by E Semrov — Se Sì: Quando capita, mi sento spaventato, preoccupato, o comunque la cosa mi crea problemi: ☐ Fortemente in disaccordo. ☐ In disaccordo. ☐ Indifferente.
177 KB – 11 Pages
PAGE – 1 ============
204JOURNAL OF PSYCHOPATHOLOGY 2018;24:204-214 © Copyright by Pacini Editore Srl OPEN ACCESS Received : April 4, 2018 Accepted : June 15, 2018 Correspondence Lorenzo Pelizza c/o CSM Petrella via E. Petrella 1/A, 42100 Reggio Emilia, Italy Ł Tel. +39 0522 339501 Ł Fax +39 0522 339523 Ł E-mail: lorenzo.pelizza@ausl.re.it Summary ObjectiveAmong current screeners for psychosis-risk mental states, the Prodromal Questionnaire-Brief (21 items) (PQ-B) is used. We aimed to assess reliability of the Italian version of the PQ-B in a young help-seeking sample. Methods Œ We included 151 individuals, aged 13-35 years, seeking help at the Reggio Emilia outpatient mental health services in a large semirural catchment area (550.000 inhabitants). Participants completed the Italian version of the PQ-B (iPQ-B) and were subsequently evaluated with the Comprehensive Assessment of At-Risk Mental States (CAARMS). We examined test-retest reliability, internal consistency and diag -nostic accuracy (i.e. sensitivity, specificity, positive and negative predictive values, and posi -tive and negative likelihood ratios) between PQ-B and CAARMS UHR-defined criteria using coefficient of stability (k), Cronbach™s alpha and Cohen™s kappa, respectively. ResultsThe iPQ-B showed excellent short term test-retest reliability (k = 0.891), high internal consist-ency ( = 0.876) and acceptable diagnostic accuracy (sensitivity = 91.4% at the proposed cut-off of ˜ 6 on total distress score).ConclusionsPsychometric properties of the iPQ-B were satisfactory. The iPQ-B is a suitable screening tool for routine use in mental health care services. Indeed, it is short and therefore easy to implement in routine assessment of early psychosis. Key wordsUltra-High Risk Ł Prodrome Ł Early Detection Ł Screening Ł Psychosis Ł Schizophrenia Ł AssessmentIntroductionSpecialist treatment for Ultra-High Risk (UHR) mental states of psychotic disorders can effectively reduce psychosis conversion rate˜ 1. However, identifying individuals with UHR remains a significant challenge˜ 2. Fo-cusing mainly on attenuated positive symptoms, McGorry et al. (2003)˜ 3 proposed the following UHR criteria: (a) Attenuated Psychotic Symptoms (APS), which represent subthreshold positive symptoms; (b) Brief Lim -ited Intermittent Psychotic Symptoms (BLIPS), which are transient positive symptoms that spontaneously disappear within 1 week; and (c) Genetic Risk and Functioning Deterioration syndrome (GRFD), a trait/state risk condition characterized by a history of psychosis in first-degree family members or a schizotypal personality disorder in the subject together with a low functioning for at least 1 month˜4. Translating the early detection/ intervention research framework into clinical care pathways relies, in part, on the recognition of these young people at the earliest point in their help- seeking trajectory˜ 5˜6.
PAGE – 2 ============
205ti-step procedure, and (b) to provide evidence-based interventions that are supposed to be effective in UHR/ FEP subjects (i.e. intensive case management, family psycho-education, individual cognitive-behavioral ther-apy, pharmacological treatment [as appropriate]). The first filtering step included a pre-clinical triage service, conducted by trained non-medical personnel, using the fiScreening Schedulefl for Psychosis (SS)˜ 17. Such triage was mainly meant to maximise appropriate referrals to the ReARMS project and avoid over-inclusion of sub -jects clearly outside the severity threshold for presumed psychosis risk spectrum. The second step included a comprehensive multidimensional battery including the iPQ-B, followed by the administration of the CAARMS to define the clinical status (i.e. psychosis risk, psychosis, or neither) and the consequent access to the ReARMS clinical-therapeutic pathways˜ 16. Complying with the declaration of Helsinki, relevant ethical approvals were locally sought for the study. Participants For the purpose of the study (i.e. field-testing the reli -ability of the iPQ-B in identifying UHR mental states), we focused on adolescent and young adult help-seekers, aged 13-35˜years, who were consecutively referred to all of child/adolescent and adult mental health servic-es of the Reggio Emilia Department of Mental Health between September 2012 and September 2017. In the present research, inclusion criteria were: (a) specialist help-seeking; (b) age between 13 and 35 years; and (c) presence of UHR criteria defined by the CAARMS (i.e. APS, BLIPS, and/or GRFD)˜ 4 at the initial assessment. Individuals who were below the CAARMS UHR thresh -old were considered as CAARMS-UHR negative cases. The exclusion criteria were modeled on the psychomet -ric approach adopted by Loewy et al. (2011)˜ 13 in the validation study of the original version of the PQ-B: (a) history of past frank psychotic episodes, either affective or schizophrenic (as described in the DSM-5)˜ 18; (b) his-tory of previous exposure to antipsychotics; (c) current substance dependence; (d) severe learning disability or known mental retardation (Intelligence Quotient <˜70); (e) neurological disease or any other medical disorder associated with psychiatric symptoms; (f) poor fluency in the Italian language; and (g) residence outside the catchment area. All these exclusion criteria have been applied after the SS administration in order to select a sample comparable to one assessed by Loewy et al. (2011)˜13.All help-seekers entering the ReARMS project agreed to participate to the research and gave their informed con -sent to the psychopathological evaluation, composed Œ˜among others˜ 16 Œ˜by the CAARMS (approved Italian translation by Raballo et al., 2013 [CAARMS-ITA])˜ 19 and the PQ-B (authorized Italian version by Preti and Rabal -Although structured interviews, such as the Com -prehensive Assessment of At-Risk Mental States (CAARMS)˜4 or the Structured Interview for Prodromal States (SIPS)˜7, can reliably diagnose UHR states˜ 8, they generally require extensive training to be administered and can take hours to be completed˜9. Therefore, an array of self-report screening tools has been developed to preselect potential UHR individuals for subsequent in-depth clinical assessment˜10. Accumulating empirical evidence suggests that these self-report instruments are sufficiently sensitive and specific to detect the ma -jority of those subjects that merit a more comprehensive evaluation for UHR or First-Episode Psychosis (FEP)˜11.The 92-item Prodromal Questionnaire (PQ-92)˜ 12 is the most commonly used screener for psychosis risk in the literature˜ 10. However, this instrument remains rather time-consuming for routine screening˜ 9. Thus, Loewy et al. (2011)˜13 developed a Brief 21-item version (PQ-B), focusing on the positive symptom items of the PQ-92, since they are the essential ones for interview-based diagnoses of symptomatic prodromal syndromes (i.e. APS and BLIPS). A cut-off of ˚˜6 on the PQ-B total dis -tress score predicted SIPS-UHR/psychosis diagnosis with high sensitivity (88%) and good specificity (68%)˜13.Overall, early intervention in young people at UHR for developing psychosis are less widespread in Italy than in other European countries˜ 14. In particular, some pilot programmes have focused specifically on early detec -tion and intervention in UHR young adults, aged 18-30 years (see Cocchi et al., 2008: fiProgramma 2000fl)˜ 15. Therefore, translating an easy and suitable self-report screening instrument (such as the PQ-B) into Italian lan -guage could lead to the implementation of specific ser-vices for UHR individuals within the framework of Italy™s National Health Service. To the best of our knowledge, no psychometric evaluation study on the PQ-B in an Ital-ian clinical sample has been reported in the literature to date. Thus, the current study was designed to test the reliability of the Italian version of the PQ-B (iPQ-B) in identifying young people at UHR of psychosis in a help-seeking community population.Materials and methodsSetting As detailed in Raballo et al. (2014)˜16, the fiReggio Emilia At-Risk Mental Statesfl (ReARMS) project is an early de -tection/intervention infrastructure implemented under the aegis of the fiRegional Project on Early Detection and Intervention in Psychosisfl in the Reggio Emilia De-partment of Mental Health. This project aims: (a) to iden -tify people with FEP and individuals at high clinical risk according to UHR criteria˜ 4 among help-seeking ado-lescents and young adults (13-35 years) through a mul -
PAGE - 3 ============
206(i.e. zooming in on prodromal experiences before the CAARMS-based interview) and the CAARMS assessors were blinded to the iPQ-B scores. lo, 2011 [iPQ-B])˜20 (Appendix˜I). While in chronological terms the iPQ-B was administered after the SS for psy -chosis, the meaning of its administration was different Appendix I The Italian version of the Brief (21 -item) Prodromal Questionnaire (iPQ -B) (Source: Loewy RL, Pearson R, Vinogradov S, et al. Psychosis risk screening with the Prodromal Questionnaire -Brief version (PQ -B). Sc hizophr Res 2011;129:42 -6). (Authori zed Italian version by Preti A, Raballo A. Studio CAPIRE. Cagliari Psychosis: Investigation on Risk Emergence , 2011). Per cortesia, indica se hai avuto i seguenti pensieri, sentimenti ed esperienze nel corso dell™ultimo mese segnando fiSìfl o fiNofl per cias cuna domanda. Non tenere conto di esperienze che si ve rificano sotto influenza di alc ol, droghe o farmaci che non ti erano stati prescritti . Se rispondi fiSìfl a una domanda indica anche quanto disagio ti ha causato quell™esperienza [quanto spiacevole è stat a per te quell™esperienza]. 1. Capita talvolta che gli ambienti abituali ti sembrino strani, confusi, minacciosi o irreali? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 2. Hai mai sentito suoni insoliti come esplosioni, schiocchi, sibili, schianti o squilli nelle tue orecchie? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 3. Le cose che vedi ti appaiono differenti dal modo in cui sono abitualmente (più luminose o più scure, più larghe o più piccole, comunque cambiate in qualche modo)? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: For temente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 4. Hai avuto esperienze con la telepatia, le forze psichiche o la predizione del f uturo? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 5. Ti sei sentito come se non avessi controllo sulle tue idee o pensieri? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 6. Hai difficoltà a spiegarti, perché fai troppe digressi oni o devi dal filo del discorso quando parli? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 7. Hai l™impressione o la convinzione di essere dotato in modo particolare o di possedere un talento speciale? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 8. Hai l™impressione che altre persone ti stiano tenendo d™occhio o parlino di te? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 9. Hai talvolta sensazioni strane sulla pelle o appena al di sotto, come insetti che camminano? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo
PAGE - 4 ============
207[SOFAS] module)˜ 4. It takes approximately 1-1.5˜hours to be administered and consists of 27˜items (each one scored in terms of frequency/duration [0-6] and intensi -ty [0-6]). Those items are clustered in seven subscales: (a) fiPositive Symptomsfl, (b) fiCognitive Change, Atten-Measures The CAARMS is a semi-structured clinical interview de -signed to cover different aspects of attenuated psycho -pathology as well as functioning (via the integrated So-cial and Occupational Functioning Assessment Scale 10. Ti capita talvolta di essere distratto all™improvviso da s uoni distanti dei quali generalmente non sei consapevole? [ai quali normalmente non presti attenzione] SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 11. Hai la sensazione che qualche persona o forza ti stia accanto anche se tu non puoi vederla? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 12. Ti preoccupi talvolta d el fatto che qualcosa nella tua mente non funzioni correttamente? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 13. Hai mai avuto la sensazione di non esistere, o che il mondo non esiste, o di essere morto? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 14. Qualche volta ti sei sentito confuso sulla natura reale o immaginaria di un ™esperienza? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 15. Hai delle idee o delle conv inzioni che altre persone troverebbero insolite o bizzarre? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 16. Senti che parti del tuo corpo sono cambiate in qualche modo , o che funzionano in modo diverso? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 17. I tuoi pensieri sono talvolta così forti che puoi quasi udirli? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 18. Ti capita di provare sfiducia o essere sospettoso riguardo alle altre persone? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 19. Hai visto oggetti insoliti come bagliori, fiamme, lampi accecanti o figure geometriche? SI NO Se Sì: Quando capita, mi sento spavent ato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 20. Hai visto cose che altri non riescono a vedere o non sembrano notare? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo 21. Capita talvolta che le persone abbiano difficoltà a capire quello che stai dicendo? SI NO Se Sì: Quando capita, mi sento spaventato, preoccupato, o com unque la cosa mi crea problemi: Fortemente in disaccordo In disaccordo Indifferente D™accordo Fortemente d™accordo
PAGE - 5 ============
208of ˚˜3 symptoms endorsed was supported13, whereas in a lower prevalence sample from a similar setting, a higher threshold of ˚˜9 was identified (albeit below 75% sensitivity)˜22.Statistical analysis Data were analyzed using the fiStatistical Package for Social Sciencefl (SPSS) 18.0 for Windows˜24. For the specific purposes of this study, the sample was dichot -omized as follows: UHR+ (i.e. those who were above CAARMS UHR threshold [that is APS, BLIPS and/ or GRFD]), and UHR- (i.e. those who are below such threshold)˜ 4. The two subgroups were compared on socio-demographic, clinical, and psychopathological parameters. Categorical data were analysed using Chi- squared test with Yates™ correction. Quantitative vari -ables were examined using the Mann-Whitney™s U test or the Student™s t-test Œ˜as appropriate˜Œ. Following the psychometric approach adopting by Kot -zalidis et al. (2017)˜25 in the validation study of the Italian version of the PQ-92 in order to compare their and our results, in the present research we measured short-term test-retest reliability of the iPQ-B over two weeks calcu -lating the coefficient of stability˜ 26 on a subsample of 15 participants who had scored ˚˜6 on the iPQ-B total dis -tress score (i.e. the best recommended original cut-off proposed by Loewy et al., 2011)˜ 13 at the baseline as-sessment. This rather short-time interval was chosen to limit the possible impact of both symptomatic changes and memory effects˜ 27. According to Heise (1969)˜ 26, we interpreted test-retest reliability coefficients as follows: ˚˜0.90 excellent reliability, 0.81-0.90 good reliability, 0.71-0.80 acceptable reliability, 0.61-0.70 questionable reliability, 0.51-0.60 poor reliability, and ˛˜0.50 unaccep -table reliability. Moreover, we examined long-term test-retest reliability of the iPQ-B calculating the coefficient of stability within all the participants who had scored ˚˜6 on the iPQ-B total distress score at the initial assessment (n˜=˜123). As additional measure of reliability, the internal consist -ency of the iPQ-B was assessed using the Cronbach™s statistics within the total sample. A score above 0.65 represented a sufficient internal consistency˜ 6. We al -so examined how each PQ-B item correlated with the recommended total score (i.e. the total distress score). Correlations less than r˜=˜0.30 indicated that the item might need to be removed from the questionnaire to make it more reliable˜ 28. Finally, we were interested in Cronbach™s alpha value if each iPQ-B item was deleted. If this score went up after item deletion, removal should be considered to ameliorate screening tool reliability of the instrument˜28.Furthermore, we investigated the concurrent validity of the iPQ-B by comparing its results to CAARMS outcomes. In the total sample, we examined diagnostic accuracy tion and Concentrationfl, (c) fiEmotional Disturbancefl, (d) fiNegative Symptomsfl, (e) fiBehavioral Changefl, (f) fiMotor/Physical Changesfl, and (g) fiGeneral Psychopa-thologyfl. The CAARMS fiPositive Symptomsfl subscale, which covers delusions, hallucinations and thought disorder, is used to determine the UHR criteria˜ 4. UHR status is defined as follows: (a) GRFD group: schizo -typal personality disorder in the subject or history of psychosis in a first-degree family member associated with 30% drop in functioning for ˛˜1˜month or chronic low functioning (the decline in functioning is estimated by subtracting the current SOFAS score from the high -est SOFAS score in the past year); (b) APS group: sub -threshold positive psychotic symptoms within the past 12˜months; and (c) BLIPS group: criteria for psychotic disorder met for <˜7 day and remitting spontaneously (i.e. without antipsychotic medication).CAARMS interviews are conducted by specialized per -sonnel including clinical psychologists and psychia-trists, who underwent collective supervision by the main author of the approved Italian translation˜ 19, who was trained at Orygen, the National Youth Research Cent -er in Melbourne, Australia. The inter-rater reliability of these assessments was ensured by regular CAARMS scoring workshops and supervision sessions.The PQ-B13 is a self-report questionnaire used to screen individuals for the risk of psychosis. It only takes approximately 4˜minutes to be completed and com -prises of 21˜items recording positive symptoms expe -rienced over the past month. For each endorsed symp-tom, responders rate whether they found it distressing or impairing, ranging from˜1 (fistrongly disagreefl) to 5 (fistrongly agreefl), with a 4 or 5 indicating distress˜ 13. The PQ-B has been adopted as a screening tool using the total number of items endorsed (fisymptom total scorefl), the number of items that are identified as distressing (fidistressing item total scorefl) (both range˜0-21), and the total distress score (range˜0-105), with the latter method recommended by Loewy et al. (2011)˜ 13. In a re -cent systematic review on psychosis risk screening us -ing the PQ in its different iterations, Savill et al. (2017)˜ 21 examined eight diagnostic accuracy studies using the PQ-B. Of these, one evaluated the number of distress -ing symptom endorsed in an UHR/psychosis-enriched sample and found a threshold of ˚˜4 distressing items as optimal cut-off˜ 22. Six studies examined the total dis-tress score for screening: in samples with a very high prevalence (~ 80%) of UHR/psychosis individuals, a total distress score of ˚˜6 was supported˜ 13, whereas in similar settings with a much lower prevalence (<˜40%), a total distress score ˚˜18 was recommended˜ 22˜23 . Final-ly, four studies adopted the total number of symptoms endorsed as cut-off: in a sample with a very high pro -portion of UHR/psychotic participants, a cut-off score
PAGE - 6 ============
209In comparison with UHR-, UHR+ individuals showed significantly higher iPQ-B scores (Tab.˜I). To calculate short-term test-retest reliability, the iPQ-B was re-ad -ministered to 15 participants who had scored ˚˜6 on total distress score at the first assessment. Their socio- demographic characteristics were comparable to those of the total sample, with a mean age of 19.94 years and a SD of 4.89 years. Eight (53%) participants were fe -males. The coefficient of stability was 0.891 for iPQ-B total distress score, indicating good to excellent short- term test-retest reliability˜ 26.To examine long-term test-retest reliability, the iPQ-B was administered over˜1 year to 123 individuals who had scored ˚˜6 on total distress score at the baseline. Their demographic features were comparable to those of the entire sample, with a mean age of 20.10 years and a SD of 5.01˜years. Sixty-three (51.2%) subjects were females. The coefficient of stability was 0.395, in -dicating unacceptable long-term test-retest reliability˜ 26.Across the total sample, the iPQ-B total distress score showed a Cronbach™s alpha of 0.876. All item-total corre -lations were higher than 0.30, with the exception of item˜ 9 (fiDo you sometimes get strange feelings on or just be -neath your skin, like bugs crawling?fl) (r˜=˜0.213) (Tab.˜II). Therefore, most item appeared to be worthy of retention, resulting in a decrease in the alpha if deleted. Exception to this was item 9, whose deletion increased Cronbach™s alpha up to the value of 0.879. Thus, removal of this item can be considered. measures (i.e. sensitivity, specificity, positive and nega -tive predictive values [PPV and NPV], and positive and negative likelihood ratios [LR+ and LR-], that balance sensitivity against specificity). As an additional measure of concurrent validity, the correspondence of positive re -sults on the iPQ-B (i.e. a total distress score ˚ 6 or, as alternatives, the recommended symptom total score ˚˜3 or item distressing total score ˚˜4) and on the CAARMS (i.e. a score ˚˜3 on at least one positive symptom item) was also examined by Cohen™s kappa statistics. Finally, to explore which iPQ-B items were likely to be more predictive of CAARMS UHR diagnosis, we em -ployed a forward stepwise logistic regression analysis, with iPQ-B item scores as independent varables and dichotomized CAARMS diagnoses (i.e. UHR- vs UHR+) as dependent variable.ResultsOver the course of the study, 151 individuals (79˜fe -males and 72˜males; mean age ±˜Standard Deviation [SD] =˜20.00˜±˜5.78) consecutively participated at the intake interview within the ReARMS protocol. Table˜I shows screening outcomes and demographic charac -teristics of the total sample and the two subgroups, i.e. UHR+ (n˜=˜70) and UHR- (n˜=˜81). No significant dif -ferences were found in terms of gender, ethnic group, mother tongue, age, years of education, and Duration of Untreated Illness (DUI)˜ 6.TABLE I. CAARMS criteria, demographic and clinical data. Total sample (n = 151) UHR-(n = 81)UHR+(n = 70)2/t/ZGender (female) Ethnic group (Caucasian) Mother tongue (Italian) Age Years of Education DUI (in weeks) iPQ-B symptom total score (range 0-21) iPQ-B total distress score (range 0-105) iPQ-B distressing item total score (range 0-21) 79 (52.3%)130 (86.1%) 138 (91.4%) 20.00 (5.78)11.34 (2.39) 69.59 (51.00) 7.31 (4.90) 24.80 (18.98) 3.85 (3.83)41 (50.6%)69 (85.2%)76 (93.8%) 20.26 (6.44)11.47 (2.40) 66.39 (54.65)5.68 (4.42)18.14 (15.95) 2.66 (3.07)38 (54.3)61 (87.1%) 62(88.6%)19.54 (4.53) 11.19 (2.38) (47.34) 9.24 (4.76) 32.65 (19.37) 5.25 (4.17) 0.2030.012 0.7350.5410.726-0.572-4.467*-4.717* -3.966**˜p˜<˜0.001.Frequencies and percentages, mean (standard deviation), chi-squared ( 2) test (with Yates correction), Student™s t test, and Mann-Whitney U test (Z) values are reported.
PAGE - 8 ============
211help-seeking sample (age between 12 and 35 years), Loewy et al. (2011)˜13 found an excellent PQ-B internal consistency with a Cronbach™s alpha of 0.853. Moreo -ver, Xu et al. (2016)31 showed an overlapping internal consistency (˜=˜0.897) in Chinese help-seeking indi -viduals (aged 15-45˜years) visiting a general mental health setting. Therefore, PQ-B appears to be reliably good in different samples and cultures. Moreover, in our sample, iPQ-B demonstrated a Cronbach™s alpha value that we consider as satisfactory internal consistency for a screener that has to come before a clinical interview˜ 6.In a recent validation study of the Italian version of the PQ-92, Kotzalidis et al. (2017)˜ 25 re-administered the instrument to 15˜individuals two weeks after first as -sessment and found excellent short-term test-retest reliability(coefficient of stability =˜0.942 for PQ-92 total score). Similarly, we found a coefficient of stability equal to 0.891, indicating a good to excellent short-term (two- week) test-retest reliability of the iPQ-B. Based on what Kotzalidis et al. (2017)˜25 suggested, we also addressed longer term test-retest reliability ad -ministering the iPQ-B over˜1 year to all the participants who had scored ˚˜6 on total distress score at first as -sessment. We found a coefficient of stability =˜0.395, indicating unacceptable long-term test-retest reliability. According to Michel et al. (2014)˜ 27, this finding sug-gests that the self-report screening questionnaire as -sessed a fluctuating condition rather than a trait char-acteristic, i.e. a condition itself that varied between test and retest. When examining these results, some methodological peculiarities of the current study shall be considered. Indeed, ReARMS is a clinically project providing evidence-based interventions that are sup -posed to be effective in UHR individuals (i.e. intensive positive regression coefficients, iPQ-B˜9 showed a nega -tive one. The percentage of correct diagnosis using this model for predicting CAARMS UHR diagnosis was 76.7%. DiscussionAim of the current was to evaluate the reliability of PQ-B in an Italian clinical sample of young people at UHR of psychosis. Introducing and promoting the routinary use of the Italian version of a validated assessment tool to detect UHR subjects in the general help-seeking popu-lation (such as the iPQ-B) could positively impact on the implementation of specific services for early detection and intervention on UHR individuals within the frame-work of Italy™s National Health Service. In the current study, we therefore examined test-retest reliability and internal consistency of the iPQ-B in consecutive young help-seekers attending all of child/adolescent and adult mental health services of the Reggio Emilia Department of Mental Health. In comparison with UHR-, UHR+ individuals showed significantly higher iPQ-B total scores. On a dimension -al level Œ˜as expected on the basis of the PQ-B item composition - these findings suggest that increasing PQ-B scores are associated with the severity of both psychotic and general psychopathology, as well as the intensity of distress related to prodromal symptoms. We found excellent reliability of the iPQ-B with re -gard to internal consistency of the total distress score (˜=˜0.876). Removal of item˜9 (fiDo you sometimes get strange feelings on or just beneath your skin, like bugs crawling?fl), which resulted in a slight increase in Cronbach™s alpha value up to 0.879, can be consid -ered. Similarly, in a comparable adolescent/young adult TABLE III. Logistic regression of dichotomized CAARMS UHR diagnoses by iPQ-B items. iPQ-B item BSEWald dfpORPQ-B5PQ-B6PQ-B8PQ-B9PQ-B13 Constant 0.195 0.189 0.401 -0.4430.237-0.2260.0890.0970.0890.116 0.135 0.2294.8493.79520.115 14.522 3.113 1.240 1111110.0280.0480.0000.0000.0450.2851.216 1.208 1.493 0.6421.268 0.775Overall model ˜t test 2 = 62.653, p = 0.000 Associated strength Cox-Snell R2 = 0.228, Negelkelke R2 = 0.319 iPQ-B: Italian Prodromal Questionnaire Œ Brief version, CAARMS: Comprehensive Assessment of At-Risk Mental States; UHR: Ultra-High Risk mental states; B: regression coefficient; SE: standard error; Wald: Wald statistic value; df: degree of freedom; p: statistical significance; and OR: odd ratio
PAGE - 9 ============
212Compared to the PQ-B cut-off of ˚˜6 on total distress score, the proposed ˚˜3 threshold on symptom total score (range˜0-21)˜ 13 slightly increased specificity value up to 28.4%, while maintaining a 91.4% sensitivity. Fi -nally, using the recommended PQ-B cut-off of˜˚˜4 on distressing item total score (range˜0-21)˜ 21, even if spec-ificity increased to 75.3%, a sensitivity value of 61.4% is quite low and means that a relevant number of people who would appropriate for early intervention services are not being identified. Among all iPQ-B items, five symptoms (i.e. iPQ-B5 [fiHave you felt that you are not in control of your own ideas or thoughts?fl], iPQ-B6 [fiDo you have difficulty getting your point across because you ramble or go off the track a lot when you talk?fl], iPQ-B8 [fiDo you feel that other people are watching you or talking about you?fl], iPQ-B9 [fiDo you sometimes get strange feelings on or just beneath your skin, like bugs or crawling?fl], and iPQB13 [(fiHave you ever felt you don™t exist, the world does not exist, or that you are dead?fl]) correctly predicted UHR- vs UHR+ diagnoses (76.7% of correct diagnostic ascription in the logistic regression model). It should be noted that iPQ-B9 showed a significant nega -tive correlation coefficient, thus making its absence full of meaning. This is in line with the above mentioned result emerging from Cronbach™s statistics, which con -cluded for the removal of iPQ-B item˜9. Although this item is intended to be an attenuated psychotic symp-tom, the results of the logistic regression suggest that on a group-level, subjects endorsing it might actually report paresthesic sensations (e.g. feelings of pins and needles is when their arms or legs fifall asleepfl) or somato-vegetative expressions of anxiety. In their early pilot study, Yung et al. (1996)˜ 32 reported that three kinds of attenuated positive symptoms (i.e. perceptual abnor -malities, suspiciousness, and delusional mood) could account for 62%, 71%, and 62%, respectively, of symp -toms that prodromal individuals experienced. Limitations Firstly, a possible limitation of this study is that the iPQ- B was completed in a population plausibly fienrichedfl for the target diagnoses, i.e. young help-seekers with clinical features of possible psychosis. Therefore, the current field-test of the iPQ-B was not meant to identify cut-offs applicable to the general population, in which the psychometric endorsement of so-called psychotic-like experiences might occasionally occur, yet with tran -sient temporal pattern, not necessarily accompanied by distress or treatment seeking, and not inevitably fol -lowed by a transition to psychosis˜2˜14 . Indeed, a certain number of false positives would be identified.Another limitation is that since the SS for psychosis˜ 17 was used in the eligibility triage for the ReARMS proto -col (i.e. before the iPQ-B administration), this is likely to case-management, family psycho-education, individual cognitive-behavioral therapy within the framework of assertive community treatment). Precisely because pro -viding the optimal treatment for the help-seekers was the main ethical mandate in our clinical setting, our treatments were not controlled (e.g. against placebo group or other treatments), but evenly delivered to all UHR participants˜ 6.In the original study validating the PQ-B, Loewy et al. (2011)˜13 observed a good to excellent concurrent validity with CAARMS diagnoses in a sample of adolescent and young adult help-seekers attending to an Early Interven-tion Psychosis (EIP) service. A cut-off of ˚˜6 on total dis -tress score had a high sensitivity (88%) and good speci -ficity (68%) in discriminating between people with UHR/psychosis and individuals without CAARMS diagnosis.With regard to the diagnostic accuracy at the proposed PQ-B cut-off of ˚˜6 on total distress score˜ 13, sensitivity in our sample (91.4%) was substantially in line with previ -ously reported for the PQ in its various versions˜ 10. How-ever, this result was much higher to that (62%) observed by Kotzalidis et al. (2017)˜25 in the validation study of the Italian version of the 92-item PQ. Moreover, at the pro -posed PQ-B cut-off of ˚˜6, our PPV (51.2%) was consist -ent with previously reported, with values ranging between 29% and 44%˜21. In particular, PPV was equal to 38% in the validation study of the Italian version of the 92-item PQ˜25. The difference between these findings may be the result of differences in selection procedures. In fact, first screening procedure in the ReARMS protocol included a triage service using the SS for psychosis˜17, which prob -ably excluded a certain amount of true negative cases. In our sample, specificity (approximately 25%) was low -er than previously reported. Indeed, specificity values were good to excellent in the original study validating the PQ-B at a cut-off of ˚˜6 on total distress score˜ 13 and in the validation study of the Italian version of the 92-item PQ˜25 (68% and 82%, respectively). Likewise, our NPV (approximately 77%) was good, but slightly lower than previously reported, with values ranging between 90% and 100%˜2˜21 . In this regards, Kotzalidis et al. (2017)˜ 25 found a NPV of 91% in the validation study of the Italian version of the 92-item PQ. The difference between these findings may be the result of the same differences in selection procedures previously mentioned. However, according to Loewy et al. (2011)˜ 13, for screen -ing purposes, greater weighting should be given to sen -sitivity over specificity as part of a two-step screening process. Indeed, low sensitivity scores mean that a cer -tain number of people who would appropriate for early intervention are not being identified. Consequently, in most cases, having a few more false positives is less of an issue than missing appropriate individuals from a clinical perspective.
177 KB – 11 Pages