Reply

We are grateful to Trew et al ^{Reference Trew, Rodda and Shergill1} for their comments on our paper. ^{Reference Moncrieff, Pillai, Marston, Lewis, Barnes and Johnson2} We agree with the need to tailor treatment decisions to individual circumstances and to taper antipsychotics very gradually if a decision is made to reduce them.

Our research showed that people who are diagnosed with schizophrenia or recurrent psychotic episodes and who experience a relapse return to their baseline level of social functioning and symptoms within months. ^{Reference Moncrieff, Pillai, Marston, Lewis, Barnes and Johnson2} Our main analysis did not, in fact, show effects on employment, but a more detailed analysis of changes in employment status suggested there may be some effect. However, this involved small numbers and needs replication.

Therefore our results suggest that, if relapse has any long-term impact, it is on the individual’s social circumstances, which most probably reflects the social disruption that relapse can produce rather than an ongoing biological process. We did not support ideas that relapse involves ‘neurological risks’, since these would be expected to be manifested by changes in symptoms and overall functional capacities.

The idea that relapse is a neurotoxic state that leads to progressive decline and treatment resistance is prevalent and has significant implications. For patients, it is frightening and conveys a nihilistic view of the prognosis of the condition. It suggests that recovery is unlikely and stability is precarious, and it may discourage patients from taking an active role in decisions about their treatment. Our results are consistent with most other follow-up studies that suggest a more optimistic picture with no evidence of permanent deterioration following relapse. ^{Reference Curson, Hirsch, Platt, Bamber and Barnes3–Reference Almond, Knapp, Francois, Toumi and Brugha5} This aligns with the views of leading researchers, that schizophrenia is not an inherently progressive condition and that recovery and stability should be expected outcomes. ^{Reference Zipursky, Reilly and Murray6,Reference Zipursky and Agid7}

Nevertheless, we do not deny that relapse can be disruptive to social and personal relationships and cause suffering and distress. Discussions with patients about ongoing antipsychotic treatment should highlight these issues, given the higher risk of relapse following reduction and discontinuation of antipsychotics. However, people should not be given the impression that relapse must be avoided at all costs in order to prevent disease progression and permanent deterioration, since there is insufficient evidence to support this idea.

We concur with the suggested limitations to the generalisability of our results, although several of these, such as lack of diversity and exclusion of patients considered to be at high risk of harm, are common to most clinical trials. In fact, we demonstrated that our sample was reasonably similar to the target population of people in secondary care with an eligible diagnosis from which it was drawn. ^{Reference Freudenthal, Marston, Stansfeld, Priebe and Moncrieff8} We agree that, since our sample consisted predominantly of people with long-term conditions, the results are not necessarily generalisable to those with first or early psychotic episodes. Although it is plausible that relapse may have a greater social and psychological impact if it occurs earlier in the course of the condition, it seems unlikely that it would have distinct neurobiological effects.