Patient navigation and posting the test kit beat everything else

Design

systematic review and network meta-analysis of 76 randomised clinical trials, GRADE assessed, with P scores and rankograms

Population

adults eligible for colorectal cancer screening across 8 intervention strategies including usual care

Primary outcome

colorectal cancer screening uptake

Effect

patient navigation RR 1.58 (95% CI 1.23-2.02) and mailed FIT outreach RR 1.36 (1.07-1.74) vs usual care; mailed FIT RR 3.12 (1.70-5.71) where baseline uptake was below 30%

A network meta-analysis of 76 randomised trials compared eight strategies for increasing colorectal cancer screening uptake: patient navigation, mailed faecal immunochemical test outreach, educational multimedia, reminder only, choice-based outreach, colonoscopy outreach, multistep approaches and usual care.

Two strategies clearly outperformed usual care. Patient navigation gave a risk ratio of 1.58 (95% CI 1.23-2.02) and mailed faecal test outreach 1.36 (1.07-1.74). Educational multimedia (1.27, 0.91-1.78) and reminders alone (1.24, 0.98-1.57) did not reach significance. Choice-based outreach and colonoscopy outreach were no better than usual care. Head to head, mailed faecal outreach beat colonoscopy outreach (1.35, 1.11-1.63) and navigation beat reminders alone (1.48, 1.14-1.94).

The subgroup finding is the one to plan around. Where baseline uptake was below 30%, mailed faecal outreach was dramatically the most effective (RR 3.12, 1.70-5.71); where baseline uptake was already 30% or above, educational multimedia performed best. That means the right intervention depends on where a service currently sits, and the instinct to offer a choice of tests — which sounds patient-centred — did nothing. Sending the kit unasked is what works when uptake is low.

• Measure your current uptake first; it determines which intervention to choose
• Below 30% uptake, post the faecal test kit rather than inviting people to request one
• Offering a choice of test did not beat usual care — the decision itself is a barrier
• Reminders alone are not enough; navigation means a person, not a message
• Certainty of evidence was moderate to low, so treat the rankings as a guide rather than a hierarchy

Why it matters

It replaces a menu of plausible screening interventions with a rule that depends on one number you already have.

Don't overread it

Moderate-to-low certainty evidence and indirect comparisons — the ranking between the top two strategies is not established.

The statistics, in plain English

A network meta-analysis compares strategies that were never tested against each other by linking them through common comparators — useful, but the indirect comparisons are weaker than the direct ones. The subgroup risk ratio of 3.12 in low-uptake settings has a wide interval (1.70-5.71) and comes from fewer trials, so the direction is dependable and the magnitude is not.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.