To screen or not to screen in first-episode psychosis

Systematic review and meta-analysis of first episode psychosis

Dr Graham Blackman, King’s College, London, UK presented the outcome from a systematic review and meta-analysis seeking to determine the prevalence of any and clinically-relevant radiological abnormalities in those presenting with first episode psychosis (FEP).

Studies published between 1978 and the present day were collected. Those included in the meta-analysis reported the frequency of intracranial abnormalities based on structural magnetic resonance imaging (MRI) assessment of FEP patients; studies reporting aggregated neuroimaging findings or patients with neurological findings were excluded.

There is no international consensus on submitting those experiencing first-episode psychotic episodes to a structural MRI examination routinely

A total of 12 valid studies were collated that provided data on 1,516 FEP patients. Suspected medical causes had been screened-out in all 12 studies through examination of patient histories, assessment and/or examination.

The prevalence of any abnormality was 26% with a corresponding number needed to scan (NNS) of 4, i.e., four patients would need to be scanned to detect one abnormality. As indicated using the I² statistic, the level of heterogeneity among the included studies was high (92%), mainly due to a single influential outlier. Clinically-relevant abnormality prevalence was 6%, with an NNS of 17. The I² in this group indicated high levels of heterogeneity, but with no outliers (76%).

The prevalence of any abnormality was 26% with a corresponding number needed to scan (NNS) of 4; clinically-relevant abnormality prevalence was 6%, with an NNS of 17

The age range covered by the selected studies was 20-61 years. Age was not a significant modifier of the presence of structural abnormality.

The most accurate estimate of prevalence of structural abnormality in FEP to date

Dr Blackman suggested the data presented gave the most accurate estimate of prevalence of structural abnormality in FEP available to date despite its limitations. For instance, the most unwell patients were excluded, classification thresholds of abnormality of MRI scans likely varies between studies and the exploration of heterogeneity between studies varies as well.

Age was not a significant modifier of the presence of structural abnormality

Nevertheless, what remains to be debated is whether the prevalence of clinically-relevant abnormality at 6% is sufficiently high to warrant routine screening of all those with FEP. What has not been ascertained through this meta-analysis is the clinical benefit gained – whether the abnormality could be treated and its outcome. This suggests cost-benefit analyses are now needed.

When asked for an answer, Dr Blackman believes that all FEP patients should be screened – MRI is a non-invasive, well-tolerated neuroimaging technique – but that economic considerations do need to be taken into consideration.