1. Neck Injury Case Intake and Root-Cause Question Framing

Why intake framing decides the whole analysis

A neck-injury incident report arrives as narrative: a collision description, occupant details, a few measured or estimated loads, and a clinical impression. If you start computing Nij or reconstructing kinematics straight from that narrative, you silently commit to a mechanism before you have tested it. The intake task is therefore not administrative — it is the point where you convert unstructured evidence into a structured root-cause question set that keeps three layers separate:

• Mechanism — what kind of relative motion and loading produced the injury (e.g., rearward head rotation with neck extension, lateral bending, axial compression with flexion).
• Loading path — how that motion was transmitted through restraint, seat, and body (belt geometry, head restraint contact, seatback deflection, torso slip).
• Tissue-level failure hypothesis — which structure is suspected to have failed and by which mode (e.g., facet capsule strain, disc shear, vertebral artery stretch, nerve-root compression).

Keeping these layers distinct prevents the classic error of treating a clinical finding as proof of a mechanism, or a measured load as proof of a tissue failure. Each layer gets its own questions, its own evidence needs, and its own falsification criteria.

From raw report to evidence inventory

Before writing any question, extract every factual claim from the report and tag it by source reliability. A practical intake pass produces four buckets:

Anything in the last bucket is a candidate for a question, not a premise. For example, "occupant was looking over the left shoulder at impact" is an inference unless a witness or video confirms it — and it changes the lateral-bending hypothesis substantially.

Building the question set

A well-formed root-cause question is answerable with specific evidence and can be shown false. Vague questions ("was the neck injured?") are useless; over-specific ones ("was the left facet capsule strained 18%?") presume the answer. Aim for questions of the form: Which [mechanism / path element / tissue] is consistent with [evidence], and what observation would rule it out?

For a rear-impact case with a low-set head restraint, a structured set might look like this:

1. Mechanism: Did the head rotate rearward relative to the torso before head-restraint contact, or was contact established early enough to limit extension? Falsified if head-to-restraint gap and pulse timing show contact before peak relative rotation.
2. Mechanism: Was there a lateral component (pre-impact head turn, oblique impact) that would shift loading from pure extension to combined extension–lateral bending? Falsified if impact direction and occupant orientation are both near-sagittal.
3. Loading path: Did the shoulder belt anchor the torso sufficiently that the head-neck complex absorbed the relative motion, or did torso ride-up reduce the effective restraint? Falsified if belt force and seatback deflection indicate a stable torso.
4. Loading path: Was the head restraint positioned to contact the occiput, or did it contact lower and act as a fulcrum? Falsified by measured backset and contact height.
5. Tissue hypothesis: Is the clinical picture more consistent with capsular/ligamentous strain from extension, or with nerve-root irritation from combined motion? Discriminated by symptom distribution and imaging, not by load alone.

Notice that each question names the evidence that would settle it. That is what makes the set usable downstream: the kinematic reconstruction and Nij calculation in later work become tests of these questions, not free-standing exercises.

Separating the layers visually

The diagram below shows how a single incident report feeds three parallel question tracks, and how each track's evidence requirements differ. The key structural point is that the tracks are not sequential: a tissue hypothesis can be revised by a mechanism finding, and a loading-path finding can eliminate a mechanism without touching the tissue question.

Worked intake example

Consider a rear-impact report with these stated facts: Δv ≈ 18 km/h, occupant restrained by a three-point belt, head restraint adjusted to its lowest position, complaint of left-sided neck pain and tingling into the left hand beginning the next morning. The intake pass yields:

• Measured: Δv, impact direction (near-sagittal), belt force trace available.
• Documented: head-restraint height setting, seatback angle, occupant height.
• Clinical: left radicular symptoms with delayed onset.
• Inference to flag: pre-impact head orientation is unknown; "looking forward" is assumed, not confirmed.

The resulting question set separates cleanly. Mechanism: was relative head rotation predominantly sagittal extension, or was there a lateral component consistent with the left-sided symptoms? Loading path: with the restraint at its lowest setting, did the head restraint contact the occiput or act below it, and did the torso remain anchored? Tissue: do the delayed left radicular symptoms favor nerve-root involvement over capsular strain, and does that change which mechanism matters most?

Note what the intake does not do: it does not compute Nij, does not reconstruct kinematics, and does not conclude a cause. It produces a prioritized, falsifiable question set — and it explicitly records that the pre-impact head orientation is the single assumption most capable of flipping the mechanism track. That flag tells the next analyst exactly where to spend effort.

Intake checklist

Use this as a repeatable gate before any quantitative analysis begins:

1. Every claim in the report is tagged measured / documented / clinical / inference.
2. At least one mechanism question, one loading-path question, and one tissue question exist, each with a stated falsification condition.
3. The single most consequential assumption is named explicitly, with the observation that would overturn it.
4. No question presumes its own answer, and no clinical finding is used as proof of mechanism.
5. Evidence gaps are listed as open items rather than filled with plausible defaults.

When these five conditions hold, the question set is ready to drive the kinematic reconstruction and load-based criterion work that follow — and any later result can be traced back to the specific question it was meant to test.