Innovations In Clinical Neuroscience

NOV-DEC 2017

A peer-reviewed, evidence-based journal for clinicians in the field of neuroscience

Issue link: https://innovationscns.epubxp.com/i/924986

Contents of this Issue

Navigation

Page 47 of 83

48 ICNS INNOVATIONS IN CLINICAL NEUROSCIENCE November-December 2017 • Volume 14 • Number 11–12 O R I G I N A L R E S E A R C H I subset are the eight items that provide the most information at trait levels less than -1. In evaluating the remission set, 37 percent of all subjects met the criteria for "Remission." Only 1 in 3 "Remitted" patients were in the Symptom Relief range (θ≤ -1); a "Remitted" patient was twice as likely to have mild or greater symptoms than to have absent/ minimal symptoms, as shown in Figure 3. The average symptom severity of a remitted patient was θ= -0.80 (SD=0.77) and the average severity of the non-remitted patients was θ=0.47 (SD=0.86). This difference was statistically significant, but with large overlap between the two groups (Figure 5). The overlap coefficient 37 between the Remitted and Non-Remitted Coefficients was 0.435, as shown in Figure 6. The PANSS symptom thresholds that correspond to "Relief " depend on the sensitivity one wishes to establish. A stricter threshold (all 8 symptoms ≤2) marks only eight percent of subjects, and 80 percent of subjects within the Relief range for the latent trait θ as shown in Table 5 and Figure 6. The overlap coefficient for this threshold for FIGURE 4. The item information was used to evaluate how different item subsets assessed patients over the entire illness range. Secondarily, we assessed which symptoms best identified those subjects within the Relief range, defined to be a latent trait level in the (-4, -1) range on the general Positive and Negative Syndrome Scale (PANSS) dimension. This subset was equivalent to those items with the largest Slope/Discrimination values, and was superior to using either the items with the largest ECVI (explained common variance per item) or those items used for Remission. However, retaining all PANSS items was superior to using subsets over all ranges, suggesting that restricting evaluations to subsets of items omits core features associated with the latent trait. FIGURE 5. Distribution of expected a posteriori scoring (EAP) symptom severity scores for remitted and non-remitted groups—When measured along the general dimension, the remitted and non-remitted groups showed considerable overlap, with a statistically significant difference in the mean θ scores (p<0.001.) Roughly 37% of all subjects met remission criteria post-treatment, but 2 out of 3 remitted patients had scores that were not in a low symptom range (θ≤ -1).

Articles in this issue

Archives of this issue

view archives of Innovations In Clinical Neuroscience - NOV-DEC 2017