Authors' reply

Following the letter by Castle, we confirm an error in the secondary outcomes reported in Fig. 2(b) of our manuscript. We have now corrected it, and repeated the literature search by adopting an additional search criterion. We have directly contacted the leading authors of the largest clinical high-risk studies conducted in the past decade to seek additional quality of life (QoL) comparisons between high-risk patients and those with first-episode psychosis. We have then repeated the meta-analysis (see Fig. 1 below), which now included 238 patients at high risk compared with 205 patients with psychosis. The final results were unchanged as compared to those reported in our original analysis. ¹ There is no meta-analytical difference between the subjective QoL of patients at high risk of psychosis and those with frank psychosis (Hedges' g= 0.211, 95% CI −0.148 to 0.571, P = 0.249; Q = 9.518, d.f. = 3, I ² = 68.48, P = 0.023). This secondary meta-analytical comparison is based on a few studies only. However, should new studies become available in the near future, and eventually show a better subjective QoL in clinical high-risk patients as compared with controls, the core finding of our analysis would still remain unchanged. Indeed, our primary aim was to show that patients clinically at high risk for psychosis have significant impairments in functioning and QoL when compared with healthy controls: patients with psychosis were used as a benchmark group for comparative purposes only.

Fig. 1 Meta-analytical comparison of quality of life between patients at high clinical risk for psychosis and patients diagnosed with frank psychosis (Comparisons).

HR, high risk; MSQoL, Modular System for Quality of Life; QLS, Quality of Life Scale; QLS-role, role functioning subscale of the Quality of Life Scale; WHOQOL-BREF, abbreviated version of the World Health Organization Quality of Life assessment.