Why transition risk to psychosis is not declining at the OASIS ultra high risk service: The hidden role of stable pretest risk enrichment
Introduction
The prevention of psychosis has become clinically feasible due to the introduction of the Ultra High Risk (UHR hereafter) construct (Fusar-Poli et al., 2015a) to reliably identify young individuals who are at heightened risk for the development of psychotic disorders (Fusar-Poli and Schultze-Lutter, 2016) - mostly schizophrenia spectrum disorders (Fusar-Poli et al., 2013a) - over the following few years (Kempton et al., 2015). Conversely, there is no evidence that individuals meeting UHR criteria are at increased risk of developing new and incidental non-psychotic disorders compared to individuals assessed for an UHR state but not meeting criteria (Fusar-Poli et al., 2016c, Webb et al., 2015). The meta-analytical prognostic accuracy of the UHR designation is considered good (AUC at 38 months = 0.9) (Fusar-Poli et al., 2015a) and comparable to other preventative approaches in medicine (Fusar Poli et al., 2014). However, declining transition risks from an UHR state to psychosis over recent years has put the field into question. The two-year risk of transition to psychosis from an initial UHR state has shifted from an early 30% (Fusar-Poli et al., 2012) to the current 20% (see Table 4 in Fusar-Poli et al. (2016a)). Declining transition risk is concerning because it can undermine the clinical significance of preventative detection (Fusar-Poli, 2017c) and treatment, yielding negative findings (Fusar-Poli, 2017b). Understanding the reason for declining transition risk is of paramount relevance to overcome these limitations. Earlier studies had suggested that declining transition risk in UHR samples may be due to the fact that treatments were more effective (treatment effect). However, with the largest randomized controlled trials in UHR individuals yielding negative findings (Fusar-Poli, 2017b), there is no strong evidence indicating that the recommended preventative treatments are effective in preventing psychosis (Morrison et al., 2012). In fact, recent studies concluded that treatment effect (Nelson et al., 2016) cannot fully account for the observed decline in transition risks. Another line of research has suggested that the declining transition risk may be due to a dilution effect (Hartmann et al., 2016) i.e. finding more false positives despite individuals meeting the initial UHR criteria (dilution effect) (Yung et al., 2007). The dilution effect was only partially explained by different clinical characteristics of the UHR samples at intake over time periods (Hartmann et al., 2016). While these studies have been entirely focused on transition risk in individuals who meet the UHR criteria (posttest risk), no studies have investigated whether the dilution effect is secondary to a change in the level of risk enrichment of the entire pool of individuals undergoing an UHR assessment (pretest risk, for explanatory details see Fusar-Poli and Schultze-Lutter (2016)). The strong association between pretest risk enrichment and observed transition risk in UHR samples is an established finding and has been confirmed in all UHR samples worldwide (Fusar-Poli et al., 2016d, Fusar-Poli et al., 2016f). Since the prognostic accuracy of the UHR criteria depends on the pretest risk of the sample to which they are being applied (Fusar-Poli, 2017a), it is possible that changes of transition risk over time may parallel changes in the underlying pretest risk of the samples undergoing UHR assessment. Furthermore, the changes in pretest risk enrichment of individuals undergoing an UHR assessment are in turn modulated by outreach campaigns of clinical services, recruitment strategies (Fusar-Poli et al., 2016f) and referral pathways (Fusar-Poli et al., 2016d).
In this study, we first hypothesized that the changes in posttest transition risk over time would be determined by changes in pretest risk over time. To test this hypothesis, we stratified the course of posttest and pretest transition risk over the same fifteen-year referral period in the entire pool of individuals who were undergoing an UHR assessment at the Outreach and Support in South London (OASIS) UHR service (Fusar-Poli et al., 2013b), and addressed the relationship between pretest and posttest risk of psychosis. Our second aim was to investigate potential sociodemographic and referral pathway factors (Fusar-Poli et al., 2016d) that may account for any changes in pretest psychosis risk enrichment over different time periods.
Section snippets
Sample
We included all non-psychotic individuals who were assessed on suspicion of psychosis risk by the OASIS UHR service (Fusar-Poli et al., 2013b). All subjects referred to the OASIS in the period 2001 to 2015 were initially considered eligible. We then excluded those who were referred but never assessed by the team, and those who were already psychotic at baseline. The remaining sample was therefore composed of all non-psychotic subjects undergoing a Comprehensive Assessment of At Risk Mental
Sociodemographic and clinical characteristics of the sample
From 2001 to 2015, a total of 1115 subjects were referred to the OASIS clinic for UHR assessment. Among them, 125 subjects did not undergo the UHR assessment and had no contact with the OASIS service. An additional 280 subjects were already psychotic at baseline (the clinical fate of these subjects is described elsewhere (Fusar-Poli et al., 2016b)). Therefore, a final sample of 710 non-psychotic subjects who underwent UHR assessment was used in the current study (Table 1).
The mean follow-up was
Discussion
To the best of our knowledge, this is the first study to test the hypothesis that the time course of posttest transition risk in UHR samples is determined by the underlying time course of pretest risk enrichment of individuals undergoing UHR assessment. Contrary to our expectations, the posttest risk of psychosis in individuals meeting the UHR criteria at the OASIS service has increased since the first pilot years (2001–2005) and remained stable over the following decade (2006–2010 and
Conclusions
The time course of transition risk to psychosis in UHR services is strictly associated with the time course of pretest risk enrichment. If the latter remains stable over time, as for the OASIS service, no declining transition risk is observed over the most recent years. Pretest risk enrichment is determined by recruitment and sampling strategies. This study confirms the need to control these factors in the UHR field.
Contributors
PFP designed the study, wrote the protocol, and managed the statistical analysis and the first draft of the manuscript. EP, CD, DO, IB, VRC, PM reviewed the manuscript and discussed the findings. All authors contributed to and have approved the final manuscript.
Role of funding sources
This work was supported in part by a 2014 NARSAD Young Investigator Award to PFP.
Conflict of interest
None.
Acknowledgements
None.
References (29)
- et al.
Outreach and support in south London (OASIS), 2001–2011: ten years of early diagnosis and treatment for young individuals at high clinical risk for psychosis
Eur. Psychiatry
(2013) - et al.
Intensive community outreach for those at ultra high risk of psychosis: dilution, not solution
Lancet Psychiatry
(2016) - et al.
Declining transition rates to psychotic disorder in “ultra-high risk” clients: investigation of a dilution effect
Schizophr. Res.
(2016) - et al.
Further examination of the reducing transition rate in ultra high risk for psychosis samples: the possible role of earlier intervention
Schizophr. Res.
(2016) - et al.
Attenuated Psychosis Syndrome: Ready for DSM-5.1?
Annu Rev Clin Psychol.
(2014) The clinical high-risk state for psychosis (CHR-P), version II
Schizophr. Bull.
(2017)Negative psychosis prevention trials
JAMA Psychiat.
(2017)Why ultra high risk criteria for psychosis prediction do not work well outside clinical samples and what to do about it
World Psychiatry
(2017)- et al.
Predicting the onset of psychosis in patients at clinical high risk: practical guide to probabilistic prognostic reasoning
Evid. Based Ment. Health
(2016) - et al.
Predicting psychosis: a meta-analysis of transition outcomes in individuals at high clinical risk
Arch. Gen. Psychiatry
(2012)
At risk for schizophrenic or affective psychoses? A meta-analysis of DSM/ICD diagnostic outcomes in individuals at high clinical risk
Schizophr. Bull.
At risk or not at risk? Meta-analysis of the prognostic accuracy of psychometric interviews for psychosis prediction
World Psychiatry
Antidepressant, antipsychotic and psychological interventions in subjects at high clinical risk for psychosis: OASIS 6-year naturalistic study
Psychol. Med.
Heterogeneity of risk for psychosis within subjects at clinical high risk: meta-analytical stratification
JAMA Psychiat.
Cited by (34)
Meta-analysis of gender differences in transition prevalence among individuals at clinical high risk of psychosis
2023, Asian Journal of PsychiatryDiagnosis, prognosis, and treatment of brief psychotic episodes: a review and research agenda
2022, The Lancet PsychiatryCitation Excerpt :BPD has been found in 2–7% of cases of first-onset psychotic disorder,59 its incidence has not been formally investigated. Because BLIPS or BIPS are defined in help-seeking CHR-P individuals,69–71 their incidence in the general non-help-seeking population is not estimable. Most studies have found a higher incidence of brief psychotic episodes in women than in men,72 but not in LMICs,48,60 in contrast to the higher incidence in men across all psychotic disorders.73
Short clinically-based prediction model to forecast transition to psychosis in individuals at clinical high risk state
2019, European PsychiatryCitation Excerpt :The pre-specified mode was applied to the first Polish CHR-P cohort. There is converging evidence that the actual level of psychosis that is observed in CHR-P individuals is mostly determined by the way these individuals have been recruited, and therefore by their risk enrichment before they undergo CHR-P assessment [8,9,69–71]. As mentioned in the Methods section, our sample is representative of CHR-P services from different countries because the observed risk enrichment is very similar to the pooled meta-analytical mean [8].
Clinical outcomes in individuals at clinical high risk of psychosis who do not transition to psychosis: A meta-Analysis
2022, Epidemiology and Psychiatric SciencesBaseline data of a sequential multiple assignment randomized trial (STEP study)
2022, Early Intervention in Psychiatry