OCD and reassurance seeking: blocking the attempted solution

A client checks the stove twice before she believes it is off, then once more from the doorway. For about ten minutes the doubt lifts. Then it comes back.

Not despite the checking. Partly because of it. Every check teaches her that uncertainty is dangerous enough to warrant action, which makes the next uncertain moment more urgent than the last.

Reassurance is the same mechanism, outsourced

“Are you sure I locked it.” “Are you sure this isn’t serious.” The logic is identical, just delegated. Relief arrives, holds for a short window, then the doubt returns needing a slightly stronger dose to produce the same effect.

Mapped carefully, most OCD presentations involve both forms, with family members participating without ever recognising that they have joined the loop.

Family accommodation as a second solution

A parent who answers the same question fifteen times a day, or a partner who takes over the task whose uncertainty the client cannot tolerate, is running an attempted solution built from love and exhaustion in roughly equal measure. Framing this as the family’s mistake helps nobody.

O’Connor’s (1983) published case of a ten-year-old with an obsessive fear of vomiting offers a gentler route. Instead of withdrawing parental reassurance overnight, the treatment introduced a concrete object the boy could carry in its place, letting the family’s involvement taper on a schedule he could tolerate.

Not every ritual is the same ritual

Nardone’s clinical model distinguishes three variants, and the distinction changes what a prescription needs to interrupt.

Preventive rituals aim to stop a feared event before it happens: checking, avoiding, verifying. Propitiatory rituals ward off a vaguer, more magical sense of danger through a fixed sequence. Reparatory rituals undo or correct something the client believes she has already done wrong.

Reassurance-seeking appears in all three, but what is being sought differs each time. Safety information in the first, general all-clear in the second, forgiveness in the third. This is a clinical classification drawn from case experience, not from trial data, and it earns its place by changing what you prescribe rather than by having been tested.

Blocking without a flat ban

Telling a client to stop checking usually fails. It asks her to win a fight she has been losing for years, and a prohibition tends to generate covert versions of the same behaviour.

Better tolerated: agree a specific, initially short delay before any check or reassurance request is honoured, with the family included where relevant. That changes the timing and shape of the ritual without demanding its disappearance. Where the checking is especially rigid, scheduling it deliberately at a set time can work where the direct request keeps failing.

What the evidence supports

Pietrabissa and colleagues (2016) published their OCD protocol as a one-group observational design and stated openly that no quantitative study had yet been conducted. The protocol was written to help close that gap, not to report that it had closed.

A later case series found significant reduction in Y-BOCS scores across eight treated patients, with five reaching clinical remission (Proietti et al., 2022). Eight patients settles nothing, and the authors do not claim otherwise.

Both papers rest on the same background fact worth passing to any referring colleague: a substantial minority of OCD patients, put at up to 40% in that case series, do not fully respond to CBT. That is the argument for having an additional option, not for replacing first-line care.

The case behind the technique is worked through in the study of an obsessional ritual broken by paradoxical prescription.

Two practical cautions. Obsessions with aggressive, sexual or blasphemous content frequently go unreported for years out of shame, and a client will not volunteer them unless asked directly and without visible reaction. And where insight is poor, so that the feared consequence is held as fact rather than as fear, both the delay technique and the paradoxical version become considerably harder to run and the case usually warrants a psychiatric opinion alongside.