Training to recognise the early signs of recurrence in schizophrenia

Many people with schizophrenia experience periods of illness followed by relatively stable periods (although symptoms of illness such as hearing voices and seeing things often remain in the background). This means that many people with schizophrenia may become unwell again and need to go back into hospital. Training in early warning signs techniques encourages people to learn, detect and recognise the early warning signs of future illness. Studies indicate that noticing even small changes in signs and symptoms of schizophrenia can often predict future illness and relapse two to 10 weeks later. Early warning training may help to prevent or delay relapse, so reducing the chances of going into hospital. Recognition of early warning signs requires detailed history taking, sometimes with additional techniques such as diary keeping, completion of questionnaires and a plan of action based on anticipated early warning signs. Training can be undertaken by the individual or be group-based, involving health professionals, family members or carers. Successful training seems to require around 12 sessions and involves therapists of high competency.

This review includes a total of 34 studies. It found that there are positive benefits of training in early warning signs. It reduces rates of relapse and re-hospitalisation (but not on time to recurrence). It should be noted that training in early warning signs was mainly used alongside other psychological therapies, so it is not entirely clear what proportion of the positive effect is due to training in early warning signs alone. Moreover, the overall quality of the evidence from these studies was judged to be very low. This means that we do not know if interventions using early warning signs, with or without additional psychological treatments, will have the same beneficial effects outside clinical trials.

Further research is required to decide whether training in early warning signs is effective on its own. Effects on quality of life, satisfaction with care, money spent, and burden of care for carers are unclear, so ideally should be known before training programmes are put into wider use. At this time, there is not enough evidence to support training in early warning signs alone.

This plain language summary was written by consumer, Ben Gray of RETHINK.

Authors' conclusions: 

This review indicates that early warning signs interventions may have a positive effect on the proportions of people re-hospitalised and on rates of relapse, but not on time to recurrence. However, the overall quality of the evidence was very low, indicating that we do not know if early warning signs interventions will have similar effects outside trials and that it is very likely that further research will alter these estimates. Moreover, the early warning signs interventions were used along side other psychological interventions, and we do not know if they would be effective on their own. They may be cost-effective due to reduced hospitalisation and relapse rates, but before mental health services consider routinely providing psychological interventions involving the early recognition and prompt management of early warning signs to adults with schizophrenia, further research is required to provide evidence of high or moderate quality regarding the efficacy of early warning signs interventions added to usual care without additional psychological interventions, or to clarify the kinds of additional psychological interventions that might aid its efficacy. Future RCTs should be adequately-powered, and designed to minimise the risk of bias and be transparently reported. They should also systematically evaluate resource costs and resource use, alongside efficacy outcomes and other outcomes that are important to people with serious mental illness and their carers.

Read the full abstract...

Schizophrenia has a lifetime prevalence of less than one per cent. Studies have indicated that early symptoms that are idiosyncratic to the person with schizophrenia (early warning signs) often precede acute psychotic relapse. Early warning signs interventions propose that learning to detect and manage early warning signs of impending relapse might prevent or delay acute psychotic relapse.


To compare the effectiveness of early warning signs interventions plus treatment as usual involving and not involving a psychological therapy on time to relapse, hospitalisation, functioning, negative and positive symptomatology.

Search strategy: 

Search databases included the Cochrane Schizophrenia Group Trials Register (July 2007 and May 2012) which is based on regular searches of BIOSIS, CENTRAL, CINAHL, EMBASE, MEDLINE and PsycINFO. References of all identified studies were reviewed for inclusion. We inspected the UK National Research Registe and contacted relevant pharmaceutical companies and authors of trials for additional information.

Selection criteria: 

We included all randomised clinical trials (RCTs) comparing early warning signs interventions plus treatment as usual to treatment as usual for people with schizophrenia or other non-affective psychosis

Data collection and analysis: 

We assessed included studies for quality and extracted data. If more than 50% of participants were lost to follow-up, the study was excluded. For binary outcomes, we calculated standard estimates of risk ratio (RR) and the corresponding 95% confidence intervals (CI), for continuous outcomes, we calculated mean differences (MD) with standard errors estimated, and for time to event outcomes we calculated Cox proportional hazards ratios (HRs) and associated 95 % CI. We assessed risk of bias for included studies and assessed overall study quality using the GRADE approach.

Main results: 

Thirty-two RCTs and two cluster-RCTs that randomised 3554 people satisfied criteria for inclusion. Only one study examined the effects of early warning signs interventions without additional psychological interventions, and many of the outcomes for this review were not reported or poorly-reported. Significantly fewer people relapsed with early warning signs interventions than with usual care (23% versus 43%; RR 0.53, 95% CI 0.36 to 0.79; 15 RCTs, 1502 participants; very low quality evidence). Time to relapse did not significantly differ between intervention groups (6 RCTs, 550 participants; very low quality evidence). Risk of re-hospitalisation was significantly lower with early warning signs interventions compared to usual care (19% versus 39%; RR 0.48, 95% CI 0.35 to 0.66; 15 RCTS, 1457 participants; very low quality evidence). Time to re-hospitalisation did not significantly differ between intervention groups (6 RCTs; 1149 participants; very low quality evidence). Participants' satisfaction with care and economic costs were inconclusive because of a lack of evidence.