When a patient has a known visual field deficit from an injury that predates the current stroke, the examiner should score item 3 on the NIHSS based on the patient's current visual field performance, not the historical baseline, and should document the preexisting deficit separately so the score reflects the acute event's contribution to the examination. The score itself is determined by what the patient demonstrates during the test, but the interpretation and any treatment decisions must account for confounding. This distinction between scoring and attribution is the core principle examiners need to understand.
Mechanism: What Item 3 Actually Measures
NIHSS item 3 is a bedside confrontation visual field test. The examiner stands at eye level with the patient, covers one eye, and asks the patient to report or point to fingers moving in each of the four quadrants while the examiner simultaneously tests both eyes open to compare visual fields. The item scores 0 for no visual field loss, 1 for partial hemianopia or quadrantanopia, 2 for complete hemianopia, and 3 for bilateral hemianopia or blindness, including cortical blindness.
The mechanism being tested is the integrity of the optic pathways from retina through the optic chiasm, optic tract, lateral geniculate nucleus, optic radiations, and occipital cortex. A deficit anywhere along this pathway produces a characteristic field defect. The NIHSS item does not localize the lesion; it only grades the functional impact on the visual field. This is why the same score can arise from a retinal lesion, a temporal lobe stroke, or an occipital infarct.
The critical mechanistic point for this scenario is that item 3 is a performance-based item, not a history-based one. The examiner scores what is observed during the examination, not what is reported about the past. A preexisting deficit that is unchanged from baseline will produce a positive score, but that positive score does not indicate new injury. The NIHSS was designed as a stroke severity scale, not a lesion-localization instrument, and its items measure current deficit regardless of cause.
One level deeper: the confrontation method itself has known limitations. It detects roughly 50 to 70 percent of field defects compared with formal perimetry, and its sensitivity depends on the examiner's technique and the patient's attention and cooperation. A preexisting deficit may make the examination harder to interpret because the examiner cannot easily distinguish the old boundary of the field from any new extension. This is why the score is only part of the clinical picture.
Quantified Context: How Often This Occurs and What the Numbers Show
Preexisting visual field deficits are more common than many examiners assume. Population-based studies of older adults find that roughly 10 to 20 percent have some form of visual field defect on formal perimetry, with glaucoma, prior stroke, diabetic retinopathy, and age-related macular degeneration among the leading causes. About 1 percent of adults over 50 have a prior stroke-related hemianopia that never fully resolved. Among patients presenting with acute stroke, a substantial minority have a documented history of prior visual injury.
The NIHSS itself has been studied for inter-rater reliability. Item 3 shows moderate agreement, with kappa values in the range of 0.4 to 0.6 in most training studies, lower than motor items but comparable to other sensory items. The lower reliability reflects the subjectivity of confrontation testing and the difficulty of distinguishing partial from complete hemianopia in an uncooperative or aphasic patient. When a preexisting deficit is present, reliability drops further because examiners disagree about whether a finding represents old or new deficit.
The clinical stakes are quantifiable. A change in NIHSS of 1 point alters the classification of stroke severity at commonly used thresholds, such as 5 or 10 points. A falsely elevated item 3 score can push a patient across a treatment threshold. Thrombolysis eligibility for alteplase and tenecteplase requires a measurable neurologic deficit, but a preexisting deficit can satisfy that criterion even when the acute event is minor. Meanwhile, a falsely low score can obscure a new postchiasmal lesion. The number matters because decisions follow it, so the examiner must be precise about what is being measured.
Comparison of Scoring Approaches
| Approach | What it scores | Advantage | Risk | Recommended |
|---|---|---|---|---|
| Score current field only | Observed deficit at examination | Standardized, reproducible | May reflect old injury | Yes, as the recorded score |
| Score difference from baseline | Increment over old deficit | Better estimates acute contribution | No validated NIHSS method | No formal role |
| Score as if no deficit | Preeexisting deficit ignored | Corrects for bias | Invalidates the item | No |
| Score 0 if language/attention impaired | Untestable assumed normal | Corrects for poor cooperation | Underestimates deficits | No, use UN if truly untestable |
| Score UN (untestable) | Cannot be assessed | Honest about limits | Requires justification | Yes, when genuinely untestable |
| Formal perimetry then score | Gold-standard field | Accurate | Not bedside, delays care | Adjunct only |
The table makes the central choice explicit. The recorded item 3 score should reflect observed performance. The examiner then documents the baseline deficit in the chart, in the imaging or clinical comment section, and in communication with the treating team. This two-track approach preserves both standardization and clinical accuracy.
Edge Cases and the Counterintuitive Detail
The counterintuitive professional caveat is this: a score of 1 on item 3 is often the most dangerous score to interpret, because partial hemianopia or quadrantanopia can represent either a small new lesion or an old stable deficit, and the confrontation test cannot reliably distinguish them. A score of 2 or 3 is more alarming but also more likely to prompt imaging review, while a score of 1 often slips through with a note of "visual field cut" and no further attribution. The examiner should flag any item 3 score of 1 or higher with an explicit comment about baseline status.
Aphasic patients present another edge case. If a patient cannot reliably report or indicate fields because of language impairment, the item should be scored UN (untestable) rather than assumed normal. The NIHSS allows scoring based on behavioral observation, such as visual tracking, but a preexisting deficit plus aphasia makes behavioral inference unreliable. The examiner should test with visual threat responses and document the limitation.
Patients with cortical blindness from a prior bilateral occipital injury present a special problem. Their score is 3 regardless of the current stroke, and a new posterior circulation event would not raise the score further. In this case, the examiner should document that item 3 is at ceiling and rely on other items and imaging to assess the acute event. The same principle applies to any patient whose preexisting deficit already maximizes the item.
Unilateral neglect can masquerade as hemianopia, and the two can coexist. The NIHSS separates them into different items, but examiners sometimes score visual field loss when the true finding is neglect. A preexisting deficit adds another layer. Careful testing of each quadrant with the examiner's hands on both sides, and observation during the neglect items, helps separate the two.
Common Misconceptions
The most common misconception is that a preexisting deficit should be subtracted from the score or that the examiner should "score to baseline." There is no validated method for adjusting NIHSS item 3 based on prior deficits, and attempts to do so introduce unreliability. The scale scores the current examination.
A second misconception is that a known old deficit should be scored as 0 because it is not attributable to the acute stroke. This conflates the score with causation. The score is a description; the attribution is a separate clinical judgment.
A third misconception is that formal perimetry should replace the bedside examination before scoring. While automated perimetry is more sensitive, it is not practical in acute stroke, requires patient cooperation, and is not the reference standard for the NIHSS. The NIHSS was validated with confrontation testing.
A fourth misconception is that the presence of any visual field deficit automatically qualifies for thrombolysis. The treatment decision depends on the total picture, time window, imaging, and contraindications. The item 3 score is one input, not a trigger.
Practical Actions for the Examiner
Test all four quadrants in each eye and with both eyes open, and record the exact pattern, such as left inferior quadrantanopia, rather than only the score. Ask the patient or family about prior visual problems, eye surgery, glaucoma, and previous stroke, and document the reported baseline before scoring. Enter the current item 3 score per the standard scale, then add a separate comment noting the preexisting deficit and its source if known. Flag any score of 1 or higher for the treating clinician with an explicit statement about whether the deficit appears old, new, or extended. Use UN only when the item genuinely cannot be assessed, and document why, such as aphasia or inability to cooperate. Communicate the ambiguity to the stroke team before treatment decisions are finalized, especially when the total score sits near a threshold.
These steps take about two minutes and prevent the most common documentation and communication failures. The examiner should say aloud and write down the phrase "preexisting visual field deficit, item 3 reflects current examination" whenever the situation applies.
Limitations and What the Evidence Cannot Tell You
The NIHSS was not designed to distinguish acute from chronic deficits, and no version of the scale provides an adjustment rule for preexisting injury. The evidence base for scoring in this scenario is thin: guidance from the original NIHSS publications and subsequent training materials emphasizes scoring the observed examination, but does not offer a validated protocol for attribution. Examiners should not expect the scale to solve a problem it was not built for.
Confrontation testing remains operator-dependent and imperfect, and a preexisting deficit reduces the examiner's ability to detect a new extension of the field cut. Imaging, particularly diffusion-weighted MRI and perfusion studies, may clarify whether the acute lesion involves the visual pathways, but imaging also has limits in posterior circulation and small cortical lesions. The score, the history, the examination, and the imaging must be reconciled together, and the examiner should be candid in the record about residual uncertainty.
This article is general educational information about NIHSS scoring, not a diagnosis, treatment plan, or substitute for clinical judgment. It does not promise any outcome, and it should not be used to make treatment decisions for an individual patient. Readers should consult a qualified clinician and their local stroke protocols for patient-specific guidance, and should contact local emergency services for any immediate medical concern.