两阶段筛选,涉及一份自我评估报告，接下来是结构化的面试，近二十年来，已被提议用于识别临床精神病高风险（CHR）。这项研究使用PRIME屏幕修订（PS-R）和前驱综合征结构化访谈（SIPS）来研究两阶段筛查的预测有效性。在这项研究中完成了两阶段筛查的566名参与者中，192个为PS-R(-),374个为PS-R(+)。在接受SIPS采访后，112被评为CHR(+)，109人被诊断患有精神病，其他345人是CHR（-）。通过SIPS评定为CHR(+)和CHR(-)者在2年内随访观察其临床结果。91名（81.3％）CHR（+）和171名（49.6％）CHR（-）个体在基线和两年随访时完成了调查。CHR(+)组精神病累积转化率为27.5%,但CHR(-)组只有1.7%，两组之间存在显着差异(对数秩检验，χ2=30.07，p<0.001）。在两阶段筛选方面，PS-R(-)/CHR(-),PS-R(-)/CHR(+),PS-R(+)/CHR(-)组被视为预期阴性(EN)，和PS-R（+）/CHR（+）组预期阳性（EP），敏感性和特异性分别为64.3%和79.1%,分别。此外，Kaplan-Meyer生存分析显示，EP组比EN组更容易转变为精神病（log-rank检验，χ2=16.702，p<0.001）。两阶段筛选确实节省了很多时间，减轻了工作量，但可能会排除一些目标个体。优化自我报告规模，形成护士-医生-面试官联盟可能会改善自我评估报告和结构化访谈的使用。
Two-stage screening, involving a self-rating report, followed by a structured interview, has been proposed for identifying clinical high risk of psychosis (CHR) for nearly two decades. This study used PRIME Screen-Revised (PS-R) and Structured Interview for Prodromal Syndromes (SIPS) to investigate the predictive validity of the two-stage screening. Of 566 participants who completed two-stage screening in this study, 192 were PS-R(-) and 374 were PS-R(+). After being interviewed with SIPS, 112 were rated as CHR(+), 109 were diagnosed with psychosis, and the other 345 individuals were CHR(-). Those who were rated through SIPS as CHR(+) and CHR(-) were followed up within 2 years to observe their clinical outcome. Ninety one (81.3%) CHR(+) and 171 (49.6%) CHR(-) individuals completed the investigation at baseline and the two-year follow-up. The cumulative conversion rate to psychosis was 27.5% in CHR(+) group, but only 1.7% in CHR(-) group, with a significant difference between the two groups (log-rank test, χ2 = 30.07, p < 0.001). In terms of two-stage screening, PS-R(-)/CHR(-), PS-R(-)/CHR(+), and PS-R(+)/CHR(-) groups were viewed as expected negative (EN), and the PS-R(+)/CHR(+) group as expected positive (EP), and the sensitivity and specificity was 64.3% and 79.1%, respectively. Furthermore, Kaplan-Meyer survival analysis showed that EP group were more likely to convert to psychosis than EN group (log-rank test, χ2 = 16.702, p < 0.001). Two-stage screening indeed saves much time and alleviates the workload, but may exclude some target individuals. Optimizing self-report scale, and forming a nurse-doctor-interviewer coalition are likely to improve the use of self-rating report and structured interview.
