Dropping a safety behavior

Safety behaviours maintain anxiety because they prevent the client from finding out the feared outcome doesn’t happen. The phone gripped in the meeting, the route home that avoids the high street, the friend brought along to the social event, the constant scanning for exits. Each behaviour gives the client short-term relief and long-term confirmation that the feared situation requires the safety behaviour to be survivable. Dropping the behaviour, in a planned, contained way, is what shifts the underlying belief.

This worksheet is the planning structure for that drop. Not a generic “expose yourself to anxiety” instruction. A specific situation, picked deliberately, with a predicted outcome written in advance and a place to record what actually happened. The framing is experiment, not endurance test. The point is data the client can use, not proving they can tough it out.

Use it after the client has identified their safety behaviours, ideally with the safety-behaviours-list and safety-behavior-tracker worksheets. Don’t run experiments before the client knows what their behaviours are, because they’ll perform the experiment without dropping the behaviour, and the data won’t mean anything.

The clinical decisions worth making with the client when they fill it in. The pitch of the experiment matters. Too easy, and the client confirms they can do hard things while bringing safety behaviours along. Too hard, and the experiment fails, the avoidance loop strengthens, and the client is more reluctant to try again. Aim for moderate distress, around 50–70 on a 0–100 SUDS scale. That’s where the learning happens.

The obstacles section matters more than therapists usually treat it. Most failed experiments fail at the planning stage, not the execution stage. The client didn’t think through the rainy-day version, the running-late version, the friend-cancelling version. Push for at least two real obstacles and a written response to each.

The clinical pattern to catch in the post-experiment write-up. Clients who completed the experiment and rated the actual distress significantly lower than predicted are generating the data the cognitive work needs. The gap is the lever. Read it together in session and let the client see they’ve just produced their own evidence. Clients whose actual distress matched the prediction usually didn’t fully drop the safety behaviour. Worth checking which version of the behaviour they kept.

In my-cbt, the worksheet is one of the bundled system templates. Assign it for a specific upcoming situation with a personal message that frames the work as data collection. The submission saves in the case file alongside the safety-behavior tracker so you can see the experiment in the context of the broader pattern.