Development and preliminary validation of the Level of Care Index (LOCI) from the Personality Assessment Inventory (PAI) in a psychiatric sample.

Abstract

Research over the last decade has been promising in terms of the incremental utility of psychometric tools in predicting important clinical outcomes, such as mental health service utilization and inpatient psychiatric hospitalization. The purpose of this study was to develop and validate a new Level of Care Index (LOCI) from the Personality Assessment Inventory (PAI). Logistic regression was initially used in a development sample (n = 253) of psychiatric patients to identify unique PAI indicators associated with inpatient (n = 75) as opposed to outpatient (n = 178) status. Five PAI variables were ultimately retained (Suicidal Ideation, Antisocial Personality-Stimulus Seeking, Paranoia-Persecution, Negative Impression Management, and Depression-Affective) and were then aggregated into a single LOCI and independently evaluated in a second validation sample (n = 252). Results indicated the LOCI effectively differentiated inpatients from outpatients after controlling for demographic variables and was significantly associated with both internalizing and externalizing risk factors for psychiatric admission (range of ds = 0.46 for history of arrests to 0.88 for history of suicidal ideation). The LOCI was additionally found to be meaningfully associated with measures of normal personality, performance-based tests of psychological functioning, and measures of neurocognitive (executive) functioning. The clinical implications of these findings and potential utility of the LOCI are discussed.

Effectively Assessing Treatment Needs in Incarcerated Seriously Mentally Ill Individuals: The Utility of the Personality Assessment Inventory—Level of Care Index

“…The results of this study add support to the validity of the LOCI as a measure that can assist clinicians in determining appropriate level of care. The findings expand on previous research in inpatient and outpatient settings (Sinclair et al, 2013(Sinclair et al, , 2015 by demonstrating the utility of the LOCI among incarcerated individuals with serious mental illness. In addition, the LOCI predicted level of psychiatric symptoms, resilience, and treatment trajectories.…”

“…With the treatment of serious mental illness shifting from long-term psychiatric treatment facilities to acute psychiatric hospitals and outpatient services, there is added pressure to quickly identify the appropriate level of care and efficiently utilize scarce mental health resources (Sinclair et al, 2013). While a myriad of assessment instruments exist that can aid in determining the level of care among psychiatric patients, completing the assessment is often challenging due to time constraints and limited resources (Sinclair et al, 2013). The use of self-report measures may serve as an alternative means to circumvent the challenge of limited resources and time (Williams, Weiss, Edens, Johnson, & Thornby, 1998).…”

“…The use of self-report measures may serve as an alternative means to circumvent the challenge of limited resources and time (Williams, Weiss, Edens, Johnson, & Thornby, 1998). Recently, Sinclair et al (2013) developed and provided evidence for the validity of the Personality Assessment Inventory’s (PAI’s) Level of Care Index (LOCI), which has been found effective in identifying individuals who may benefit from increased levels of care, such as inpatient hospitalization (Sinclair et al, 2015).…”

Using the Personality Assessment Inventory to Identify ADHD-Like Symptoms

“…Stability of the clinical scales has also been demonstrated, with a test–retest reliability coefficient of .83 (Morey, 1991). Following the method used by Sinclair and colleagues (2013), for those scales that contain subscales, the subscales as opposed to the full scales were used as a means of increasing precision. This resulted in a total of 43 PAI scales being used.…”

Object relations and real life relationships: A cross method assessment

“…Another point worth noting regards the normal aspect of the sample. Most of the studies done with the SCORS used pathological samples (e.g., Sinclair, Slavin-Mulford, Antonius et al, 2013;Stein, Siefert, Stewart & Hilsenroth, 2011;Stein, Slavin-Mulford, Sinclair, Siefert & Blais, 2012). In these samples, OR levels are predicted and indeed shown to be significantly lower and consequently pathological, therefore the differences in OR measures are more evident for all subscales.…”