What works in loss prevention

Connected Insurance Hub grades published evidence on what risk technology and prevention do to crashes, injuries and insured losses. The register holds 125 findings, 50 of them from UK sources, each tiered by who produced it and read against the Five Gates. This page shows the headline reading. The full register, with sources, designs, caveats and guidance for underwriters and risk managers, is open to members.

What needs to be true for an intervention to reduce loss

Interventions that reduce crashes, injuries or claims share six conditions. Interventions that fail usually miss one or more of them.

Interventions by strength of evidence

Grouped by strength of evidence. The records measure different outcomes with different designs, so a single league table would mislead.

Show
A further 28 interventions have only vendor-produced, anecdotal or behaviour-only evidence so far. They are listed in the members' register with the evidence each would need. Join free to see them or open the register.
Advisory

Evidence and pilot design review, £2,950

An independent read of the evidence behind a technology proposition, and a pilot design that produces evidence an underwriter can use.

See the reviewDiscuss your opportunity →

Gradings describe study design and evidence, and say nothing about any organisation's intent. They are Connected Insurance Hub's reading for discussion, and not a recommendation to any insurer on pricing or terms. Submit evidence or a correction. Register version 0.10.2, 2026-11-12.