LONGITUDINAL FIT ADHERENCE ANALYSES USING ITALIAN DATA

Aims To analyze patterns of participation among people invited for FIT screening and to assess the impact of screening history on the main performance parameters of the test.

Methods Screening protocol: biennial single sample FIT (Oc Sensor, Eiken Co. Japan), with a positivity cut-off set at 20 µg f-Hb/gr faeces (100 ng/ml). The kit is delivered at the pharmacies

Outcomes: Adherence, positivity rate (PR), detection rate (DR), positive predictive value (PPV).

Individuals were classified as Consistent attenders (attended all rounds invited); Inconsistent attenders (attended some but not all rounds invited); Never attenders (attended no rounds invited), based on their screening history prior to index round.

We analyzed participation rates, PR, DR and PPV for advanced neoplasia (AN- CRC+ advanced adenomas) at the fourth screening round, stratifying by screening history.

Results The participation in the fourth round was predicted by the numbers of previous tests performed: subjects who attended more recent invitations were more likely to respond, within each sub-group of subjects having attended the same number of previous invitations.

PR, and DR showed an inverse association with the number of previous tests preformed: PR: 10.3%, 8.1%, 6.1%, 4.9% for never attenders, subjects having attended 1, 2 and 3 previous invitations respectively. AN DR: 3.1%, 1.9%, 1.3%, 0.9% for never attenders, subjects having attended 1, 2 and 3 previous invitations respectively.

The proportion of subjects who respond to all screening invitation varies across different programs.

It was around 49% after 5 rounds in a program showing a high participation rate in each round

Conclusions The number of previous negative examinations and possibly the interval between them are associated with the subsequent likelihood of a positive FIT result, as well as with the AN PPV and DR at subsequent rounds. The group of inconsistent attenders may include subjects at different risk