The results of Penttilä et al’s meta-analysis emphasised the importance of the duration of untreated psychosis (DUP) in long-term recovery from schizophreniform illness. ^{Reference Penttilä, Jääskeläinen, Hirvonen, Isohanni and Miettunen1} Timely initiation of effective treatment has been demonstrated to improve outcome, but the modality of treatment is currently under much debate. Robust evidence exists for the efficacy of antipsychotic medication ^{Reference Leucht, Arbter, Engel, Kissling and Davis2} but recent studies have proposed psychological interventions, specifically cognitive-behavioural therapy (CBT), as an alternative first-line treatment.

In a recent randomised controlled trial, CBT was used as a single intervention, instead of conventional antipsychotic treatment. ^{Reference Morrison, Hutton, Wardle, Spencer, Barratt and Brabban3} To our complete surprise, one of the exclusion criteria was treatment with antipsychotic drugs. We wonder how ethical approval was granted, despite Tiihonen et al’s robust demonstration of reduced mortality over a considerable follow-up period for patients receiving antipsychotic medication. ^{Reference Tiihonen, Lannqvist, Wahlbeck, Klaukka, Niskanen and Tanskanen4} We feel that this will set a dangerous precedent of offering psychological treatment as an alternative to evidence-based treatment. In a clinical setting, adherence to drug treatment is already a significant issue and there is potential to reinforce the idea that antipsychotic medication is harmful and unnecessary. We feel that this would further disadvantage an already vulnerable group of patients.

This issue has recently received a fair degree of coverage in the media, with articles such as Freeman & Freeman’s piece in The Guardian fuelling long-held popular beliefs that antipsychotics are ineffective and in fact damaging to health. ^{Reference Freeman and Freeman5} Given the well-documented drawbacks of antipsychotic drugs, it is understandable that patients and professionals will invest hope in non-drug alternatives. However, a large meta-analysis with over 3000 participants shows at best a small effect size for CBT. ^{Reference Jauhar, McKenna, Radua, Fung, Salvador and Laws6} In reference to Penttilä et al’s paper, we would be interested to read subgroup analyses of specific first-line treatments and wonder if outcomes would differ between modalities.

While we would endorse any treatment, drug or non-drug based, that is proven to reduce DUP, it is vital that we do not lose sight of the fact that antipsychotics are the only evidence-based first-line therapy in psychotic illness.