Two-step Cluster Analysis Application to a Sample of Psychiatric Inpatients at Psychiatric Service of Diagnosis and Care

Abstract

IntroductionRecent findings demonstrated significant overlaps among major psychiatric disorders on multiple neurocognitive domains. However, it is not clear which are the cognitive functions that contribute to this phenomenon.ObjectivesTo find the optimal clustering solution using the two-step cluster analysis on a sample of psychiatric patients.AimsTo classify into subgroups a cross-diagnostic sample of psychiatric inpatients on the basis of their neurocognitive profiles.MethodsSeventy-one patients with psychotic, bipolar, depressive and personality disorders hospitalised at Psychiatric Diagnosis and Care Service of Bufalini Hospital of Cesena participated in the study. The symptomatology was assessed using Health of the Nation Outcome Scales-Roma and Brief Psychiatric Rating Scale. Cognitive functions were evaluated using Tower of London, Modified Wisconsin Card Sorting Test, Judgment and Verbal Abstract Tasks test, Raven matrices, Attentional Matrices, Stroop Test and Mini Mental State Examination. Two-step cluster analysis was conducted using the standardized scores of each neurocognitive test.ResultsTwo groups were obtained:– group 1, with good cognitive performances;– group 2, with almost all subjects having impaired cognitive performances.Executive functions and attention are the major determinants of the cluster solution. The clusters did not differ on socio-demographic correlates. Different diagnoses were equally distributed amongst the clusters.ConclusionsTwo-step cluster analysis was useful in identifying subgroups of psychiatric inpatients with different cognitive functioning, overcoming other cluster techniques limitations. According to former literature, these results confirm a continuum of severity in cognitive impairment across different psychiatric disorders.Disclosure of interestThe authors have not supplied their declaration of competing interest.

Transdiagnostic neurocognitive subgroups and functional course in young people with emerging mental disorders: a cohort study

“…Notably, more than half of the participants with a primary psychotic disorder were allocated to the normal range or intermediate subgroups (supplementary Table 5), highlighting that within-diagnosis neurocognitive heterogeneity may be obscured by diagnosis-level comparisons, which tend to report a gradient of worst impairment in psychotic disorders, followed by bipolar and depressive disorders. 13,20 However, consistent with other transdiagnosticor cross-diagnosticstudies, 17,20,21 participants with psychotic disorders were overrepresented in the global impairment subgroup. Biological factors such as brain abnormalities or genetic risk for neurocognitive impairment may be important factors for such individuals with global impairment, as reported in several studies of neurocognitively impaired subgroups with schizophrenia.…”

“…Our observation that primary diagnostic groups were distributed across the three cluster groups is consistent with previous work in schizophrenia and bipolar disorder 20 and in a transdiagnostic inpatient sample. 21 In the current study, around one-quarter of the global impairment cluster had a primary depressive disorder and another quarter had a primary bipolar or anxiety disorder. Notably, more than half of the participants with a primary psychotic disorder were allocated to the normal range or intermediate subgroups (supplementary Table 5), highlighting that within-diagnosis neurocognitive heterogeneity may be obscured by diagnosis-level comparisons, which tend to report a gradient of worst impairment in psychotic disorders, followed by bipolar and depressive disorders.…”

“…[14][15][16][17] Recently, evidence of similar subgroups have been shown within samples of participants with depressive 18 and bipolar disorders, 19 and notably, across broader samples comprised of people with multiple major diagnostic groups. [20][21][22][23] Taken together, these findings of homogeneous subgroups within diagnostic groups suggest that neurocognitive impairment may follow a continuum of severity distributed across mental disorders, with data-driven subgroups potentially representing a more useful level of analysis as regards neurocognition and associated factors.…”

Home Occupant Archetypes

“…This study utilized the TwoStep cluster analysis approach. Although there are several clustering approaches, this method has traditionally been used in marketing, for customer segmentation, and gained recent popularity in health-related research, especially for the exploration of health behaviours, eating disorders, and alcoholism in the homeless, among others (Ambrosini et al, 2017;Dietrich, Schuster, & Connor, 2014;Fleury, Grenier, & Bamvita, 2015;Pugh & Waller, 2017;Zaretzky, Flatau, Spicer, Conroy, & Burns, 2017). The advantage of the TwoStep analysis over other types of clustering approaches is that it allows for the segmentation of both categorical and continuous variables in a simultaneous manner, allowing for a minimal degree of data preparation and handling for the analysis.…”

Questionnaire testing, validating, and preliminary results

“…This study utilized the TwoStep cluster analysis approach. Although there are several clustering approaches, this method has traditionally been used in marketing, for customer segmentation, and gained recent popularity in health-related research, especially for the exploration of health behaviours, eating disorders, and alcoholism in the homeless, among others (Ambrosini et al, 2017;Dietrich, Schuster, & Connor, 2014;Fleury, Grenier, & Bamvita, 2015;Pugh & Waller, 2017;Zaretzky, Flatau, Spicer, Conroy, & Burns, 2017). The advantage of the TwoStep analysis over other types of clustering approaches is that it allows for the segmentation of both categorical and continuous variables in a simultaneous manner, allowing for a minimal degree of data preparation and handling for the analysis.…”