Early intervention for psychosis: better outcomes in the short term?


What does that old adage say? Prevention is better than cure. But don’t take these words at face value – let’s discuss the evidence.

Early intervention (EI) is a model whereby people presenting with a first episode of psychosis are offered tailored, multi-disciplinary outpatient care, often involving a psychiatrist, a care coordinator, and psychological input as well as occupational and other support (Underwood R, 2016).

EI services are now widespread across the world, including the UK. The NHS is one of the few mental health services that has been subject to Government quality targets, as described here.

In psychosis, EI is known to improve symptom severity, reduce the number of hospitalisations, and boost functioning in society in terms of work and study. However, these effects appear short lived and have not yet been demonstrated beyond five years from a first episode of psychosis (FEP) (Correll CU et al. 2018; Shields G, 2016).

It remains unclear whether early intervention in psychosis, while effective in the short term, can produce lasting positive outcomes in the longer term (i.e., over five years). Using data from 547 individuals enrolled in a randomised controlled trial in Denmark 20 years prior, Hansen et al. (2023) compared EI in psychosis with treatment as usual (TAU).

Are early intervention services for psychosis effective in the long term?

Are early intervention services for psychosis effective in the long term?


This study (Hansen et al. 2023) is the 20-year follow-up of a randomised controlled trial that compared 2-year early intervention of psychosis against mental health community treatment as usual (TAU). The original randomised controlled trial was reported by Thorup et al. (2005).

In summary, the original study randomly selected 547 individuals with first-episode schizophrenia spectrum disorder from the catchment areas of Aarhus and Copenhagen, Denmark between 1998 and 2000 to receive one of the two treatments, EI or TAU.

After the 2-year intervention of EI or TAU, all individuals were followed up with standard community care, and offered to participate in several follow-up studies and interview, including one at 20-years post-treatment. This follow-up included assessments of:

  • Diagnosis status using the schedule for clinical assessment in neuropsychiatry (SCAN 2.1).
  • Positive (e.g., hallucinations or delusions) and negative (e.g., lack of motivation or slow movements) symptoms of psychosis using the Scale for Assessment for Positive Symptoms (SAPS) and the Scale for Assessment of Negative Symptoms (SANS).
  • Ability to do everyday tasks using the Global Assessment of Functioning scale (GAF-F) and the Personal and Social Performance scale.
  • Cognitive functioning using the Brief Assessment of Cognition in Schizophrenia.
  • Quality of life using the World Health Organization Quality of Life-BREF.

Further, data was obtained from Danish national registries about psychiatric service use (e.g., hospitalisations and secondary care contacts, deaths), living conditions (e.g., homeless hostels, supported accommodation), and vocational status.


30% of the 547 patients in the original study took part in the 20 year follow-up post-treatment (164 individuals). There were significant differences identified between the treatment groups of those who took part in the 20-year follow-up before any intervention took place. Specifically, the 2-year EI group were younger overall, tended to be female, had more total years of education, reported being more engaged in work, and showed better performance on measures of general functioning and overall symptoms. This may affect the quality of the results that are based on self-reported questionnaires only. When comparing treatment groups, however, no differences were found in terms of positive symptoms, negative symptoms, global functioning levels, and clinical recovery. Yet it is worth noting that sample sizes were relatively small for each group.

Importantly, all patients were followed-up on using data linkage to national health registries. It was reported that 17% of the original cohort (64 participants) passed away before the 20-year follow-up, despite the mean age being just 46 years. This again highlights the dramatically decreased life expectancy of patients with psychosis (Wallace J, 2017).

Further, thanks to the excellent Danish national registers, the researchers were able to compare the two treatment groups in this study with the whole population (excluding those who had passed away and those who had left the country) on other outcomes of interest. They found:

  • Mortality due to any cause was slightly lower in the EI group (1%) than in the treatment as usual group (15.1%), however, these differences were not statistically significant.
  • No difference in the rate of psychiatric hospitalisation per year among the EI group compared to the TAU group 10 to 20 years after randomisation.
  • No difference in the amount of outpatient secondary care received by the EI group compared to the TAU group.
  • No difference in employment status (24-28% across groups at 20 years) between the EI and the treatment as usual groups.
No long-term differences were found at the 20-year follow-up between patients who received early intervention service or community treatment as usual with regards to mortality, hospitalisation, and employment status.

No long-term differences were found at the 20-year follow-up between patients who received early intervention service compared with community treatment as usual, with regards to mortality, hospitalisation, and employment status.


The authors conclude:

In this 20-year follow-up study of a randomized clinical trial, we did not find any long-term differences among individuals receiving 2 years of EI compared with treatment as usual.

This study is interesting as there seems to be no differences after 20 years of treatment. The authors argue that further research is needed to explore how to prolong the positive short-term outcomes of EI. However, it has been noted that a Danish trial did not find a benefit in extending EI from 2 to 5 years (Albert N, et al. 2017).

What is also striking about this study, as the authors correctly underline, is that:

The mortality rate in our population was 14.1% compared with 3.8% for same-aged individuals in the background population during this period. This is an alarming finding that calls for improved prevention and treatment of physical comorbidities and suicidal behavior among individuals with schizophrenia.

Strengths and limitations

This is an excellent, unique study, due to multiple reasons.

In terms of the strengths, the study is a randomised clinical trial which is a strong research design that is unlikely to be influenced by any differences between groups before any treatment is received. Randomised clinical trials are considered to be the best design for studies testing a treatment, like early intervention in this case. Further, the study authors have been able to follow up the same patients (minus a significant number those who were not available at follow up) for over 20 years. Finally, this study also includes a great number of clinical and non-clinical measures, including those obtained by national registers overcoming issues of loss to follow-up. This allowed the researchers to answer a wide range of questions.

However, the study’s greatest strength (i.e., 20-year follow-up) is also its greatest weakness. Due to the length of time that had passed between follow-ups and the inevitable change to participants’ circumstances, the authors were only able to interview 30% of the original sample at the 20-year follow-up. This very high level of drop-out means the findings are unreliable.

Another important weakness of the study is that the people that were not followed up (because they did not respond) were on average doing significantly worse to start with, as well as showing differences in several demographic characteristics compared to the rest of the cohort. It is therefore possible that if these people had been included in the study – in other words, if they did not drop out – the findings might have been quite different.

That being said, linkage with Danish registries allowed the authors to study outcome measures often difficult to collect information on, such as secondary care contact, hospitalisation, and death. This is an important strength to acknowledge as data linkage is often quite difficult to achieve in practice.

Those who did not attend the follow-up 20-years later were on average doing significantly worse to begin with. It is possible that the results would be quite different were follow-up information on these individuals available.

Those who did not attend the follow-up 20-years later were on average doing significantly worse to begin with. It is possible that the results would be quite different were follow-up information on these individuals available.

Implications for practice

I believe this study has one major implication for clinical practice – it highlights the urgent need to improve the survival rates of patients with psychosis.

It is unacceptable that 13-14% of people with a FEP in this study passed away by the 20-year follow-up, especially given the average age for these patients was just 46. This rate is likely to be the result of multiple factors, including death by suicide (3.5%). We know that the main cause of death in these patients is cardiovascular-related, and more research is needed to investigate why this is the case, and what to do about it.

In terms of early intervention for psychosis – this has been demonstrated in previous research to be effective in the short term, as discussed in the introduction, and should continue to operate. However, we need to find better, more effective long-term treatment options.

Mortality rates remain too high in patients with psychosis and more effective long-term treatment options are urgently needed.

Mortality rates remain too high in patients with psychosis and more effective long-term treatment options are urgently needed.

Statement of interests

Emanuele has no conflicts of interest to report. He heard the first author of this paper speak at the Schizophrenia International Research Society (SIRS) conference in Toronto in May 2023. He is employed by Imperial College London and conducts research, among other things, on health outcomes in psychosis.


Primary paper

Hansen HG, Starzer M, Nilsson SF, Hjorthøj C, Albert N, Nordentoft M. (2023) Clinical Recovery and Long-Term Association of Specialized Early Intervention Services vs Treatment as Usual Among Individuals With First-Episode Schizophrenia Spectrum Disorder: 20-Year Follow-up of the OPUS Trial. JAMA Psychiatry.2023;80(4):371–379. doi:10.1001/jamapsychiatry.2022.5164

Other references

Underwood R. Charting the evolution of early intervention in psychosis. The Mental Elf, 14 April 2023.

Gov.uk Mental health services: achieving better access by 2020. 8 Oct 2014.

Correll CU, Galling B, Pawar A, et al. (2018). Comparison of Early Intervention Services vs Treatment as Usual for Early-Phase Psychosis: A Systematic Review, Meta-analysis, and Meta-regression. JAMA Psychiatry 2018 75 555–565.

Thorup A, Petersen L, Jeppesen P, et al. (2005) Integrated treatment ameliorates negative symptoms in first episode psychosis—results from the Danish OPUS trial. Schizophrenia Research 2005 79(1) 95-105.

Shields G. Early intervention in psychosis services: better outcomes, improved costs. The Mental Elf, 21 December 2016.

Wallace J. Life expectancy in schizophrenia and years of potential life lost. The Mental Elf, 12 April 2017.

Albert N, Melau M, Jensen H, et al. (2017) Five years of specialised early intervention versus two years of specialised early intervention followed by three years of standard treatment for patients with a first episode psychosis: randomised, superiority, parallel group trial in Denmark (OPUS II). BMJ 2017 356 (i6681).

Photo credits

Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+