In hospital, delirium and time pressure often make direct cognitive testing impractical, so clinicians rely on a proxy, an informant, to estimate a patient's baseline. The 16-item IQCODE has cutoffs for mild cognitive impairment and dementia; the quicker 8-item AD8 only had a single cutoff for "cognitive impairment." This study bridged them.

Using 264 hospitalised older adults and their proxies, the authors built a crosswalk between the two tools. They derived AD8 cutoffs: a score of 2 or more (equivalent to an IQCODE of 3.2) indicates mild cognitive impairment, and a score of 4 or more (IQCODE 3.5) indicates dementia.

For practice this makes the brief AD8 more useful at the bedside, letting a two-minute informant interview distinguish mild cognitive impairment from dementia rather than just flag "impairment." It supports recognising pre-existing cognitive decline in an acutely unwell older patient, which in turn aids delirium detection and discharge planning.

Crosswalk study linking the 8-item AD8 to the 16-item IQCODE in 264 hospitalised older adults.
An AD8 score of 2 or more corresponds to mild cognitive impairment.
An AD8 score of 4 or more corresponds to dementia.
Lets a brief informant interview stage cognition, not just flag impairment.
Supports recognising baseline cognitive decline for delirium detection and discharge planning.

Why it matters

It upgrades a two-minute tool so a busy ward team can distinguish mild impairment from dementia when direct testing is impossible.

The statistics, in plain English

A crosswalk translates scores between two scales using the same sample, so the AD8 cutoffs are only as representative as these 264 acute-care patients; they are a practical bedside aid, not a diagnostic test for dementia.