction Differences Change the Performance Picture

GWTG Stroke includes detailed data elements and measures designed to make performance comparable. Its rigor gives the registry value and makes abstraction exacting:

• Overlapping fields may represent the same clinical fact in different ways

• A value entered inconsistently can trigger a data quality issue

• Definitions and measure logic evolve as guidelines change

• Abstractors need to keep current while applying the specification consistently across cases

GWTG Stroke relies on training, validation and automated checks to catch errors before they reach the registry. Even so, research comparing local entries against independent physician review has found variability in specific history elements. The specification can be exact. The judgment applying it still varies.

A difference of minutes can affect whether a case appears to meet a time-based benchmark. A documentation conflict can change which patients appear eligible for a measure. Abstraction shapes how the organization understands the care it delivered, with consequences for certification, benchmarking, and quality improvement.

Why Manual Stroke Chart Abstraction Delays the Feedback Teams Need

Stroke data becomes valuable when it helps improve the next patient's care.

That requires quality teams to move from measurement to investigation:

• If door-to-needle performance changes, they need to identify where the delay entered the workflow

• If one site outperforms another, they need to understand what is different

• If an outlier case exposes a process gap, they need time to bring the right clinical teams together and address it

Manual abstraction competes with that work for the same specialized staff. As case volume grows, abstraction consumes more of the week. Backlogs put more distance between the patient encounter and the insight, so performance improvement becomes retrospective when it should be responsive.

How AI-Guided Chart Abstraction Changes Where Expertise Is Applied

Layer Health uses AI to read the full record and surface the evidence relevant to each registry variable. For a time-based measure, that can mean identifying candidate timestamps across notes, showing where each one came from, and proposing the answer that best fits the specification.

Layer Health makes my work more efficient by surfacing variables buried deep in the documentation—details I haven't reached yet. This saves me time and enhances the completeness and accuracy of my abstractions.

Stroke Nurse Abstractor, Intermountain Health

The abstractor still reviews the evidence and makes the final determination. They begin with the relevant parts of the chart already in view, resolve any ambiguity and confirm the answer.

At White Plains Hospital, a Comprehensive Stroke Center participating in both AHA Get With The Guidelines registries, Layer Health cut stroke abstraction time by double digits and held validated accuracy above 98%, including on the fields hardest to locate manually. The team worked through its backlog and absorbed rising volume without the additional registry spend it had planned.

See the proof: How White Plains Hospital Cut Registry Abstraction Time in Half Without Sacrificing Accuracy

Recovered Capacity Returns Quality Staff to Stroke Improvement Work

Faster stroke abstraction gives the health system more flexibility to address backlogs, increase case coverage or absorb growth within its existing registry spend.

For hospitals building toward stroke center certification, the constraint is often capacity rather than clinical capability. Faster abstraction lets an existing quality team take on a new registry without hiring a dedicated team or outsourcing the function.

Faster abstraction can also return experienced nurses and quality specialists to the work that needs their judgment most, including:

• Reviewing cases that missed a benchmark

• Tracing the source of delays across departments

• Coaching teams on documentation

• Testing whether a new protocol improved performance