936357 research-article2020 ASMXXX10.1177/1073191120936357AssessmentGutiérrez et al. Article Assessment Personality Disorders in the ICD-11: 2021, Vol. 28(3) 759­–772 © The Author(s) 2020 Spanish Validation of the PiCD and Article reuse guidelines: sagepub.com/journals-permissions the SASPD in a Mixed Community DOI: 10.1177/1073191120936357 https://doi.org/10.1177/1073191120936357 journals.sagepub.com/home/asm and Clinical Sample Fernando Gutiérrez1,2* , Anton Aluja3,4* , José Ruiz5 , Luis F. García4,6, Miguel Gárriz7 , Alfonso Gutiérrez-Zotes8 , David Gallardo-Pujol5,9, Maria V. Navarro-Haro10,11 , Miquel Alabèrnia-Segura5, Joan Ignasi Mestre-Pintó12,13, Marta Torrens12,14, Josep M. Peri1, Bárbara Sureda1 , Joaquim Soler14,15, Juan Carlos Pascual14,15, Gemma Vall4,16, Natalia Calvo14,17, Marc Ferrer14,17, Joshua R. Oltmanns18, and Thomas A. Widiger18 Abstract The International Classification of Diseases–11th revision (ICD-11) classification of personality disorders is the official diagnostic system that is used all over the world, and it has recently been renewed. However, as yet very few data are available on its performance. This study examines the Personality Inventory for ICD-11 (PiCD), which assesses the personality domains of the system, and the Standardized Assessment of Severity of Personality Disorder (SASPD), which determines severity. The Spanish versions of the questionnaires were administered to a community (n = 2,522) and a clinical sample (n = 797). Internal consistency was adequate in the PiCD (α = .75 to .84) but less so in the SASPD (α = .64 and .73). Factor analyses suggested a unidimensional or bidimensional structure for severity, while revealing that the personality trait qualifiers are organized into four factors: negative affectivity, detachment, dissociality, and a bipolar domain of disinhibition–anankastia. The mutual relationships between traits and severity were analyzed, as well as the ability of the whole system to identify clinical subjects. Although further improvements are required, the results generally support the use of the PiCD and the SASPD and help substantiate the new ICD-11 taxonomy that underlies them. Keywords personality disorder, personality pathology, ICD-11, severity, PiCD, SASPD The two main diagnostic systems for personality disorders diagnostic system in use all over the world, it is surprising have started a journey toward an evidence-based dimen- that these substantial reforms have not prompted more sional taxonomy. The Diagnostic and Statistical Manual attention. of Mental Disorders–Fifth edition (DSM-5; American The ICD-11 personality disorder taxonomy encompasses Psychiatric Association, 2013) fell far short of this target: a trait part, composed of five domains of personality psy- The new taxonomic system was left pending until more evi- chopathology plus the borderline specifier, and a severity dence becomes available, and it still merged dimensions and part which determines whether the subject is or not disor- categories in equal proportions. In contrast, the International dered. The trait part is based on a broad revision of the lit- Classification of Diseases–11th revision (ICD-11) classifi- erature which identified the main features contributing to cation has gone much farther by offering a wholly dimen- personality disturbance (Mulder et al., 2011; Tyrer et al., sional system and by discarding traditional categories 2011). These features were finally established as negative almost completely (World Health Organization, 2018). The affectivity, detachment, dissociality, disinhibition, and exception is the retention of the borderline specifier, which anankastia. Although the last two domains are presented is not part of the evidence-based model but whose removal separately in the classification, they have generally been was deemed a huge loss by a portion of clinicians (Tyrer conceived in the literature as the two poles of a single con- et al., 2019). In view of the fact that the ICD is the official struct (Tyrer et al., 2014). The ICD trait part is similar to the 760 Assessment 28(3) DSM-5, except that it does not consider psychoticism as a (Bach & Anderson, 2018; McCabe & Widiger, 2020; personality trait. Although there are no official instruments Oltmanns & Widiger, 2019), and it has proved capable of that measure this model, the Personality Inventory for predicting expert judgment on personality disorder sever- ICD-11 (PiCD) is the first published measure specifically ity, as well as various measures of well-being and maladap- designed for this purpose. The PiCD has shown adequate tation (Bach & Anderson, 2018). reliability and structural validity in its original version Overall, this is an entirely new taxonomy that, as its (Oltmanns & Widiger, 2018, 2019), as well as good con- authors state, “has to be tested critically and thoroughly” current validity with the Big Five, the DSM-5, and other (Mulder & Tyrer, 2018, p. 30). However, because of the personality psychopathology models (Crego & Widiger, short time interval since their publication, only a dozen 2019; Oltmanns & Widiger, 2019, 2020; Somma et al., studies are available on the properties of the PiCD or the 2020). As for the borderline specifier, it was only recently SASPD, and further evidence is urgently needed. For added and so is not included in most studies, including this example, the PiCD has been studied in nonclinical sub- one (Reed, 2018). jects (Carnovale et al., 2020; McCabe & Widiger, 2020; The severity part consists of a single dimension that Somma et al., 2020) and in patients assessed via the qualifies the subject as having a normal personality or as Internet (Oltmanns & Widiger, 2018, 2019, 2020), and suffering a mild, moderate, or severe personality disorder. thus additional examination of its reliability and validity That is to say that, in the ICD-11, it is severity that deter- in clinical samples is required. The structure of the model mines personality disorder diagnosis, whereas traits are also needs closer scrutiny. On the one hand, although optional qualifiers with a more secondary role. The idea of domains are allegedly based on the literature (Mulder a general severity continuum has been under investigation et al., 2011; Tyrer et al., 2011), they have undergone non- for some time, and its central role in the taxonomy is trivial changes since then in their number and nature thought to simplify the diagnostic process and to empha- (Crego & Widiger, 2019; Kleindienst et al., 2017): The size the degree of disturbance (Crawford et al., 2011; Tyrer early traits “emotional/unstable” and “anxious/dependent” & Johnson, 1996). As of yet, the Standardized Assessment were collapsed into negative affectivity around 2014, dis- of Severity of Personality Disorder (SASPD) is the only inhibition did not appear until 2015, and borderline was instrument that has been developed to measure this con- added at the very last moment (Reed, 2018; Tyrer et al., tinuum (Olajide et al., 2018). However, this self-reporting 2014; Tyrer et al., 2015). On the other hand, the original tool is more aligned with the initial emphasis of the ICD on structure (Oltmanns & Widiger, 2018) has only been repli- harm to others and to occupational roles (Tyrer et al., 2015) cated in students (Carnovale et al., 2020) or using an infor- than with the DSM-5, in which the focus was on problems mant version (Bach et al., 2020), and in no case were five with the self and interpersonal bonds. Despite this, the independent domains found. All three studies instead sup- SASPD correlates by between .58 and .73 with measures of ported a four-factor solution with disinhibition and anan- DSM-5 level of functioning, indicating significant overlap kastia located at the opposite poles of a single dimension. 1 Hospital Clinic, Barcelona, Spain 2 IDIBAPS, Barcelona, Spain 3 University of Lleida, Lleida, Spain 4 Biomedical Research Institute of Lleida, Lleida, Spain 5 University of Barcelona, Barcelona, Spain 6 Autonomous University of Madrid, Madrid, Spain 7 Institut de Neuropsiquiatria i Addiccions, Barcelona, Spain 8 IISPV-Universitat Rovira i Virgili, Reus, Spain 9 Institute of Neurosciences, Barcelona, Spain 10 University of Zaragoza, Zaragoza, Spain 11 Aragón Health Research Institute, Zaragoza, Spain 12 IMIM Hospital del Mar Medical Research Institute, Barcelona, Spain 13 Universitat Pompeu Fabra, Barcelona, Spain 14 Universitat Autònoma de Barcelona, Spain 15 Hospital de la Santa Creu i Sant Pau, Barcelona, Spain 16 GSS–Hospital Santa Maria, Lleida, Spain 17 Vall d'Hebron University Hospital, Barcelona, Spain 18 University of Kentucky, Lexington, KY, USA *Both authors contributed equally to this study Corresponding Author: Fernando Gutiérrez, Personality Disorder Unit, Institute of Neuroscience, Hospital Clínic de Barcelona, Villarroel 170, Barcelona, 08036, Spain. Email: fguti@clinic.cat Gutiérrez et al. 761 A paucity of data is also true for severity, as the proper- education. The clinical sample consisted of 797 outpa- ties of the SASPD have only been studied in one commu- tients, 70.7% women, with mean age 41.7 years (SD = nity sample (McCabe & Widiger, 2020) and in a few small 13.6, range 15 to 82). They were consecutively referred to or mid-sized clinical samples (Bach & Anderson, 2018; the mental health units of six hospitals in Catalonia, Spain. Olajide et al., 2018; Oltmanns & Widiger, 2019). Thus, we Patients were clinically diagnosed at their respective cen- do not know whether the SASPD, which seeks to reflect the ters, with the main diagnoses including mild to moderate combined impact of five different traits, really forms a sin- affective disorders (37.0%), anxiety or phobic disorder gle dimension or in fact forms several. Thus far, neither the (22.1%), mixed affective and anxious disorder (8.7%), only study on its structure (Rek et al., 2019) nor the evi- substance-related disorder (8.1%), eating disorder (5.8%), dence on its predecessor, the Standardized Assessment of and other disorders (4.3%) each with a frequency below Personality–Abbreviated Scale (Bach et al., 2019), have 3%. No categorical diagnoses of personality disorder were particularly supported unidimensionality. We also need to made. Based on the proposed cutoffs for the SASPD, cali- test in larger community and clinical samples how far sever- brated against expert judgment (Olajide et al., 2018), ity overlaps with personality traits, as the evidence avail- 23.4% of our clinical sample would have at least a mild able up to now suggests blurred boundaries (McCabe & personality disorder (SASPD ≥ 8), of whom 10.1% would Widiger, 2020; Oltmanns & Widiger, 2019). It is even also exceed the threshold for a moderate or severe person- unclear whether severity works differently for each trait, ality disorder (SASPD ≥ 10). Combined sample size and whether measuring severity provides any advantage allowed a rate of 55.3 participants per trait in factor analy- over simply assessing domains. Finally, the diagnostic util- sis. The study was approved by the ethical committees of ity of the joint trait and severity parts also needs to be tested the respective centers. in studies that directly compare community and clinical samples. In this process, population norms have to be estab- lished, and nonarbitrary diagnostic thresholds should be set Instruments based on the prediction of maladaptive outcomes (Bagby & The PiCD (Oltmanns & Widiger, 2018) is a 60-item self- Widiger, 2020; Herpertz et al., 2017). report measuring the five domains of the dimensional ICD- We expect to find a four-factor personality structure with 11 personality model (World Health Organization, 2018): good psychometric properties, and a partly independent, negative affectivity, detachment, dissociality, disinhibition, unidimensional severity continuum able to accurately detect and anankastia. Each domain has 12 items rated from 1 participants with mental health problems. Given that vali- (strongly disagree) to 5 (strongly agree). The PiCD was dation in other languages and cultures is what gives the translated by three Spanish native speakers who were famil- classification a truly universal scope, and that only an iar with the constructs being measured and worked inde- Italian PiCD (Somma et al., 2020) and a Danish and a pendently. Semantic and conceptual equivalence were German SASPD (Bach & Anderson, 2018; Rek et al., 2019) given priority throughout the adaptation process. The trans- are available currently, we hope our study will further our lators then agreed on a common version, which was blindly knowledge of the ICD-11 system and help make it available back-translated by an English native speaker and compared to a wider range of researchers. with the original. Discrepancies were consensually resolved by the translators and the original authors of the instrument. Before final approval, a pilot study was undertaken in 20 Method participants from the target population to ensure under- standability and feasibility. The above procedures followed Participants the standard recommendations for adapting questionnaires The community sample consisted of 2,522 volunteers, (Epstein et al., 2015). 59.2% women, with mean age 39.8 years (SD = 19.0, The SASPD (Olajide et al., 2018) was modeled after the range 16 to 92). Participants were undergraduates and their Standardized Assessment of Personality–Abbreviated Scale relatives and acquaintances recruited from three different (Moran et al., 2003). Its nine items reflect possible unde- universities in Catalonia and Madrid, Spain. They answered sired consequences of each of the five ICD-11 domains: anonymously and did not receive any compensation for Items 4, 6, and 9 are linked to negative affectivity (losing participating. This sample was representative of the gen- temper, worrying, and feeling helpless); Items 1 and 3 to eral Spanish population in terms of age (43.4 years; www. detachment (avoiding people and lacking friends); Items 2 ine.es) and level of education: In a subsample of 801 par- and 8 to dissociality (distrusting others and being callous); ticipants, about one third had completed primary and lower Item 7 to anankastia (being excessively organized); and secondary education, one quarter upper secondary and Item 5 to disinhibition (being impulsive). The translation post-secondary education, and the remaining 40% tertiary procedures described above were applied to the SASPD. 762 Assessment 28(3) Data Analysis curves. SPSS 22 was used for all analyses except Factor 10.9.02 (Ferrando & Lorenzo-Seva, 2017) for EFA, and R Descriptive statistics for the PiCD and the SASPD were 3.6.0 (R Core Team, 2018) for IRT analysis (package “ltm”; obtained in the clinical and community samples separately. Rizopoulos, 2006), CFA, and ESEM (package “lavaan”; Internal reliability was examined through Cronbach’s Rosseel, 2012). We report how we determined our sample alphas (α) and corrected item-scale correlations (ri-s). To size, all data exclusions, all manipulations, and all measures determine whether similar amounts of information are pro- in the study. vided along the scales, we conducted an item response the- ory (IRT) analysis using Samejima’s graded response model (Toland, 2014). Results The structure of the PiCD was examined at the item level Descriptives, Distribution, and Reliability of the using a range of factor analytical methods in line with cur- rent recommendations (Hopwood & Donnellan, 2010; PiCD and the SASPD McCrae et al., 1996; Osborne, 2014). Given that the PiCD Descriptive statistics of the PiCD and the SASPD in the structure has been defined previously, confirmatory factor community and clinical samples are provided in Table 1. analysis (CFA) was conducted first. Model fit was exam- When calculated for each sex (Supplementary Table S1, ined through the comparative fit index (CFI), Tucker–Lewis available online), raw scores in the community sample can index (TLI), root mean square error of approximation be taken as the population norms for computing T scores. (RMSEA), and standardized root mean square residual Under this metric, the clinical sample was 1 SD above the (SRMR). Values above .95 for CFI and TLI, below .06 for mean in negative affectivity and the SASPD, around ½ a RMSEA, and below .08 for SRMR, were considered a good SD in detachment, disinhibition, and dissociality, and on a fit (Brown, 2015). As a significant number of items showed par in anankastia. Comparisons with the previous literature at least moderate skewness (>|1|) or kurtosis (>1), all anal- suggested that the Spanish population, either clinical or yses were based on the polychoric correlation matrix (Flora nonclinical, was slightly lower than others in detachment et al., 2012) and the robust diagonally weighted least and dissociality, and that our clinical sample was some- squares estimator. what higher in negative affectivity (Supplementary Figure Fit eventually proved to be poor, so we moved on to an S1; Supplementary Tables S2 and S3, available online). exploratory factor analysis (EFA) approach. Solutions with Distributions of all the scales were within the commonly different numbers of factors from one to seven were succes- accepted limits of ±2 regarding skewness and zero-cen- sively retained. We used the Hull method (Lorenzo-Seva tered kurtosis (frequency distributions in Supplementary et al., 2011) and parallel analysis based on minimum rank Figure S2, available online). factor analysis (Timmerman & Lorenzo-Seva, 2011) to The PiCD domains showed acceptable internal reliabili- decide on the most appropriate number of factors. Solutions ties of α = .80 to .86 in the community and .75 to .83 in the were rotated to direct oblimin except the four- and five- clinical sample. Corrected item-scale (ri-s) correlations factor solutions, which were rotated to Procrustes targeted were good overall, but reflected the presence of some less at the previously published solutions (Oltmanns & Widiger, consistent items. For example, Item 26 (“Changes in my 2018). Additionally, these solutions were independently mood are unrelated to what is happening in my life”) was obtained in two random partitions of the sample, and then poorly correlated with its corresponding negative affectiv- used as targets in an exploratory structural equation model ity domain in both samples (ri-s = .09 and .10, respectively; (ESEM) analysis to examine fit in a cross-validation design. Supplementary Table S4, available online), and Item 23 (“I Target loadings <.10 were freely estimated. In a more feel pretty much the same all the time”) showed ri-s = .10 demanding design, the same procedure was applied to the and .11 with detachment. With regard to the SASPD, inter- community and clinical samples. Replicability between nal reliability was lower (α = .64 in the community and solutions was additionally examined through Tucker’s con- .73 in the clinical sample, respectively), with Items 5 to 9 gruence coefficients (Φ), with Φ ≥ .85 indicating fair simi- showing ri-s below .30 in either sample. Furthermore, 12 larity and Φ ≥ .95 equivalence (Lorenzo-Seva & ten Berge, items of the PiCD and the SASPD showed skewness and/ 2006). or kurtosis greater than ±2 (Supplementary Table S4, We examined the mutual relationships between the PiCD available online). For the most part, these were items of domains and the SASPD through Pearson’s correlations (r) pathological content that obtained low scores as expected and multiple linear regression. We also studied the ability of only in the community sample. Four items, however, the PiCD and SASPD to predict in a logistic regression showed skewed distributions in both samples (PiCD Items whether the subject belongs to the clinical sample (case- 9, 24, and 54, and SASPD Item 8). They all reflected ness). Sensitivity (Se) and specificity (Sp) were calculated extreme forms of dissociality, such as indifference or for the most predictive dimensions, and the best cutoff points amusement about others’ suffering, which were not com- were estimated through receiver operating characteristic mon in either sample. Table 1. Mean (SD), Distribution Shape, Internal Reliability, Item-Scale Correlations, and Group Differences Between the Community and Clinical Samples. Community sample (n = 2,522) Clinical sample (n = 797) Student’s t ri-s ri-s M SD Skew Kurt α M Range M SD Skew Kurt α M Range t p Cohen’s d Negative affectivity 31.9 8.1 0.298 −0.206 .83 0.49 0.09-0.69 42.3 8.4 −0.313 −0.391 .83 0.49 0.10-0.66 −31.2 <.001 1.27 Detachment 25.7 7.5 0.535 −0.008 .82 0.48 0.10-0.65 29.4 8.8 0.282 −0.417 .82 0.48 0.11-0.63 −11.8 <.001 0.47 Dissociality 21.8 6.9 1.150 1.810 .82 0.47 0.32-0.64 23.8 7.5 0.560 0.136 .79 0.44 0.27-0.58 −7.0 <.001 0.28 Disinhibition 24.2 7.5 0.665 0.263 .84 0.49 0.30-0.62 27.8 8.5 0.902 0.607 .83 0.48 0.29-0.60 −11.7 <.001 0.46 Anankastia 39.5 7.2 −0.208 0.074 .80 0.44 0.28-0.60 40.1 7.2 −0.204 0.340 .75 0.39 0.24-0.54 −2.2 .030 0.08 SASPD total 5.7 3.2 1.297 2.942 .64 0.32 0.23-0.42 8.8 4.4 0.681 0.179 .73 0.40 0.21-0.52 −21.2 <.001 0.88 Note. Skew = skewness; Kurt = zero-centered kurtosis; ri-s = corrected item-scale correlations; SASPD = Standardized Assessment of Severity of Personality Disorder. 763 764 Assessment 28(3) Table 2. Intercorrelations Between the PiCD and the SASPD in the Community (Below the Diagonal, n = 2,522) and the Clinical Sample (Above the Diagonal, n = 797). Negative affectivity Detachment Dissociality Disinhibition Anankastia PiCD total SASPD Negative affectivity — .36** .33** .47** .02 .76** .63** Detachment .21** — .23** .33** .08* .71** .48** Dissociality .18** .26** — .50** −.14** .65** .43** Disinhibition .36** .22** .44** — −.46** .65** .49** Anankastia .20** .15** −.09* −.46** — .13** −.13** PiCD total .71** .66** .61** .56** .28** — .66** SASPD .44** .30** .27** .26** .02 .46** — Note. Correlations ≥|.30| are in bold type. PiCD = Personality Inventory for ICD-11; SASPD = Standardized Assessment of Severity of Personality Disorder. *p < .05. **p < .01. Graded response model analyses indicated that the the entire hierarchical structure, solutions with different SASPD and three PiCD domains—detachment, dissociality, numbers of factors from one to seven were successively and disinhibition—provided much more information at retained and rotated to direct oblimin. The exceptions were higher levels of the latent trait: 62.8% in the upper half ver- the four- and five-factor structures, which underwent sus 28.2% in the lower half, on average (Supplementary Procrustean rotations targeted at the originally published Figure S3, available online). By contrast, negative affectiv- solutions (Tables 6 and S1 in Oltmanns & Widiger, 2018). ity showed the opposite pattern in the clinical sample, and Hull and parallel tests suggested four and five factors anankastia in both samples. respectively. The four-factor solution (Table 3, left) explained 45% of the variance. The second to fourth fac- tors faithfully reproduced the dissociality, detachment, and Factor Structure of the PiCD negative affectivity domains (r = .95, .97, and .93 respec- The PiCD domains were not orthogonal, as shown by the tively), whereas the first factor was a bipolar dimension Pearson’s intercorrelations in Table 2. In the community with disinhibition (.82) and anankastia (−.85) at opposite sample, the highest associations were between disinhibition poles. Absolute intercorrelations between factors ranged and negative affectivity (r = .36), dissociality (.44), and from .02 to .36, with mean .23. As for the five-factor solu- anankastia (−.46). In the clinical sample, the highest corre- tion, the second to fourth factors also reproduced the dis- lations were between disinhibition and negative affectivity sociality, detachment, and negative affectivity domains (.47), detachment (.33), dissociality (.50), and anankastia (r = .96 to .99), but anankastia and disinhibition could not (−.46), while negative affectivity also correlated with be recovered as separate domains (Table 3, right). Instead, detachment (.36) and dissociality (.33). the disinhibition–anankastia factor split into two dimen- The Kaiser–Meyer–Olkin (KMO) index of sampling sions which may be called “impulsivity versus caution” adequacy was .93 (very good) in the whole sample, and the and “carelessness versus perfectionism,” and showed a Bartlett's test of sphericity was highly significant (69565.9, poorer replicability between samples than the bipolar fac- degrees of freedom = 1,770, p < .001). Given that the PiCD tor (Φ = .93 and .86). Targeting the original structure with structure had been examined before, CFA was applied first. 12 items per domain instead of the empirically based solu- However, neither the five-factor solution with 12 items per tion did not produce different results. Figure 1 represents domain nor the previously reported four- and five-factor the complete hierarchical structure from one to five fac- solutions (Oltmanns & Widiger, 2018) produced an optimal tors and the correspondence of each factor with the five fit: CFI = .77 to .83, TLI = .76 to .82, RMSEA = .10 to .11, original PiCD domains (one- to three-factor solutions in and SRMR = .10 to .11. This was also true when the com- Supplementary Table S6; correlations in Supplementary munity and the clinical samples were analyzed separately Table S7, available online). Additional factors beyond the (Supplementary Table S5, available online). fifth were moderately replicable but barely interpretable Poor fit is common when applying CFA to personality (Supplementary Table S8, available online). structures, due to the strict requirement of simple structures To cross-validate the four- and five-factor solutions, and zero cross-loadings (Ferrando & Lorenzo-Seva, EFAs were repeated in two randomly split samples. Each 2017; Hopwood & Donnellan, 2010). Therefore, following solution was then used as a target for ESEM analysis in the current recommendations, we applied an EFA approach complementary sample. Both the four-factor (CFI = .97, (McCrae et al., 1996; Osborne, 2014). In order to examine TLI = .97, RMSEA = .05, SRMR = .05) and the Gutiérrez et al. 765 Table 3. Four- and Five-Factor Solutions of the PiCD in the Whole Sample (N = 3,319). Four factors Five factors Items F4.1 F4.2 F4.3 F4.4 F5.1 F5.2 F5.3 F5.4 F5.5 picd15_ak −.71 .09 .25 −.19 −.79 .15 .00 −.13 .00 picd35_ak −.67 .33 −.15 .25 −.19 .22 .03 .21 −.60 picd5_ak −.66 .25 −.08 .18 −.26 .16 .06 .16 −.50 picd55_ak −.64 −.14 .20 .04 −.61 −.13 .07 .08 −.11 picd20_ak −.61 .40 −.09 .24 −.16 .29 .08 .21 −.56 picd45_ak −.61 .06 .22 .18 −.65 .12 −.01 .24 −.02 picd12_dn .58 −.12 .23 .08 .05 .05 −.08 .14 .66 picd7_dn .53 .04 .24 .06 .12 .16 .02 .10 .51 picd27_dn .53 −.02 .18 .19 .06 .14 −.10 .24 .58 picd60_ak −.53 −.05 .15 −.02 −.62 −.01 −.06 .04 −.01 picd47_dn .50 .06 −.09 .50 .58 .04 .07 .45 −.03 picd25_ak −.47 −.09 .28 .02 −.60 −.02 .04 .09 .10 picd32_dn .46 .09 −.11 .34 .55 .06 .06 .29 −.04 picd57_dn .44 .00 .28 .12 .06 .12 .06 .16 .46 picd50_ak −.44 .10 −.26 .20 −.20 .07 −.20 .20 −.32 picd30_ak −.40 −.06 .37 −.33 −.65 .02 .08 −.26 .23 picd42_dn .36 .17 .22 .04 −.01 .29 −.01 .08 .45 picd40_ak −.31 .01 .18 .22 −.41 .08 −.02 .27 .07 picd52_dn .30 .30 .09 .11 .15 .35 .02 .12 .19 picd59_ds .00 .78 −.07 .00 .00 .81 −.12 .00 −.02 picd44_ds −.06 .76 −.05 .03 −.06 .79 −.12 .04 −.02 picd54_ds .04 .72 .18 −.26 .02 .72 .16 −.26 .02 picd14_ds −.17 .69 −.22 −.02 −.08 .70 −.24 −.02 −.13 picd49_ds −.03 .68 −.07 .05 −.02 .70 −.11 .05 −.04 picd19_ds −.26 .64 −.09 .07 .00 .58 .03 .04 −.33 picd34_ds −.08 .64 −.03 −.07 −.01 .63 −.02 −.08 −.10 picd9_ds .05 .60 .18 −.12 .00 .62 .13 −.12 .05 picd39_ds .00 .59 .20 −.10 −.04 .60 .14 −.10 .04 picd24_ds −.02 .55 .27 −.22 −.13 .57 .17 −.20 .11 picd4_ds .09 .46 −.08 .27 .18 .46 −.03 .25 −.10 picd29_ds .05 .31 .01 .20 .03 .34 −.03 .20 .03 picd37_dn .27 .31 .11 .05 −.04 .43 −.11 .09 .37 picd22_dn .13 .29 .25 .04 −.02 .34 .14 .06 .18 picd43_dt −.07 −.15 .75 .01 −.04 −.22 .81 .00 −.02 picd13_dt −.07 −.19 .74 .07 −.04 −.25 .79 .06 −.02 picd33_dt −.05 .12 .68 .06 .02 .06 .75 .04 −.07 picd28_dt −.07 −.28 .67 .08 −.07 −.32 .69 .09 .02 picd58_dt .01 −.02 .63 .23 −.01 −.04 .63 .24 .03 picd3_dt −.02 .08 .57 .21 .10 .01 .68 .19 −.12 picd8_dt −.06 .26 .51 −.12 .02 .20 .59 −.14 −.08 picd38_dt .04 .03 .51 .17 −.06 .05 .44 .18 .12 picd53_dt .00 .30 .47 −.18 .01 .26 .49 −.19 −.00 picd18_dt .06 .21 .45 .07 .05 .19 .46 .07 .03 picd48_dt .03 .27 .39 −.02 .10 .22 .47 −.04 −.06 picd10_ak −.26 −.20 .31 .04 −.43 −.14 .10 .09 .16 picd26_na .04 .06 .25 .05 .00 .06 .23 .05 .05 picd41_na .06 −.09 .13 .78 .01 −.02 .03 .80 .07 picd16_na −.16 −.09 .13 .76 −.09 −.06 .07 .77 −.08 picd56_na .00 .07 .03 .75 .05 .11 −.01 .75 −.05 (continued) 766 Assessment 28(3) Table 3. (continued) Four factors Five factors Items F4.1 F4.2 F4.3 F4.4 F5.1 F5.2 F5.3 F5.4 F5.5 picd1_na −.03 .04 −.11 .72 .11 .05 −.07 .70 −.16 picd11_na .05 .01 .02 .72 .07 .06 −.02 .72 −.01 picd46_na .00 −.14 .17 .69 −.05 −.08 .07 .71 .06 picd21_na −.08 −.13 .20 .68 −.07 −.09 .13 .69 −.01 picd36_na −.10 −.40 −.17 .65 −.10 −.34 −.25 .67 −.00 picd2_dn .39 .11 −.25 .61 .65 .05 .01 .54 −.24 picd51_na .08 −.07 .25 .60 .02 −.02 .17 .61 .08 picd31_na .00 −.09 .02 .56 .08 −.09 .05 .55 −.08 picd17_dn .46 .06 −.15 .55 .62 .02 .06 .49 −.12 picd6_na −.39 .24 −.05 .50 −.15 .22 −.02 .50 −.30 picd23_dt −.13 .08 .24 −.39 −.16 .06 .21 −.37 .03 McDonald’s ω .87 .86 .85 .89 .87 .85 .87 .89 .68 Φwhole-comm 1.00 1.00 1.00 .99 .98 1.00 .99 .99 .97 Φwhole-clin .98 .99 .99 .99 .96 .99 .97 .99 .94 Φcomm-clin .97 .98 .97 .97 .93 .98 .94 .97 .86 Φwhole-Oltmanns .82 .92 .93 .95 .84 .91 .91 .90 .29 Φwhole-Bach .94 .86 .69 .67 .72 .87 .92 .71 .80 Φwhole-Carnovale .96 .89 .89 .96 — — — — — Note. Loadings λ ≥ |.30|, McDonald’s Omega coefficients ω ≥ .70, and Tucker’s coefficients Φ ≥ .95 are in bold type. Items are sorted following the four-factor solution. na = negative affectivity; dt = detachment; ds = dissociality; dn = disinhibition; ak = anankastia; comm = community sample; clin = clinical sample; Oltmanns = Oltmanns and Widiger (2018); Bach = Bach et al. (2019); Carnovale = Carnovale et al. (2020). five-factor solutions (CFI = .98, TLI = .98, RMSEA = .04, significant (4414.5, degrees of freedom = 36, p < .001). SRMR = .04) showed a good fit. Replicability between Hull and parallel tests coincided in suggesting a one-factor samples was also good, with Φ = .99 across the four-factor solution, and only the one- and two-factor solutions showed solution and Φ = .97 to 99 in the five-factor solution. In a good replicability between the community and clinical sam- more demanding design, the same approach was applied ples (Φ = .96 to .99; Table 4). The former included all items to the community and clinical samples. Both the four- and gave support to a unidimensional construct, whereas (CFI = .95, TLI = .94 to .95, RMSEA = .06, SRMR = the latter separated social maladaptation (dislike of being .06, congruence Φ = .96 to .98) and the five-factor solu- with others, lacking friends, mistrust, and indifference to tions (CFI = .97, TLI = .96 to .97, RMSEA = .04 to .05, other’s feelings) from nonsocial maladaptation (losing tem- SRMR = .05, congruence Φ = .90 to .97; Supplementary per, acting impulsively, and worrying). However, the struc- Table S9, available online) showed acceptable fit and repli- ture was problematic in some aspects: Two out of nine cability. An additional random-intercept analysis (Carnovale items, reflecting orderliness and neediness, did not load on et al., 2020; Maydeu-Olivares & Coffman, 2006), aimed at any factor, and ESEM analyses favored two or three factors controlling for acquiescent responding or other biases, over one (Supplementary Table S9, available online). barely improved fit. Furthermore, the method factor only explained 6.0% and 4.9% of the variance of the four- and Criterion Validity five-factor solutions when added to the CFA models, and 2.5% and 2.6% in the case of ESEM models. Concerning the mutual relationships between personality traits and severity, all PiCD domains significantly correlated with the SASPD, in both the community (range r = .26 to Factor Structure of the SASPD .44) and the clinical samples (.43 to .63). The exception was CFA showed poor fit for the unidimensional structure of the anankastia, which showed nil (.02) or negative (−.13) cor- SASPD in the whole sample (CFI = .88, TLI = .84, relation with severity respectively (Table 2). When we pre- RMSEA = .12, SRMR = .09; Supplementary Table S5, dicted the SASPD from the PiCD through linear regression, available online). In the subsequent EFA, the KMO index the higher coefficients were found for negative affectivity was .75 indicating mediocre sample adequacy (Kaiser & in both samples (b = .406 and .456, respectively, p < Rice, 1974), but the Bartlett’s test of sphericity was highly .001; Supplementary Table S10, available online). The Gutiérrez et al. 767 Figure 1. Hierarchical structure of the PiCD. Note. All between-level correlations ≥.50 are reported. Grey-shaded areas indicate the correspondence with the originally proposed domains. The complete correlation matrix is provided in Supplementary Table S7 (available online). PiCD = Personality Inventory for ICD-11. contribution of detachment and dissociality was smaller, that important contribution to the prediction, indicating that the of disinhibition was negligible, and anankastia rather acted as domains are unipolar with regard to maladaptation. Finally, a protective factor (−.101 and −.116, respectively, p < .001). the interactions between traits and severity, represented Next, caseness (i.e., belonging to the clinical sample) by the product of each domain multiplied by the SASPD, was used as an external criterion to examine whether sever- barely contributed at all to the prediction of caseness when ity was able to predict maladaptation better than and beyond added to the original variables, and did not improve the personality traits. Although all variables differed to some prediction when used instead of them. extent between the community and clinical samples (Table 1, Receiver operating characteristic curves gave similar right), only negative affectivity and the SASPD (odds ratio results (Supplementary Figure S4, available online). Areas = 1.15 and 1.06, respectively, p < .001) were substantially under the curve were .81 for negative affectivity and .72 for predictive in logistic regression (Supplementary Table S10, the SASPD, with all other variables having limited predic- available online). Introducing the disinhibition–anankastia tive ability. The best cutoffs for these two variables were bipolar factor instead of its constituent domains did not obtained by applying Youden’s index (Se + Sp − 1 = max). change the results, as their maladaptive potential seems The resulting thresholds were ≥37 for negative affectivity, marginal in this case. Nor did quadratic terms make any which gave Se = .76 and Sp = .73, and ≥7 for the SASPD, 768 Assessment 28(3) Table 4. One- to Three-Factor Solutions of the SASPD in the Whole Sample (N = 3,292). One factor Two factors Three factors Items F1.1 F2.1 F2.2 F3.1 F3.2 F3.3 saspd1_dt .68 .80 −.01 .05 .78 −.03 saspd2_ds .62 .53 .16 .06 .53 .14 saspd3_dt .67 .83 −.04 .05 .81 −.06 saspd4_na .62 .04 .82 .06 .05 .78 saspd5_dn .45 −.08 .72 .03 −.09 .73 saspd6_na .44 .16 .36 .98 .03 .04 saspd7_ak .33 .18 .21 .18 .16 .12 saspd8_ds .50 .43 .13 −.29 .50 .28 saspd9_na .35 .14 .28 .10 .13 .24 McDonald’s ω .77 .75 .68 — .76 .73 Φwhole-comm .99 1.00 .98 .93 .99 .97 Φwhole-clin .99 .98 .98 .99 .98 .98 Φcomm-clin .96 .97 .99 .94 .96 .99 Note. Loadings λ ≥ |.30|, McDonald’s Omega coefficients ω ≥ .70, and Tucker’s coefficients Φ ≥ .95 are in bold type. na = negative affectivity; dt = detachment; ds = dissociality; dn = disinhibition; ak = anankastia; comm = community sample; clin = clinical sample. which gave Se = .66, Sp = .68 (Supplementary Table S10, and with the literature: Whereas the bipolar domain of dis- available online). inhibition–anankastia showed at least partial agreement with prior solutions (Φ = .82 with Oltmanns & Widiger, 2018; .94 with Bach et al., 2019; .96 with Carnovale et al., Discussion 2020), impulsivity and carelessness did not (.29 to .86; In this study, we describe Spanish adaptations of the PiCD Table 3). The PiCD has also resulted in a four-factor struc- and the SASPD, which assess the trait and severity parts of ture whenever it has been analyzed together with other the ICD-11 classification of personality disorders, respec- questionnaires (Crego & Widiger, 2019; McCabe & tively. The PiCD proved to be reliable and showed a consis- Widiger, 2020; Oltmanns & Widiger, 2018, 2020; Somma tent factorial structure, whereas the SASPD seems to require et al., 2020). It is of note that whereas a bipolar construct of some psychometric refinement. Even so, both instruments disinhibition–anankastia has found support beyond the were able to detect the clinical cases with reasonable accu- PiCD (Harkness et al., 2012; Widiger & Simonsen, 2005; racy. The general conclusions are that they can be used for Wright et al., 2012), other models with a broader or simply diagnosis in clinical settings with due caution, and that the a different definition of anankastia have identified it as a underlying ICD-11 taxonomy is supported. Some points separate factor (De Clercq et al., 2006; Gutiérrez et al., deserve further comment. 2014; Kushner et al., 2011). During their development both the DSM-5 and the ICD- Another point is that the PiCD domains are far from 11 have wavered with regard to whether disinhibition and being orthogonal. This may be a drawback, as the massive anankastia should be kept together or apart (Skodol, 2011; overlap between traditional diagnostic categories was a Tyrer et al., 2014), but it is the former option which is sup- leading cause of their replacement by the current classifica- ported by the evidence (Mulder et al., 2011; Wright et al., tion. However, the two cases differ in essential respects. 2012). Our results also confirm this view, given that nega- On the one hand, a moderate overlap appears to reflect the tive affectivity, detachment, and dissociality are reproduced true nature of personality traits, and typically results in as they appear in the taxonomy, whereas disinhibition and broader superordinate factors forming hierarchical struc- anankastia form the opposite extremes of a single bipolar tures (Condon & Mroczek, 2016; Markon et al., 2005). For dimension, resulting in a four-factor structure. The attempt example, average intercorrelations of .19 have been reported to retain five factors does not bring us any closer to a mean- in Livesley’s personality model, .36 in the DSM-5 dimen- ingful five-factor system, because disinhibition and anan- sional taxonomy, and .20 in the Big Five (Gutiérrez et al., kastia cannot be recovered as separate domains. Instead, 2020; Saucier, 2002). This means that the PiCD domains two narrower bipolar domains emerge, reflecting “impul- are well within the usual range (average r = .28) but the sivity” and “carelessness.” Furthermore, although both DSM categories are not (.63; Lenzenweger et al., 2007). On models fit the data about equally well in ESEM, the four- the other hand, overlap between personality domains forms factor solution was more congruent between the samples an integral part of a repeatedly replicated structure (Markon Gutiérrez et al. 769 et al., 2005), whereas categories are arbitrary conglomer- would have benefited from assessing other dimensional ates of traits whose structure has never been confirmed models (Ofrat et al., 2018) to investigate the concurrent (Muñoz-Champel et al., 2018). validity of the ICD-11 classification. This task is currently Finally, not all PiCD domains appear equally maladap- underway. Finally, as stated above, caseness is a useful but tive at first sight. Only negative affectivity, and to a lesser not particularly specific criterion of personality disorder, extent detachment and dissociality, can predict the SASPD, and it should be taken as an indicator of general psycho- and only negative affectivity substantially predicts case- pathological morbidity. More varied and specific indicators ness. Disinhibition and anankastia play little or no role in of personality pathology need to be tested in future studies. either case. These differences between domains are not Insofar as depression and anxiety symptoms may be over- obvious when the correlations are examined (Table 2; represented in our clinical sample, the results also need Oltmanns & Widiger, 2019; Somma et al., 2020) but emerge replication in other clinical populations. in multiple regression (Supplementary Table S9, available In sum, the PiCD and—with some reservations—the online). There are several possible explanations for these SASPD proved to be adequate in their Spanish versions, results. They may indeed indicate that some domains are far and can be used with confidence in clinical and research less maladaptive than others, or that their harmfulness is settings. Even so, the current ICD-11 classification is “on partly due to their association with negative affectivity (Vall probation only” (Tyrer et al., 2019, p. 496) and needs fur- et al., 2015). Certainly, the pervasive presence of negative ther refinement. Minor inconsistencies in the exact number affectivity in most psychopathology has long been known and nature of personality domains need to be solved; how- (Claridge & Davis, 2001; Widiger & Oltmanns, 2017). ever, evidence on this point is rapidly accumulating and it is However, the results may also highlight that the SASPD unlikely to be a problem. Contrarily, the most consequential gives more weight to negative affectivity than to any other issue—who has a disorder and how severe—seems as elu- domain, particularly disinhibition and anankastia, as men- sive as ever. There is still disagreement on whether diagno- tioned in the Method section. They may also reflect that sis should be based on the intensity of traits, the impact on caseness does not specifically identify personality disor- functioning, the decline in well-being, or the probability of ders. Even if we assume that needing clinical care is a com- psychopathology, disability, or death (Zimmerman et al., mon complication of maladapted personalities, and that 2018). It is not even clear whether different domains might many of our patients probably present a personality disor- require distinct definitions of severity. We hope that making der (Beckwith et al., 2014), our clinical sample was actually the PiCD and the SASPD available to a broader range of seeking professional help for a variety of mental problems; researchers will promote much-needed conceptual and as these problems are predominantly depressive and anxiety empirical advances in this area. disorders, negative emotionality must have been favored over all other domains (Claridge & Davis, 2001; Vall et al., Declaration of Conflicting Interests 2015). This bias may also explain why severity, as defined The author(s) declared no potential conflicts of interest with by the SASPD, is a worse predictor of caseness than nega- respect to the research, authorship, and/or publication of this tive affectivity, even if the poor reliability of the SASPD article. may also have played a nontrivial role. Nevertheless, all the above should not obscure the fact that constructs such as Funding “severity” or “dysfunction,” which increasingly form the basis for diagnosis, are still poorly understood and riddled The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This with ambiguities. For example, whereas the trait parts of work was supported by project PI15/00536, part of the Plan Estatal the ICD-11 and DSM-5 have a strong empirical basis and de Investigación Científica y Técnica y de Innovación 2013-2016, are quite similar with each other (Crego & Widiger, 2019; financed by the ISCIII Subdirección General de Evaluación and Somma et al., 2020), the severity/dysfunction parts still the cofinanced by the European Regional Development Fund need considerable research investment (Bach & Anderson, (ERDF, “A way to build Europe”; PI: F. Gutiérrez). 2018; Oltmanns & Widiger, 2019). This means that careful consideration is in order before we can completely trust the ORCID iDs new systems. Fernando Gutiérrez https://orcid.org/0000-0002-6662-1796 Our results should be interpreted in the light of several Anton Aluja https://orcid.org/0000-0002-7865-0287 limitations. First, we lack the borderline specifier, which José Ruiz https://orcid.org/0000-0002-8694-6176 was incorporated to the system as a sixth domain after this Miguel Gárriz https://orcid.org/0000-0002-0467-9490 study began. Even if this decision might be more pragmatic Alfonso Gutiérrez-Zotes https://orcid.org/0000-0002-4858 than scientific (Mulder & Tyrer, 2018; Tyrer et al., 2019), -1487 the role and possible redundancies of borderline traits Maria V. Navarro-Haro https://orcid.org/0000-0001-9726-3250 within the system warrant examination. Second, this study Bárbara Sureda https://orcid.org/0000-0001-8256-3613 770 Assessment 28(3) Supplemental Material Epidemiology, 68(4), 435-441. https://doi.org/10.1016/j. jclinepi.2014.11.021 Supplemental material for this article is available online. Ferrando, P. J., & Lorenzo-Seva, U. (2017). Program FACTOR at 10: Origins, development and future directions. Psicothema, References 29(2), 236-240. https://doi.org/10.7334/psicothema2016.304 American Psychiatric Association. (2013). Diagnostic and statis- Flora, D. B., LaBrish, C., & Chalmers, R. P. (2012). Old and new tical manual of mental disorders, 5th ed. Author. ideas for data screening and assumption testing for exploratory Bach, B., & Anderson, J. L. (2018). Patient-reported ICD-11 per- and confirmatory factor analysis. Frontiers in Psychology, 3, sonality disorder severity and DSM-5 level of personality Article 55. https://doi.org/10.3389/fpsyg.2012.00055 functioning. Journal of Personality Disorders, 34(2), 231- Gutiérrez, F., Ruiz, J., Peri, J. M., Gárriz, M., Vall, G., & Cavero, M. 249. https://doi.org/10.1521/pedi_2018_32_393 (2020). Toward an integrated model of pathological personal- Bach, B., Christensen, S., Kongerslev, M. T., Sellbom, M., & ity traits: Common hierarchical structure of the PID-5 and the Simonsen, E. (2020). Structure of clinician-reported ICD-11 DAPP-BQ. Journal of Personality Disorders, 34(Supplement personality disorder trait qualifiers. Psychological Assessment, C), 25-39. https://doi.org/10.1521/pedi_2019_33_431 2(1), 50-59. http://dx.doi.org/10.1037/pas0000747 Gutiérrez, F., Vall, G., Peri, J. M., Gárriz, M., & Garrido, J. Bach, B., Kongerslev, M. T., & Simonsen, E. (2019). Prevalence M. (2014). A hierarchical model of normal and abnormal and structure of self-other problems in SAPAS screening personality up to seven factors. Comprehensive Psychiatry, for personality disorder in a national sample. Personality 55(2), 326-335. https://doi.org/10.1016/j.comppsych.2013 and Mental Health, 14(2), 175-185. https://doi.org/10.1002/ .08.022 pmh.1470 Harkness, A. R., Finn, J. A., McNulty, J. L., & Shields, S. M. Bagby, R. M., & Widiger, T. A. (2020). Assessment of the ICD- (2012). The Personality Psychopathology–Five (PSY-5): 11 dimensional trait model: An introduction to the special Recent constructive replication and assessment literature section. Psychological Assessment, 32(1), 1-7. https://doi. review. Psychological Assessment, 24(2), 432-443. https:// org/10.1037/pas0000785 doi.org/10.1037/a0025830 Beckwith, H., Moran, P. F., & Reilly, J. (2014). Personality dis- Herpertz, S. C., Huprich, S. K., Bohus, M., Chanen, A., Goodman, order prevalence in psychiatric outpatients: A systematic lit- M., Mehlum, L., Moran, P., Newton-Howes, G., Scott, L., & erature review. Personality and Mental Health, 8(2), 91-101. Sharp, C. (2017). The challenge of transforming the diagnos- https://doi.org/10.1002/pmh.1252 tic system of personality disorders. Journal of Personality Brown, T. A. (2015). Confirmatory factor analysis for applied Disorders, 31(5), 577-589. https://doi.org/10.1521/pedi_2017 research (2nd ed.). Guilford Press. _31_338 Carnovale, M., Sellbom, M., & Bagby, R. M. (2020). The Hopwood, C. J., & Donnellan, M. B. (2010). How should the Personality Inventory for ICD-11: Investigating reliabil- internal structure of personality inventories be evaluated? ity, structural and concurrent validity, and method vari- Personality and Social Psychology Review, 14(3), 332-346. ance. Psychological Assessment, 32(1), 8-17. http://dx.doi. https://doi.org/10.1177/1088868310361240 org/10.1037/pas0000776 Kaiser, H. F., & Rice, J. (1974). Little Jiffy, Mark IV. Educational Claridge, G., & Davis, C. (2001). What’s the use of neuroticism? and Psychological Measurement, 34(1), 111-117. https://doi. Personality and Individual Differences, 31(3), 383-400. org/10.1177/001316447403400115 https://doi.org/10.1016/S0191-8869(00)00144-6 Kleindienst, N., Lyssenko, L., & Bohus, M. (2017). Validating Condon, D. M., & Mroczek, D. K. (2016). Time to move beyond the proposed ICD-11 domains: A comment on Mulder et al. the Big Five? European Journal of Personality, 30(4), 311-312. (2016). Personality and Mental Health, 11(4), 211-212. https://doi.org/10.1002/per.2060 https://doi.org/10.1002/pmh.1371 Crawford, M. J., Koldobsky, N., Mulder, R., & Tyrer, P. (2011). Kushner, S. C., Quilty, L. C., Tackett, J. L., & Bagby, R. M. (2011). Classifying personality disorder according to severity. The hierarchical structure of the Dimensional Assessment of Journal of Personality Disorders, 25(3), 321-330. https://doi. Personality Pathology (DAPP-BQ). Journal of Personality org/10.1521/pedi.2011.25.3.321 Disorders, 25(4), 504-516. https://doi.org/10.1521/pedi Crego, C., & Widiger, T. A. (2019). The convergent, dis- .2011.25.4.504 criminant, and structural relationship of the DAPP-BQ Lenzenweger, M. F., Lane, M. C., Loranger, A. W., & Kessler, and SNAP with the ICD-11, DSM-5, and FFM trait mod- R. C. (2007). DSM-IV personality disorders in the national els. Psychological Assessment, 32(1), 18-28. https://doi. comorbidity survey replication. Biological Psychiatry, 62(6), org/10.1037/pas0000757 553-564. https://doi.org/10.1016/j.biopsych.2006.09.019 De Clercq, B., De Fruyt, F., Van Leeuwen, K., & Mervielde, I. Lorenzo-Seva, U., & ten Berge, J. M. F. (2006). Tucker’s con- (2006). The structure of maladaptive personality traits in child- gruence coefficient as a meaningful index of factor similar- hood: A step toward an integrative developmental perspective ity. Methodology, 2(2), 57-64. https://doi.org/10.1027/1614 for DSM-V. Journal of Abnormal Psychology, 115(4), 639- -2241.2.2.57 657. https://doi.org/10.1037/0021-843X.115.4.639 Lorenzo-Seva, U., Timmerman, M. E., & Kiers, H. A. L. (2011). Epstein, J., Santo, R. M., & Guillemin, F. (2015). A review The Hull method for selecting the number of common factors. of guidelines for cross-cultural adaptation of question- Multivariate Behavioral Research, 46(2), 340-364. https:// naires could not bring out a consensus. Journal of Clinical doi.org/10.1080/00273171.2011.564527 Gutiérrez et al. 771 Markon, K. E., Krueger, R. F., & Watson, D. (2005). Delineating of the ICD-11 trait model. Psychological Assessment, 32(1), the structure of normal and abnormal personality: An integra- 60-71. https://doi.org/10.1037/pas0000763 tive hierarchical approach. Journal of Personality and Social Osborne, J. W. (2014). Best practices in exploratory factor Psychology, 88(1), 139-157. https://doi.org/10.1037/0022- analysis. CreateSpace Independent Publishing. 3514.88.1.139 R Core Team. (2018). R: A language and environment for sta- Maydeu-Olivares, A., & Coffman, D. L. (2006). Random intercept tistical computing. R Foundation for Statistical Computing. item factor analysis. Psychological Methods, 11(4), 344-362. http://www.R-project.org https://doi.org/10.1037/1082-989X.11.4.344 Reed, G. M. (2018). Progress in developing a classification of McCabe, G. A., & Widiger, T. A. (2020). A comprehensive com- personality disorders for ICD-11. World Psychiatry, 17(2), parison of the ICD-11 and DSM-5 section III personality 227-229. https://doi.org/10.1002/wps.20533 disorder models. Psychological Assessment, 32(1), 72-84. Rek, K., Thielmann, I., Henkel, M., Crawford, M., Piccirilli, L., https://doi.org/10.1037/pas0000772 Graff, A., Mestel, R, & Zimmermann, J. (2019). A com- McCrae, R. R., Zonderman, A. B., Costa, P. T. J., Bond, M. H., & prehensive evaluation of the Standardized Assessment of Paunonen, S. V. (1996). Evaluating replicability of factors in Severity of Personality Disorder (SASPD) in clinical and non- the Revised NEO Personality Inventory: Confirmatory factor clinical German samples. Manuscript submitted for publica- analysis versus Procrustes rotation. Journal of Personality and tion. https://psyarxiv.com/unqhm/ Social Psychology, 70(3), 552-566. http://dx.doi.org/10.1037 Rizopoulos, D. (2006). ltm: An R package for latent variable /0022-3514.70.3.552 modeling and item response theory analyses. Journal of Moran, P., Leese, M., Lee, T., Walters, P., Thornicroft, G., & Statistical Software, 17(5), 1-25. https://doi.org/10.18637/ Mann, A. (2003). Standardised Assessment of Personality– jss.v017.i05 Abbreviated Scale (SAPAS): Preliminary validation of a brief Rosseel, Y. (2012). lavaan: An R Package for structural equation screen for personality disorder. British Journal of Psychiatry, modeling. Journal of Statistical Software, 48(2), 1-36. https:// 183(3), 228-232. https://doi.org/10.1192/bjp.183.3.228 doi.org/10.18637/jss.v048.i02 Mulder, R. T., Newton-Howes, G., Crawford, M. J., & Tyrer, P. Saucier, G. (2002). Orthogonal markers for orthogonal factors: J. (2011). The central domains of personality pathology in The case of the Big Five. Journal of Research in Personality, psychiatric patients. Journal of Personality Disorders, 25(3), 36(1), 1-31. https://doi.org/10.1006/jrpe.2001.2335 364-377. https://doi.org/10.1521/pedi.2011.25.3.364 Skodol, A. E. (2011). Scientific issues in the revision of personal- Mulder, R. T., & Tyrer, P. (2018). Diagnosis and classification of ity disorders for DSM-5. Personality and Mental Health, 5(2), personality disorders. Current Opinion in Psychiatry, 32(1), 97-111. https://doi.org/10.1002/pmh.161 27-31. https://doi.org/10.1097/YCO.0000000000000461 Somma, A., Gialdi, G., & Fossati, A. (2020). Reliability and con- Muñoz-Champel, A., Gutiérrez, F., Peri, J. M., & Torrubia, R. struct validity of the Personality Inventory for ICD-11 (PiCD) (2018). Personality disorders are not as we thought: Hierarchical in Italian adult participants. Psychological Assessment, 32(1), factor structure at the criterion level. Journal of Personality 29-39. https://doi.org/10.1037/pas0000766 Assessment, 100(1), 16-29. https://doi.org/10.1080/00223891 Timmerman, M. E., & Lorenzo-Seva, U. (2011). Dimensionality .2017.1278700 assessment of ordered polytomous items with parallel anal- Ofrat, S., Krueger, R. F., & Clark, L. A. (2018). Dimensional ysis. Psychological Methods, 16(2), 209-220. https://doi. approaches to personality disorder classification. In W. J. org/10.1037/a0023353 Livesley, & R. Larstone (Eds.), Handbook of personality dis- Toland, M. D. (2014). Practical guide to conducting an item orders. theory, research, and treatment (2nd ed., pp. 72-87). response theory analysis. Journal of Early Adolescence, Guilford Press. 34(1), 120-151. https://doi.org/10.1177/0272431613511332 Olajide, K., Munjiza, J., Moran, P., O’Connell, L., Newton- Tyrer, P., Crawford, M., Mulder, R., Blashfield, R., Farnam, A., Howes, G., Bassett, P., Akintomide, G., Ng, N., Tyrer, P., Fossati, A., Kim, Y.-R., Koldobsky, N., Lecic-Tosevski, D., Mulder, R., & Crawford, M. J. (2018). Development and Ndetei, D., Swales, M., Clark, L. A., & Reed, G. M. (2011). psychometric properties of the Standardized Assessment The rationale for the reclassification of personality disorder in of Severity of Personality Disorder (SASPD). Journal of the 11th revision of the International Classification of Diseases Personality Disorders, 32(1), 44-56. https://doi.org/10.1521/ (ICD-11). Personality and Mental Health, 5(4), 246-259. pedi_2017_31_285 https://doi.org/10.1002/pmh.190 Oltmanns, J. R., & Widiger, T. A. (2018). A self-report mea- Tyrer, P., Crawford, M., Sanatinia, R., Tyrer, H., Cooper, S., sure for the ICD-11 dimensional trait model proposal: The Muller-Pollard, C., Christodoulou, P., Zauter-Tutt, M., Personality Inventory for ICD-11. Psychological Assessment, Miloseska-Reid, K., Loebenberg, G., Guo, B., Yang, M., 30(2), 154-169. https://doi.org/10.1037/pas0000459 Wang, D., & Weich, S. (2014). Preliminary studies of the Oltmanns, J. R., & Widiger, T. A. (2019). Evaluating the assess- ICD-11 classification of personality disorder in practice. ment of the ICD-11 personality disorder diagnostic system. Personality and Mental Health, 8(4), 254-263. https://doi. Psychological Assessment, 31(5), 674-684. https://doi.org/ org/10.1002/pmh.1275 10.1037/pas0000693 Tyrer, P., & Johnson, T. (1996). Establishing the severity of per- Oltmanns, J. R., & Widiger, T. A. (2020). The Five-Factor sonality disorder. American Journal of Psychiatry, 153(12), Personality Inventory for ICD-11: A facet-level assessment 1593-1597. https://doi.org/10.1176/ajp.153.12.1593 772 Assessment 28(3) Tyrer, P., Mulder, R., Kim, Y.-R., & Crawford, M. J. (2019). implications. World Psychiatry, 16(2), 144-145. https://doi. The development of the ICD-11 classification of personal- org/10.1002/wps.20411 ity disorders: An amalgam of science, pragmatism, and poli- Widiger, T. A., & Simonsen, E. (2005). Alternative dimensional tics. Annual Review of Clinical Psychology, 15(1), 481-502. models of personality disorder: Finding a common ground. https://doi.org/10.1146/annurev-clinpsy-050718-095736 Journal of Personality Disorders, 19(2), 110-130. https://doi. Tyrer, P., Reed, G. M., & Crawford, M. (2015). Classification, org/10.1521/pedi.19.2.110.62628 assessment, prevalence, and effect of personality disor- World Health Organization. (2018). International statistical clas- der. Lancet, 385, 717-726. https://doi.org/10.1016/S0140- sification of diseases for mortality and morbidity statistics 6736(14)61995-4 (11th Revision). https://icd.who.int/browse11/l-m/en Vall, G., Gutiérrez, F., Peri, J. M., Gárriz, M., Ferraz, L., Baillés, E., Wright, A. G. C., Thomas, K. M., Hopwood, C. J., Markon, K. E., & Obiols, J. E. (2015). Seven basic dimensions of personality Pincus, A. L., & Krueger, R. F. (2012). The hierarchical structure pathology and their clinical consequences: Are all personali- of DSM-5 pathological personality traits. Journal of Abnormal ties equally harmful? British Journal of Clinical Psychology, Psychology, 121(4), 951-957. https://doi.org/10.1037/a0027669 54(4), 450-468. https://doi.org/10.1111/bjc.12091 Zimmerman, M., Morgan, T. A., & Stanton, K. (2018). The sever- Widiger, T. A., & Oltmanns, J. R. (2017). Neuroticism is a fun- ity of psychiatric disorders. World Psychiatry, 17(3), 258-275. damental domain of personality with enormous public health https://doi.org/10.1002/wps.20569