Reading focus
Syllabus: How Do Psychotic Experiences Develop and Persist?
Contrast the help-seeking samples and transition threshold with Fisher and colleagues’ childhood cohort.
Why you’re reading it. This shows why recruitment and follow-up must accompany any interpretation of clinical risk.
A question to consider. Could the transition estimate be transferred to children answering a population questionnaire? Identify two concrete differences that would need resolving first.
Why this reading
Clinical high-risk research begins with selected people seeking help, not a random sample of everyone who has an unusual experience. This meta-analysis examines how transition estimates change with follow-up and study composition. It is useful for learning what a risk estimate depends on before deciding where it can be applied.
Before you read
Clinical high risk is a research and clinical assessment category defined by specific instruments and recruitment processes. Transition here means meeting an instrument’s psychosis threshold; it is not always a recorded diagnosis of schizophrenia. Cumulative risk counts events over time. A hazard describes their rate among those still at risk, while survival analysis accounts for incomplete observation under assumptions.
Read for
- Compare recruitment and eligibility with the outcome definition. Which features enrich these samples for risk before follow-up begins?
- Distinguish estimates pooled separately at each follow-up from the reconstructed Kaplan–Meier curve. Different study sets can contribute to different time points.
- Inspect attrition assumptions, sensitivity analyses and the diminishing number followed for the longest periods. What uncertainty remains beyond a pooled confidence interval?
Consider afterwards
- Why would applying the three-year transition estimate to an unselected school population be unjustified?
- If two services have different transition rates, what information would distinguish recruitment differences from an effective intervention?
Editorial context
The review includes 130 reports from 74 cohorts, with substantial heterogeneity and uneven geographical coverage. Study-level moderators cannot establish individual causal predictors. Transition is only one outcome: not transitioning does not necessarily mean that distress or functional difficulties have resolved. The paper’s proposed extension of care requires an intervention comparison; the prognostic synthesis itself is not that test.