OBJECTIVE To develop a typology of criminal defendants found incompetent to stand trial using data-driven classification techniques and validate it against forensically relevant outcomes. HYPOTHESES We hypothesized that discrete groups of defendants determined to be incompetent exist that can be identified in the structure of observed clinical, demographic, and criminological data. We also expected that class membership would be differentially associated with competency restoration. METHOD We coded hospital records for 492 consecutive male criminal defendants committed to a secure hospital for competency restoration between 2013 and 2017 (mean [M] age = 38.7 years, standard deviation [SD] = 14.2; 61.0% White, 34.2% Black, 2.6% Hispanic, 2.2% "Other"). Clinical, demographic, and criminological data were analyzed using latent class analysis. Validation analyses modeled competency restoration outcomes as a function of class membership. RESULTS An 8-class solution best fit the data and included 3 discrete classes of patients with psychotic disorders (Class 2, n = 74; Class 3, n = 78; Class 6, n = 68), as well as classes characterized by intellectual limitations without comorbid psychosis (Class 4, n = 54), comorbid psychosis and intellectual limitations (Class 1, n = 41), mood disorders (Class 5, n = 80), older adults with neurocognitive disorders (Class 8, n = 59), and chronic instability (Class 7, n = 38). The restoration rate in the overall sample was 87.8%, and Classes 1-7 showed restoration rates similar to the overall sample, ranging from 82.9% to 100%. The restoration rate of Class 8 was 66.1%, and this was the only class to show significantly lower odds (odds ratio [OR] = 0.181, 95% confidence interval [CI: 0.093, 0.353], p < .001) and hazards (hazard ratio [HR] = 0.511, 95% CI [0.361, 0.724], p < .001) of restoration. CONCLUSION Older adults with neurocognitive disorders admitted for competency restoration are at increased risk of failed restoration. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

中文翻译：

---

住院能力恢复入院的刑事被告人类型学的发展和验证：潜在类别分析。

目的开发一种不适合使用数据驱动分类技术进行审判的刑事被告人类型，并针对法医相关结果进行验证。假设我们假设存在被确定为不称职的离散被告人群体，可以在观察到的临床，人口统计学和犯罪学数据的结构中对其进行识别。我们还期望班级成员资格与能力恢复存在差异。方法我们对2013年至2017年之间承诺恢复安全能力的安全医院的492名连续男性刑事被告的医院记录进行了编码（平均[M]年龄= 38.7岁，标准差[SD] = 14.2； 61.0％白人，34.2％黑人，2.6 ％西班牙裔美国人，2.2％“其他”）。临床，人口，并使用潜在类别分析法分析犯罪数据。验证分析建模的能力恢复结果作为班级成员资格的函数。结果8类解决方案最适合数据，包括3种精神病患者的离散类别（2类，n = 74； 3类，n = 78； 6类，n = 68），以及以智力障碍为特征的类没有合并症的限制（4级，n = 54），合并症和智力限制（1级，n = 41），情绪障碍（5级，n = 80），患有神经认知障碍的老年人（8级，n = 59） ）和慢性不稳定症（7类，n = 38）。总体样本的恢复率为87.8％，1-7级的恢复率与总体样本相似，范围为82.9％至100％。第8类的恢复率为66.1％，并且这是唯一显示赔率（赔率[OR] = 0.181，95％置信区间[CI：0.093，0.353]，p <.001）和危险（危险比[HR] = 0.511，95％）明显更低的类别CI [0.361，0.724]，p <.001）。结论接受能力恢复的神经认知障碍老年人的恢复失败风险增加。（PsycInfo数据库记录（c）2021 APA，保留所有权利）。