An item bank is a group of items (questions) that measures a subjective construct such as ‘activity limitation’. An item bank usually contains far more items than would normally be found in a paper-pencil questionnaire. The items are calibrated on the same scale (like a ruler) using Item Response Theory methods.
A CAT system administers the items from the calibrated item bank. Using an algorithm and information from the person’s previous answers, the CAT iteratively selects the next item that most closely matches the person’s level of the construct estimated at that point in the test.
For example, if a person has no difficulty with ‘cooking a meal’, the next selected item would be a more difficult task such as ‘reading the newspaper’. However, if ‘cooking a meal’ is a challenge, an easier task like ‘watching TV’ will be presented.
This process continues until the CAT obtains a precise estimate of the person’s level of the construct. Even though each person may answer different items, scores are comparable as the items belong to the same calibrated item bank.
Optimising PROM assessment in healthcare using CAT
As healthcare moves from a volume-based to value-based system, measuring the impact of disease and treatments from the patient’s perspective using high-quality PROMs is essential.
Use of PROM data in healthcare can help guide clinical care and personalize disease management, support shared decision-making, and improve patient experience.
For PROM assessment to succeed in busy clinics, it needs to be valid, reliable, and fast, and easily integrated into electronic medical records.
Our CATs offer the potential for innovative, feed-forward models of PROM assessment in healthcare.